Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Conservation genetics of whitebark pine (Pinus albicaulis Engelm) in British Columbia Krakowski, Jodie 2001

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

Item Metadata


831-ubc_2002-0133.pdf [ 5.77MB ]
JSON: 831-1.0099640.json
JSON-LD: 831-1.0099640-ld.json
RDF/XML (Pretty): 831-1.0099640-rdf.xml
RDF/JSON: 831-1.0099640-rdf.json
Turtle: 831-1.0099640-turtle.txt
N-Triples: 831-1.0099640-rdf-ntriples.txt
Original Record: 831-1.0099640-source.json
Full Text

Full Text

CONSERVATION GENETICS OF W H I T E B A R K PINE (Pinus albicaulis  Engelm.) IN BRITISH C O L U M B I A  by JODIE KRAKOWSKI B.Sc. (For. Sci.), The University of British Columbia, 1998 A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F THE REQUIREMENTS FOR T H E DEGREE O F Master o f  Science  in T H E FACULTY O F G R A D U A T E S T U D I E S Faculty of Forestry ( D e p a r t m e n t of Forest Science)  W e accept this thesis as complying with the required standard  T H E U N I V E R S I T Y O F BRITISH C O L U M B I A December 2001 ©Jodie Krakowski, 2001  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  University  of  British  Columbia,  for  this or  thesis  reference  thesis by  this  for  his thesis  and  study.  scholarly  or  her  for  purposes  gain  of  &-€ST  The University of British Vancouver, Canada  Date  DE-6  (2/88)  T^o  ^CU^i^c Columbia  '>o\  P <Z  requirements  shall  that  agree  may  representatives.  financial  the  I agree  I further  permission.  Department  of  It not  be is  that  the  Library  permission  granted  by  understood be  for  allowed  an  advanced  shall for  the that  without  make  it  extensive  head  of  my  copying  or  my  written  ABSTRACT Pinus albicaulis  Engelm. is a keystone subalpine species found throughout mountainous  regions of western North America. Population genetic investigations in British C o l u m b i a using isozymes (17 populations, 12 loci) extracted from bud tissue revealed that the species has high levels of observed and expected heterozygosity compared to other pine species (0.213 and 0.262, respectively). Isozyme analysis (two populations, ten loci) using maternal gametopyte tissue a n d e m b r y o s extracted from seed elucidated that biparental inbreeding, and possibly selfing, is c o m m o n (mean multilocus outcrossing rate = 0.73, mean single-locus outcrossing rate = 0.69). There is moderate population substructuring ( F  S T  = 0.061), typified by the clumped distribution of  trees, influencing gene flow, although seed distribution by Clark's nutcracker appears to be the overriding factor influencing genetic patterns. There were few rare alleles found a n d genetic distances between populations were small (Nei's 1978 distance ranged from 0.006 to 0.134 a n d Cavalli-Sforza a n d Edwards' (1967) chord distance from 0.086 to 0.297). Genetic distances w e r e weakly related to physical distances between populations (Mantel test, p = 0.036). O b s e r v e d heterozygosity w a s significantly negatively correlated with longitude ( R = 0.295) a n d latitude ( R = 2  2  0.357). Population genetic parameters were consistent with other studies suggesting northerly postglacial recolonization f r o m refugia in the Washington and Oregon Cascades a n d several more northern refugia in the Rockies, including the possiblity of a refugium near Roger's Pass, BC. Nearly all populations were observed to have Cronartium  ribicola Fisch. (white pine blister rust)  infections, mortality of trees of all ages w a s often present (due to various causes), a n d regeneration w a s often sparse or absent. A conservation strategy w a s developed based on the results of these investigations, concurrent with the priorities and recommendations of other agencies involved with whitebark pine conservation. Priorities included continuing surveys of natural stands in order to identify and monitor putatively resistant trees, collecting s e e d f r o m all available s e e d sources and especially these selected individuals, establishing c o m m o n garden  ii  tests to assess adaptive variation and screen for disease resistance, establishing field trials in natural habitats with a variety of hazard ratings for blister rust, developing appropriate seed a n d scion transfer guidelines, and maintaining a cooperative exchange in terms of materials and research with other jurisdictions involved in whitebark pine conservation. Future research may involve isolation of any specific resistance mechanisms, genetic transformation or cross-breeding of susceptible individuals, and bulk propagation of resistant individuals or families via rooting cuttings or somatic embryogenesis. In the longer term, breeding strategies involving controlled crosses of putatively restitant parents in order to produce hardy and disease resistant planting stock for a variety of hazard-rated sites should be instituted. Due to the extremely long generation time of this species, it is critical that conservation measures begin immediately.  iii  TABLE OF CONTENTS ABSTRACT  ii  LIST OF TABLES  vi  LIST OF FIGURES  vii  LIST OF APPENDICES  viii  ACKNOWLEDGEMENTS  ix  Chapter 1 - Introduction and Objectives  1  1.1 Introduction  1  1.1.1 Whitebark pine: autecology of a keystone species  1  1.1.2 Mating system  8  1.1.3 Genetic diversity  9  1.1.4 Conservation of whitebark pine  11  1.2 Thesis Objectives  13  Chapter 2 - Mating system  14  2.1 Introduction  14  2.1.1 Objectives  16  2.2 Methods and Materials  16  2.2.1 Field collections  16  2.2.2 Genetic analysis  17 Laboratory  17 Analysis  18  2.3 Results  19  2.4 Discussion  23  2.4.1 Mating system of stone pines  23  2.4.3 Sources of error  28  2.4.4 Comparison with other populations  29  2.5 Conclusion  30  Chapter 3 - Genetic diversity in British Columbia 3.1 Introduction  31 31  3.1.1 Importance of genetic diversity for whitebark pine  31  3.1.2 Genetic diversity and population structure  31  3.1.3 Objectives  35  3.2 Methods and materials  35 iv  3.2.1 Field collections  35  3.2.2 Genetic analysis  37  3.3 Analysis  38  3.4 Results  39  3.4.1 Removing loci  39  3.4.2 Genetic diversity statistics  40  3.4.3 Wright's F-statistics  42  3.4.4 Mean statistics by geographic region  44  3.4.5 Physical versus genetic distances  45  3.4.6 Geographic patterns in genetic diversity  48  3.5 Discussion  51  3.5.1 Patterns of genetic diversity  51  3.5.2 Wright's F-statistics  56  3.5.3 Sources of error  -58  3.6 Conclusions  60  Chapter 4 - A conservation strategy for whitebark pine in British Columbia  61  4.1 Introduction  61  4.1.1 Justification for conservation  :  61  4.1.2 Climate change  62  4.1.3 G e n e conservation  66  4.1.4 White pine blister rust  70  4.2 Whitebark pine conservaton strategy for British Columbia  71  4.2.1 Introduction  71  4.2.2 Short-term goals  71  4.2.3 Medium-term goals  74  4.2.4 Long-term goals  75  4.2.5 Regional conservation priorities  77  4.3 Conclusions  80  Literature Cited  81  Appendix I - Table of allele frequencies by population  95  Appendix II - Buffer recipes  98  Appendix III - Locations of all populations sampled  .100  Appendix IV - Zymograms  102  Appendix V - Tables of genetic diversity and Wright's F-statistics  105  v  LIST O F T A B L E S Table 2 . 1 . List of e n z y m e loci screened for mating systems analysis  17  T a b l e 2.2. N u m b e r of families with pairs of loci in linkage disequilibrium  20  Table 2.3. Log-likelihood G-test on segregation ratios of polymorphic loci  20  Table 2.4. Estimates of t at the population level  21  T a b l e 2.5. Outcrossing d a t a for stone pines (subsection Cembrae) a n d other pines  24  Table 3 . 1 . List of sample locations s u m m e r 2000  37  Table 3.2. List of loci scored for isozyme analysis  '.  38  Table 3.3. Electrophoretic buffer systems  38  Table 3.4. Genetic diversity statistics for all populations  42  Table 3.5. Wright's F-statistics for 17 populations  43  T a b l e 3.6. Genetic diversity statistics for subdivided population groupings  44  Table 3.7. Physical most parsimonious distances between populations  45  Table 3.8. Nei's 1972 a n d 1978 (unbiased) genetic distances  46  T a b l e 3.9. Cavalli-Sforza & Edwards (1967) arc and chord distances  47  Table 3.10. S u m m a r y of whitebark pine genetic data by study  52  Table A.1.2. Ovule and pollen allele frequencies for Mannning and Baldy  97  T a b l e A . 2 . 1 . Extraction buffer recipe  99  Table A.2.2. Morpholine electrode and gel buffer recipes  99  Table A.2.3. Tris-citrate electrode and gel buffer recipes  99  T a b l e A.2.4. Ridgeway electrode a n d gel buffer recipes  99  Table A . 3 . 1 . List of populations and sampling locations  101  Table A . 5 . 1 . S u m m a r y of genetic parameters by locus  106  Table A.5.2. S u m m a r y of Wright's F-statistics  106  Table A.5.3. Genetic diversity statistics for the other 17 populations combined by locus  107  vi  LIST OF FIGURES Figure 2 . 1 . Location of sampling sites for mating systems study  17  Figure 2.2. Frequency distribution of family outcrossing rates for all loci for Manning  23  Figure 2.3. Frequency distribution of family outcrossing rates for all loci for Mt. Baldy  23  Figure 3 . 1 . Genetic diversity sampling locations  36  Figure 3.2. W a g n e r tree produced by rooting at midpoint of longest path  48  Figure 3.3. Regression of observed heterozygosity on latitude  49  Figure 3.4. Regression of observed heterozygosity o n longitude  49  Figure 3.6. Regression of expected heterozygosity on longitude  ,  50  Figure 3.7. Regression of Wright's F on latitude  50  Figure 3.8. Regression of Wright's F on longitude  51  Figure A . 3 . 1 . M a p of sampling locations  102  Figure A . 4 . 1 . Z y m o g r a m s  104  vii  LIST OF APPENDICES Appendix I - Table of allele frequencies by population  95  Appendix II - Buffer recipes  98  Appendix III - Locations of all populations sampled  100  Appendix IV - Z y m o g r a m s  103  Appendix V - Tables of genetic diversity and Wright's F-statistics supplemental to the text  105  viii  ACKNOWLEDGEMENTS I w o u l d like to express my gratitude first and foremost to my supervisor, Dr. Sally Aitken, for inviting m e to participate in this project, one about which many of my friends a n d colleagues have admitted their, well...envy for both the adventures in the wilds of beautiful B.C. a n d practical application for conservation. It has been a real pleasure working in such a supportive environment with a door that w a s always open and her assurance that none of my questions were stupid. Her confidence never flagged in m e . Next I would like to thank Dr. Carol Ritland for her invaluable help in the laboratory: her wealth of practical experience and positive attitude, plus all of our other, less academic discussions. Also, thanks to Dr. Kermit Ritland, Hugh W e l l m a n , Charles C h e n , Dilara Ally a n d Yanik Berube for their assistance with isozymes both in the lab and analysis. Emily Pritchard w a s always a source of motivation with both her perseverance and success in her own academic and all of her other pursuits. In the field, the amazing energy, experience (and culinary skills) of Petra Heppner a n d also Christopher Stephenson made our adventures safe, successful and f u n . T o everyone else that saved us f r o m getting lost a n d finding the always elusive whitebark pine, too many to list, I w o u l d also like to thank. S o m e especially helpful information w a s provided by Elizabeth C a m p b e l l , Jon Stuart-Smith, Stephan Zeglen (B.C.MoF, Nanaimo) and Ray C o u p e (B.C.MoF, Smithers). Glen Davidson (B.C.MoELP, Williams Lake) and his summer intern took us into the stunning wilderness of Tsy'los Provincial Park, saving us days of aimlessly driving around on a maze of unsigned logging roads in the middle of nowhere. Thanks also to Doug Thorburn and I rina a n d J i m for their incredible hospitality and hike to J u m b o Pass, Denys Bell of Babine Forest Products a n d Huckleberry Mines for directions in the logging road maze around Houston, the Line Creek Mine in S p a r w o o d for access and escort, and everyone else w h o w a s so helpful at the many c o m p a n y a n d Ministry offices. John Lavery collected samples on one of his rare days off in McBride.  ix  Drs. Mike Meagher and Dave Edwards collected the seed for the mating system study, a n d J u d y Miller, Provincial Parks, Okanagan District, also provided support. Shirley Barnes painstakingly germinated the seeds and collected isozyme data, which Dr. Youssry El-Kassaby w a s generous enough to let m e use for this study. Dr. Alvin Yanchuk provided input to help m e focus my project. Dave Kolotelo (B.C.MoF, Vancouver) provided a valuable f o r u m for whitebark pine enthusiasts by organizing a symposium so that w e could learn from each other a n d develop new conservation strategies. Dr. J o h n Worrall inspired so many students and foresters with the fascinating story of the commercially marginal whitebark pine and its symbiotic relationship with the Clark's nutcracker. I would like to acknowledge the late Dennis Reynolds for sparking my ongoing interest in genetics as the mysterious link between the chemical and the biological, the animate and the inanimate, and for giving m e the confidence to continue with my education. Forest Renewal B C , the BC Forest Genetics Council and the Centre for Forest G e n e Conservation all provided generous financial support for this project. I w o u l d especially like to thank Christine Chourmouzis, w h o w a s so consistently supportive she helped m e beyond all reasonable expectations with my computer deficiencies a n d always m a d e time for my incessant confusion, making sure I kept everything in perspective, if not under control.  x  CHAPTER 1 - INTRODUCTION AND OBJECTIVES "A tree is a tree. Now how many more do you need to look at?" - Ronald Reagan, 1965  1.1  INTRODUCTION  1.1.1 WHITEBARK PINE: AUTECOLOGY OF A KEYSTONE SPECIES Whitebark pine (Pinus albicaulis  Engelm.) is a high-elevation conifer, typically found f r o m the  subalpine to timberline (Achuff 1989; Arno a n d Hoff 1989; Callaway 1998; Douglas a n d Bliss 1977). It ranges f r o m central British Columbia and Alberta south to t h e Sierra Nevadas, f r o m 5 5 ° N to 3 7 ° N , along t h e Cascade a n d Coast ranges and t h e Rocky Mountains. T h e species is subdivided into eastern a n d western populations (Ogilvie 1990), separated at t h e closest point (in southern British Columbia) by 100 kilometres (Arno a n d Hoff 1989, 1990; M c C a u g h e y a n d Schmidt 1990). It survives o n ridgetops a n d exposed talus slopes, enduring extreme abiotic conditions (Perkins a n d S w e t n a m 1996) such as wind dessication, high ultraviolet exposure, freezing temperatures a n d a very short growing season (Arno a n d Hoff 1 9 8 9 , 1 9 9 0 ; Campbell 1998; M c C a u g h e y a n d Schmidt 1990). Whitebark pine is t h e only m e m b e r of the stone pines (subgenus Strobus, section Strobi, subsection C e m b r a e ) (Critchfield 1986; Price et al. 1998) in North America (Bruederle et al. 1998; Goncharenko et al. 1992; Krutovskii et al. 1995), although t h e phylogeny a n d t a x o n o m y of this group is still unresolved (Bruederle etal. 1998, 2 0 0 1 ; Krutovskii etal. 1995; Liston etal. 1996; Politov a n d Krutovskii 2 0 0 1 , unpublished data; Price etal. 1998). T h e geographic isolation of whitebark pine from t h e other stone pines a n d ambiguous results of previous studies using various markers a n d characteristics led to s o m e contention regarding its alliance with t h e other stone pines. Recently, several studies (Krutovskii et al. 1995; Liston et al. 1996; Price et al. 1998) have found support for t h e monophyly of t h e stone pines based o n c p D N A sequences. This h a s  1  provided support for recognition of this debated taxonomic group originally defined by morphological characteristics (Little and Critchfield 1969; Critchfield 1986; Axelrod 1986). This subsection includes several haploxylon five-needled pines found throughout Eurasia a n d Northern Europe which feature heavy, wingless seeds (Arno and Hoff 1989; Critchfield 1986), indehiscent cones (Arno a n d Hoff 1990; Krutovskii et al. 1995; T o m b a c k 1986) and a mutualistic association with birds of the Nucifragia  genus (Arno and Hoff 1990; Bruederle et al. 1998;  Jorgensen a n d Hamrick 1997; Krutovskii etal. 1995; T o m b a c k 1982; T o m b a c k a n d Linhart 1990), nutcrackers which facilitate seed dispersal (Callaway 1998; Lanner 1982; Stuart-Smith 1998). P. albicaulis  has coevolved with the Clark's nutcracker (N. columbiana  Wilson) (Critchfield 1986;  T o m b a c k 1982) to the point w h e r e the tree species is completely reliant on the nutcracker for dispersing its seeds, which also provides ideal conditions for germination and establishment (Hutchins and Lanner 1982). It has been estimated that the nutcrackers c o n s u m e approximately o n e third of the seeds they cache annually (Tomback 1982); the consumption rate is nearly 1 0 0 % for small m a m m a l s which also cache the seeds (Arno and Hoff 1989). Despite its narrow geographic range, P. albicaulis  is a m e m b e r of a variety of plant communities  along its latitudinal gradient and grows primarily in association with subalpine larch (Larix lyallii Pari.), subalpine fir (Abies lasiocarpa  [Hook.] Nutt.), Engelmann spruce (Picea engelmannii  Parry  ex Engelm.), limber pine (Pinus flexilis James) and lodgepole pine (Pinus contorta var. latifolia Dougl. ex Loud.) (Achuff 1989; Arno and Hoff 1989, 1990; Campbell 1998; M c C a u g h e y a n d Schmidt 1990; Ogilvie 1990; Perkins and S w e t n a m 1996). T h e large, heavy, high-fat a n d nutrientrich s e e d s (Lanner 1982,1986; Lanner and Gilbert 1994; T o m b a c k 1982; T o m b a c k a n d Linhart 1990) serve as a key food source for a wide variety of animals (Arno and Hoff 1990; Keane and Arno 1993; Lanner and Gilbert 1994), including other birds, red squirrels (Tamariscus (Arno a n d Hoff 1989), black bears (Ursus amencanus)  hudsonicus)  (Mattson and Reinhart 1997), grizzly bears  (Ursos arctos hornbilus) (Mattson and Reinhart 1997; McCaughey and Schmidt 1990) a n d many  2  small m a m m a l s (Arno a n d Hoff 1989; Kendall a n d Arno 1990; M c C a u g h e y 1994; M c C a u g h e y a n d Schmidt 1990). O n average, per seed values for whitebark pine in t h e U.S. a r e : dry weight, 0.09g; 18 percent protein; 21 percent carbohydrate; 52 percent fat; a s well a s being high in m a n y a m i n o acids, fatty acids a n d minerals. C o n e crop abundance has been linked to population cycles a n d behaviour of nutcrackers, squirrels a n d grizzly bears (Mattson a n d Reinhart 1997), a s well as t h e animals with which they interact (Bruederle etal. 1998; Kendall a n d Arno 1990; Keane a n d Arno 1993; T o m b a c k et al. 1995). For these reasons, it has been suggested that whitebark pine be regarded a s a keystone or a n umbrella species (Tomback er al. 2 0 0 1 ; Campbell 1998; Stuart-Smith 1998), a species w h o s e health a n d ecosystem status is integrally linked t o , a n d a n overall indicator of, t h e health a n d survival of other species and communities (Callaway 1998; Ledig 1988; Mattson a n d Reinhart 1997; Primack 1998). Currently, whitebark pine is considered threatened by several agencies in British Columbia since it is under direct pressure from a number of environmental a n d anthropogenic threats, although it is not formally listed as such under C O S E W I C (Yanchuk a n d Lester 1996; Forest Health Committee of B.C. 1999; L. Pedersen, B.C. Chief Forester 1998, o p . cit. Kieran 1998). Since it already exists at the upper altitudinal periphery of its fundamental ecological niche, t h e potential for future global climate change may have a serious impact o n t h e survival of not only whitebark pine but all of t h e biotic communities of which it is a component ( N a m k o o n g 1992). Based o n predictions of doubled atmospheric carbon dioxide within t h e next century (Bradshaw and McNeilly 1 9 9 1 ; Huntley 1991), s o m e climate modelling projections forecast imminent w a r m i n g of northern a n d high altitude areas by a n annual m e a n of three t o six degrees Celsius (Bradshaw a n d McNeilly 1 9 9 1 ; Huntley 1 9 9 1 ; Running a n d Nemani 1 9 9 1 ; U S E P A 2 0 0 0 ; W a t s o n etal. 1997). This is associated with unknown changes in moisture regimes, although winter s n o w p a c k is likely to melt sooner than currently (IPCC 2001 a,b; Franklin et al. 1991). Since the majority of moisture in whitebark pine habitat occurs a s snow and much of t h e annual soil moisture is received a s  3  snowmelt throughout the warmer portion of the year (McCaughey a n d Schmidt 1990; Ogilvie 1990) , global climate change will drastically alter the hydrology (McCaughey a n d Schmidt 1990; Perry et al. 1 9 9 1 ; Running and Nemani 1991) and abiotic conditions of current whitebark pine habitat, including a significant change in the length of the growing season (Running a n d Nemani 1 9 9 1 ; W a t s o n etal. 1997). Whitebark pine currently serves an important ecological role in the hydrology of m o n t a n e a n d headwater systems by intercepting snow and serving as a moderator of snowmelt (Arno and Hoff 1 9 8 9 , 1 9 9 0 ; Keane and Arno 1993; McCaughey 1994); a decrease in snowpack m a y lead to critical growing season moisture deficits and higher incidence and severity of fires in high elevation areas (Perry et al. 1 9 9 1 ; Watson et al. 1997). Slope stability in whitebark pine habitat, typically steep, montane areas with shallow, rocky soils, will also be affected (Keane and Arno 1993) as these complex factors interact to alter the survival and establishment of living trees as well as the size and longevity of snags. A s future climate change alters the abiotic character of whitebark pine ecosystems, the trees themselves may no longer be optimally adapted to the sites they currently occupy (Franklin et al. 1991) . T h e y m a y then be even more susceptible to competition from other species with which they currently coexist at lower altitudes; above the timberline, they are the only trees present (Arno a n d Hoff 1990; Campbell 1998). Since generations are too long to evolve adaptive traits at a sufficient rate to keep pace with rapid climate change, the only means available to whitebark pine trees for long-term species survival would be to migrate (Delcourt and Delcourt 1998; Huntley 1991). This must occur via dispersal by Clark's nutcrackers. Typical seed dispersal distances of this bird have been g a u g e d at one to three kilometres, although distances over twenty kilometres have been recorded (Tomback a n d Linhart 1990; Vander Wall and Balda 1977), a n d dispersal direction is not related to prevailing winds (Tomback 2001). These distances may ensure that whitebark pine  4  populations could adapt to a new regime of climatic zonation a s a result of the observed rates of climatic c h a n g e (McCaughey a n d Schmidt 2001). It is unknown whether the birds would cache the seeds in such a manner a s to extend the species range northward (Sedjo a n d Solomon 1988) at a rate approximating that of the location of suitable habitat, which is expected to move 150 to 550 k m north, or 150 to 550 m upwards in elevation within the next century (Franklin et al. 1 9 9 1 ; Rogers ef al. 1999; U S E P A 2 0 0 0 b ; W a t s o n  etal. 1997). This is not an impossibility, however: since the most recent glaciation, the range of whitebark pine has been expanding northward to reoccupy its historic range throughout the mountainous areas of western North America (Jorgensen a n d Hamrick 1997; M c C a u g h e y a n d Schmidt 2 0 0 1 ; T o m b a c k 2001), a n d the only means by which this might occur, given t h e closed cones a n d wingless seeds is by nutcrackers (Baker 1990). T h e critical question is whether the current projections of the increased rate of climate change a n d corresponding ecosystem change can be o v e r c o m e by the rate of northward migration of nutcrackers (Tomback a n d Linhart 1990). Climate change notwithstanding, whitebark pine is under many other immediate threats throughout its range (Callaway 1998). White pine blister rust (Cronartium  ribicola), a n introduced  fungal pathogen which infects many white pines, is a virulent disease of this species (Arno a n d Hoff 1990; Hoff etal. 1980,1994), causing extremely high mortality (Arno a n d Hoff 1989; Bruederle  etal. 1998; M c C a u g h e y a n d Schmidt 1990), especially in Montana (Hoff a n d Hagle 1990; Keane and Arno 1993; M c C a u g h e y 1990; T o m b a c k etal. 1995). Although mortality appears t o be less severe in C a n a d a , infection rates are still high and have no obvious constraints to their future expansion throughout the entire range of whitebark pine (Campbell 1998; Stuart-Smith 1998). Research programs to locate a n d develop genetic resistance t o this disease are under w a y in the United States (Arno a n d Hoff 1990; Hoff 1984,1986; Hoff a n d Hagle 1990; Hoff etal. 1994; J o r g e n s e n a n d Hamrick 1997; T o m b a c k et al. 1995), but no programs have been initiated in C a n a d a t o date. There are many other pathogens which cause injury a n d mortality in this species,  5  including limber pine dwarf mistletoe [Arceuthobium  cyanocarpum)  (Arno a n d Hoff 1989; Hoff a n d  Hagle 1990; Mathiasen a n d Hawksworth 1988), but none have had the severity of impact in B C that blister rust has had. T h e detrimental effects of blister rust are exacerbated by the longevity of whitebark pine: a g e s of 4 0 0 to 500 are not u n c o m m o n (Ogilvie 1990), a n d krummholtz specimens of over 1700 years have been found (Perkins a n d S w e t n a m 1996). Insects such a s the mountain pine beetle (Dendroctonus  ponderosae  Hopkins) (Arno a n d Hoff  1990; Baker etal. 1 9 7 1 ; Keane a n d Arno 1993; McCaughey a n d Schmidt 1990; Perkins a n d S w e t n a m 1996) a n d the ambrosia beetle Ips pini Say have also caused mortality in whitebark pine stands (Arno a n d Hoff 1989): however, their effects have been more serious in the United States than in C a n a d a to date (Tomback et al. 1995). T h e severity of the impacts of both insects a n d pathogens has been attributed to fire suppression policies in North America (Arno a n d Hoff 1990). Whitebark pine has evolved with a medium-intensity fire regime with a 5 0 to 3 5 0 year return pattern (Campbell 1998; McCaughey a n d Schmidt 1990), which would effectively kill competing trees a n d seeds in the substrate, enabling whitebark pine to play a role a s both a pioneer a n d climax species, both a s a true serai climax a n d a fire-maintained subclimax (Arno a n d Hoff 1989; Callaway 1998; Campbell 1998). Since the advent of fire suppression, successional c h a n g e s have gone unchecked a n d competition-induced mortality is c o m m o n (Arno a n d Hoff 1989; T o m b a c k et  al. 1995). Slow-growing whitebark pines are outcompeted by subalpine fir a n d Engelmann spruce, leaving a g e class gaps and weakening surviving trees (Campbell 1998). In addition, an a b u n d a n c e of old lodgepole pine stands, resulting from fire suppression, are a significant factor causing mortality in coexisting whitebark pine via mountain pine beetle infection (Kendall a n d Arno 1990; M c C a u g h e y 1994). Whitebark pine has developed a unique population genetic structure a s a result of bird-mediated s e e d dispersal (Jorgensen a n d Hamrick 1997; Bruederle etal. 1998; Tani etal. 1998); adaptation to this symbiotic mechanism has been accompanied by selection for morphological a n d  6  phenological adaptations which reflect t h e mutualistic association with nutcrackers. Unlike most coniferous species, stone pine cones lack the schlerenchyma in t h e female cones which c a u s e t h e m to o p e n a n d release their seeds upon maturity (McCaughey 1994; M c C a u g h e y a n d Schmidt 1990). T h e cones are situated at branch tips in the top of the crown tree so they are difficult to locate from the ground but easily visible from above (Lanner 1982; Furnier etal. 1987). Clark's nutcrackers chisel the immature cone scales apart in July a n d August (McCaughey a n d Schmidt 1990; T o m b a c k 1982), and collect up to 150 seeds at a time in a sublingual p o u c h , a unique adaptation of Nucifragia  species (Arno and Hoff 1990; McCaughey 1994; Lanner 1982; T o m b a c k  1982; T o m b a c k a n d Linhart 1990). T h e seeds are then cached o n e to three centimetres deep solitarily or more often in groups of up to fifteen seeds (Bruederle etal. 1998; T o m b a c k 1982,1986; T o m b a c k etal. 1995), and a rock or cone is placed o n top (Tomback a n d Linhart 1990). T h e s e groupings, combined with typical nutcracker behaviour of returning to the s a m e caching a r e a for several flights, results in a complex population genetic structure (Tomback 2001). Typical dispersal distances range from several hundred metres to five kilometres (Arno a n d Hoff 1990), although distances of over 2 0 kilometres have been reported. Nutcrackers c a n collect  '  100,000 seeds annually (McCaughey 1994) a n d find up to 30,000 cached seeds each year ( T o m b a c k 1982), ensuring a year-round nutritive food supply. They can remember t h e locations of cached seeds for up t o three years, and forgotten seeds typically germinate gradually over t h e course of up to three years (McCaughey 1994). Many of the embryos in the mature cones are immature (Arno a n d Hoff 1989; L e a d a m 1986; McCaughey and Schmidt 1990; Pitel a n d W a n g 1990), a n d seeds continue to mature after caching. T h e resulting delayed germination m a y be another adaptation reflecting t h e coevolution of the stone pines a n d nutcrackers: t h e trees have been subject to thousands of generations of selection for large seed size through bird preference, a n d t h e large seeds require o n e to three years to reach embryo maturity. T h e high s e e d weight m a y also reflect t h e optimal nutrition requirements for t h e embryo to survive in cold climates.  7  1.1.2 MATING SYSTEM T h e mating system, or degrees of outcrossing (mating between unrelated individuals) a n d inbreeding (selfing a n d mating a m o n g relatives), a species typically exhibits is a critical factor both influencing a n d influenced by factors such a s genetic structure, population density a n d distribution, a n d g e n e flow. T h e interrelatedness and sometimes opposing effects of these parameters m a k e s it difficult to isolate t h e effects of genotype a n d environment, although clearly it is their interaction which results in t h e expressed mating system. Conifers (including most pine species) are generally highly outcrossing (Hamrick etal. 1992). T h e unique demographics, ecology a n d dispersal of stone pines m a y exert selection pressures on the genes controlling mating system parameters t o such a n extent that they may differ from other, wind-dispersed pines. Rogers et al. (1999, p. 75) wrote, "whitebark pine is considered to have a largely outcrossing mating system, yet there is little local or empirical information to support t h e theory." This view is also held by Krutovskii etal. (1995). However, based o n the unique adaptations a n d population structure, it w a s suspected that it would likely feature a high degree of inbreeding (Tomback a n d Schuster 1994; Krutovskii etal. 1995). T h e isolation of populations a n d their low density (i.e., their sporadic distribution in subalpine and timberline areas) would likely contribute t o inbreeding (Mitton 1992).  This is supported by the hypothesis of kin selection which facilitates survival a n d  establishment of related genotypes while outcompeting or hindering other unrelated individuals. Traits such a s root grafting, clumping a n d multiple stem formations all support s o m e degree of increasing t h e fitness of relatives, perhaps at the expense of the interacting individuals (Tomback a n d Linhart 1990). T h e complex traits resulting from t h e coevolution of whitebark pine a n d t h e Clark's nutcracker, especially the adaptations concerning seed dispersal a n d establishment, appear to promote kin selection a n d increase the degree of mating a m o n g related individuals. Krutovskii a n d his colleagues ( 1 9 9 4 , 1 9 9 5 ) have documented varying levels of inbreeding a m o n g t h e Eurasian stone pines (and o n e single population of P. albicaulis);  8  t h e results show  substantially higher inbreeding in stone pines than in other taxonomic subdivisions of the genus Pinus with the exception of P. maximartinezii  which exists as a single, isolated population of  relatively small size in Mexico (Ledig etal. 1999) and other bird-dispersed pines, primarily the pihon pines. H u m a n seed herbivory in the case of P. maximartinezii,  and bird s e e d caching in the  cases of pinon and whitebark pines (Richardson 2001), generate unique population structure and selection pressures, and would explain the relatively high inbreeding found based on their impact o n dispersal and regeneration patterns.  1.1.3 GENETIC DIVERSITY Several studies have examined genetic diversity of this species, but even the most comprehensive in terms of area covered, Jorgensen and Hamrick (1997), did not include any samples f r o m British Columbia. O n e other study examined populations along the B.C.-Alberta border (Stuart-Smith 1998), but none have yet looked at genetic variation across the large and topographically complex province of B.C. Historical events, including the several glacial cycles in the last 100,000 years, have undoubtedly left a genetic signature o n whitebark pine. Its long generation time, often 100 years, suggests in some cases that the effects are still evident (Richardson 2001). Glacial events during the Pleistocene era reduced many.conifers' ranges in North A m e r i c a and Eurasia and in s o m e cases have led to genetic bottlenecks. T h e s e events contributed to different patterns of genetic structure and diversity in those areas. Typically, the heterozygosity of conifers decreases slightly with increasing latitude as a result of postglacial recolonization (Millar a n d Westfall 1992). Since conifers, and specifically whitebark pine, have such long generation times, these populations have often not yet returned to a state of genetic equilibrium with respect to drift, migration and selection. Although whitebark pine exists in an ecologically peripheral habitat, global w a r m i n g could potentially force species to migrate to more northerly latitudes as well as higher elevations. Since whitebark pine already exists at the the tops of many mountains, the only realistic option in terms of  9  the overall species range is to migrate northward, in addition to upward elevational range expansion w h e r e physically possible. Conifers generally have very high levels of heterozygosity, a s measured both by isozymes a n d other molecular markers (Hamrick etal. 1992). There m a y be differences between observed (a direct count of t h e heterozygous individuals at each locus, H ) a n d expected (calculated using 0  allele frequencies, H ) heterozygosity at the individual a n d population levels. A c o m m o n index of e  the difference is t h e inbreeding coefficient F, where F = 1 -r]JH . e  This index estimates t h e relative  excess or deficiency of heterozygotes in the actual population c o m p a r e d to Hardy-Weinberg equilibrium (HWE). Differences between H a n d H can arise f r o m empirical causes such a s 0  e  localized anomalies in allele frequencies, errors, or deficiencies in t h e model used to calculate t h e frequencies, which typically oversimplifies real factors that interact in complex, often unquantified, ways. While inbreeding typically reduces t h e observed heterozygosity, especially for selfing organisms, s o m e primarily inbreeding species have fairly high expected heterozygosity, suggesting that m a n y generations of inbreeding effectively purged deleterious alleles (many of which w o u l d have been recessives found at low frequencies in t h e population) a n d further inbreeding resulted in no additional reduction of heterozygosity (Kirkpatrick and Jarne 2000). It is therefore possible to have s o m e w h a t contradictory results regarding genetic diversity a n d inbreeding at first glance. Heterozygosity, a s measured by isozymes, has also been demonstrated t o increase with a g e in conifers a s individuals homozygous for deleterious or lethal alleles die during embryonic a n d juvenile life stages a n d are likely to be outcompeted (Bush a n d S m o u s e 1992). Since genetic diversity a n d patterns in whitebark pine are dependent o n Clark's nutcrackers, and to a lesser degree small mammals, there is a host of interesting evolutionary a n d ecological questions which m a y be posed. Will t h e nutcrackers be able to survive in more northerly environments? Will seeds result in established seedlings, or will they be outcompeted by other  10  plants which m a y occupy the s a m e ecological niche in the changing environment? Will t h e communities of other species which have evolved around whitebark pine be able t o carry on their ecological roles in future climates? H o w will these communities change over time? T h e s e questions m a y determine the success of whitebark pine both as a species a n d a s a keystone m e m b e r of the timberline t o subalpine community, both in the short a n d long term, with respect t o natural a n d anthropogenic cycles of climate change.  1.1.4 CONSERVATION OF WHITEBARK PINE Currently, B.C. has a Protected Areas Strategy under which conservation of natural resources, ecosystems a n d unique features are protected by law within a network of parks, wilderness areas and ecological reserves (Ecological Reserves Program 1993; Province of B.C. 1996). Although all ecosystem types are supposed to be represented in this system, high elevation ecosystems are over-represented in terms of area protected (Ecological Reserves Program 1993; Province of B.C. 1996). This is the natural habitat of whitebark pine, a n d large contiguous areas of current a n d potential future habitat are already under protection. In C a n a d a , there are vast tracts of wilderness which are seldom encountered by humans a n d not under immediate threat of development (Achuff 1989; Y a n c h u k a n d Lester 1996). T h e preservation of landscape-level processes a n d dynamics has been d e e m e d essential for the conservation of adequate levels of variation, as this approach takes into account the metapopulation structure a n d dynamics of most species a n d their associated communities (Delcourt a n d Delcourt 1998), and provides s o m e long-term security in t h e event of future uncertainty (Erikkson et al. 1993; Noss 1990). In the United States, this is not t h e case: although much of the whitebark pine habitat is in extremely remote areas, m u c h of it is under pressure of development: primarily road construction, resource extraction, heavy tourism a n d recreational ski areas (Cole a n d Landres 1996). Since whitebark pine ecosystems are areas of heavy wildlife use, sensitive hydrology a n d a host of other non-timber values, including aesthetics and traditional values (Arno a n d Hoff 1 9 8 9 , 1 9 9 0 ) ,  11  preserving t h e habitat a n d the inherent genetic variation of whitebark pine in its natural habitat is likely to be a more difficult task in the United States, with its large human population a n d mixed-use wilderness areas. T h e current paradigm of delineating evolutionarily significant units, or E S U s (Moritz 1994), poses additional problems: should each region be considered a n ESU? W h a t about special adaptations such a s blister rust resistance, other rare alleles, or interacting g e n e complexes (Williams etal. 1995), which serve a s insurance against future c h a n g e (Erikkson etal. 1993; Lande a n d Barrowclough 1987; Yanchuk a n d Lester 1996)? Identifying a n d legislating protection which conserves t h e inherent variability in t h e species will be difficult a n d costly (Hard 1995; Y a n c h u k a n d Lester 1996), but m a y be a necessary step in order t o o v e r c o m e t h e short-term threat of inbreeding to the health of the species (Erikkson etal. 1993; Millar a n d Westfall 1 9 9 2 ; N a m k o o n g 1992). In-situ conservation strategies must be designed to take into account t h e potential for future changes in areas such a s land-use, public opinion, genetic bottlenecks, natural catastrophes a n d climate change (Achuff 1989; Franklin etal. 1 9 9 1 ; Ledig 1986; Millar a n d Westfall 1992; N a m k o o n g 1992; Yanchuk a n d Lester 1996). Developing comprehensive ex-situ collection of the genetic resources of the species is not likely to b e feasible, given t h e expense of collecting in remote areas a n d preserving a n d cataloguing t h e material (McDonald a n d Hoff 2 0 0 1 ; Millar and Westfall 1992; Yanchuk a n d Lester 1996). T h e remote nature of t h e populations a n d irregular nature of whitebark pine cone crops (Arno a n d Hoff 1989; M c C a u g h e y a n d Schmidt 1990; W e a v e r a n d Forcella 1986) m a k e collecting s e e d a costly a n d uncertain proposition. Repeat visits to potential cone collecting sites are required each season as c o n e s must be c a g e d in early summer to prevent seed predation, which c a n otherwise lead t o total loss. Large seed size, embryo immaturity a n d low germination percentage of whitebark pine, coupled with its susceptibility to fungal pathogens in storage which would further reduce t h e viability of stored s e e d , would also add to operational difficulties (McCaughey a n d Schmidt 1994; M c C a u g h e y a n d T o m b a c k 2001). Various techniques to artificially enhance germination rates a n d  12  embryo development have been attempted, with moderate to significant success (Leadam 1986; Pitel and Wang 1990). Regular viability testing, which is essential for  ex-situ  conservation in seed  banks, is expensive, time-consuming and uses up valuable seed. Establishment of  ex-situ  collections in living genetic archives or clone banks is also very expensive. 1.2 THESIS  OBJECTIVES  In light of the paucity of data regarding whitebark pine's mating system, and levels and patterns of genetic diversity in British Columbia, several objectives for this study were established: 1. To determine the mating system of whitebark pine; 2. to quantify the level and patterns of genetic diversity in whitebark pine in B.C. and to compare these results with those of related studies from other geographic areas; and 3. based on the results of the preceding objectives and existing frameworks, to propose a conservation strategy for whitebark pine. It is hoped that the results of this study can be combined with other efforts currently underway to establish a feasible, fact-based management plan to mitigate the current decline of whitebark pine ecosystems in this region.  13  CHAPTER 2 - MATING SYSTEM "People make the mistake of talking about 'natural laws.' There are no natural laws. There are only temporary habits of nature." - Alfred North Whitehead, 1910  2.1 INTRODUCTION T h e nature of a species' mating system is both a reflection a n d a result of the evolutionary forces influencing that species and the ecological niche it occupies. While there is certainly s o m e degree of the "chicken and egg" argument regarding the influence mating systems have o n other life history traits, the mating system (specifically, the relative degrees of selfing vs. outcrossing a species exhibits) can be influenced by factors such as density (Clegg 1980; Mitton e r a / . 1981), which could be altered by human intervention (Gooding 1998). Many models and techniques have been utilized for the analysis of mating systems; their accuracy varies with sampling design, availability of materials (in terms of seasonality, resource allocation a n d conservation requirements) a n d assumptions involved. Assaying seeds using molecular markers is likely the most accurate w a y to determine the mating system of wind-pollinated species, and specifically conifers. Genetic information is available for both the mother via the seed megagametophyte, a n d embryo, a n d the genotype of the pollen parent can be inferred from differences between the two (Shaw  etal.  1981). This procedure  permits estimations of the degree of outcrossing, based on the degree of similarity between the embryo's t w o parents (Jarne and Charlesworth 1993). While conifers are typically outcrossing (Hamrick et al. 1991,1992), there are s o m e clear exceptions (e.g., Ledig et al. 1999). Mixed mating models can b e used to analyze mating systems and to accurately detect levels of inbreeding or selfing.  Studies have shown that propogule dissemination, individual male or female  fitness, a n d reproductive phenology all affect the results by altering factors influencing the rates of outcrossing a n d selfing (Clegg 1980; Hamrick and Allard 1972; Richards 2000; S h a w et al. 1981). Aborted and empty seeds may also reflect products of selfing and since it is impossible to perform  14  genetic analysis on t h e m , these missing data would consequently lead to underestimation of selfing (Stettler a n d Bradshaw 1994). Population genetics theory and empirical studies suggest that genetic bottlenecks, while causing an immediate reduction in heterozygosity, could also serve to purge the gene pool of recessive deleterious alleles, thereby facilitating a greater level of inbreeding (Kirkpatrick a n d Jarne 2000). T h e resultant increased inbreeding may not extend to complete selfing, however, since there w o u l d be no masking of deleterious or lethal alleles in later generations (Jarne a n d Charlesworth 1993) and would therefore be restricted to matings a m o n g relatives. Polyembryony in pines has also been found to act as an early selection agent against homozygotes a n d can impact the degree of selfing (Hedrick et al. 1999). A s a result of the population structure caused by related individuals growing in clumps, there is potential for a high degree of inbreeding in this species (Tomback a n d Schuster 1994; Bruederle et al. 1998). Pollen flow is more likely to occur between individuals within a clump, given the short reproductive w i n d o w and physical proximity of the individuals. Pinus species d o not possess SI (self-incompatibility) genes, which would effectively promote heterosis (Tomback a n d Linhart 1990; Politov and Krutovskii 1994), and help explain the high numbers of aborted and empty seeds which are found in cones (Stettler and Bradshaw 1994), although another explanation is a high genetic load. Latta and Ritland (1994) have proposed that a stable mixed mating system is possible w h e r e b y strongly deleterious alleles are purged by selfing and mildly deleterious alleles, subject to w e a k e r selection pressure, can be carried at a fairly constant genetic load. This model did not incorporate more complex permutations of mating such as biparental inbreeding, however, that m a y be more c o m m o n in empirical situations. It is likely, however, under similar constraints, that the key results of the model would be similar but more gradual if mating a m o n g other classes of relatives w a s included.  15  Nothing about t h e mating system of whitebark pine is currently known. Estimates of relatedness within a n d a m o n g tree clumps have been calculated (Tomback a n d Schuster 1994; Furnier etal. 1987), but outcrossing rates have not. Mating systems analysis has been conducted for other stone pines (Krutovskii etal. 1995; Politov a n d Krutovskii 1994), a n d it is likely that whitebark pine has similar levels of consanguineous mating a n d selfing to those species, since they share demographic structural patterns a s a result of nutcracker dispersal.  2.1.1 OBJECTIVES Knowledge of t h e mating system of a species is important for formulating a n effective m a n a g e m e n t strategy. Obtaining this information will fill an important information g a p for whitebark pine research. T h e objectives of this section are to: 1. Obtain quantitative estimates of single-locus and multilocus outcrossing rates of whitebark pine, and 2. C o m p a r e these data with the mating systems of other stone pines.  2.2  METHODS  AND  MATERIALS  2.2.1 FIELD COLLECTIONS All t h e information contained herein is adapted from Meagher a n d Edwards (1997); s e e Figure 2 . 1 . T e n to 2 0 cones per tree were collected from two populations o n October 1 , 1 9 9 7 .  Sampling  sites w e r e located in E.C. Manning Provincial Park (with appropriate permits) in t h e m e a d o w s at the terminus of t h e Blackwall road (49°06'12"N, 120°45'40"W, 2 0 0 0 - 2 0 4 0 m elev.) a n d around t h e ridges a b o v e t h e ski facilities at Mount Baldy (49°10'N, 119°15'W, 2 1 0 0 - 2 2 0 0 m elev.). Trees s a m p l e d w e r e at least five metres apart; cones were collected from multiple stems in a clump if a tree appeared to be multi-stemmed. Blister rust (Cronartium  ribicola) incidence did not impact  sampling decisions. If cones were partially d a m a g e d by birds, they w e r e still collected if t h e majority of t h e cone appeared intact. Twenty-five trees were sampled at Manning, a n d 3 0 at Mount Baldy.  16  Figure 2.1. Location of sampling sites for mating systems study.  2.2.2 GENETIC ANALYSIS Laboratory Thirty filled seeds per tree for both populations were dissected to isolate the haploid megagametophyte and diploid embryo tissues. These samples w e r e then subjected to isozyme analysis via starch gel electrophoresis during the s u m m e r of 1998. Five e n z y m e s y s t e m s w e r e assayed and ten scorable loci were detected (see Table 2.1) using slightly modified buffers detailed in Mitton et al. (1977). Table 2.1. List of enzyme loci screened for mating systems analysis. E.C. number Enzyme Locus Enzyme name Pgi Phbsphoglucose isomerase .1,2 Pgm Phosphoglucomutase 1,2 6-Phosphogluconic dehydrogenase 6Pg Isocitrate dehydrogenase Idh 1,2,3,4 Mdh Malate dehydrogenase  17 Analysis Individual genotypic data were entered into a Microsoft E x c e l ™ spreadsheet. T h e s e data w e r e assessed for linkage disequilibrium utilizing a heterogeneity-G test, a modification of the % test 2  (Sokal a n d Rohlf 1995) for each segregating pair of alleles. Linkage prevents Mendelian segregation and may obscure other genetic effects, so strongly linked loci should be excluded f r o m mating systems analyses. Based on haploid genotypic data from the maternal a n d e m b r y o tissue, an analysis w a s first conducted in order to assess the populations for linkage disequilibrium. Pairs of segregating alleles were compared and subjected to a heterogeneity-G test (Sokal a n d Rohlf 1995). T h e software application Popgene V.3.2 (Yeh etal. 1999) w a s utilized to determine linkage disequilibrium following Ohta's (1982) method by performing an analysis using c o m p o n e n t s of variance. This approach w a s developed to elucidate effects of population structure a n d gene flow by calculating and then adjusting for linkage disequilibrium. No consistent patterns of linkage disequilibrium were found in the loci examined, thus all loci w e r e retained and the genotypic data were assessed for inbreeding levels and other genetic parameters using the program M L T R (Ritland 1989,1990). This program uses genotypic or allelic frequency d a t a to calculate estimates of inbreeding at the family and population level via bootstrapping to a specified confidence level using a mixed mating model (i.e., both selfing a n d outcrossing are a s s u m e d to occur within the population). M L T R can estimate both selfing a n d biparental inbreeding, a s well as other statistics correlating the relative proportions of inbreeding between parents a n d offspring, but utilizes the assumption that progeny are either products of selfing (t = 0) or complete outcrossing (t = 1) (Ritland 1990). If data f r o m both parents a n d the offspring are not available, an inference technique using paternity exclusion is e m p l o y e d . A s e e d , or starting number is selected by the user, which is then the starting point for the bootstrapping estimates; a higher number of bootstrapped estimates will likely give a more accurate estimate of the parameters, assuming maternal fitness and potential paternal genotypes are equal a m o n g all parents, loci are not linked, and are selectively neutral (Clegg 1980; G o o d i n g 1998).  18  Each family and population w a s assessed for single-locus and multilocus estimates of outcrossing, observed measures of all parameters, and 100 bootstrapped Newton-Raphson iterations w e r e used to generate error estimates. Multilocus estimates of t (t ) are m u c h more m  informative than single-locus estimates (t ), however, since they give an integrated estimate based s  o n the total of all the information collected, and thus have far more degrees of f r e e d o m a n d statistical power, a n d they are more robust to violations of assumptions inherent in models which calculate t (Young et al. 2000). It is useful to estimate t since the difference between t s  m  a n d t can s  provide a n estimate of the amount of biparental inbreeding: if no significant difference is f o u n d , then most inbreeding likely results from selfing, but if the difference is significantly higher than zero, then m u c h of the inbreeding could be accounted for by mating a m o n g relatives (Gooding 1998). For any given data set (i.e., a family or population), the multilocus estimate of outcrossing will always b e higher due to the robustness of the data to violations of assumptions a n d thereby w o u l d provide an estimate which would reflect more outcrossing than the single-locus estimate which w o u l d tend to be biased toward more selfing (Shaw etal. 1981).  2.3 RESULTS Only heterozygous loci w h e r e it is possible to detect segregation can be used in linkage analysis, so Mdh1, which w a s monomorphic in both populations and 6Pg2, which w a s monomorphic in one population, were not included in the analysis. Table 2.2 s h o w s the results of the heterogeneity-G test in which each pair of segregating loci is tested for segregation distortion. No consistent trends were found and no pairs of alleles showed linkage w h e n analyzed following Sokal a n d Rohlf (1995). All pairs of alleles thus appeared to segregate according to random Mendelian patterns. A summary of the numbers of trees which exhibited segregation distortion at e a c h pair of loci is in Table 2.3. T h e results of the heterogeneity-G test, presented in Table 2.3, reveal that no loci significantly deviated f r o m expected random segregation patterns despite several families showing isolated instances of disequilibrium in Table 2.2. Interestingly, w h e n all loci were c o m b i n e d , the results of 19  both the pooled and heterogeneity tests were highly significant (at a = 0.05 a n d 0 . 0 1 , respectively), indicating that although no individual locus deviated from random segregation, the cumulative effect throughout both populations did show s o m e systematic bias towards the c o m m o n allele (1). Table 2.2. Number of families with pairs of loci in linkage disequilibrium (Manning and Baldy populations combined, megagametophytes only) Pgii Pgi2 Idh Pgm 6Pg1 6Pg2 Mdh1 Mdh2 Mdh3 Mdh4  2 0 0 1 0 0 4 0 5  Pgi2  Idh  Pgm  6Pg1  6Pg2  Mdh1  Mdh2  Mdh3  -  0 0 2 0 0 1 1 3  -  0 2 0 0 2 0 4  -  0 0 0 0 0 0  -  0 0 4 0 1  -  0 0 0 0  -  0 0 0  -  0 2  0  Table 2.3. Log-likelihood G-test on segregation ratios of polymorphic loci for combined populations Manning and Baldy. Locus Alleles No. of Observed Ratio Pooled G Heterogeneity G Detected Trees of Alleles (goodness of fit) (test of independence) 15 225:205 0.930568 14.64466 1,2 Pgii 138:132 0.133344 Pgi2 9 2.693182 1,2 2 Idh 30:30 0 0.266864 1,3 Pgm 17 0.197641 1,3 258:248 24.84116 1 16:14 0.133432 0 1,3 6Pg1 6Pg2 22 1.847554 1,3 330:296 25.32880 Mdh1 1 0 Mdh2 2 0.018822 1,2 31:20 2.391295 Mdh3 30 459:423 1.469796 25.47460 1,3 17 257:237 0.809937 21.44237 Mdh4 1,3 1605:3349  All  5.770842*  4636.929"  Significant at a == 0.05 * Significant at a = 0.01  Furnier et al. (1986) detected slight linkage disequilibrium between several pairs of loci in whitebark pine; however, the lowest recombination rate (r) for any pairwise test w a s 0.35 (for Adh:Pgi2),  which w a s not markedly lower than 0.50, the value representing completely unlinked  loci (Falconer a n d Mackay 1996; Haiti and Clark 1997; Bruederle er al. 1998). None of the pairs of loci exhibiting linkage disequilibrium in that study were involved in similar patterns in this study as different enzymes were investigated in both studies, and w h e r e the s a m e loci w e r e a s s e s s e d , similar patterns were not observed. While no systematic bias w a s discovered a m o n g the ten loci analyzed, overall distributions of allele frequencies deviated significantly (p < 0.05) f r o m the 20  expected segregation under assumptions of random mating (x goodness of fit test). In the case of 2  6Pg1, only one family w a s polymorphic, and only two families were polymorphic for Idh and Mdh2. T h e low number of heterozygous trees a n d total observations for these loci unfortunately decrease the statistical precision of the tests in these cases. T h e relatively small number of loci m a y have also obscured any extant linkage. Table 2.4. Estimates of t at the population level. Single-locus (SL) and multilocus (ML) estimates are equivalent, except when all loci are combined. Standard errors of the mean in parentheses. Locus  Mt. Baldy  Manning  Pqi1 Pgi2 Pgm Idh Mdh1 Mdh2 Mdh3 Mdh4 6Pg1 6Pg2 Combined SL Combined ML  0.762 (0.109) 0.888 (0.083) 0.396 (0.282) 0.621 (0.085) 0.000 (0.000) 0.646 (0.089) 1.319 (0.952) 0.897 (0.107) 0.913 (0.114) 0.000 (0.000) 0.735 (0.048) 0.736 (0.042) 0.001 (0.014) 0.082 (0.052) 0.208 (0.082) 30 853  0.493 (0.230) 0.777 (0.150) 0.123 (0.057) 0.709 (0.155) 0.000 (0.000) 0.758 (0.055) 0.614(0.385) 0.759 (0.069) 0.684 (0.160) 0.294 (0.237) 0.650 (0.061) 0.722 (0.054) 0.068 (0.025) 0.074 (0.046) 0.148 (0.063) 25 750  r  t  r  D  No. of families No. of observations  Single tree estimates of t varied from 0 to 1 for Mt. Baldy and Manning. Arithmetic m e a n s (± SE) w e r e 0.550 (+ 0.013) for the former and 0.519 (± 0.014) for the latter. Individual trees varied considerably in their estimated outcrossing rates by locus and there w a s wide variation a m o n g trees within populations. T h e single-locus and multilocus estimates of t for Mt. Baldy w e r e nearly identical (0.735 and 0.736, respectively), while for Manning they were different (0.650 a n d 0.722, respectively), suggesting slight biparental inbreeding based on t - t in the latter population, while m  s  accounting for most inbreeding in the former by selfing. T h e difference between single- a n d multilocus outcrossing rates w a s not statistically significant in either population at the 0.05 significance (a) level. t and t were not statistically different w h e n comparing all families using a m  s  paired t-test at a = 0.05. With respect to individual loci outcrossing rates, excluding fixed alleles  21  (which by definition have t = 0), the minimum for Mt. Baldy w a s 0.396 for Pgm and the m a x i m u m 1.319 (which is effectively 1.000, since numbers greater than one are a statistical artifact of the estimation algorithm, and biologically impossible) for Mdh3.  For Manning, the m i n i m u m value w a s  0.123 for Pgm a n d the m a x i m u m w a s 0.777 for Pgr/2. Manning had generally lower t values, but Mt. Baldy had an additional fixed allele (6Pg2) which lowered the combined rate for the overall population. Excluding this locus, Baldy and Manning each had multilocus t values of 0.716 and 0.546, respectively. All loci had differing estimates of the outcrossing rate (t) except for Mdh1 which w a s fixed in both populations. In most cases, t for individual loci were within o n e standard error of each other, although this w a s not so for 6Pg2, which had two alleles in Manning (t = 0.294 ± 0.237) a n d w a s fixed at Mt. Baldy (t = 0 ± 0). 6Pg1 had only one segregating family and Idh a n d Mdh2 had only two, possibly leading to lower estimates at these loci. T h e standard error for Mdh3 w a s quite large relative to the m e a n t value, since this w a s a highly heterozygous locus and there w a s considerable variation both a m o n g and within families. T h e statistic r represents the correlation between parental and progeny values of t in a t  population (Ritland and Jain 1 9 8 1 ; Ritland and El-Kassaby 1985; Ritland 1990), a n d this value w a s slightly higher for Baldy, although the results were not significantly different. r is the correlation of p  progeny, representing the chance that two randomly chosen progeny are full sibs (Ritland 1990). For Baldy, this value w a s 0.208, or almost 21 %, and for Manning 0.148, or 1 5 % . T h e s e values w o u l d double for the probabilities of randomly drawing half sibs, supporting a strongly structured population comprised of individuals with varying degrees of relatedness, but often sharing a parent or grandparent. Displayed in graphical format, family (single-tree) outcrossing estimates for all loci c o m b i n e d reveal a bimodal distribution for both populations (Figures 2.2, 2.3). There w a s a very w i d e range of outcrossing rates for both populations, although there w a s a gap for the category of t = 0.90-0.99  22  in both populations. O n e family in each population appeared t o have nearly complete selfing (t = 0.00-0.09), a n d o n e family in Manning and two at Mt. Baldy appeared t o be completely outcrossing (t > 1.00), although the estimates for individual loci vary within those families.  S?3-  C  Oi  3 0-2 V i_ Li.  0-.09 .10-.19 .20-.29 .30-.39 .40-.49 .50-.59 .60-.69 .70-.79 .80-.89 .90-.99 1.00+  Outcrossing rate (t) Figure 2.2. Frequency distribution of family outcrossing rates for all loci for Manning.  u c  0) 3  CT —  TT 0-.09  .10- 19  —  .20-29  .30-.39  T  .40-.49  . 50-.59  —  .60-.69  —  .70-.79  .80-.89  .90-.99  1.00+  Outcrossing rate (t) Figure 2.3. Frequency distribution of family outcrossing rates for all loci for Mt. Baldy.  2.4 DISCUSSION 2.4.1 MATING S Y S T E M O F S T O N E PINES Krutovskii a n d others ( 1 9 9 4 , 1 9 9 5 ) have determined the mating systems of other stone pines (subsection Cembrae), but have not done so for whitebark pine. Table 2.5 includes outcrossing 23  rates of the stone pines. Although data are lacking for the dwarf Eurasian Pinus  pumila,  outcrossing rates of stone pines are lower than many other pines, ranging f r o m 0.686 (P. to 0.974 (P. koraiensis) koraiensis)  for multilocus estimates and from 0.693 (P. albicaulis)  for single-locus estimates. While P. koraiensis  cembra)  to 0.936 (P.  w a s generally outcrossing, the other  species all exhibited significant levels of inbreeding. This is not surprising given their similar habitat types and life history characteristics, all reliant on nutcrackers for seed dispersal. Table 2.5. Outcrossing data for stone pines (subsection Cembrae) and other pines Taxonomic group  Species  t (std. error)  t (std. error)  s  m  Pinus albicaulis 0.693 (0.055) 0.729 (0.048) subsection P. cembra 0.707 (0.045) 0.686 (0.025) P. koraiensis Cembrae 0.936 (0.051) 0.974 (0.058) P. pumila n/a n/a P. sibirica 0.862 (0.054) 0.894 (0.057) P. ponderosa subsection Ponderosae 0.933 (0.052) 0.960 (0.030) P. Jeffrey? subsection Ponderosae 0.911 (0.081) 0.935 (0.021) P. contorta subsection Contortae 0.974 (0.016) 0.926 (0.034) P. monticola subsection Strobi 0.925 (0.056) 0.977 (0.023) P. maximartinezii subsection Cembroides 0.816 0.761 'this study; Krutovskii etal. 1995; Mitton etal. 1981; "Furnier and Adams 1986; Perry and Dancik 1985; EI-Kassaby etal. 1987; Ledig et al. 1999; n/a data not available for this species 1  2  2  2  2  3  5  6  7  2  3  5  6  7  Murawski a n d others (1994) found that selective logging decreased the multilocus outcrossing rate of a tropical canopy tree by 1 8 % ; Gooding (1998) found a value of 0.642 for p o n d e r o s a pine in an area under pressure from harvesting and urban development, c o m p a r e d to the 0.960 found by Mitton a n d others (1981) in areas not subject to the s a m e impacts. In light of the increased inbreeding caused by human impact, it is advisable that the effects of human intervention be carefully considered in whitebark pine ecosystems. T h e d o c u m e n t e d history of coevolution between Nucifragia influenced the mating system of subsection Cembrae  spp. and stone pines m a y have  compared to pines with wind-dispersed  s e e d s . T h e caching of groups of related individuals together and their subsequent germination and synchronized phenology would probably lead to greater opportunity for self-pollination a n d mating a m o n g relatives (biparental inbreeding) than other pines. This would account for the relatively high inbreeding found in this study, but not for the wide range of outcrossing estimates for individual  24  families. Other studies have documented similar levels of inbreeding a m o n g bird-dispersed pines in other taxonomic categories: ponderosa and maxipinon pines. T h e most likely explanation for the range and distribution of outcrossing coefficients found a m o n g families is that outcrossing rate is influenced by many genes which directly a n d indirectly affect the mating system, leading to a continuous, rather than a discrete, distribution. G e n e s affecting the mating system could impact factors such as male and female fecundity a n d fertility, pre- a n d postzygotic barriers to fertilization (especially in the case of inbred individuals), a n d reproductive phenology (e.g., timing and duration of gamete production and receptivity) (Jarne and Charlesworth 1993). T h e bimodal shape of the outcrossing rate distribution exhibited in both populations could be the result of diversifying selection acting differentially both o n the loci and families. S o m e families clearly experience a very high level of inbreeding and possibly e v e n selfing, while s o m e appear to be primarily outcrossers. If selection, environment and their interaction had similar effects across families and loci, the curve would be normally distributed; instead, the individual trees generally appear to have differential responses resulting in s o m e primarily outcrossing and others primarily inbreeding. The relatively low number of generations since glaciation w o u l d partially explain the persistence of families with intermediate outcrossing rates, since H W E has not yet been reached, both in terms of time a n d the inherent instability of a bimodal character distribution within populations. This distribution may reflect the recolonization of the species' range in the B.C. southern interior f r o m two refugial populations, o n e primarily outcrossing and the other primarily selfing. During the several ice ages during the past 100,000 years, whitebark pine populations fluctuated and were relegated to genetically bottlenecked refugia during glacial m a x i m a and expanding its range via founder effects in the interglacial periods. T h e bottlenecked populations m a y have developed a greater tolerance to selfing, w h e r e a s m o r e continuous populations retained higher outcrossing rates; as the refugial populations e x p a n d e d and e x c h a n g e d genetic material, the more gradual bimodal distribution could have developed.  25  Selection for different temporal and spatial levels of inbreeding has been found in other conifers, namely Scots pine (Hedrick  etal.  1999), western white pine (El-Kassaby  pine (Perry and Dancik 1985), ponderosa pine (Mitton  etal.  etal.  1994), lodgepole  1981) and Sitka spruce (El-Kassaby  1994). T h e s e differences have been found not only a m o n g families a n d provenances, but also a m o n g crown strata within individual trees (El-Kassaby 1994). Mating system differentiation must therefore operate at very fine scales and be exerted by a multitude of environmental a n d genetic factors. It is difficult to verify whether these differences are the direct result of selection since there are so many complex factors involved and it would take many years to conduct controlled tests in conifers to this effect. Hedrick a n d others (1999) have suggested that populations have differential susceptibility to inbreeding depression. T h e number of lethal equivalents a m o n g individuals a n d populations w o u l d differ with many of the factors suggested in the preceding discussion. Populations m a y also have different intensities of selection acting on those factors, as well as direct selection against lethal equivalents. T h e y postulated that polyembryony acted as a mechanism to effectively increase the tolerable genetic load in the event of inbreeding since the cost of producing offspring is lower since two or more proembryos are simultaneously produced, and in the event that o n e has a high genetic load, the remaining embryo(s) would likely still be successful. Karkainen a n d others (1999) using controlled pollination experiments, determined that individual  Pinus sylvestris trees  have differing  levels of tolerance to inbreeding depression, and that although the early effects of selfing w e r e that the vast majority of seeds were aborted, maternal genotype w a s the dominant factor determining fitness, measured by seed set. It is certainly possible that whitebark pine could manifest variation in selfing tolerance in a similar fashion, both at the individual and population level, although in this study unfortunately it w a s not possible to determine whether aborted seeds were primarily the products of selfing. Given the genetic architecture of most whitebark pine populations in B C , the pollen pool available to most maternal parents likely consists of a high proportion of self a n d  26  related pollen, a n d given the widespread occurrence of aborted, empty, a n d underdeveloped s e e d , it is likely that a large proportion of these types of seeds are the result of selfing. O n e type of selection applicable in this instance and explored in the context of limber pine (Pinus flexilis James) by Schuster and Mitton (1991), is kin selection. This p h e n o m e n o n occurs w h e n it is advantageous, in terms of overall survival of genotypes, for related individuals to facilitate each others' reproduction and success at the potential temporary expense of the fitness of the individuals involved (Slatkin 1987). This would allow for a higher level of inbreeding, often associated with reduced individual fitness, to be compensated for by higher overall survival of related families (Jarne and Charlesworth 1993); models suggest that in s o m e cases, mating a m o n g relatives may decrease the genetic load over time by purging deleterious alleles to a point w h e r e s o m e inbreeding can be tolerated without further reducing the fitness of the population (Kirkpatrick and Jarne 2 0 0 0 ) . T h e c o m m o n occurrence of root grafting a n d chemical transfer between individual related genotypes would also facilitate consanguineous mating by transferring photosynthates a m o n g grafted individuals for increased overall survival (Tomback and Linhart 1990. For the Manning populations, the difference between the single- and multilocus outcrossing rates w a s 0 . 0 0 1 , indicating that the inbred offspring were most likely the products of selfing. Figure 2.2 s h o w s one family in the category of t = 0-0.09, which suggests that one family has consistently high levels of selfing. For Mt. Baldy, t - t w a s 0.068, suggesting that biparental inbreeding, as m  s  o p p o s e d to selfing, is the more c o m m o n mechanism of inbreeding in this population, although one family in this population also showed nearly complete selfing (Figure 2.3). T h e correlated mating statistic, r„ w a s under 1 0 % for both populations, with a mean of 0.078, reflecting a 7 . 8 % correlation between the outcrossing rates of parents and their offspring. r w a s relatively high (0.178 average p  for both populations) in whitebark pine; El-Kassaby and Jaquish (1996) reported a value of 0.082 for western larch (Larix occidentalis  Nutt.).  27  2.4.3 SOURCES OF ERROR While isozymes are a n accepted a n d tested method of inferring mating system, s o m e uncertainties remain. Statistical power obviously increases with t h e n u m b e r o f loci used (Hamrick a n d El-Kassaby 1987) a n d t h e sample size. T h e sample size in this study, t e n loci a n d up to 3 0 s a m p l e s per maternal parent (the cone collectors assumed each clump represented o n e parent, which m a y not be the case), is generally accepted a s adequate (Sokal a n d Rohlf 1995), provided the loci s h o w Mendelian inheritance, each parent has equivalent fitness compared to other parents in t h e gametic pool a n d all maternal parents have identical outcrossing rates, a n d t h e loci are selectively neutral (Mitton 1992). While t h e loci appear t o follow Mendelian patterns of inheritance and other studies have not found strong effects of selection o n t h e loci used in this study with respect to fitness of pine trees, Figures 2.2 and 2.3 reveal that each maternal parent has a different tendency towards outcrossing; these results are consistent with those El-Kassaby a n d others (1987) found in western white pine (Pinus monticola).  T h e pollen pool contributions a n d relative  receptivity of ovules were not tested in this study, but often these assumptions are violated in other species (Hedrick etal. 1999; S h a w etal. 1981). Outcrossing rates in western white pine have been found to vary from year t o year (El-Kassaby et al. 1993), a n d this m a y also be t h e case for whitebark pine. T h e only w a y to verify this is to take samples f r o m t h e s a m e families for several years a n d assay t h e seeds for the same loci. Although t h e use of multilocus outcrossing estimates is more robust to violations of t h e statistical assumptions than single-locus estimates, there m a y still be s o m e inherent errors d u e to the effects of selection, genotype by environment interaction, a n d temporal variation in outcrossing that w a s not measured in this study. In addition, high genetic loads associated with individuals which are t h e products of selfing m a y have resulted in early postzygotic barriers to e m b r y o survival, causing a n underestimation in the number of inbred individuals.  28  A d a m s (1992) suggested that for paternity analysis using haploid tissues, only 12 polymorphic isozyme loci are required for 9 0 % confidence, and 13 loci using diploid tissues; for 9 9 % confidence, 2 3 haploid markers would be required. If each stem in a clump that w a s collected f r o m actually w a s a different genotype, then estimates of outcrossing were overestimated a s the inclusion of different individuals would lead to inflated measures of genetic diversity. A s detailed in Chapter 3, there are s o m e subjective aspects to isozyme interpretation a n d analysis related to the laboratory conditions and analysis methods (Gillet 1993). T h e program M L T R , developed by K. Ritland, has been used extensively to estimate inbreeding parameters a n d is generally accepted a s an effective tool. Occasionally, outcrossing coefficients (t) > 1.00 are calculated. While this is a statistical artifact of the calculation, a value greater than one is biologically impossible, although it could also be interpreted as a type of assortative mating for obligate outcrossers (Young et al. 2000).  2.4.4 COMPARISON WITH OTHER POPULATIONS Due to the limited nature of this mating system study (only two geographically close populations w e r e assessed), the results of this study may not be directly extrapolated to the entire species. While they d o provide a good approximation for populations in the southern Coast Mountains, the scope of this analysis is likely too small to extrapolate much beyond that. O n e other possibility is that these two populations actually could be parts of the s a m e metapopulation: they are close enough to exchange s o m e genetic material. Studies of pollen flow (Latta and Mitton 1997) show that while the vast majority of propogules are disseminated close to the parent, m a n y viable pollen grains can be transported considerable distances by prevailing winds. Seed dispersal could also account for this p h e n o m e n o n since although Clark's nutcrackers have been observed caching seeds 22 k m from the seed source (Vander Wall and Balda 1977), black bears in the a r e a also c o n s u m e large numbers of seed and have very large h o m e ranges, and could possibly disseminate  29  u n c h e w e d seeds that pass intact through their digestive tracts over thousands of hectares via their droppings.  2.5  CONCLUSION Whitebark pine has a fairly high level of inbreeding compared with other pines (mean multilocus  outcrossing rate = 0.73), but these estimates are within the range of those of other bird-dispersed stone a n d pinon pines. For the two populations tested, each had a bimodal distribution of outcrossing rates a m o n g families; t w a s highly variable a m o n g families, from nearly zero to complete outcrossing. This distribution may reflect population genetic structure facilitated by bird s e e d caching, as well as differential individual tree responses to selection acting upon a complex group of genes impacting genetic load, reproductive fitness, and consequently, mating system.  30  CHAPTER 3 - GENETIC DIVERSITY IN BRITISH COLUMBIA "Nature has good intentions, of course, but as Aristotle once said, she cannot carry them out." -Oscar Wilde, 1891  3.1  INTRODUCTION  3.1.1 IMPORTANCE OF GENETIC DIVERSITY FOR WHITEBARK PINE Maintaining the genetic diversity of whitebark pine is critical for the long-term survival of high elevation ecosystems of which it is a keystone species (Bradshaw and McNeilly 1 9 9 1 ; Erikkson et al. 1993). Although whitebark pine communities typically have low timber value, they have extremely high values in other areas (Watson etal. 1997): watershed protection, slope stability, wildlife habitat, aesthetics (Arno and Hoff 1990), First Nations cultural heritage a n d biodiversity, to n a m e a few. Natural ecosystems are regarded as a reservoir of genetic diversity w h i c h ensures future e c o s y s t e m stability (Boyle 1992; Millar and Westfall 1992). It is therefore essential that entire intact ecosystems b e protected in order to preserve evolutional processes a n d linkages of interdependent species (Boyle 1992; Leopold 1933; Ledig 1986,1988; Millar a n d Westfall 1992). T h e extremely slow growth rate of whitebark pine lends extra weight to the c o n s e q u e n c e s of the decisions that must be m a d e now in terms of genetic conservation: any impact that m a n a g e m e n t strategies h a v e m a y take d e c a d e s t o appear, a n d centuries t o remedy should they b e t h e w r o n g o n e s (Brussard 1990; Bradshaw and NcNeilly 1 9 9 1 ; T o m b a c k et al. 2 0 0 1 ; Cole a n d Landres 1996; Ledig 1986).  3.1.2 GENETIC DIVERSITY AND POPULATION STRUCTURE A s a consequence of nutcracker seed caching, whitebark (and limber) pine trees often grow in cohorts which contain related individuals (Linhart and T o m b a c k 1985; T o m b a c k a n d Schuster 1994; Bruederle etal. 1998). Trees may grow monopodially, but a multi-stemmed growth f o r m is quite c o m m o n (McCaughey 1994; McCaughey and Schmidt 1990; Ogilvie 1990; T o m b a c k and Schuster 1994; W e a v e r and Forcella 1986). On exposed ridges, a krummholtz f o r m of the tree exists (Ogilvie 1989; T o m b a c k 1986), and it also reproduces vegetatively by layering (Arno a n d  31  Hoff  1989,1990; McCaughey 1994; Rogers etal. 1999). Due to s e e d caching a n d abiotic  influences on growth form, these multi-stemmed trees m a y be a single or several individuals (Linhart a n d T o m b a c k  1985; Furnier et al. 1987; Weaver a n d Jacobs 1990). It is impossible to tell  by observation since the trees are often grafted together; genotypic analysis is the only w a y t o determine the identity of the individual stems within a multi-stemmed clump (Tomback a n d Linhart  1990). Stands typically contain related individuals, from full-sibs to half-sibs (Brussard 1990; Furnier et al. 1987; Jorgensen a n d Hamrick 1997; Rogers etal. 1999; Schuster a n d Mitton 1991;  Tomback  and Linhart 1990; Chapter 2, this study), a n d there is no apparent pattern of relatedness a m o n g stands (Furnier et al.  1987; Bruederle er al. 1998; Rogers et al. 1999). This is likely d u e t o many  nutcrackers caching seeds randomly throughout their home ranges, w h e r e each s e e d cache is likely to contain s o m e related individuals, although the placement of the caches themselves is essentially random (Tani et al.  1998; T o m b a c k a n d Schuster 1994).  In m a n y studies involving isozymes, populations coalesce into regional groups, reflecting a n overall gradient of relatedness throughout the species' range (Bruederle er al. 1998;  Jorgensen  and Hamrick 1997; Yandell 1992). T h e western portion of the species range, found along the Rocky Mountains, displays only o n e third of the genetic variability of the eastern populations. This is probably a result of the recolonization of the species range northward by populations that survived in glacial refugia (Axelrod 1986; Baker 1990; Richardson 2001) that w e r e m o r e abundant in the eastern portion of the range. Founder effects (i.e., the founding of populations f r o m a small number of individuals) d u e to subsequent recolonization via bird-mediated seed dispersal m a y have been o n e cause of the low level of population differentiation a n d high gene flow (i.e., low F a n d high  S T  N J found in northern populations (Hard 1995; Jorgensen a n d Hamrick 1997).  G e n e flow occurs between local populations via wind-pollination (Brussard  1990), but  interpopulation pollen flow is limited between regional or distant local populations by factors such  32  as w i n d dessication of pollen a n d phenological differences (Arno a n d Hoff 1989; Hamrick et al. 1992; Jorgensen a n d Hamrick 1997, personal observation). Most pollen drift occurs within populations (Brussard 1990) and seeds are typically dispersed randomly within several kilometres of t h e parents (Schuster er al. 1989). G e n e flow patterns of whitebark pine thus generate a population structure that encompasses t h e majority of t h e population genetic variation a m o n g individuals (Bruederle etal. 1998; Yandell 1992), but low differentiation a m o n g populations (Gregorius a n d Baradat 1992; Hamrick etal. 1991,1992; Jorgensen a n d Hamrick 1997; Krutovskii et al. 1995; Schuster et al. 1989). Other studies have found clear regional differentiation, a n d low differentiation a m o n g populations within regions (Jorgensen a n d Hamrick 1997; Yandell 1992; Stuart-Smith 1998): these characteristics m a y reflect t h e northward, radiative range expansion from refugial populations of the species following t h e most recent glaciation (Baker 1990; Ellstrand 1992). Founder effects resulting from nutcracker caching also influence the population genetic structure by creating a stepwise northward migration pattern where the mixture of genotypes is fairly heterogeneous a m o n g a n d within populations, but the mixture of genotypes within clumps reflects a high degree of relatedness (Jorgensen a n d Hamrick 1997; Latta a n d Mitton 1997; T o m b a c k a n d Schuster 1994; Richardson 2 0 0 1 ) . Many studies have s h o w n in conifers that the percentage of heterozygotes increases significantly with a g e from embryos to mature individuals (e.g., Bush a n d S m o u s e 1992; Politov a n d Krutovskii 1994). Several studies have found that mature conifers in general, a n d whitebark pine in particular, have excess heterozygotes (Politov a n d Krutovskii 1994; Bruederle etal. 1998; Stuart-Smith 1998; Rogers etal. 1999; Koelewijn etal. 1999). This p h e n o m e n o n could b e interpreted a s t h e result of overdominance: increased fitness of heterozygotes relative to homozygotes, but this hypothesis has not been specifically tested.  Selection against deleterious  or lethal alleles which are more frequently expressed in homozygotes, especially those which are  33  the products of selfing, is consequently manifested a s selection against homozygotes, a n d especially inbred individuals (Gregorius a n d Baradat 1992; Krutovskii etal. 1995; Fu a n d Ritland 1994; W a n g a n d Hill 1999; Morgan 2001). Both heterosis a n d genetic load, t h e former involving overdominance a n d t h e latter dominance, have been implicated in t h e cause of inbreeding depression in plants. While t h e ideal means of elucidating the root cause would involve multigeneration controlled crossing experiments a n d QTLs linked to deleterious alleles (Fu a n d Ritland 1996; Charlesworth a n d Charlesworth 1999), it is also possible to draw s o m e conclusions f r o m studies using isozymes based o n Wright's inbreeding coefficient F (where F = 1 - HJH ), e  which  can assess pre-existing inbreeding levels (Ledig and others 1997, 2000). T h e theory of kin selection could explain t h e c o m m o n occurrence of grafting which occurs between roots or stems, indicating that the tissues are often compatible a n d allelopathic interactions seldom occur between related individuals (Tomback a n d Linhart 1990; T o m b a c k a n d Schuster 1994; W e a v e r a n d Jacobs 1990); grafting has been noted in a variety of stress-tolerant conifers in ecologically severe conditions (Tomback a n d Linhart 1990). Another explanation which has been offered to support overdominance is that heterozygous individuals have increased fitness in extreme environments d u e to their inherently greater potential for adaptation a n d evolution (Lande a n d Barrowclough 1987). Most of t h e genetic analysis performed o n whitebark pine has been concluded using allozymes; extending this technique to B C populations facilitates comparisons a m o n g studies. For this purpose, isozymes are ideal markers: codominant, polymorphic, with clear alleles (Gregorius a n d Baradat 1992), they generate reproducible a n d reliable results, a n d are relatively inexpensive in terms of labour a n d equipment (Cruzan 1998). El-Kassaby (1991) a n d others have calculated that the proportion of genie variation within a n organism detectable by isozymes m a y be < 0.01 %, a n d that allozyme diversity is generally not associated with adaptive traits (Berg a n d Hamrick 1997; Bush etal. 1987). T h e number of loci is limited d u e to the nature of protein expression (Parker et  34  al. 1998), a n d in s o m e instances their neutrality has been questioned (Bush a n d S m o u s e 1992; M a r k o v a etal. 2000). Since they are based o n fundamentally different portions of t h e g e n o m e (coding v s . noncoding, respectively), results from isozymes a n d other molecular markers based o n noncoding regions such as microsatellites are not directly comparable (Petkau etal. 1997). Traditional analytic measures developed for isozymes m a y not be applicable to microsatellites a s their mutational mechanisms differ a n d their mutation rates differ by so m u c h . T h u s , it is difficult to m a k e direct meaningful comparisons between D N A markers such a s c p D N A (which has a unique m o d e of inheritance) or microsatellites and isozyme data.  3.1.3 OBJECTIVES T h e results of a genetic analysis can reveal many types of information, depending o n t h e initial objectives a n d sampling design. This study attempts to fill in existing information gaps by focusing on populations throughout B.C., encompassing the northernmost range limits of whitebark pine. Objectives of this study are: 1. to calculate basic genetic diversity statistics (expected a n d observed heterozygosity, alleles per locus, etc.) for populations encompassing the entire range of whitebark pine in B.C.; 2. to calculate Wright's F-statistics a n d compare them with the results from t h e mating systems analysis in Chapter 2; 3. to identify patterns of genetic diversity in B.C. whitebark pine; a n d 4 . to c o m p a r e t h e results of t h e above with those found in studies of whitebark pine f r o m other geographic areas, a n d attempt to explain similarities or differences.  3.2  METHODS  AND  MATERIALS  3.2.1 FIELD COLLECTIONS B e t w e e n M a y a n d August 2000, bud samples were collected from 2 9 populations, including 2 6 populations f r o m throughout the native range of whitebark pine around B.C., three f r o m t h e Alberta Rockies a n d o n e in the Washington Cascades near B.C. Of these 2 9 , 1 7 were successfully a s s a y e d in this study (See Figure 3.1 a n d Table 3.1 for populations analyzed in this study, see  35  Appendix III for a list of all populations sampled.) O n e bud w a s collected per tree along with a sample of the previous year's needles from approximately 30 trees per population. Trees w e r e s a m p l e d a m i n i m u m of ten metres apart, and only one stem w a s sampled per clump. Blister rust incidence a n d size did not influence sampling decisions, providing the tree w a s large enough to survive removal of the sample. Samples were then w r a p p e d in aluminum foil, labelled a n d stored in a portable liquid nitrogen container until they were stored in a -80°C freezer.  Figure 3.1. Genetic diversity sampling locations.  36  Table 3.1. List of sample locations summer 2000; PP = provincial park, NP = national park, Ck = creek, Mt = Mount; Mtn = mountain, Lk = lake, R = river. Pop NTS 1:50,000 Latitude (N) Elevation Location Area Mapsheet Longitude (W) (m) # 1 Hudson Bay Mtn Smithers Smithers 93U14 54°56'25" 127°19'15" 1850 2 Higgins Creek Babine Mtns PP Driftwood Ck 93L715 54°54'20" 126°46'55" 1600 3 Sweeney Lake Houston Newcombe Lk 93E/14 53°45'25" 127°12'35" 1630 4 Heckman Pass Tweedsmuir PP TusulkoR93C/12 52°32'20" 125°48'40" 1600 Perkins Peak 5 Chilcotin Tatla Lk 92N/15 51°50'45" 124°59'10" 1700 6 Tchaikazan R Ts'yl-os PR Tchaikazan R 920/4 51°12'00" 123°39'30" 1600 7 Yalakom R Lillooet Big Bar 920/1 51°04'50" 122°27'05" 1900 8 D'arcy D'arcy Birkenhead Lk 92J/10 50°31'15" 122°34'35" 1910 9 Van Horlick Ck Lillooet Duffy Lk 92J/8 50°16'20" 122°14'45" 2000 10 Whistler Mtn Whistler Whistler 92J/2 50°03'45" 122°56'00" 1700 11 Lime Lookout Clinton Clinton 92P/4 51°05'25" 121°39'55" 1980 12 Hart's Pass Okanogan USGS 1:24,000 Slate Peak 48°42'30" 120°41'00" 2050 (Washington, U.S.A.) National Forest N4837.5W12037.5/7.5 13 Kootenay Pass Stagleap PP Salmo 82F/3 49°05'10" 117°02'30" 1940 14 Jumbo Pass Purcell Mts Duncan Lk 82K/7 50°20'20" 116°38'00" 2060 15 Stanley Glacier Kootenay NP Mt Goodsir 82N/1 51°11'10" 116°04'40" 1850 16 Paget Peak Yoho NP Lk Louise 82N/8 51°26'50" 116°21'55" 2240 17 Mt Edith Cavell Jasper NP Amethyst Lks 83D/9 52°42'00" 118°03'30" 1750 18* Blackwall Peak Manning PP Manning Park 92H/2 49°05'35" 120°45'35" 2000 19* Mt Baldy Grand Forks Grand Forks 92I/4 49°10'20" 119°15'25" 2150 * Maternal megagametophyte tissue from seeds analyzed from these populations 3.2.2  GENETIC ANALYSIS  S a m p l e s w e r e removed from the freezer, placed on ice and ground with t w o drops of grinding buffer developed by S. Barnes, modified slightly from Mitton (1977; see Appendix II for details). T h r e e running buffer systems were used (Table 3.2). Buds were dissected to remove bud scales which m a y have contained secondary c o m p o u n d s that can interfere with isozyme analysis. If the current year's needles had grown past ~ 3 m m , then they were removed and stored for future use and the remaining bud tissue w a s used. Samples were ground with a handheld Conair Z i p w h i p ™ eggbeater o n ice, then 1.5mm x 6 m m chromatographic paper wicks w e r e immediately inserted to absorb the supernatant. Twelve per cent (w/v) starch gels with 5 % (w/v) sucrose were utilized for electrophoresis. Gels w e r e first run at half voltage for 30 minutes in a refrigerator at 4 ° C , after which the wicks w e r e r e m o v e d , a spacer inserted into the gel adjacent to the gel plate wall to enhance the continuity of the matrix, a n d an icepack w a s a d d e d to the top of the gel. Then the voltage w a s increased to full power a s the system ran for the duration in the refrigerator.  37  Enzyme Aat (= Got) Dia Est Gdh  G6pdh Idh Lap Mdh Pgm  6Pgd Pgi Sod Prx Adh Sdkh Fdp  System # 1 2 3  Locus 1 1 3 1 1 1 1,2 1,2,3 1 1 2 1 2,3 1 1,2 1  Table 3.2. List of loci scored for isozyme analysis. Buffer system Enzyme name 2,3 Aspartate aminotransferase 1,3 Diaphorase 1 Esterase 2 Glutamate dehydrogenase 2 G lucose-6-phosphate dehydrogenase 3 Isocitrate dehydrogenase 2 Leucine aminotransferase 1 Malate dehydrogenase 1 Phosphoglucomutase 3 6-Phosphogluconate dehydrogenase 3 Phosphoglucose isomerase 3 Superoxide dismutase Peroxidase 1,2 1 Alcohol dehydrogenase 1 Shikimate dehydrogenase 2 Fructose-1,6-diphosphatase  Buffer system Morpholine Lithium borate Tris citrate  pH 8.0 8.3 7.0  Type continuous discontinuous continuous  Run time 3.5 hrs 6 hrs 3.5 hrs  E.C. number  Reference (Clayton and Tretiak 1972) (Ridgeway et al. 1970) (Stuart-Smith 1998)  Loci were scored a s follows: the most c o m m o n allele w a s scored a s 1, a n d other alleles w e r e designated sequentially as they appeared (see Appendix IV for clarification a n d z y m o g r a m s ) . All gel slices were scored immediately, a n d then fixed with a 1:5:5 mixture of glacial acetic acid:water:methanol, w r a p p e d in plastic a n d stored in dark, cold conditions for further verification.  3.3  ANALYSIS Individual genotypic data were entered into a Microsoft E x c e l ™ spreadsheet. Data w e r e then  formatted for use in BIOSYS-2 (Swofford etal. 1997) a n d Popgene V.3.2 (Yeh etal. 1999), w h e r e loci with significant amounts of missing data were excluded from t h e analysis to avoid skewing t h e results a n d to ensure that t h e programs would process the data optimally. T h e s e loci w e r e  Prx-1,  Est-7, Est-2, Mdh-4, Pgi-1, Dia, Adh, Pgm, Aat, G6pdh and Sod. For t h e t w o populations used in the mating systems study, maternal diploid genotypes were inferred using t h e megagametophyte haplotype data, assuming that under Mendelian segregation for t w o alleles at a locus, the likelihood of assuming the correct maternal genotype f r o m t h e observed haplotypes w a s 1 - (1 - p ) , where n = the number of progeny scored, a n d p = t h e n  38  frequency of allele 1 (Hartl a n d Clark 1994). For 30 progeny, this would translate into a probability very close to unity. T h e s e genotypes were then used to conduct an analysis of the genetic diversity parameters for these populations with the s a m e methods as mentioned a b o v e . All ten loci w e r e used for these analyses. In cases w h e r e populations have low genetic differentiation, such as whitebark pine, w h e r e nearly all of Nei's (1972) genetic identities are > 0.9 (Yandell 1992), Cavalli-Sforza a n d Edwards' (1967) chord distance can effectively highlight similarities and differences. Its converse, the arc distance, is a similar measure measured along a different multidimensional surface. Matrices of physical a n d genetic distances were compared using a Mantel test (Manly 1985) with the Mantel Nonparametric Test Calculator V.2.00© (Liedloff 1999). All pairwise combinations of geographic a n d genetic distances are simultaneously assessed using Monte Carlo simulations, a n d the relationship between the two is determined based on T y p e I error (a) limtations.  3.4  RESULTS  3.4.1 REMOVING LOCI Although they w e r e consistently present aross populations and had fairly high heterozygosities, the decision w a s m a d e to remove esterase (Est) and peroxidase (Prx) loci for several reasons. First, they represent a series of complex products (Dukharev 1978) and are difficult to clearly score (Juo a n d Stotzky 1973; Lieu et al. 2001). S e c o n d , they can be produced by multiple metabolic pathways, a n d results f r o m one locus to the next revealed with a single staining procedure m a y reflect differing physiological processes or environmental influences (Copes 1978; Mayberry a n d Feret 1977). Third, P / x m a y be influenced at the constitutive level by the presence a n d severity of the  Cronartium ribicola fungus  in that not only may the tree's o w n metabolism respond by altering  the level of peroxidases, but the fungus itself may also produce or secrete other peroxidase forms that w o u l d appear w h e n tissues of the tree were analyzed (Adorada et al. 2 0 0 0 ; G a y a n d T u z u n 2000a,b; M a a a n d Liao 2000; Tyagi et al. 2000). It is also possible that the fungal hyphae m a y  39  secrete other e n z y m e s that would be detectable by electrophoresis and m a y interfere with or confound the interpretation of the resulting bands, although this would be more likely w h e n analyzing needle tissue, and depend on the degree of infection.  3.4.2 GENETIC DIVERSITY STATISTICS T h e standard genetic diversity statistics and their estimates, such as expected a n d observed heterozygosity, number of alleles per locus, etc., can provide valuable information in characterizing the genetic architecture of populations. While various studies may employ differing sampling strategies or laboratory techniques, the standardized and accepted methods of analysis a n d interpretation for isozyme data facilitate the augmentation of existing data sets with new information, as well as enabling scientists to make comparisons between species a n d taxonomic groups. T h e r e w a s o n e private allele found within the two populations analyzed for mating systems:  6Pg2-3 in  Manning (see Appendix I for pollen and ovule allele frequencies). This m a y not actually  reflect a true private allele, however; it may only indicate sampling error or be the result of small sample size of other populations since these data were generated from the inferred maternal genotypes of 55 trees. Distributions of alleles within loci tended to have o n e c o m m o n allele with frequencies > 0.85. Berg and Hamrick (1997) advocate the use of no percentage criterion w h e n calculating genetic diversity statistics such as alleles per locus (A) and percentage of polymorphic loci (P), in order to avoid artificially low heterozygosity estimates. For populations w h e r e there m a y be m a n y loci with very low frequencies of alleles other than the most c o m m o n one, using restrictive criteria could also significantly alter the character of the population in terms of genetic description. Comparison of the observed and expected heterozygosities of the mature parent trees for Manning reflects a homozygote deficiency, or heterozygote excess of 11 %, in contrast to the selfing implied and resulting excess of homozygotes in seeds found in the analysis in chapter 2 (Table 3.4). Conversely, the opposite situation applied for Baldy: homozygote excess of 11 %  40  c o m p a r e d to expected values, concurring with the results in chapter 2 that most of the inbreeding in this population w a s d u e to biparental inbreeding. T h e heterozygote imbalances w e r e not statistically different from zero (paired t-test, a = 0.05). T h e s e two populations therefore d o not statistically deviate from H W E . Genetic variability statistics in the other 17 populations are also presented in Table 3.4. Y a l a k o m , located in the Coast Mountains near the middle of the latitudinal distribution within B.C., had the lowest number of alleles per locus (A), averaging 1.6, and Edith, in the Rockies near the northern portion of the range, had A = 1.7. The highest allelic diversity w a s 2.6, f o u n d in Manning, and 2.5 in Baldy f r o m the maternal seed genotypes; both populations are in the southernmost portion of the Coast Mountains, the former contiguous with the eastern portion of the range, a n d the latter a n outlying population further to the east. Yalakom had the lowest proportion of polymorphic loci (P) at 5 0 % , Edith w a s the next lowest at 5 8 . 3 % . Paget in the northern Rockies a n d Perkins in the north Coast Mountains both had the highest value of P at 9 1 . 7 % ; Manning had 90%. O b s e r v e d heterozygosity averaged 0.213; Hudson, in the northern Coast Mountains had the lowest value (0.119) and the northern Washington population the highest (0.289). Other populations in the northwestern portion of the range also had low H , while those in the south w e r e e  markedly higher. Expected heterozygosity averaged 0.257 across all populations, with a m i n i m u m at Manning in the southern Coast Mountains (0.184); Perkins had the highest value (0.312). Populations in the Rocky Mountains did not exhibit clear trends with respect to heterozygosity. Over all 19 populations, H w a s greater than H in all but the populations at Tchaikazan, e  0  W a s h i n g t o n a n d Manning, and the difference w a s statistically non-significant in Tchaikazan a n d Manning. In most cases, H w a s greater than H by at least 5 0 % , and sometimes by a multiple of e  0  o n e or more. T h e strongest heterozygote deficiency occurred at the northwest extent of the species range, with the exception of the Sweeney population.  41  Table 3.4. Genetic diversity statistics for all populations. Standard errors of the mean in parentheses. N = mean sample number; A = mean alleles per locus, no criterion; P = percentage of polymorphic loci, no criterion; H = expected heterozygosity; H = observed heterozygosity. Manning and Baldy statistics based on ten loci. e  0  Population  Pop#  N  Hudson Higgins Sweeney Heckman Perkins Tchaikazan Yalakom D'arcy Van Horlick Whistler Lime USA Kootenay Jumbo Stanley Paget Edith Manning Baldy  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19  18.1 (0.6) 17.8 (2.4) 19.9 (1.5) 25.8 (2.0) 23.8 (2.2) 23.6 (2.2) 23.3 (1.6) 23.8 (0.8) 20.4 (1.7) 29.1 (0.4) 30.0 (0.0) 17.0 (0.0) 20.3 (1.3) 31.1 (1.5) 29.3 (0.4) 27.9 (0.9) 16.3(1.5) 750 (0.0) 853 (0.0) 23.2" (1.3)  .  Mean  A 1.8 1.9 1.8 1.8 2.0 1.8 1.6 1.8 1.8 1.9 2.0 1.8 1.9 2.0 2.1 2.2 1.7  P  (0.2) (0.1) (0.2) (0.2) (0.1) (0.2) (0.2) (0.1) (0.2) (0.2) (0.2) (0.2) (0.1) (0.2) (0.2) (0.2) (0.2)  2.6 (0.2) 2.5 (0.2)  1.9 (0.058)  75.0 83.3 66.7 75.0 91.7 75.0 50.0 75.0 75.0 75.0 83.3 75.0 83.3 83.3 83.3 91.7 58.3 90.0  80.0 69.5 (2.4)  H 0.119 (0.051) 0.169 (0.070) 0.192 (0.070) 0.127 (0.048) 0.205 (0.081) 0.264 (0.100) 0.147 (0.072) 0.277 (0.103) 0.167 (0.054) 0.229 (0.083) 0.253 (0.083) 0.289 (0.112) 0.224 (0.081) 0.214(0.078) 0.284 (0.080) 0.274 (0.088) 0.180 (0.084) 0.184 (0.068) 0.243 (0.065) 0.213 (0.012) 0  H 0.237 (0.057) 0.298 (0.065) 0.210(0.058) 0.265 (0.060) 0.312(0.062) 0.262 (0.066) 0.194 (0.066) 0.309 (0.066) 0.229 (0.052) 0.301 (0.066) 0.300 (0.054) 0.260 (0.067) 0.284 (0.064) 0.243 (0.059) 0.291 (0.064) 0.262 (0.056) 0.204 (0.062) 0.204 (0.050) 0.218(0.057) 0.257 (0.009) e  'Not including populations 18 and 19.  3.4.3 WRIGHT'S F-STATISTICS Since the t w o populations included in the mating system study were analyzed using different tissues a n d a different (although partially overlapping) set of loci, as well as a haploid data set, the salient results will not be included with those from the other populations. T h e statistical power of the analysis for these populations is such that only a brief summary of the findings will be presented here. Further detail can be found in Appendix V. For both Manning and Baldy, all of the m e a n fixation indices were not significantly different f r o m zero (t-test, a = 0.05). The proximity of the two populations to each other may explain the relatively similar results. It is in fact possible that they are representative of two sub-samples of o n e metapopulation, although the small sample sizes (25 trees for Manning, 30 for Baldy) w o u l d prohibit drawing any definite conclusions in this regard. F  42  S T  between the two populations w a s  -0.024, but w a s not statistically different from zero. This slightly negative value implies low subpopulation differentiation, which is also supported by the allele frequencies in Table 3.4. Values for F, a n d F, , respectively, were -0.025 and 0.008, reflecting negligible effects of s  T  inbreeding with respect to heterozygosity of mature individuals in these two populations. For the results presented in Table 3.5, all populations in the Selkirk, Purcell, Cariboo a n d Rocky Mountains are designated to be within the Rockies for analytical and interpretive purposes. With respect to the fixation index F| , the mean value for all populations of 0.345 s h o w s a pronounced S  effect of inbreeding a m o n g individuals within populations. T h e high standard deviation reflects the high variability a m o n g loci. Fdp w a s not fixed, but had u n c o m m o n alleles at very low frequency, resulting in a high F,  s  and F, (> 0.999 for both statistics). Nine loci had positive values, m a n y with T  values > 0.5, suggesting a significant decrease in heterozygosity a m o n g inbred individuals within e a c h subpopulation at these loci. Three loci had F, less than zero, although the value for Skd2 s  w a s -0.017, very close to zero. Table 3.5. Wright's F-statistics for Allele Mdh1 Mdh2 Mdh3 Pgm Skd1 Skd2 Fdp Gdh Lap1 Lap2 Idh Pgi2 Mean  17 populations; standard errors of the mean in parentheses. F 0.540 0.091 0.369 -0.237 0.704 -0.004 1.000 0.279 0.940 0.757 0.470 -0.250 0.388 (0.118)  Fis 0.538 0.085 0.349 -0.395 0.697 -0.017 1.000 0.194 0.936 0.739 0.396 -0.379 0.345 (0.129)  1T  F 0.004 0.006 0.031 0.113 0.023 0.013 0.071 0.105 0.0770.067 0.123 0.093 0.061 (0.012) S T  F| values were similar in magnitude a n d sign to F, , with an equally wide range of values. T  s  Overall, for all populations combined, inbred individuals express a 3 8 . 8 % decrease in heterozygosity relative to expectation under panmixis.  43  Perhaps the most informative of these three measures, F  S T  had an overall value of 0 . 0 6 1 , which  is within Wright's (1931) subjective category boundaries of 0.050 to 0.150 for populations with moderate levels of population differentiation. Values for loci ranged from 0.004 for Mdh1 to 0.123 for Idh; all loci were therefore within the low to moderate range of population differentiation a n d there w e r e no outstanding anomalies which affected the overall m e a n . T h e low standard deviation c o m p a r e d to the other F-statistics confirms this.  3.4.4  M E A N STATISTICS BY G E O G R A P H I C REGION  While the north/south divisions of population were somewhat arbitrary, due to the lack of definitive information on glacial refugia, general trends are still apparent (Table 3.6). Southern a n d northern populations had equal mean numbers of alleles per locus, and Coastal populations had a significantly lower number than other subdivisions. T h e Rockies had the highest proportion of polymorphic loci, and the Coast the lowest, although the value w a s very similar to the southern populations a n d the standard errors overlap between all subdivisions. Expected heterozygosity w a s highest in the south and the Coast Mountains, a n d lowest in the Rockies. Observed heterozygosity w a s lowest in the north (0.202) and Coast Mountains (0.203), and substantially higher in the Rocky Mountain populations (0.235): a difference of 1 4 % between east a n d west. Genetic variability w a s lowest in the Coast Mountains and highest in the Rockies, but the differences a m o n g groupings w a s not substantial; expected heterozygosity had slightly higher variability than observed. Wright's inbreeding coefficient F w a s very high (0.231) in the Coast Mountains which had the lowest H , and low (0.085) in the Rockies, which had the highest 0  H . There w a s a heterozygote deficiency of 0.223 (22%) in the north and 0.151(15%) in the south. 0  Table 3.6. Genetic diversity statistics for subdivided population groupings; standard errors of the mean in parentheses; all abbreviations as in Table 3.4. Group  Pop#  Coast Mtns 1-12 Rockies 13-17 North 1-6, 15-17 South 7-14  # of Pops 12 5 9 8  N 21.9 25.0 22.5 23.1  (1.5) (2.8) (1.6) (2.3)  A 1.8 2.0 1.9 1.9 44  (0.03) (0.09) (0.06) (0.05)  P 75.0 80.0 77.8 75.0  (2.9) (5.7) (3.7) (3.9)  H (0.012) (0.016) (0.013) (0.014) e  0.265 0.257 0.260 0.265  H (0.017) (0.019) (0.020) (0.018) 0  0.203 0.235 0.202 0.225  3.4.5 PHYSICAL VERSUS GENETIC DISTANCES Physical and genetic distances often are not directly related. This m a y certainly be the case for whitebark pine in B.C.'s complex topography, which m a y s h o w genetic patterns of relatedness along major drainages or mountain ranges, while physically adjacent mountain ranges m a y be closer. Bird seed dispersal m a y generate a stepwise dispersal pattern w h e r e adjacent populations are genetically closer to each other than to more distant populations, but other factors m a y obscure these patterns, if they d o exist. While populations 1 (Hudson) a n d 2 (Higgins) appear very close to each other, they are located on opposite sides of a major drainage. S o m e populations, such as the northern W a s h i n g t o n population, are physically closer to populations located in different mountain ranges than to those within the s a m e range (Table 3.1). Table 3.7. Physical most parsimonious distances between populations (km). Population numbers as in Table 3.1. 1 2 4 3 5 6 7 8 11 9 10 12 13 14 15 16 2 25 3  140  130  4  265  270  145  5  360  360  235  6  450  445  340  135  115  7  520  515  410  280  200  80  95  8  560  555  445  305  220  110  55  9  595  590  480  345  255  145  80  40  10  595  590  470  330  240  145  110  55  11  540  530  440  310  240  125  50  90  95  145  12  815  815  710  565  550  370  295  260  220  240  290  13  920  930  920  710  635  520  435  425  390  430  415  240  14  845  825  760  645  580  475  395  405  375  430  350  315  160  15  845  825  780  345  615  505  435  455  430  410  385  385  235  85  16  795  775  730  630  575  475  405  430  410  465  355  400  265  105  50  17  640  620  590  510  475  390  340  380  370  435  295  455  395  250  215  60  165  Nei's (1972) genetic distance w a s not highly informative, but is included here for comparisons with other studies (Table 3.8). Populations are too genetically similar to accurately g a u g e relative genetic distances using this measure, exhibited by the fact that many pairs of populations have genetic distances of 0.000 and very few are > 0 . 1 . This may reflect the close relationships a m o n g 45  all populations following postglacial recolonization f r o m few refugia, but m o r e fine-scale relationships are difficult to detect using this m e t h o d . Nei's (1978) u n b i a s e d distance, a d j u s t e d for population s a m p l e size, is slightly improved for this purpose. Using this m e a s u r e , t h e populations most closely related to each other w e r e 7 (Yalakom) a n d 9 (Van Horlick), w h i c h are s e p a r a t e d by 8 0 kilometres; 12 (Washington) a n d 17 (Edith), separated by 4 5 5 k m ; a n d 16 (Paget) a n d 1 7 , 1 6 5 k m apart. H e c k m a n (4) s h o w e d t h e greatest genetic separation f r o m any other populations: genetic distances b e t w e e n 4 and 7, 8 (D'arcy) a n d 9 w e r e highest, 0.132, 0.118 a n d 0.134, respectively, but all of t h e s e populations w e r e within the Coast Mountains a n d not physically v e r y distant. Table 3.8.  Nei's 1972 genetic distance for 17 populations (above diagonal) and unbiased (1978) genetic distance (below diagonal). Population numbers as in Table 3.1.  Pop  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  1  *****  0.004 0.018 0.071 0.038 0.025 0.051 0.050 0.059 0.057 0.033 0.049 0.046 0.028 0.057 0.028 0.043  2  0.026 ***** 0.023 0.072 0.032 0.037 0.066 0.052 0.065 0.069 0.043 0.051 0.055 0.039 0.053 0.030 0.040  3  0.026 0.044 ***** 0.050 0.022 0.016 0.059 0.048 0.068 0.036 0.012 0.026 0.033 0.013 0.019 0.022 0.023  4  0.080 0.093 0.058 ***** 0.043 0.067 0.122 0.109 0.126 0.092 0.077 0.081 0.080 0.093 0.070 0.082 0.096  5  0.048 0.056 0.032 0.053 ***** 0.000 0.039 0.020 0.034 0.011 0.016 0.008 0.020 0.018 0.009 0.005 0.015  6  0.036 0.061 0.027 0.078 0.012 ***** 0.012 0.031 0.012 0.002 0.014 0.005 0.013 0.000 0.014 0.000 0.010  7  0.061 0.090 0.068 0.132 0.050 0.025 ***** 0.050 0.000 0.024 0.048 0.018 0.010 0.019 0.055 0.010 0.016  8  0.059 0.075 0.057 0.118 0.030 0.042 0.060 ***** 0.047 0.023 0.027 0.019 0.029 0.026 0.032 0.022 0.023  9  0.068 0.088 0.077 0.134 0.045 0.024 0.006 0.056 ***** 0.022 0.057 0.025 0.020 0.021 0.055 0.012 0.025  10 0.065 0.091 0.044 0.100 0.020 0.013 0.033 0.031 0.030 ***** 0.015 0.007 0.015 0.009 0.020 0.011 0.020 11  0.041 0.064 0.019 0.084 0.025 0.024 0.057 0.036 0.066 0.023 ***** 0.008 0.014 0.014 0.016 0.014 .0.12  12 0.059 0.074 0.035 0.090 0.018 0.017 0.029 0.029 0.035 0.016 0.016 ***** 0.003 0.011 0.016 0.002 0.000 13 0.055 0.078 0.042 0.089 0.031 0.025 0.021 0.039 0.030 0.024 0.023 0.013 ***** 0.014 0.017 0.008 0.003 14 0.035 0.059 0.019 0.100 0.027 0.010 0.027 0.033 0.028 0.016 0.020 0.019 0.022 ***** 0.015 0.006 0.012 15 0.065 0.074 0.026 0.078 0.018 0.024 0.064 0.040 0.063 0.027 0.023 0.025 0.026 0.022  ***** 0.018 0.021  16 0.035 0.051 0.029 0.089 0.014 0.006 0.019 0.030 0.020 0.018 0.021 0.010 0.016 0.012 0.024 *****  0.000  17 0.053 0.064 0.033 0.106 0.027 0.022 0.027 0.033 0.035 0.029 0.021 0.007 0.017 0.021 0.029 0.008  *****  T h e m i n i m u m Cavalli-Sforza a n d Edwards' (1967) chord distances w e r e f o u n d b e t w e e n 6 (Tchaikazan) a n d 12 (Washington), w h i c h are located in the Coast Mountains 3 7 0 k m apart; a n d 6 a n d 16 (Perkins), located in different mountain ranges (Table 3.9). T h e m a x i m u m c h o r d distance w a s b e t w e e n 4 ( H e c k m a n ) a n d 8 (D'arcy), which concurs m o r e or less with t h e results f r o m Nei's distance statistics.  46  Table 3.9. Below diagonal: Cavalli-Sforza & Edwards (1967) arc distance; above diagonal: CavalliSforza & Edwards (1967) chord distance. Pop  1  1  *****  2  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  0.132 0.182 0.227 0.184 0.171 0.178 0.236 0.175 0.211 0.197 0.206 0.187 0.177 0.220 0.162 0.223  4  0 133 ***** 0.194 0.233 0.151 0.189 0.230 0.230 0.185 0.219 0.211 0.208 0.216 0.200 0.216 0.165 0.224 ***** 0.176 0.159 0.124 0.202 0.222 0.195 0.159 0.151 0.143 0.169 0.140 0.172 0.140 0.159 0 184 0 231 0.179 * **** 0.194 0.193 0.273 0.297 0.241 0.234 0.243 0.228 0.230 0.242 0.245 0.213 0.275  5  0 187  0.160 0 196 ***** 0.111 0.195 0.158 0.144 0.113 0.148 0.127 0.154 0.145 0.152 0.102 0.172  6  0 172  0.125 0 196 0.111  7  0 180  0.206 0 278 0.198 0.144 ***** 0.215 0.106 0.177 0.208 0.153 0.153 0.145 0.213 0.146 0.158  8  0 240  0.225 0 311 0.159 0.187 0.218 ***** 0.196 0.147 0.181 0.159 0.194 0.152 0.187 0.154 0.153  9  3  ***** 0.143 0.185 0.115 0.105 0.146 0.086 0.133 0.097 0.146 0.090 0.146  0 178  0.200 0 245 0.145 0.115 0.107 0.198 ***** 0.142 0.195 0.143 0.148 0.138 0.194 0.115 0.169  10 0 213  0.160 0 239 0.114 0.106 0.179 0.148 0.143 ***** 0.130 0.109 0.129 0.126 0.149 0.102 0.156  11 0 199  0.152 0 251 0.149 0.146 0.210 0.182 0.197 0.131  12 0 208  0.144 0 234 0.127 0.087 0.154 0.161 0.144 0.109 0.132 *****  13 0 188  0.170 0 233 0.155 0.134 0.154 0.196 0.149 0.129 0.111 0.123 ***** 0.116 0.132 0.114 0.157  14 0 178  0.141 0 247 0.146 0.097 0.146 0.153 0.138 0.126 0.124 0.110 0.116 ***** 0.122 0.106 0.125  15 0 224  0.173 0 249 0.153 0.147 0.217 0.189 0.196 0.150 0.154 0.144 0.132 0.122 ***** 0.125 0.171  16 0 164  0.141 0 217 0.102 0.090 0.147 0.154 0.116 0.102 0.133 0.095 0.115 0.106 0.126 ***** 0.117  17 0 226  0.160 0 288  ***** 0.131 0.111 0.124 0.153 0.133 0.158 0.122 0.109 0.143 0.095 0.102  Physical and chord distances were compared with a nonparametric Mantel test (Manly 1985; Jorgensen a n d Hamrick 1997) using the program Mantel V.2.00©. (Liedloff 1999). Results from 10,000 random permutations were significant at the p = 0.05 level adjusted for the number of pairwise tests, indicating that physical and genetic distances are correlated in this species in B.C. Using a standardized Mantel Z-score, the critical value w a s 1.645 for a = 0.05 b a s e d o n 136 pairwise comparisons; the test statistic, g (also termed the standard normal coefficient) w a s 1.872 resulting in a p-value of 0.0355. T h e Pearson correlation coefficient, r, for the t w o matrices w a s 0.24, indicating that 2 4 % of the data in each matrix is explained by the other; conversely, 7 6 % of the relationship between physical and chord distance could not be explained by either factor. A d e n d r o g r a m based on five iterations of the W a g n e r distance procedure using the CavalliSforza and Edwards (1967) chord distance revealed in each iteration that there w e r e only slight correlations with spatial and genetic distances (final iteration shown in Figure 3.2). While s o m e populations consistently grouped together, such a s Hudson and Higgins, a n d S w e e n e y a n d H e c k m a n , groupings of populations, especially a m o n g mountain ranges, were w e a k . Edith (in t h e  47  east) a n d Lime (in the west), for example, are both found in clades which contain populations f r o m the opposite, western a n d eastern portions of the species range, respectively. In t h e various iterations, the positions a n d branch lengths of most populations a m o n g clades w a s variable.  0.00  0.02  Distance from root 0.07 0.09  0.05  0.12  0.14  *************************** Hudson (1) ********** ********* * ******************************* Higgins (2) ******* * * * * * * * * * * * * * * * * t! * * * * * * * * * * * * * * Yalakom (7) ** ********** ** ******************* Van Horlick (9) ** * * * * * * * * * * * * * * * * * * * * * Tchaikazan (6) * * * * * * * * * * ** * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * D'arcy (8) * * ***** ******* **************************** Edith (17) ** * ** ******************** USA (12) ** ** ************************* Perkins (5) ****** *** ************************* Whistler (10) ******** * * * ****************** Paget (16) * * * * *** ** * * * * * * * * * * * * * * * * * * * * * * Lime (11) * * ***** * * ** ********************** Kootenay (13) * * ** * ***************************************** Stanley (15) * ' * * ********************* Jumbo (14) * * *********************** Sweeney (2) ********* ************** *************************************** Heckman(4) r  o.o6"~~ ""~ao2~~~ ~~ otd5"" "~"oToy"~ ""~broy"" "~~6riy~ ~~~ o!i4 +  +  +  +  +  -+  Figure 3.2. Wagner tree produced by rooting at midpoint of longest path (after optimization) by BIOSYS-2. 3.4.6  G E O G R A P H I C P A T T E R N S IN GENETIC DIVERSITY  There were surprisingly strong correlations between geographical variables a n d heterozygosity (Figures 3.3 to 3.8). Correlations with observed heterozygosity were statistically significant at a significance level of 0.05, a n d showed the strongest trends (for latitude: R = 0.357, p = 0 . 0 1 1 ; for 2  longitude: R = 0.295, p = 0.024). Regressions on expected heterozygosity were w e a k a n d not 2  statistically significant (for latitude; R = 0.039, p = 0.450; for longitude; R = 0.000, p = 0.992). 2  2  Genetic diversity generally increased towards the south a n d east.  48  0.30  y =-0.0186x+ 1.1692  • •  8  •  0.25  N  •  § CD  0.20  -o  0.15  R = 0.3571 2  • •  •  •  •  CD  •  •  $  •  0.1 O H  O  48.00  50.00  52.00  54.00  56.00  Degrees Latitude, N Figure 3.3. Regression of observed heterozygosity on latitude.  0.30  •  co  • *  8) 0.25 o £ 0.20 CD  X "S  y = -0.0077x +1.1563  •  •  •»  •  •  R = 0.295 2  • ^  •  0.15  —  •  •  CD -Q 0.10 H O 115.00  120.00  125.00  130.00  Degrees Longitude, W  Figure 3.4. Regression of observed heterozygosity on longitude.  0.35  y = -0.0042x + 0.4679 R = 0.0385 2  §  0.30  o  0.25  CD  • •  * * *  X "O  •  • •  •4—' CD  •  • •  0.20  CD o 0.15 S" 48.00 LU  50.00  52.00  54.00  Degrees Latitude, N  Figure 3.5. Regression of expected heterozygosity on latitude.  49  56.00  0.35  y = - 2 x 1 0 x + 0.2655 ^ R = 7x10" 5  2  to o  6  0.30  O) N O  0.25  L_  •  CD CD  X  •  0.20  T3  B O  CD Q. X  0.15  UJ  115.00  120.00  125.00  130.00  Degrees Longitude, W Figure 3.6. Regression of expected heterozygosity on longitude. Correlations between F, Wright's index of heterozygote deficiency measured by 1-(Ho/H ), also e  s h o w e d strong and significant patterns in relation to geographic variables (Figures 3.7 a n d 3.8). T h e regression coefficient (R ) for F with latitude w a s 0.296 (p = 0.024) a n d with longitude 0.368 ( p 2  = 0.010). Heterozygote deficiency increased with increases in both longitude and latitude, indicating the strongest heterozygote deficiencies in the north and west (i.e., the Coast Mountains), and the tendency towards small heterozygote excesses in the south a n d eastern regions of B.C.  0.6  • •  •  0.4  •  F 0.2 •  i  0.0  y = 0.0559X - 2.693  •  R = 0.2958 2  -0.2 48.00  50.00  52.00  54.00  Degrees Latitude (N) Figure 3.7. Regression of Wright's F on latitude.  50  56.00  0.6 • •  0.4  •  •  F 0.2 0.0  •  •  •  y = 0.0286X - 3.298 R = 0.3676 2  -0.2 115.00  120.00  125.00  130.0Q  Degrees Longitude (W) Figure 3.8.  3.5  Regression of Wright's F on longitude.  DISCUSSION  3.5.1 PATTERNS OF GENETIC DIVERSITY A s Stuart-Smith (1998) pointed out, bird-mediated seed dispersal would likely obscure genetic patterns stemming f r o m founder effects. This is due to the fact that birds have been d o c u m e n t e d to randomly cache seeds with respect to collection and deposition locations; i.e., a bird would be no more likely to cache seeds f r o m the s a m e stand near each other than near caches f r o m other stands. This would lead to the integrated mosaic of genotypes that has been found by other researchers (e.g. Bruederle etal. 1998; Furnier etal. 1986; Miller a n d Westfall 1992; Rogers etal. 1999) in which individuals within clumps are likely to be related to each other, but genotypes between clumps show no isolation by distance relationships a n d tend to be randomly distributed. T h e data f r o m this study correlate well with the results of both Stuart-Smith (1998) a n d Yandell (1992), although again the standard deviations reflect the relatively small sample size a n d number of loci. Bruederle etal. (1998) and Jorgensen and Hamrick (1997) had lower estimates of H , in e  the order of 0.15 and 0.10, respectively for their studies, although the standard deviations overlap with ranges in this study. T h e geographic ranges of studies and laboratory conditions, including loci analyzed, may account for s o m e of the differences (Conkle 1971).  51  Table 3.10. Summary of whitebark pine genetic data by study. Superscripts indicate criterion percentage. Study  # of  Area  pops 1  9  Yellowstone NP  # of  Analysis  loci  level  19 Population  A  %P  H  0  H or 0  F  F  IS  H, 1.66 1.6  no  or  ST  F  G  ST  38.6  0.148  0.152  0.016  25 85  0.088  0.092 0.102 0.267  0.041  0.025  0.026  95  2  30  Entire range but BC  20 Population Species  3  14  Great Basin  13 Population  4  29  BC/AB Rockies  16 Population Species 21 Hierarchical  95  no  1.6" 48.8  0.191  0.204  0.060  0.043 0.034 0.143  0.088  0.064  1.64 50.2 0.218 0.211 -0.033 2.06 56.3 0.218 0.224 -0.035 0.030 0.062 0.027 5 3 MonoLk, CA 5.6- 0.033- 0.026- ~ -0.3 - 0 . 3 -0.055 0.00419.1 0.089 0.087 0.334 6 17 BC 12 Population 1.9 69.5"°0.212 0.257 0.154 0.207 0.061 0.168 1 - Bruederle etal. 1998; 2 - Jorgensen and Hamrick 1997; 3 - Yandell 1992; 4 - Stuart-Smith 1998; 5 - Rogers etal. 1999; 6 - this study  T h e results of the province-wide isozyme analysis reflect heterozygosity statistics similar to those calculated by Stuart-Smith (1998) w h o performed an analysis based on populations along the B.C.-Alberta border (overlapping with the current study), as well as Yandell's (1992) study of Great Basin populations, none of which were assayed here. Results from this a n d the t w o aforementioned studies, however, have heterozygosities up to twice as high as those found in s o m e other studies (Bruederle et al. 1998; Jorgensen and Hamrick 1997). Reasons for these discrepancies likely include the tissue sampled and the sampling season, whereby different loci a n d levels of e n z y m e activity are expressed at different phenological stages, thus making expression s o m e w h a t influenced by sampling date. Adh, an e n z y m e involved in leaf abscission a n d dormancy, is one e x a m p l e of such a locus; Prx, involved in eliminating toxins a n d w a s t e is another. Other studies have s h o w n that there can be differences in the loci expressed a m o n g tissue types; while most other studies have used needle tissue, this study has used bud tissue that includes primordial needles, meristem tissue and possibly the following year's primordial reproductive tissues. Thus, isozyme studies examining different tissues or loci for the s a m e organism could have differing results. Studies involving northerly populations generally have higher overall heterozygosity, while those focusing on unglaciated and southerly areas have lower heterozygosity, in contrast to results s h o w n in Figures 3.3 and 3.5. This may reflect differences in laboratory procedures, buffer  52  systems, statistical methods and resolution of bands following staining, but it m a y also reflect actual differences. O n e w a y to resolve these discrepancies is to compare values for different studies w h e r e the s a m e populations were sampled. Jorgensen a n d Hamrick (1997) s a m p l e d o n e population which w a s also sampled in this study, Mt. Edith Cavell. T h e s e results w e r e quite different to Jorgensen and Hamrick (1997), yet similar to what Stuart-Smith (1998) f o u n d : both observed and expected heterozygosities were substantially higher in this study (0.204 a n d 0.180, respectively) than the former (0.080 and 0.088, respectively). T h e s a m e differences w e r e found between populations analyzed by Yandell (1992) and Jorgensen and Hamrick (1997): heterozygosity values were two to three times higher in Yandell's study. Stuart-Smith attributed the differences between Jorgensen and Hamricks' results and the other studies primarily to differences in laboratory technique and resolution of c o m m o n alleles in electrophoresis; this possibility w a s supported by Hamrick (J. Hamrick, U. of Athens, GA, Depts. of Botany and Evolution, pers. c o m m . 2001). T h e use of the 9 5 % criterion in the definition of statistical parameters w a s also implicated, a n d he demonstrated that the low 2 5 % level of polymorphism could be increased to 8 5 % by employing a 9 9 % criterion. Since the data in this study concur fairly closely with those of StuartSmith (1998), I a m inclined to agree with his hypotheses, especially since similar buffer systems w e r e used for this and his study. Differences found in statistics and distribution for the s a m e alleles a m o n g studies could be c a u s e d by selection acting on non-neutral loci linked to those analyzed in this study or drift c a u s e d by founder effects affecting allele frequencies. Latta and Mitton (1997) found very different dispersal patterns and selection regimes for different markers measured for limber pine, a n d several researchers emphasize the selective role that Clark's nutcracker plays for bird-dispersed species (Carsey and T o m b a c k 1994; Vander Wall and Balda 1977; Schoettle a n d Rochelle 2 0 0 0 ) . T h e relatively recent glaciation and long generation time would also indicate that whitebark pine allele frequencies w o u l d not yet approach expected equilibrium values in many c a s e s . Since each  53  study analyzed different loci to calculate population estimates of genetic parameters, the inclusion in o n e study of several loci which depart from the typical pattern of allele distribution w o u l d skew the results f r o m the s a m e population. A n interesting relationship between expected and observed heterozygosity is revealed in this study. It is apparent f r o m the geographic relationship between regional blister rust mortality (highest in the southern Rockies (Campbell 1998; Stuart-Smith 1998; pers. observ.) a n d Wright's F (Figures 3.7 and 3.8) that the areas w h e r e mortality from the disease is highest s h o w a striking trend towards heterozygote excess, while those where the pathogen is either absent or has a less severe impact on the population tend towards heterozygote deficiency. This could possibly be interpreted as evidence supporting selection against homozygous genotypes, or conversely, that populations with a higher proportion of heterozygotes are able to withstand or tolerate the effects of white pine blister rust via the wider available range of biochemical mechanisms conferred by increased genetic diversity. A causal basis for this apparent correlation between heterozygosity a n d degree of blister rust infection would need to be verified with a study specifically investigating the relationship between heterozygosity and resistance on an individual-tree basis. Peripheral populations did not appear to show any striking genetic patterns, although there w e r e higher heterozygosities in the south and east (Table 3.6). Lesica and Allendorf (1995) suggest that there m a y be s o m e selective advantage in those populations in terms of future adaptations, since selection pressures differ in peripheral or extreme environments, but lower heterozygosity and a loss of rare or private alleles would be two consequences if they underwent genetic bottlenecks or w e r e recently diverged (Nei etal. 1975). Certainly, both options are possible, although the high heterozygosity of nearly all populations would counter arguments for a genetic bottleneck. T h e general lack of private alleles may reflect repeated founder effects resulting f r o m bird s e e d caching instead, a n d the high heterozygosity could result from multiple founding events in the s a m e area, which has been d o c u m e n t e d (Tomback and Schuster 1994; Richardson 2001).  54  While their ranges d o not generally overlap in C a n a d a in terms of elevation, whitebark a n d limber pine share many similarities, especially morphologically. T h e two are generally indistinguishable in the absence of cones, which is the key differentiating characteristic in their taxonomies (Little a n d Critchfield 1969; Critchfield 1986). Both species also have corvids in c o m m o n as their primary agents of seed dispersal (Tomback a n d Linhart 1990). W h i l e the c o n e s of limber pine d o open upon maturity, leaving a wider spectrum of opportunities for s e e d dispersal, birds remain a key factor influencing limber pine gene flow and population structure. Exact results f r o m studies conducted on limber pine cannot be directly applied to whitebark pine, however general deductions regarding topics such as the role of birds influencing genetic neighbourhoods or g e n e flow using different measures a n d markers may be extrapolated w h e n the circumstances are similar for both species. Schoettle a n d Rochelle (2000) found strong selection effects on morphology a n d phenology of limber pine across a wide ecological amplitude. Schuster etal. (1989) found strong elevational differences for the limber pine, and that there w a s fairly high gene flow via s e e d . Latta a n d Mitton (1997) discovered strong differences in gene flow and selection pressures a m o n g male a n d female g a m e t e s using different molecular markers. T h e elevational d i n e s apparent in limber pine m a y not apply directly to whitebark pine due to whitebark pine's more restricted nature in terms of elevation, but it is likely that the two species share s o m e selection pressures and similarities in effective pollen a n d seed gene flow patterns that explain s o m e of the large and small scale population differentiation. T h e patterns of genetic diversity found in this study confirm reconstructed biogeographic patterns of postglacial recolonization (Richardson 2001), reflecting fairly recent founder effects. T h e higher levels of heterozygosity in the south and east imply that bird-mediated s e e d dispersal generally progressed northwesterly, and that there were either more refugia in the east, or that there w e r e s o m e refugia in the Rockies at more northerly latitudes, concurring with a hypothesized  55  refugium east of Roger's Pass, just outside Glacier NP (B. Richardson, U S D A For. Serv., pers. comm.). W h e n subjected to the Mantel test (Jorgensen and Hamrick 1997; Manly 1985) to g a u g e isolation by distance, a statistically significant correlation between physical a n d genetic distance w a s revealed. This does provide s o m e support for the repopulation of habitat by founder effects f r o m glacial refugia, reflecting a stepping-stone model. Very little, if any, cross-Cordilleran migration w o u l d have been likely: the closest distance between known whitebark pine provenances in the Coast and Rocky Mountain ranges (approximately 60 km) is still greater than the furthest recorded nutcracker caching distance (22 k m , cf. Vander Wall and Balda 1977. T h e complex life history factors affecting whitebark pine interact across many scales a n d exert different influences both at different life stages and ecological stages (Bruederle etal. 2001). T h e s e factors, which are all affected by selection, include dispersal of pollen and seeds, ecological micro a n d macro-scale effects, successional stage and life history characteristics of the species. M a n y trends w o u l d only be obvious using a highly specific and intensive sampling s c h e m e across several spatial scales. Selection is generally stronger, although heritability is low for fitness a n d adaptive traits, which generally involve many genes (Merila and Sheldon 1999). Acknowledging the subtle a n d complex interplay a m o n g adaptive traits, Vida (1994) suggested that there m a y b e more adaptive traits than neutral, calling into question the broad applicability of allozyme markers. Further research on adaptive traits is clearly necessary in this species in order to understand the physiological aspects of intra- and interpopulation variation to develop effective conservation measures.  3.5.2 WRIGHT'S F-STATISTICS T h e various F-statistics developed by Wright (1931,1951,1965) can be used to infer the prior history of inbreeding within populations as well as to gauge hierarchical levels of population differentiation. T h e s e statistics use the estimated reduction in heterozygosity a m o n g inbred  56  individuals attributed to inbreeding as a gauge of the degree of relatedness a m o n g individuals within a n d a m o n g populations (Haiti and Clark 1997). Assumptions critical to these calculations include (1) that all loci follow Mendelian inheritance and (2) that populations are in Hardy-Weinberg equilibrium. W h i l e other studies have demonstrated the general applicability of the former condition (Furnier etal. 1986), the latter is almost certainly not true: nearly all of the areas currently populated by whitebark pine in B.C. have been glaciated as recently as 8,000 years ago, a n d s o m e even more recently. If the average generation time is approximately 80 years and s o m e areas have been recolonized from a few glacial refugia near Roger's Pass as well as several others in the northern U.S., then approximately 100 generations have passed since the most recent glacial event. Most populations w o u l d likely have originated exhibiting severe founder effects due to bird caching, a n d in such environmentally extreme habitats are subject to constant environmental selection pressure. In mountainous topography, extremely fine micro-scale differentiation could conceivably result d u e to environmental variability in terms of snowpack, temperature gradients, moisture availability and intra- a n d interspecies competition. This fine scale differentiation, if not s w a m p e d by local gene flow, could be quantified using systematic sampling. Recently, many populations have also been experiencing severe selection pressure from the introduced white pine blister rust pathogen. It is nearly impossible, even under the most liberal assumptions, to a s s u m e that any of these populations w o u l d be experiencing conditions approaching genetic equilibrium. T o s u m up, many of the critical underlying assumptions of Wright's F-statistics a n d other calculations involving the preconditions of Hardy-Weinberg equilibrium (such a s heterozygosity calculations) are violated to s o m e degree by this species, particularly in the northern portion of its range. Assumptions made in estimating genetic parameters for unglaciated areas in the southern portion (the Great Basin, portions of the Sierra Nevada and Yellowstone NP, s o m e other refugia) may be more accurate than the (glaciated) northern portion (Yandell 1992; Jorgensen a n d Hamrick  57  1997; Bruederle et al. 1998) due to the confounding effects of repeated bottlenecks a n d population size fluctuations. Despite assurances that these formulae are fairly robust with respect to violations of many of the assumptions required, it is unlikely that the degree of divergence can be o v e r c o m e to provide highly accurate data. However, since these are the generally accepted universal m o d e s of expressing and calculating genetic diversity statistics, and since other formulae are not generally used, I have employed the standard formulae, but with an a d d e d caveat. T h e fact that 9 4 % of population genetic variation w a s found a m o n g individuals within populations ( F  S T  = 0.061 for 12 loci) concurs with fairly high gene flow implied by the results of  other studies (Stuart-Smith 1998; Jorgensen and Hamrick 1997; Yandell 1992), a n d is typical of m a n y conifers (Hamrick and others 1991,1992). Fixation indices (F, ) w e r e fairly high (> 0.3) for s  most loci assayed in this study. T h e most likely explanation for a positive F, using neutral markers s  (i.e., isozymes) is identity by descent, or that the individuals assessed for allele frequencies share a c o m m o n ancestor; a high F, indicates a strong correlation between alleles within subpopulations s  relative to those found within that subpopulation under random mating (Wright 1965). This correlates fairly well with the level of inbreeding found in the mating systems analysis (Chapter 2, this study). F, values followed generally the s a m e trends and variability as F, , showing that for T  s  e a c h locus the effects of inbreeding a m o n g individuals within the total population w e r e similar to those within subpopulations.  3.5.3 SOURCES OF ERROR A s Stuart-Smith (1998) pointed out, a W a h l u n d effect, w h e r e separate populations are c o m b i n e d a n d analyzed as a single population (Hartl and Clark 1997) m a y obscure population genetic patterns a n d account for higher than expected heterozygosity values. In whitebark pine, this m a y occur if a single population sample w a s taken from individuals representing two different populations. O n e case w h e r e this may occur is where ecological factors create a physical disjunction in the environment, causing two or more populations adjacent to each other to  58  superficially appear as one population. This situation can occur w h e r e part of a population spans a n a r e a disturbed by fire or avalanche, creating a successional disjunction, or w h e r e o n e population is spread over a n area large enough to be characterized by different slope aspects, soil or geomorphological types. This also could result if the local environment caused a disjunction in reproduction by affecting phenology, e.g., elevation affecting the dates at which physiological threshhold values of climatic variables are reached. Given the longevity of the species, a W a h l u n d effect cannot be ruled out in s o m e cases, although the majority of populations w e r e s a m p l e d along a single slope face within a fairly narrow elevational range (< 100 m) and appeared to f o r m a single continuous population. O n e other possible source of error w a s scoring. While many alleles w e r e clearly detectable a n d distinguishable f r o m each other, there were s o m e cases w h e r e the distinctions w e r e less obvious. Attempts w e r e m a d e to mitigate this problem by including samples f r o m several different populations o n the s a m e gel to facilitate comparison, and preparing several runs using the s a m e buffer mixture a n d keeping t h e m frozen until used. Double runs were often conducted so that m a n y samples w e r e processed using the exact s a m e grinding and running buffer systems a n d stains. All gels w e r e fixed and then rescored following the completion of all initial runs to ensure consistency within a n d a m o n g populations with respect to scoring methodology and interpretation. El-Kassaby (1991) suggests a sample size of 40-60 for isozymes to accurately measure patterns of genetic diversity w h e r e allele frequencies approach 0.5; the lack of sufficient sample size m a y have obscured s o m e of the relationships in this study although most c o m m o n allele frequencies (p) in the majority of populations were > 0.8. Populations with very high proportions of trees of infected by blister rust often had very w e a k staining or did not produce scorable results, and those loci and populations w e r e d r o p p e d f r o m the analysis. Consequently, there may be s o m e inherent bias in the results of this study in that s o m e genotypes or alleles may have been underrepresented as a result of this. Since it appears likely  59  that blister rust would have s o m e metabolic effect on the e n z y m e expression of a n infected tree or vice versa, trees more susceptible to blister rust may have resulted in s o m e systematic e n z y m e signature or allelic expression (which may or may not have been detectable by electrophoretic technique). O n e possible result of this interaction is the possibility of selection on a n electrophoretically detectable locus. T h e results shown both in this study in Figures 3.7 a n d 3.8, a n d by Stuart-Smith (1998) w h o found an extremely strong correlation between F, a n d white pine s  blister rust, d o lend support to this hypothesis, although isozyme loci are generally considered neutral. It is therefore possible that s o m e of the populations which were heavily infected with blister rust m a y have been less heterozygous, or more monomorphic, or expressed specific alleles which w e r e not detectable during the course of this study. This would account for s o m e of the high heterozygosity values calculated for the other populations which s h o w e d only minimal to moderate blister rust infection.  3.6  CONCLUSIONS  For whitebark pine populations encompassing the species' range throughout B C , observed heterozygosity averaged 0.213, and expected 0.257. Populations and individual loci w e r e highly variable, and the majority (94%) of genetic variability occurred a m o n g individuals. Populations w e r e moderately differentiated ( F  S T  = 0.061); there w a s a statistically significant but w e a k isolation  by distance effect, a n d populations within major mountain ranges were more genetically similar than a m o n g mountain ranges. Observed heterozygosity w a s highest in the south a n d east, but trends w e r e w e a k for expected heterozygosity. Effects of glaciation, founder effects caused by avian dispersal and high levels of inbreeding would all contribute to the observed patterns a n d selection pressures involved in maintaining t h e m .  60  CHAPTER 4 - A CONSERVATION STRATEGY FOR WHITEBARK PINE IN BRITISH COLUMBIA "We're getting environmental Band-aids when we need intensive care." - Mike Harcourt, 1989 4.1  INTRODUCTION  4.1.1 JUSTIFICATION FOR CONSERVATION T h e r e are a s many reasons to conserve or preserve as there are opinions; there are similarly as m a n y a r g u m e n t s to the contrary. Randall (1986) lists several reasons in favour of conservation which m a y apply in this instance: inherent existence value of a species, anthropocentric altruism, a n d the role a species plays within an ecosystem, including generating oxygen a n d providing a carbon sink (Oldfield 1984; Salwasser 1990; Ledig etal. 1998). Other arguments put forward include concepts such as species rights, potential future importance, economic value, the value of the knowledge gained from a species, preserving a "natural order", a reverence for life for its o w n sake, a nature-centured empathy, and a theistic model (Callicott 1986; Lovejoy 1986; Leopold 1933; Ledig 1988). Clearly, all of these share s o m e subjective and even emotional component, making t h e m practically impossible to rank or quantify. In the modern context, decisions are often reduced to a fiscal scale with stakeholders or interest groups presenting their relative rankings to influence conservation decisions (Falk 1991). T h e key flaw in this methodology is that many, or sometimes most, of the values related to conservation arguments have no definable monetary value or price, a n d there is often no w a y to place such a value on abstract or preexisting ecological or subjective functions (Oldfield 1984; Hanemann 1986; Lande 1999). Whitebark pine is a species of little immediate financial value w h e n harvested; it w o u l d be costly to create infrastructure to facilitate harvesting and generally of poor f o r m . T h e real value of this species is evident w h e n whitebark pine trees are left intact within their indigenous ecosystems. Production of wildlife food, aesthetics, soil anchoring, a foundation for high-elevation succession, keystone of subalpine biodiversity, meltwater channelization, insect and fungal habitat a n d  61  microclimate modification at the timberline are all functions which whitebark pine has in its natural setting (Arno a n d Hoff 1990). In British Columbia, f whitebark pine ecosystems are currently under less threat (re: levels of blister rust infection) than in southern Alberta or the Intermountain regions of the United States (Campbell and Antos 2000; Keane etal. 1990; Wilson and Stuart-Smith 2000; T o m b a c k etal. 2001). Most of the land in B.C. suitable for current and future habitat is o w n e d by the C r o w n , i.e., the Province. While a large area of high-elevation ecosystems is already preserved in a contiguous, large network of national and provincial parks, s o m e is also allocated under both shorta n d long-term tenured licences for resource extraction such as logging a n d mining. A tiny c o m p o n e n t is privately o w n e d , primarily for outdoor recreational purposes. Whitebark pine has recently received unprecedented attention from the provincial government d u e to its threatened status (Yanchuk'and Lester 1996; Yanchuk 2001), a n d even the province's chief forester has recognized its precarious predicament (Pederson 1998, op. cit. Kieran 1998). T h e Rocky Mountain National Parks are in the midst of formulating a conservation a n d e c o s y s t e m restoration strategy (Wilson and Stuart-Smith 2000; Rob Walker, Chief Ecologist, Parks C a n a d a , Rocky Mountain National Parks, pers. c o m m . , 2000). Based on knowledge gained f r o m ecological studies (Campbell 1998; Campbell and Antos 2000; Stuart-Smith 1998), provincial protected areas policy (Sawicki 2000) and the current genetic study, a feasible conservation strategy for the province can now be created with a sound scientific basis which adheres to a priori goals such as protecting biodiversity through maintaining all components of fully functioning ecosystems ( B C M o E L P 2000).  4.1.2 CLIMATE CHANGE T h e current buzzwords "global climate change" imply catastrophic changes but actually provide little information. Global climate has historically been in a constant state of flux; the current rate of change, however, has led concerned citizens and scientists to press for more information and  62  action (Jackson and Overpeck 2000). Agencies such as the United States Environmental Protection A g e n c y (USEPA), the United Nations-sanctioned International Panel on Climate C h a n g e (IPCC) a n d m a n y other regional, national and international organizations have been attempting to clarify a n d quantify the nature, scope and scale of this dramatic change. Most of the scientific endeavours to define the magnitude and potential effects of climate c h a n g e on the biosphere and specific regions of it involve General Circulation Models (GCMs), which attempt to model the potential magnitude and impact of global carbon allocation on the biosphere according to several scenarios. While these models are steadily involving more complex variables a n d able to produce more specific and varied effects, they still leave huge g a p s in our knowledge. Such models only provide hypothetical scenarios and many of the critical input variables and feedback mechanisms are still unknown or too complex to model (Shafer et al. 2001). No matter which body or study is consulted, the final projections still emphasize the t r e m e n d o u s uncertainty associated with the estimates (USEPA 2001 b; IPCC 2001 a,b). T h e s e levels of uncertainty are so large that many estimates involving even 8 0 % confidence intervals still e n c o m p a s s both positive and negative temperature change scenarios. Attempting to project the future climate of areas that currently and in future could support whitebark pine ecosystems is even more difficult. Hydrological regimes are difficult to model (including estimates of snow accumulation and melting patterns) a n d complex topography 'throw a w r e n c h ' into the most sophisticated modelling systems. Satellite data has helped ameliorate s o m e of these problems, but very little empirical climate data is available from high elevation sites, w h e r e there is a paucity of recording stations (Prentice et al. 2000). While the impact of climate change is widely expected to be greater in magnitude in the northern latitudes and higher elevations, the nature of the change is poorly understood. Planning for future ecological m a n a g e m e n t of whitebark pine is hence made even more complicated, with a generation time approaching the  63  century mark, and drastic changes expected to be evident within 50 years, thus decisions m a d e now are even more critical (Peters I11986). Most projections for montane and cordilleran regions of the Pacific Northwest estimate a m e a n temperature increase of two to six degrees Celsius within the next century ( U S E P A 2 0 0 0 , 2001 a; W a t s o n et al. 1997; IPCC 2001 a). Mean summer temperatures are expected to remain the s a m e or increase, while s u m m e r drought is likely to increase as snowpack in cordilleran areas melts faster, creating more water stress and increasing the likelihood of associated disease a n d insect susceptibility. This may result in a longer growing season of up to three w e e k s in the subalpine a n d alpine in terms of temperature, but this may be limited by moisture availability a n d increased frequency of extreme events, such as early and late growing season frosts (Easterling etal. 2 0 0 0 ) . Peterson et al. (1990) have already found that within the last 150 years, and especially within the last 3 0 or so, tree radial growth at the timberline of whitebark pine a n d other species has drastically increased, possibly a s a result of a longer growing season due to w a r m e r temperatures. Wildfires are generally predicted to increase in frequency and severity as a result of climate c h a n g e (Perry et al. 1 9 9 1 ; Keane etal. 1999; Keane and and Arno 1993; U S E P A 2000; IPCC 2001 a,b). T h e effects o n precipitation regime vary with the predictive model used, but generally s u m m e r s in mountain habitats will be drier, although winters are expected to be warmer, resulting in d e c r e a s e d mortality of seasonal pathogens and insects from lighter winter kills (Ayres a n d Lombardero 2000; Simberloff 2000). Species' ranges are expected to shift generally northward and upward, causing an increase in the timberline and shifting species ranges north (Watson etal. 1997, IPCC 2001 a,b; Peters I11986; Shafer et al. 2 0 0 1 ; Davis a n d S h a w 2001). T h e rate of climatic change, however, is not necessarily expected to be matched by the availability of suitable substrates or s e e d b e d s . Species m a y therefore not be able to migrate at the rate the climate shifts (Davis and S h a w 2 0 0 1 ; T o m b a c k etal. 2001); moreover, earlier life stages are less resilient to extreme climate events than mature  64  trees. Such off-site populations will be suffering from maladaptation or competition f r o m other species better adapted to the new environmental conditions, especially "weedy" species (Simberloff 2 0 0 0 ) . Most ecological communities as they currently exist will likely not retain their character in the next century (Huntley 1991). Micro- and macrofauna associated with m a n y of the plant species will also experience similar migrational difficulties as their hosts or habitat a n d food sources m a y be maladapted to their current habitat as climatic conditions change (Jackson a n d Overpeck 2 0 0 0 ; Ledig and Kitzmiller 1992). While historical shifts both in species composition and elevation of timberline a n d the historical range pf whitebark pine have been dated using palynology and other techniques (e.g., Luckman and Kearney 1986; Kearney and Luckman 1983), the future remains uncertain. For whitebark pine, it is uncertain whether the Clark's nutcracker will shift its range at a rate consistent with climatic change. Birds have been observed caching seeds in sites outside of whitebark pine's current range (Carsey and T o m b a c k 1994), and the potential for s e e d dispersal to ideal sites is certainly enhanced with a mutualist, long-distance seed disperser. It is difficult to predict whether animal behaviour will shift in a similar manner to other ecological factors (Peters I11986). A s the ecological character of current whitebark pine habitat alters, however, it is likely that other species may be able to outcompete it: lodgepole pine, subalpine fir and Engelmann a n d white spruce (Picea glauca (Moench.) Voss) being the prime candidates in B.C. O n sites w h e r e fires m a y be far more frequent and severe, subalpine larch (Larix lyallii Pari.) m a y also increase its range a n d density. S o m e research has been conducted specifically regarding whitebark pine ecosystems by Keane and Arno (1993) in Montana's Glacier National Park. Besides the startling prediction that there will be no glaciers left in this U.S. national monument in 70 years, the nature a n d location of ecosystems containing whitebark pine w e r e projected to shift upwards and northwards, with severe s u m m e r droughts and consequent increases in catastrophic wildfires accompanying these changes  65  (Keane etal. 1999; W a t s o n etal. 1997; IPCC 2001a). Shafer a n d others (2001) have also suggested that species currently found o n the windward (western) side of North American mountain ranges m a y also shift to the leeward (eastern) side. If fire suppression programs currently in place are discontinued, t h e potential for whitebark pine to survive in these areas m a y be e n h a n c e d a s populations of t h e alternate herbaceous blister rust host would b e periodically diminished in t h e area. If current fire suppression regimes continue, they will drastically increase in e x p e n s e (and likely decrease in success), and whitebark pine ecosystems will rapidly be replaced by more closed-canopy subalpine fir-Engelmann spruce ecosystems. Subalpine m e a d o w s will also b e c o m e more scarce a s they will be subject to constant tree recruitment without moderately frequent fires to maintain t h e position of the timberline (Keane a n d Arno 1993; Keane e t al. 1990; Kendall a n d Keane 2001). Observations substantiating this p h e n o m e n o n have been m a d e over m u c h of t h e southern portion of the range of whitebark pine aleady (Murray et al. 2 0 0 0 ) . Eighty t o one hundred percent mortality is expected for whitebark pine stands in southeast B.C. (Campbell 1998; Campbell a n d Antos 2000; Kendall a n d Keane 2001), although there w e r e no evident correlations with either weather or climate (Campbell 1998; Kendall a n d Keane 2 0 0 1 ) .  4.1.3 GENE CONSERVATION A n effective gene conservation strategy is the key to conserving adaptive variation (Aitken 2000). T h e potential for evolution is quantified in large part by genetic diversity, which is roughly analogous to intra- a n d interpopulation measures of heterozygosity. Since trees are sessile a n d can only migrate during the seed dispersal stage (although portions of t h e g e n o m e m a y also be dispersed during pollination, this does not provide a n opportunity for range expansion), t h e long generation time of whitebark pine makes the option of in situ adaptation t o climate c h a n g e a critical c o m p o n e n t of a n y conservation strategy for the species. By targeting highly diverse populations a s well a s genetically unique ones for special status or active management, t h e most efficient use of resources c a n be achieved.  66  Many unique alleles are only detectable in DNA sequences, or D N A markers b a s e d o n these sequences, since many mutations are functionally neutral or serve to inactivate a protein. Many mutations also will only appear at the most fundamental level of D N A since several different s e q u e n c e s m a y produce the s a m e RNA or protein based on genetic redundancy of codons. Many mutations are either detrimental to, or have no effect o n , an individual's fitness in the current climate or in certain environments and are found at infinitesmally low frequencies (Holsinger et al 1 9 9 1 ; Caughley 1994). S o m e researchers have suggested that if resources or time are limiting, it m a y be most prudent to capture the majority of c o m m o n alleles instead, based o n the distribution of rare alleles and the diminishing returns inherent in expending resources to capture t h e m (Falk 1 9 9 1 ; Brown and Briggs 1991), especially given the neutral (i.e., non-adaptive) nature of allozyme variation (El-Kassaby -1982; Karhu et al. 1996). Determining the genetic composition of individuals can be done at m a n y levels of resolution (Dekker-Robertson et al. 2 0 0 1 ; Bruederle 1998); however, the finer the resolution, the more expensive the technique. Heterozygosity and fitness are not always linked (Frankham 1995), although heterozygosity may provide s o m e genetic insurance against future environmental c h a n g e (Huennke 1 9 9 1 ; V i d a 1994). T h e most cost-effective technique would be to visually survey individuals and populations and gauge the myriad complex relationships between a n d a m o n g genes, individuals and the environment by the physical and phenological adaptations they display in their native habitat. This will not provide much specific information about the actual genetic composition, and none at all about genes or alleles which are not expressed under those conditions. With respect to the critical conservation issue of locating individuals which m a y be resistant or highly tolerant of white pine blister rust, however, visual assessments w o u l d actually provide such genetic information as well as capturing the most important genetic adaptations for the present as well as the future (Ledig 1988).  67  While ideally every individual a n d genotype of every species could be conserved in order to preserve the optimal adaptive potential for future generations, this is impossible a n d unnecessary in reality. A target effective population size (N ) can be used as a surrogate for the proportion of e  genetic variability captured in a population, a n d estimates can be approximated f r o m heterozygosity statistics in the ratio of roughly H being proportional t o the inverse of 2 N . Precise e  e  calculations of N are difficult in any case for a hermaphroditic organism with overlapping e  generations, a mixed mating system, unknown exact population size, variable risk a n d susceptibility to mortality in different areas, long-distance multi-directional vector-mediated seed dispersal, w i n d pollination, a n d subdivided population structure (Cain etal. 2 0 0 0 ; Nunney 1999; N o m u r a 1999; Hartl a n d Clark 1994). A n exact estimate would involve extremely complicated models a n d formulae which are still being developed a n d tested (Cain etal. 2 0 0 0 ; C a s e a n d Taper 2000). T h e stepping-stone population structure of whitebark pine results in relatedness a m o n g nearby populations a n d would result in spatially differentiated requirements for N , depending o n e  the relatedness of adjacent populations. Yanchuk (2001), following recent suggestions that a traditionally suggested size of 500 m a y not be sufficient (e.g., Lynch 1996) has suggested a target size of 1000 for N , roughly corresponding to a census adult population size of 5 0 0 0 for B.C. e  conifer species. T h e inbreeding found in this study m a y necessitate a slightly higher c e n s u s . population, since Yanchuk w a s assuming an inbreeding coefficient of < 0 . 1 , so a census population o n t h e order of 6000 m a y be more appropriate for in situ conservation purposes. A better actual number would consider the frequency of any disease resistance genes or genotypes in each conservation area, since a key goal of in situ conservation for whitebark pine w o u l d be to capture a threshhold percentage (one to five percent has been suggested by Hoff et al. 2 0 0 1 ) of those genes. While the results f r o m Chapter 2 indicated that whitebark pine exhibits substantial inbreeding relative t o other pines, although within a range typical for stone pines, results f r o m Chapter 3  68  revealed a fairly high level of heterozygosity. Most of the genetic variation w a s found a m o n g individuals within populations, and there were geographical trends with both heterozygosity and heterozygote deficiency. Based on these observations, effective population sizes for whitebark pine conservation m a y not be strongly impacted by current levels of inbreeding. This m a y be d u e in part to the high interpopulation gene flow, thus a conservation strategy should incorporate more than o n e population f r o m each general area in order to maintain gene flow and minimize potential inbreeding. Although heterozygosity w a s lowest in the north, it m a y not b e necessary t o c o m p e n s a t e for the lower N in this region by conserving a greater number of individuals relative to e  the south since the populations in the south are at a far greater risk of immediate mortality f r o m white pine blister rust. A s will be explained in the subsequent section, conservation efforts in the south should not b e g e a r e d towards a gross number a s m a y be appropriate in the north; the pressing problem of blister rust engenders a conservation strategy targeted towards specific individuals or families, and will require more aggressive management. Since future environmental conditions are not known, the best strategy may be to conserve a s w i d e a s p e c t r u m a s possible of genotypes (Frankham 1995; Ledig 1986,1988; Ledig a n d Kitzmiller 1992). T h e most cost-effective conservation strategy would undoubtedly be to provide for  in situ  reserves containing as many individuals as possible which would allow for migration to habitat suitable in the future (Peters I11986; Aitken 2000; Ledig 1986). Making comprehensive  ex situ  g e r m p l a s m collections, seed orchards or seed collections would likely be too expensive due to the extremely long period of time required for propogule production and costs of seed collection, processing and storage. S o m e ancillary e x  situ or inter situ collections may be instituted as a  complementary measure for areas where the creation of natural reserve areas is either not feasible or the local populations are under immediate threat of extirpation (Yanchuk 2001). T h e most effective option would probably be cryogenic seed or germplasm storage in those c a s e s , although the effects of long-term seed storage on germplasm viability remain largely u n k n o w n for this  69  species a n d its seeds often deteriorate rapidly in conventional types of storage (Ledig 1988; Bonner 1990; D. Kolotelo 2000, B C M o F Seed a n d Cone Officer, pers. c o m m . ) .  4.1.4 WHITE PINE BLISTER RUST In terms of species conservation, it is probably most effective to target those populations most at risk for special conservation efforts, thereby providing trickle-down benefits t o other ecosystem c o m p o n e n t s ( N a m k o o n g 1992). While B.C. is fortunate to have fairly large, contiguous regions occupied by whitebark pine ecosystems, populations in the south appear to be especially susceptible to blister rust a n d mountian pine beetle attacks. Intensive surveys currently being conducted by the B.C. Ministry of Forests (Zeglen 1999, 2000) are attempting to locate individuals throughout the province which appear to be resistant or tolerant of blister rust. Identified trees will be subject t o further screening in the future; however, it is critical to institute screening programs c o m b i n e d with site hazard assessment in order to correctly assess the resistance of the stock a n d determine suitable outplanting sites (Kendall a n d Keane 2 0 0 1 ; McDonald a n d Hoff 2001). Other five-needle pines affected by this disease have been extensively studied. A single-gene m e c h a n i s m for resistance has been found in both western white pine (Pinus monticola Dougl. ex D. Don) a n d sugar pine (Pinus lambertiana Dougl.) (Devey etal. 1995), but different genes for each species. Adaptive traits such as the multigenic bark reaction resistance to white pine blister rust are likely to be more robust in the long term than a single-gene trait. T h e complexities of a hostpathogen system are also not always fully understood, a n d recently virulent strains of the blister rust have been found in areas where there w a s strong selection d u e to a high frequency of the single-gene resistance, which the pathogen can now overcome in both sugar a n d limber pine in s o m e regions. T h e stability a n d efficacy of the multigene tolerance or bark reaction m e c h a n i s m has a trade-off in that there will likely be s o m e growth impact on the tree caused by the disease (Hoff 1984; Kojwang 1994). These contradictory effects on fitness conferred by t h e multigenic disease resistance mechanism m a y be an artifact of the variation inherent in the complex of genes  70  acting to cause this reaction. Should a screening process for a naturally-occurring resistance gene be developed, it would be ideal for whitebark pine conservation since individuals can b e screened in a non-destructive manner to identify resistant genotypes.  4.2  WHITEBARK  PINE CONSERVATON  STRATEGY  FOR BRITISH  COLUMBIA  4.2.1 INTRODUCTION While a conservation strategy for such a widespread species which is a c o m p o n e n t of so m a n y different ecological communities cannot focus around a single species or geographical area (Lovejoy 1986), this plan is intended as one link within an integrated framework including the Rocky Mountain National Parks Strategy (2000) and initiatives undertaken within the United States Intermountain region (Kendall 1994). A plan for conserving whitebark pine, while nominally centred a r o u n d a single species, also ensures the health of a large number of different species a n d taxonomic groups given its keystone role. Conservation of the keystone species can have ancillary effects to conserve the nature, structure and functioning components of the representative e c o s y s t e m (Tomback et al. 2001). Given the potential implications for global climatic change, it is probably best to institute a conservation strategy which provides insurance against the widest possible range of future scenarios ( H a n e m a n n 1986). Despite the enormous uncertainty around the nature and scale of the effects of these predicted changes, there are still a few parameters which remain relatively certain: that current climatic zones will shift upward in elevation and northward in latitude. However, the magnitude of changes in hydrological regimes and differential rates of migration, extinction of different species and ecotypes, and even opportunities for speciation, remain too uncertain to predict.  4.2.2 SHORT-TERM GOALS O n e step which has been nearly completed in B.C. is the intensive surveys of whitebark pine populations collecting data o n growth and blister rust (Zeglen 1999, 2000). A key goal of these 71  surveys is to identify putatively resistant individuals for future management. In the U.S., a system of blister rust hazard ratings for the range of whitebark pine has also been instituted based on ecological a n d biogeographical factors (Hoff et al. 2001). Extending a n d adapting this system to B.C. w o u l d facilitate blister rust management as well as improving whitebark pine outplanting success for any rehabilitation efforts. Another facet of active management could include inter- a n d in situ measures to c o m p l e m e n t the ex situ program (Brown a n d Briggs 1 9 9 1 ; Ledig 1986; Millar a n d Westfall 1992). This could include s e e d collection, germination a n d outplanting o n suitable sites, or sites expected to b e suitable within 50-100 years (Ledig a n d Kitzmiller 1992). This option is extremely labour-intensive and expensive. In years with cone crops, branches with cones in suitable stands must be c a g e d in the early to mid-summer, as soon as access permits, in order to protect seeds f r o m predation. T h e s e stands must be revisited in the autumn for cone collections; this step m a y be more or less complicated depending on whether individual genotypes or mother trees are to be identified or not. T h e c o n e s must then be transported to a processing facility. Specialized procedures must be followed t o extract a n d identify filled seeds, stratify a n d germinate t h e m , a n d then return t h e m t o appropriate locations for outplanting. Whitebark pine has been shown to be highly susceptible to a large variety of pathogenic seed a n d cone storage fungi (Vujanovic et al. 2001). While ideally the seedlings w o u l d be returned to their original provenances, s o m e seed transfer m a y b e permitted, depending o n the nature of the conservation and land use decision, climate c h a n g e projections, or data f r o m c o m m o n garden experiments. O n e trade-off to take into account in this case is the removal of potential sources of natural regeneration as well as wildlife food; many times the number of seeds must actually be taken compared to the number of seedlings required for t h e experiments d u e to the typically very low germination rate. Virtually nothing is currently known about the extent a n d nature of adaptive variation in whitebark pine in B.C. O n c e seed has been procured, c o m m o n garden experiments should be  72  established to assess important adaptive traits such as budburst date, early growth rates a n d biomass allocation. Differences a m o n g populations and regions could then be used to define appropriate seed a n d scion transfer guidelines. Seedlings grown for these tests could be utilized for further data collection, including screening for white pine blister rust resistance in highly controlled conditions. Co-operation a m o n g regions w h e r e whitebark pine grows would benefit all stakeholders: jurisdictions could share technology and facilities, and develop joint solutions while reducing duplication. Active, ecosystem-based management may also play a role in this conservation strategy (Cole a n d Landres 1996; Hoff et al. 2001). O n e facet of this currently being explored in the Rocky Mountain National Parks and in the U.S. Intermountain region is the establishment of controlled burns in whitebark pine ecosystems. A century of fire suppression has altered these ecosystems to the degree that their fundamental characteristics have c h a n g e d dramatically. T h e r e is an increased risk of high-severity fires as fuel loads are not periodically burned out by light surface fires; the advanced age of many trees also harbours increased susceptibility to attack by mountain pine beetle (Dendroctonus  pondorosae  in the understory is favourable for Cronartium  Hopkins) and increased occurrence of Ribes spp.  nbicola.  T h e lack of surface fires has also allowed  for succession and soil development on the talus and thin forest floor, eliminating potential s e e d b e d s for pine and facilitating growth of competing species such as Engelmann spruce engelmannii  Parry ex Engelm.), mountain hemlock (Tsuga medensiana  subalpine fir (Abies lasiocarpa  (Picea  (Bong.) Carr.) a n d  (Hook.) Nutt.). Periodic moderate to light surface fires also  eliminate Ribes, the alternate host of the blister rust pathogen, and may moderate local spore densities (Keane etal. 1993). A regime of controlled burns and follow-up monitoring should be instituted in selected areas in B.C. to restore whitebark habitat. Lightning-caused wildfires in remote areas w h e r e h u m a n habitation a n d activity would not be at risk should not be suppressed (Wilson and Stuart-Smith  73  2000). Wildfires of small size that could be controlled at areas that interface with h u m a n use or habitation should be allowed to burn, if possible. Nutcrackers have been frequently observed caching pine seeds in recently burned areas, which provide ideal competition-free mineral soil seedbeds for germination (Tomback 1986). Providing the largest possible suitable areas for seedling establishment would help maintain more genotypes as a larger number of c a c h e d seeds w o u l d germinate (Tomback et al. 2001). In cases where a resistant genotype is f o u n d , it m a y be desirable to limit fires to low intensities in those areas, or to suppress local fires until t h e cones have been collected.  4.2.3 MEDIUM-TERM GOALS In the case w h e r e a unique genotype has been located, such as a rust-resistant tree, it m a y be desirable to also attempt vegetative propogation or tissue culture, a n d it is definitely worth the extra effort of intensive surveys a n d repeated site visits to ensure a supply of s e e d from these trees (Hoff et al. 2 0 0 1 ) . O n e possibility is the potential for grafting selected individuals onto western white pine rootstock, also in the taxonomic section Strobi, in order to reduce operational time a n d increase vigour a n d growth rates (Arno a n d Hoff 1989). Planting seeds or seedlings outside their native provenances may be desirable if short-term extirpation of the species in a region is a more pressing threat than maintaining t h e indigenous genotypes (Conway 1986). This could also be useful if a consensus has been reached to facilitate species migration through planting northwards or upwards into areas it d o e s not occur at present, a s an ameliorating measure against the impacts of climate change (Peters I11986). In light of the observed a n d anticipated rates of climatic change, I believe that planting seedlings in a steppingstone pattern to the north of their current provenances by up to several hundred kilometres in s o m e cases w o u l d be acceptable a n d that survival would considerably outpace mortality. Long-term reconstructions of historical climate a n d pollen data show a more contiguous distribution of highelevation species during w a r m periods, including whitebark pine (Prentice etal. 2 0 0 0 ) . 74  Nutcrackers also do cache seeds outside of the current species' range, implying that a northward shift facilitated by planting could encourage the birds to expand their range northwards. Once seed transfer zones have been established, appropriate selections could be made during local mast years, assuming the climate will change at the rates and magnitude outlined by various projections (IPCC 2001a; USEPA 2000; Shafer etal. 2001). Trees growing in high-snowpack areas typically survive and regenerate best in clumps, where a microclimate effect causes earlier snowmelt and shallower snowpacks around clumps (Klinka and Chourmouzis 2001). Survival will probably be best if seedlings are planted in clumps adjacent to stumps or coarse woody debris on south-facing, well-drained regosols or calcareous sites where mature potential trees which may be potential competitors are scarce or absent. The possibility that planted seeds may be eaten before they germinate makes the extra effort and expense of planting seedlings worth while. 4.2.4 L O N G - T E R M G O A L S  Once suitable individuals have been located, it may be possible to establish and maintain a few small orchards. It is possible to maintain a representative sample of at least 90% of the most common alleles, i.e., the majority of the genetic potential of a population, with a collection of as few as 10-50 individuals, provided they are collected from a wide range of habitats (Brown and Briggs 1991), especially given the relatively high heterozygosity and fairly low F found in this study. One ST  seed orchard could be established per transfer zone in order to keep the genetic character of each regional population. Priority should of course be given to individuals putatively resistant or tolerant to white pine blister rust. Collecting scion material from these individuals and using controlled pollination would also ensure that the resistance or tolerance characteristics (the most desirable features to conserve) of the parent trees would be expressed. Establishing a series of range-wide seed orchards for whitebark pine should be a long-term objective due to the enormous expense of collecting and maintaining such an orchard, since it is not a commercially valuable species (Brussard 1990). While only a small number of individuals 75  w o u l d b e needed to provide genetic resources, the effort required to produce sufficient suitable adults will be very costly a n d take years. A suitable location for such a collection w o u l d also be difficult to find since existing seed orchard complexes in British Columbia are currently located in climates which are likely not suitable for the growth a n d reproduction of whitebark pine. Seedlings have been successfully grown in nurseries; however, it is not certain that the trees would survive t o reproductive a g e or receive their chilling requirements in any habitat type but the rather extreme o n e s they now occupy. While public sentiments in B.C. currently run counter t o genetic engineering of forest trees, the potential to mitigate the impact of white pine blister rust a n d maintain the health of whitebark pine ecosystems m a y be greatly aided by such technology. They are included in this "long-term goals" section since the technology m a y take many years of costly a n d intensive research to be successful a n d active genetic manipulation of noncommercial forest species is a low priority at this time c o m p a r e d to in situ techniques a n d more conventional strategies. In the event that these techniques do b e c o m e feasible, s o m e of the applications are listed here. Somatic embryogenesis and tissue culture could rapidly reproduce tolerant or resistant individuals for outplanting in t w o to three years, instead of waiting dozens of years for seed to grow a n d reach sexual maturity. In t h e event that a similar single-gene resistance mechanism is isolated for whitebark pine to the pathogen as w a s found for sugar and western white pines, it m a y be temporarily beneficial until a breeding strategy for producing resistant seedlings is operational. While the comparatively rapid mutation rate of the pathogen m a y respond to the selection regime imposed by introducing this gene, it could be carefully applied depending on the hazard rating of the outplanting a r e a . In the long t e r m , a multigenic form of either tolerance or resistance would be more beneficial since it w o u l d be m u c h more difficult for the pathogen to develop a virulent mutant in such a case (Hoff et  al. 2001).  76  4.2.5 REGIONAL CONSERVATION PRIORITIES W h e n selecting appropriate populations to target for conservation, heterozygosity w o u l d be the simplest criterion, aside from blister rust resistance. In terms of heterozygosity, Perkins, D'arcy, Whistler a n d Lime were particularly high. Paget and Stanley, although they exhibited lower heterozygosity, had slightly higher mean numbers of alleles per locus (there w a s no real statistical difference for this measure), and more polymorphic loci. Perkins w a s high in both heterozygosity a n d allelic richness. For a northwestern seed transfer or conservation quadrant, this population may serve as a genetic reservoir. For the northeast, Stanley may be a suitable population, although the trees in this particular population were on average quite y o u n g . Conservation activities should involve both Paget and Stanley, the easiest populations to sample in t e r m s of accessibility, or other populations in this area. In Paget there are many trees of reproductive a g e but they are suffering heavily from blister rust. This would also tie in with the Rocky Mountain National Parks conservation strategy (Wilson and Stuart-Smith  2000). In the southwest, Whistler  a n d Lime w o u l d be suitable candidates with abundant individuals, cone-bearing adults and regeneration, as well as relatively low blister rust impact. In the southeast, I w o u l d not suggest a single population should serve as a source population for conservation efforts. Due to the extreme severity of blister rust in this region, selections should be m a d e on an individual-tree basis following comprehensive screening, and resistant genotypes, if available, could even be imported f r o m the bordering states of Idaho or Montana which share similar climate and topography. T h e symbiosis between whitebark pine and Clark's nutcracker is a necessary prerequisite to the shift of the species' range northwards, the bird being the key agent of seed dispersal. While the nutcracker's range can extend beyond that of the tree, it is unknown how dramatically it will shift with a c h a n g e in climate, and it is also uncertain whether suitable habitat will occur if c a c h e locations shift northward. Assuming that the bird's behaviour will remain consistent w h e n looking for cache locations, optimal germination sites should continue to be a large proportion of bird seed  77  caches. However, that is no guarantee that this will remain the case, or that other species which are c o m p o n e n t s in current whitebark pine ecosystems will accompany this migration (Fisher a n d Myres 1980; T o m b a c k 2001). Propogule and gamete dispersal by water and w i n d , as most of the rest of the high-elevation species such as mosses, flowers, shrubs a n d lichens exhibit, will likely result in a far slower migration rate upwards and northwards. Despite t h e shorter generation times of these other organisms, it is unlikely that subalpine and timberline communities, as w e know t h e m , will remain intact as climate shifts, as historical evidence supports (Davis a n d S h a w 2 0 0 1 ) . Introduced and fast-growing, or weedy, species may have an advantage as conditions rapidly c h a n g e due to short generation times and generally high fecundity (Simberloff 2 0 0 0 ) . Most species in the g e n u s Ribes, which are the secondary host of the blister rust pathogen, could also easily expand their range or increase in abundance, providing a vector for the disease w h e r e it m a y currently be absent. T h e fungus itself could also evolve and spread in a similar manner (Ayres a n d Lombardero 2000). T h a t said, the most effective insurance policy for high-elevation ecosystems is to set aside a contiguous land base as large as possible, representing as many biogeoclimatic regions a s possible, that would support the migration of these species and minimize the effects of fragmentation (Yanchuk 2 0 0 1 ; Millar and Westfall 1992; Lester and Y a n c h u k 1996). In B.C., this is still a real possibility, especially at the northern species boundary. In California, W y o m i n g a n d Montana, a mosaic of land tenures and uses such as roads, resource extraction, tourist facilities and development have led to a highly fragmented range over much of the mountainous topography. T h e largely undeveloped, unpopulated and unroaded mountains of B.C. provide the opportunity for this critical first step in conservation. Currently, the status of whitebark pine conservation in most of B.C. with respect to future climate scenarios is quite good. Despite the spread of blister rust, large tracts of suitable habitat remain unthreatened by human impact. Vast areas of contiguous wilderness are already protected  78 •  in parks; although s o m e regions do not provide adequate protection yet, they are generally remote, unroaded a n d inaccessible, providing potential habitat for future range shifts a n d a rich spectrum of different habitats. Many of these areas are allowed t o burn if wildfires occur a n d provide ideal germination sites for whitebark pine. T h e species has been designated a high conservation priority, a n d s o m e resources have been allocated to gather baseline information a n d develop a n d implement strategies. T h e highest priority should be allocated to identifying rust-resistant individuals. This work has already been initiated. Conducting site hazard assessments for white pine blister rust a n d collecting s e e d f r o m as many trees as possible both from these selected individuals a s well a s targeted populations within regions should be the next most pressing items. Conducting c o m m o n garden experiments in order'to determine adaptive variation a n d delineate seed transfer zones, followed by outplanting of two- to three-year-old seedlings should be conducted, along with periodic monitoring of both the c o m m o n garden experiments a n d the outplanting test sites. All other activities should be assigned lower priorities, a n d continue if resources allow. O n g o i n g m a n a g e m e n t including letting wildfires burn in whitebark pine habitat, should be allowed to continue, a n d to be applied throughout the species' range in B.C., w h e r e no h u m a n livelihoods are threatened. Although blister rust is likely to decimate whitebark pine across its northern range in t h e future, it is anticipated that enough resistant or tolerant genotypes will remain to ensure its survival in t h e north, although at severely reduced numbers (Hoff etal. 2001). Estimates of natural resistance or tolerance rates between o n e a n d five percent have been given based o n observations of t h e effects of blister rust on natural populations (Lanner 1996; Hoff etal. 2001). T h e magnitude of this impact will d e p e n d largely upon human intervention at the present time; given t h e incredibly rapid spread of t h e fungal pathogen, there is no time to lose if our efforts are to m a k e a positive difference (Hoff et al. 2001).  79  4.3  CONCLUSIONS Due to the effects of fire suppression, introduced fungal disease a n d mountain pine beetle,  whitebark pine ecosystems in B.C. are under threat of drastically reduced population density and e v e n complete extirpation in many southern populations within one generation. It is critical that active range-wide conservation measures be taken as soon as possible. T h e s e m e a s u r e s include: (1) screening natural populations for individuals which may be resistant or tolerant to white pine blister rust, (2) collecting seed from accessible targeted populations and supplementing natural regeneration in suitable sites by planting seedlings, except in the southern Rockies w h e r e targeting disease-resistant individuals should be paramount, (3) developing appropriate s e e d transfer zones within B.C. and across jurisdictions to facilitate the transfer of the most suitable materials by implementing c o m m o n garden experiments to quantify adaptive variation of s o m e important traits, (4) returning wildfires in high-elevation ecosystems to the historical frequencies, a n d (5) collecting s e e d a n d scion materials f r o m putatively resistant individuals and establishing  ex situ breeding  populations in suitable areas. Genetic engineering may have s o m e long-term potential for beneficial effects, but should be considered a lower priority unless no natural resistance to blister rust exists. Sharing research and resources with other jurisdictions included in whitebark pine's range is critical. Extending the range of whitebark pine to the north in anticipation of accelerated rates of climate change may ensure a seed supply for natural regeneration and adaptation in the next century.  80  LITERATURE CITED Achuff, P.L. 1989. Old-growth forests of the Canadian Rocky Mountain national parks. Nat. Areas J . 9(1): 12-25. A d a m s , W . T . 1992. G e n e dispersal within forest tree populations. New For. 6: 2 1 7 - 2 4 0 . A d o r a d a , D.L., C L . Biles, C M . Liddell, S. Fernandez-Pavia, K.O. W a u g h and M.E. W a u g h . 2 0 0 0 . Disease development and enhanced susceptibility of w o u n d e d pepper roots to Phytophthora capsici. Plant Path. 49(6): 719-926. Aitken, S.N. 2000. Plenary address: conserving adaptive variation in forest ecosystems. J . Sust. For. 10(1/2): 1-12. A m o s , W . T w o problems with the measurement of genetic diversity and genetic distance. Pp. 75100 in Genetics a n d the extinction of species. L.F. Landweber and A . P . D o b s o n , eds. Princeton Univ. Press, Princeton, NJ. Arno, S.F. and R.J. Hoff. 1989. Silvics of whitebark pine (Pinus albicaulis). Rep. INT-253. O g d e n , Ut. 11 pp.  USDAFS Gen. Tech.  Arno, S.F. and R.J. Hoff. 1990. Pinus albicaulis Engelm. Pp. 268-279 in Silvics of North A m e r i c a , V o l . 1 . Conifers. Burns, R.M. and B.H. Honkala, technical co-ordinators. Agriculture Handbook 654, U S D A For. Serv. Washington D . C Axelrod, D.l. 1986. Cenozoic history of s o m e western American pines. A n n . Missouri Bot. G a r d . 73: 5 6 5 - 6 4 1 . Ayres, M.P. a n d M.J. Lombardero. 2000. Assessing the consequences of global c h a n g e for forest disturbance f r o m herbivores and pathogens. Sci. Total Env. 262: 263-286. Baker, B.H., G.D. A m m a n and G.C. Trostle. 1 9 7 1 . Does the mountain pine beetle change hosts in mixed lodgepole and whitebark pine stands? U S D A F S Res. Note I N T - 1 5 1 . O g d e n , UT. 6 pp. Baker, R.G. 1990. Late quaternary history of whitebark pine in the Rocky Mountains. Pp. 40-48 in Proc. s y m p o s i u m on whitebark pine ecosystems: ecology a n d management of a high-mountain resource. B o z e m a n , MT., USA. March 2 9 - 3 1 , 1 9 8 9 . Beaulieu, J . and J.-P. Simon. 1994. Inheritance and linkage relationships of allozymes in Pinus strobusL. Silv. Genet. 43(4): 2 5 3 - 2 6 1 . Berg, E.E. a n d J.L. Hamrick. 1997. Quantification of genetic diversity at allozyme loci. C a n . J . For. Res. 2 7 : 415-424. Bonner, F T . 1990. Storage of seeds: potential and limitations for germplasm conservation. For. Ecol. Mgmt. 3 5 : 35-43. Boyle, T.J.B. 1992. Biodiversity of Canadian forests: current status and future challenges. For. C h r o n . 68(4): 444-452. Bradshaw, A . D . and T. McNeilly. 1 9 9 1 . Evolutionary response to global climatic c h a n g e . A n n . Bot. 67(Supplement 1): 5-14. British Columbia Ministry of Environment, Lands and Parks. 2000. Guiding principles for conservation: B C Parks' commitment to conservation.  81  Brown, A . H . D . and J.D. Briggs. 1 9 9 1 . Sampling strategies for genetic variation in ex situ collections of endangered plant species. Pp. 99-119 in Genetics a n d conservation of rare plants. D.A. F a l k a n d K.E. Holsinger, Eds. Oxford Univ. Press, N.Y., NY. Bruederle, L.P., D.F. Tomback, K.K. Kelly and R.C. Hardwick. 1998. Population genetic structure in a bird-dispersed pine, Pinus albicaulis (Pinaceae). C a n . J . Bot. 7 6 : 83-90. Brussard, P.F. 1990. The role of genetic diversity in whitebark pine conservation. Pp. 315-318 in Proc. s y m p o s i u m on whitebark pine ecosystems: ecology and m a n a g e m e n t of a high-mountain resource. B o z e m a n , MT., USA. March 2 9 - 3 1 , 1989. Bucci, G., M. Anzidei, A. Madaghiele and G.G. Vendramin. 1998. Detection of haplotypic variation and natural hybridization in haplensis-comp\ex pine species using chloroplast simple s e q u e n c e repeat (SSR) markers. Moi. Ecol. 7: 1633-1643. Bush, R.M. a n d P.E. S m o u s e . 1992. Evidence for the adaptive significance of allozymes in forest trees. New For. 6: 179-196. Bush, R.M., P.E. S m o u s e and F T . Ledig. 1987, T h e fitness consequences of multiple-locus heterozygosity and growth rate in pitch pine (Pinus rigida Mill.). Evol. 41 (4): 787-798. Cain, M.L., B.G. Milligan and A . E . Strand. 2000. Long-distance seed dispersal in plant populations. A m . J . Bot. 87(9): 1217-1227. Callaway, R.M. 1998. Competition and facilitation on elevation gradients in subalpine forests of the northern Rocky Mountains, USA. Oikos 82: 561-573. Callicott, J.B. 1986. O n the intrinsic value of nonhuman species. Pp. 110-137 in Preservation of species: the value of biological diversity. B.G. Norton, Ed. Princeton Univ. Press, Princeton, N.J. 305 pp. C a m p b e l l , E.M. 1998. Whitebark pine forests in British Columbia: composition, dynamics a n d the effects of blister rust. M.Sc. Thesis. Univ. of Victoria, B.C . 136 pp. Campbell, E.M. and J.A. Antos. 2000. Distribution and severity of white pine blister rust and mountain pine beetle on whitebark pine in British Columbia. C a n . J . For. Res. 3 0 : 1051-1059. Carsey, K S . a n d D.F. Tomback. 1994. Growth form distribution and genetic relationships in tree clusters of Pinus flexilis, a bird-dispersed pine. Oecologia 98: 4 0 2 - 4 1 1 . C a s e , T . J . and M.L. Taper. 2000. Interspecific competition, environmental gradients, gene flow, and the coevolution of species' borders. A m . Nat. 155(5): 584-605. Caughley, G. 1994. Directions in conservation biology. J . A n i m . Ecol. 34: 215-144. Cavalli-Sforza, J.L. a n d D.W. Edwards. 1967. Phylogenetic analysis: models and estimation procedures. Evol. 21): 550-570. Clayton J.B. a n d E.M. Tretiak. 1972. Differences in allozyme variation in Atlantic herring. J . Fish Res. Board C a n . 35: 1167-1170. Clegg, M.T. 1980. Measuring plant mating systems. Bioscience 30(12): 814-818. Cole, D.N. 1990. Recreation in whitebark pine ecosystems: d e m a n d , problems, a n d m a n a g e m e n t strategies. Pp. 305-309 in Proc. symposium on whitebark pine ecosystems: ecology a n d m a n a g e m e n t of a high-mountain resource. Bozeman, MT., USA. March 2 9 - 3 1 , 1 9 8 9 .  82  Cole, D.N. and P.B. Landres. 1996. Threats to wilderness ecosystems: impacts a n d research needs. Ecol. Appl. 6(1): 168-184. C o n w a y , W . 1986. Can technology aid species preservation? Pp. 263-268 in Biodiversity. Wilson, E d . National A c a d . Press, W a s h . D.C.  E.O.  Critchfield, W . B . 1986. Hybridization and classification of the white pines (Pinus section Strobus). T a x o n 35(4): 647-656. C r u z a n , M.B. 1998. Genetic markers in plant evolutionary ecology. Ecol. 79(2): 4 0 0 - 4 1 2 . Davis, M.B. and R.G. Shaw. 2 0 0 1 . Range shifts and adaptive responses to quaternary climate c h a n g e . Science 292: 673-679. Day, R.J. 1967. Whitebark pine in the Rocky Mountains of Alberta. For. C h r o n . 43(3): 2 7 8 - 2 8 2 . Dekker-Robertson, D., L.P. Bruederle and G.E. Howe. 2 0 0 1 . Implications of genetic structure. Pp. 96-110 in Whitebark pine communities: ecology and restoration. D.F. T o m b a c k , S.F. Arno a n d R.E. Keane, eds. Island Press, Washington D.C. Delcourt, P.A. a n d H.R. Delcourt. 1998. Paleoecological insights on conservation of biodiversity: a focus o n species, ecosystems and landscapes. Ecol. Appl. 8(4): 921-934. Devey, M.E., A. Delfino-Mix, B.B. Kinloch, Jr. and D.B. Neale. 1995. R a n d o m amplified polymorphic DNA markers tightly linked to a gene for resistance to white pine blister rust in sugar pine. P N A S . 92: 2066-2070. Douglas, G.W. and L.C. Bliss. 1977. Alpine and high subalpine plant communities of the north C a s c a d e s range, Washington and British Columbia. Ecol. M o n o g . 4 7 : 113-150. Dukharev, V.A. 1978. Composition of the esterases in Pinus sylvestris in uneven-aged stands. Lesovedenie. 6: 45-50 (Abstr.) op. cit. Forestry Abstracts 4 0 : 4 5 6 0 . Easterling, D.R., G.A. Meehl, C. Parmesan, S.A. C h a n g m a n , T.R. Karl a n d L.O. Mearns. 2 0 0 0 . Climate estremes: observations, modelling, and impacts. Science 2 8 9 : 2 0 6 8 - 2 0 7 4 . Echt, C.S., L.L. DeVerno, M. Anzidei and G.G. Vendramin. 1998. Chloroplast microsatellites reveal population genetic diversity in red pine, Pinus resinosa Ait. Mol. Ecol. 7: 307-316. Ecological Reserves Program, Planning and Conservation Services. 1993. Guide to ecological reserves in British Columbia. March 1993. B.C. Parks, Victoria, B.C. 12 pp. El-Kassaby, Y.A. 1982. Associations between allozyme genotypes and quantitative traits in Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco). Genetics 1 0 1 : 103-115. El-Kassaby, Y.A. 1984. Variation in the outcrossing rate a m o n g crown strata of Sitka spruce. Silv. Genet. 32(2-3): 232-237. El-Kassaby, Y.A. 1 9 9 1 . Genetic variation within and a m o n g conifer populations: review a n d evaluation of methods. Pp. 61-76 in Biochemical markers in the population genetics of forest trees. S. Fineschi, M.E. Malvolti, F. Cannata and H.H. Hattemer, eds. S B P A c a d . Pub. bv, T h e Hague, Netherlands. El-Kassaby, Y.A. and B. Jaquish. 1996. Population density and mating pattern in western larch. J . Hered. 82: 438-443. El-Kassaby, Y.A., M.D. Meagher and R. Davidson. 1993. Temporal variation in the outcrossing rate in a natural stand of western white pine. Silv. G e n . 42(2-3): 131-135. 83  El-Kassaby, Y.A., M.D. Meagher, J . Parkinson and F T . Portlock. 1987. Allozyme inheritance, heterozygosity and outcrossing rate a m o n g Pinus monticola near Ladysmith, British C o l u m b i a . Hered. 58: 173-181. Ellstrand, N.C. 1992. G e n e flow a m o n g seed plant populations. New For. 6: 2 4 1 - 2 5 6 . Erikkson, G., G. N a m k o o n g and J.H. Roberds. 1993. Dynamic gene conservation for uncertain futures. For. Ecol. Mgmt. 62: 15-37. Falconer, D.S. a n d T.F.C. Mackay. 1996. Introduction to quantitative genetics. 4 G r o u p Ltd., Harlow, U.K.  t h  ed. Longman  Falk, D.A. 1 9 9 1 . Joining biological and economic models for conserving plant genetic diversity. Pp. 209-222 in Genetics and conservation of rare plants. D.A. Falk and K.E. Holsinger, Eds. Oxford Univ. Press, N.Y., NY. Fisher, R.M. and M T . Myres. 1979. A review of factors influencing extralimital occurrences of Clark's nutcracker in C a n a d a . C a n . Field Nat. 94(1): 4 3 - 4 1 . Forest Health Committee of British Columbia. 1999. Forest Health Committeee meeting minutes J u n e 2 3 , 1 9 9 9 . Williams Lake, B.C. 21 pp. F r a n k h a m , R. 1995. Conservation genetics. A n n . Rev. Genet. 2 9 : 305-327. Franklin, J.F., F.J. S w a n s o n , M.E. H a r m o n , D.A. Perry, T A . Spies, V . H . Dale, A. M c K e e , W.K. Ferrell, J.E. Means, S.V. Gregory, J.D. Lattin, T.D. Schowalter a n d D. Larsen. 1 9 9 1 . Effects of global climatic change on forests in northwestern North America. Northwest Env. J . 7: 2 3 3 254. Furnier, G.R. a n d W.T. A d a m s . 1986. Mating system in natural populations of Jeffrey pine. A m . J . Bot. 73(7): 1002-1008. Furnier, G.R., P. Knowles, M.A. Aleksiuk and B.P. Dancik. 1986. Inheritance and linkage of allozymes in seed tissue of whitebark pine. C a n . J . Genet. Cytol. 2 8 : 601 -604. Furnier, G.R., P. Knowles, M.A. Clyde and B.P. Danick. 1987. Effects of avian dispersal on the genetic structure of whitebark pine populations. Evol. 4(3): 607-612. Gay, P.A. a n d S. T u z u n . 2 0 0 0 a . Involvement of a novel peroxidase isozyme and lignification in hydathodes in resistance to black rot disease in cabbage. C a n . J . Bot. 78(9): 1144-1149. . 2000b. Temporal and spatial assessment of defense responses in resistant a n d susceptible cabbage varieties during infection with Xanthomonas campestris pv. campestris. Phys. Molec. Plant Path. 57(5): 201-210. Gillet, E.M. 1996. Qualitative inheritance analysis of isoenzymes in haploid gametophytes: principles and a computerized model. Silv. Genet. 45(1): 8-15. Goldstein, D.B. a n d D.D. Pollock. 1997. Launching microsatellites: a review of mutation processes and methods of phylogenetic inference. J . Hered. 88: 335-342. Goldstein, D.B., A . R . Linares, L.L. Cavalli-Sforza and M.W. Feldman. 1995. A n evaluation of genetic distances for use with microsatellite loci. Genetics 139: 4 6 3 - 4 7 1 . Goncharenko, G.G., V . E . Pauditov and A . E . Silin. 1992. Population structure, gene diversity, a n d differentiation in natural populations of Cedar pines (Pinus subsect. Cembrae, Pinaceae) in the U S S R . PI. Syst. Evol. 182: 121-134.  84  G o o d i n g , G.D. 1998. Genetic variation and mating system of ponderosa pine in the Willamette Valley of O r e g o n . M.Sc. Thesis, Oregon State Univ., Corvallis, O R . 93 pp. Gregorius, H.-R. and P. Baradat. 1992. A commentary on current approaches to forest population genetics. New For. 6: 409-420. Hamrick, J.L. a n d R.W. Allard. 1972. Microgeographical variation in allozyme frequencies in Avena barbata. P N A S . 69(8): 2100-2104. Hamrick, J.L., M.-J. W . Godt and S.L. Sherman-Broyles. 1992. Factors influencing levels of genetic diversity in w o o d y plant species. New For. 6: 95-124. Hamrick, J.L., M.-J. W . Godt, D.A. Murawski and M.D. Loveless. 1 9 9 1 . Correlations between species traits a n d allozyme diversity: implications for conservation biology. Pp. 75-99 in Genetics and conservation of rare plants. D.A. Falk and K.E. Holsinger, Eds. Oxford Univ. Press, N.Y., NY. H a n e m a n n , W . M . 1986. Economics and the preservation of biodiversity. Pp. 193-199 in Biodiversity. E.O. Wilson, Ed. National A c a d . Press, W a s h . D.C. Hard, J.J. 1995. A quantitative genetic perspective on the conservation of intraspecific diversity. A m . Fish. Soc. S y m p . 17: 304-326. Hartl, D.L. and A . G . Clark. 1997. Principles of population genetics. 3 Sunderland, Mass. 542 pp.  r d  e d . Sinauer Assoc. Inc.,  Hedrick P.W., O. Savolainen and K. Karkkainen. 1999. Factors influencing the extent of inbreeding depression: an example from Scots pine. Hered. 82: 441-450. Hendrick, P.W. 1999. Perspective: highly variable loci and their interpretation in evolution a n d conservation. Evol. 53(2): 313-318. Hoff, R.J. 1984. Resistance to Cronartium ribicola in Pinus monticola: trees. U S D A F S Res. Note INT-343. O g d e n , UT. 7 pp.  higher survival of infected  Hoff, R.J. 1986. Inheritance of the bark reaction resistance mechanism in Pinus monticola infected by Cronartium ribicola. U S D A F S Res. Note I N T - 3 6 1 . O g d e n , UT. 9 pp. Hoff, R.J. a n d S.K. Hagle. 1990. Diseases of whitebark pine with special emphasis on white pine blister rust. Pp. 179-190 in Proc. symposium on whitebark pine ecosystems: ecology a n d m a n a g e m e n t of a high-mountain resource. Bozeman, MT., USA. March 2 9 - 3 1 , 1 9 8 9 . Hoff, R.J., D.E. Ferguson, G.I. McDonald and R.E. Keane. 2 0 0 1 . Strategies for managing whitebark pine blister rust. Pp. 346-366 in Whitebark pine communities: ecology a n d restoration. D.F. Tomback, S.F. Arno and R.E. Keane, eds. Island Press, W a s h i n g t o n D.C. Hoff, R.J., S.K. Hagle a n d R.G. Krebill. 1994. Genetic consequences and research challenges of blister rust in whitebark pine forests. Pp. 118-126 211 in Proc. international w o r k s h o p on subalpine stone pines and their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5 - 1 1 , 1 9 9 2 . Holsinger, K.E., R.J. Mason-Gamer and J . Whitton. 1999. Genes, d e m e s , and plant conservation. Pp. 23-46 in Genetics and the extinction of species. L.F. Landweber and A . P . D o b s o n , eds. Princeton Univ. Press, Princeton, NJ.  85  H u e n n e k e , L.F. 1 9 9 1 . Ecological implications of genetic variation in plant populations. Pp. 31-44 in Genetics a n d conservation of rare plants. D.A. Falk and K.E. Holsinger, Eds. Oxford Univ. Press, N.Y., NY. Huntley, B. 1 9 9 1 . How plants respond to climate change: migration rates, individualism a n d the c o n s e q u e n c e s for plant communities. A n n . Bot. 67(Supplement 1): 15-22. Hutchins, H.E. a n d R.M. Lanner. 1982. The central role of Clark's nutcracker in the dispersal a n d establishment of whitebark pine. Oecologia 55: 1 9 2 - 2 0 1 . Intergovernmental Panel on Climate Change. 2001a. a n d vulnerability. 15 pp.  Climate c h a n g e 2 0 0 1 : impacts, adaptation,  Intergovernmental Panel on Climate Change. 2001b. Third assessment report: s u m m a r y for policymakers. 18 pp. Jackson, S T . and J T . Overpeck. 2000. Responses of plant populations a n d communities to environmental changes of the late Quaternary. Pp. 194-220 in D.H. Erwin a n d S.L. W i n g , eds. Deep time: paleobiology's perspective. Paleobiology 2 6 (Supplement).. Jarne, P. a n d D. Charlesworth. 1993. T h e evolution of the selfing rate in functionally hermaphrodite plants and animals. A n n . Rev. Ecol. Syst. 2 4 : 441-466. J o r g e n s e n , S . M . and J.L. Hamrick. 1997. Biogeography and population genetics of whitebark pine, Pinus albicaulis. C a n . J . For. Res. 2 7 : 1574-1585. Juo, P.-S. a n d G . Stotzky. 1973. Electrophoretic analysis of isozymes f r o m seeds of Pinus, Abies, a n d Pseudotsuga. C a n . J . Bot. 5 1 : 2201-2205. Karhu, A., P. Hurme, M. Karjalainen, P. Karvonen, K. Karkkainen, D. Neale a n d O. Savolainen. 1996. Do molecular markers reflect patterns of differentiation in adaptive traits of markers? Theor. A p p l . Genet. 93: 2 1 5 - 2 2 1 . Karp, A., P.G. Isaac and D.S. Ingram, eds. 1998. Molecular tools for screening biodiversity. L o n d o n : C h a p m a n and Hall. 498 pp. Keane, R.E. a n d S.F. Arno. 1993. Rapid decline of whitebark pine in western Montana: evidence f r o m 20-year remeasurements. West. J . Appl. For. 8(2): 44-47. Keane, R.E., S.F. Arno, J.K. Brown and D.F. Tomback. 1990. Modelling stand d y n a m i c s in whitebark pine (Pinus albicaulis) forests. Ecol. Model. 5 1 : 73-95. Keane, R.E., P. Morgan a n d J.D. White. 1999. Temporal patterns of ecosystem processes on simulated landscapes in Glacier National Park, Montana, USA. Landscape Ecology 14: 3 1 1 329. Kendall, K.C. 1994. Whitebark pine conservation in North American national parks. Pp. 302-307 in Proc. international workshop on subalpine stone pines and their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5 - 1 1 , 1992. Kendall, K.C. a n d S.F. Arno. 1990. Whitebark pine - an important but endangered wildlife resource. Pp. 264-273 in Proc. symposium on whitebark pine ecosystems: ecology a n d m a n a g e m e n t of a high-mountain resource. Bozeman, MT., USA. March 2 9 - 3 1 , 1989. Kieran, B. 1998. This just in f r o m Geneva. Canoe C N e w s , Sep. 9 , 1 9 9 8 . kieran.html  86  Kimura, M. a n d J.F. Crow. 1964. T h e number of alleles that can be maintained in a finite population. Genetics 49: 725-738. Kirkpatrick, M. and P. Jarne. 2000. The effects of a bottleneck on inbreeding depression and the genetic load. A m . Nat. 155(2): 154-167. Klinka, K. a n d C. Chourmouzis. 2 0 0 0 1 . The Mountain Hemlock Z o n e . U B C Faculty of Forestry, Vancouver, B C . 187 pp. Krutovskii, K.V., D.V. Politov and Y.P. Alkhutov. 1995. Isozyme study of population genetic structure, mating system and phylogenetic relationships of the five stone pine species (subsection Cembrae, section Strobi, subgenus Strobus). Pp. 270-304 in Population genetics a n d conservation of forest trees. Baradat, P., W.T. A d a m s and G . Muller-Starck, eds. S P B A c a d e m i c Publishing, the Netherlands. ; . 1994. Genetic differentiation a n d phylogeny of stone pines based on isozyme loci. Pp. 19-30 in Proc. international workshop o n subalpine stone pines and their environment: the status of our knowledge. St. Moritz, Switzerland. S e p . 5 - 1 1 , 1992. Lande, R. 1999. Extinction risks from anthropogenic, ecological a n d genetic factors. Pp. 1-22 in Genetics and the extinction of species. L.F. Landweber and A . P . D o b s o n , eds. Princeton Univ. Press, Princeton, NJ. Lande, R. a n d G.F. Barrowclough. 1987. Effective population size, genetic variation, a n d their use in population genetic management. Pp. 87-124 in Viable popuations for conservation. Soule, M.E., E d . Cambridge Univ. Press, Cambridge, Mass., U.S.A. Lanner, R.M. 1982. Adaptations of whitebark pine for seed dispersal by Clark's nutcracker. C a n . J . For. Res. 12: 391-402. Lanner, R.M. 1996. Made for each other: a symbiosis of birds and pines. Oxford Univ. Press. N.Y., N.Y. 160 pp. Lanner, R.M. and B.K. Gilbert. 1994. Nutritive value of whitebark pine seeds, a n d the question of their variable dormancy. Pp. 206-211 in Proc. international workshop on subalpine stone pines a n d their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5 - 1 1 , 1 9 9 2 . Latta, R.G. a n d J.B. Mitton. 1997. A comparison of population differentiation across four classes of gene marker in limber pine (Pinus flexilis James). Genetics 146: 1153-1163. L e a d a m , C L . 1986. Seed dormancy in three Pinus species of the inland mountain west. Pp. 117124 in Proc. conifer tree seed in the inland mountain west symposium. Missoula, MT. A u g . 5-6, 1985. Lecerteau, E. and A . E . Szmidt. 1999. Properties of A F L P markers in inheritance and genetic diversity studies of Pinus sylvestris L. Hered. 82: 252-260. Ledig, F T . 1986. Conservation strategies for forest genetic resources. For. Ecol. Mgmt. 14: 7 7 90. Ledig, F T . 1988. T h e conservation of diversity in forest trees: w h y and how should g e n e s be conserved? Bioscience 38(7): 471-479. Ledig, F T . and J . H . Kitzmiller. 1992. Genetic strategies for reforestation in the face of global climate c h a n g e . For. Ecol. Mgmt. 50: 153-169.  87  Ledig, F T . , J.J. Vargas-Hernandez and K.H. Johnsen. 1998. T h e conservation of forest: case histories f r o m C a n a d a , Mexico, and the United States. J . For. 95: 3 2 - 4 1 . Ledig, F T . , M T . Conkle, B. Bermejo-Velasquez, T. Eguiluz-Piedra, P.D. Hodgskiss, D.R. J o h n s o n a n d W . S . Dvorak. 1999. Evidence for an extreme bottleneck in a rare Mexican pinyon: genetic diversity, disequilibrium, and the mating system in Pinus maximartinezii. Evol. 53(1): 91-99. Leopold, A. 1933. T h e conservation ethic. J . For. 3 1 : 634-643. Lesica, P. and F.W. Allendorf. 1995. W h e n are peripheral populations valuable for conservation? C o n s . Biol. 9(4): 753-760. Lester, D T . and A . D . Yanchuk. 1996. A survey of the protected status of conifers in British C o l u m b i a : in situ gene conservation. B C M o F , Research Report No. 4, Victoria, B C . Levene, H. 1949. O n a matching problem arising in genetics. A n n . Math. Stat. 2 0 : 91-94. Lewontin, R.C. 1984. Detecting population differences in quantitative characters as o p p o s e d to g e n e frequencies. A m . Nat. 123(1): 115-124. Lieu, G.Y., C C . C h e n , G.F. Yuan and C.Y. C h e n . 2 0 0 1 . A taxonomic study of the g e n u s Rhizopus by isozyme patterns. Nova Hedwigia. 72(1-2): 231-239 (Abstr.) Linhart, Y . B . a n d D.F. Tomback. 1985. Seed dispersal by nutcrackers causes multi-trunk growth f o r m in pines. Oecologia. 67: 107-110. Liston, A., W.A. Robinson, E.R. Alvarez-Buylla and D. Pinero. 1996. Pine trees: a comparison of internal transcribed spacer region and chloroplast phylogenies of Pinus (Pinaceae). Abstracts of the A I B S Meetings, Seattle, W a . A u g . 3 - 8 , 1 9 9 6 . A m . J . Bot. 83(6): 175 (Abstract). Lovejoy, B.K. 1986. the interconnectedness of species and their preservation. Pp. 156-161 in Biodiversity. E.O. Wilson, Ed. National A c a d . Press, W a s h . D.C. ' Lu, M.-Z., A . E . Szmidt and X.-R. W a n g . 1995. Inheritance of R A P D fragments in haploid a n d diploid tissues of Pinus sylvestris (L.). Hered. 7 4 : 583-589. M a a , W . C . J , and S.C. Liao. 2000. Culture-dependent variation in esterase isozymes a n d malathion susceptibility of diamondback moth, Plutella xylostella L. Zool. Studies 39(4): 3 7 5 386. Manly, B.F.J. 1985. Multivariate statistical methods: a primer. C h a p m a n Hall, N.Y., N.Y. 159 pp. Mathiasen, R.L. and F.G. Hawksworth. 1988. Dwarf mistletoes on western white pine a n d whitebark pine in northern California and southern Oregon. For. Sci. 34(2): 4 2 9 - 4 4 0 . Mattson, D.J. a n d D.P. Reinhart. 1997. Excavation of red squirrel middens by grizzly bears in the whitebark pine zone. J . Appl. Ecol. 34: 926-940. Mayberry, J.S. and P.P. Feret. 1977. Peroxidases in developing acorns a n d seedlings of albaL. Silv. Genet. 2 6 : 218-223.  Quercus  M c C a u g h e y , W . W . 1994. The regeneration process of whitebark pine. Pp. 179-187 in Proc. international workshop on subalpine stone pines and their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5 - 1 1 , 1 9 9 2 . M c C a u g h e y , W . W . and W . C . Schmidt. 1990. Autecology of whitebark pine. Pp. 85-96 in Proc. s y m p o s i u m on whitebark pine ecosystems: ecology and management of a high-mountain resource. B o z e m a n , M T , USA. March 2 9 - 3 1 , 1989.  88  M c C a u g h e y , W . W . and W . C . Schmidt. 2 0 0 1 . Taxonomy, distribution a n d history. Pp. 29-40 in Whitebark pine communities: ecology and restoration. D.F. Tomback, S.F. Arno a n d R.E. Keane, eds. Island Press, Washington D.C. McDonald, G.I. and R.J. Hoff. 2 0 0 1 . Blister rust: an introduced plague. Pp. 196-220 in Whitebark pine communities: ecology and restoration. D.F. Tomback, S.F. Arno and R.E. Keane, eds. Island Press, Washington D.C. Meagher, M.D. a n d D.G. Edwards. 1997. 1997 cone collections report. 2 pp. plus appendices. Unpublished report. Millar, C.I. a n d R.D. Westfall. 1992. Allozyme markers in forest genetic conservation. N e w For. 6: 347-371. Mitton, J.B. 1992. T h e dynamic mating systems of conifers. New For. 6: 197-216. Mitton, J.B., Y . B . Linhart, J.B. Hamrick, and J.S. B e c k m a n , J.S. 1997. Observations on the genetic structure and mating system of ponderosa pine in the Colorado Front Range. Theor. Appl. Genet. 51(1): 5-13. Mitton, J.B., Y . B . Linhart, M.L. Davis and K.B. Sturgeon. 1 9 8 1 . Estimation of outcrossing in ponderosa pine, Pinus ponderosa Laws., from patterns of segregation of protein polymorphisms a n d f r o m frequencies of albino seedlings. Silv. G e n . 30(4-5): 1 1 7 - 1 2 1 . Mitton, J.B., Y . H . Linhart, S.W. Sturgeon and J.P. Farrar. 1977. Protein polymorphisms detected in mature needle tissue of ponderosa pine. A m . J . Bot. 4 2 : 1154-1157. Moritz, C. 1994. Defining 'evolutionarily significant units' for conservation. Trends Ecol. Evol. 9(10): 373-375. Murawski, D.A., I.A.U.N. Gunatilleke and K.S. B a w a . 1994. T h e effects of selective logging o n inbreeding in Shorea megistophylla (Dipterocarpaceae) from Sri Lanka. C o n s . Biol. 8(4): 9 9 7 1002. Murray, M.P., S.C. Bunting and P. Morgan. 2000. Landscape trends (1753-1993) of whitebark pine {Pinus albicaulis) forests in the west Big Hole Range of Idaho/Montana, U.S.A. Arctic, Ant. a n d Alpine Res. 32(4): 412-418. N a m k o o n g , G. 1992. 438-443.  Biodiversity - issues in genetics, forestry a n d ethics. For. C h r o n . 68(4):  Nei, M. 1972. Genetic distance between populations. A m . Nat. 106(949): 2 8 3 - 2 9 1 . Nei, M. 1973. Analyis of gene diversity in subdivided populations. P N A S . 70(12): 3321-3323. Nei, M. 1977. F-statistics and analysis of gene diversity in subdivided populations. A n n . H u m . Genet., Lond. 4 1 : 225-233. Nei, M. 1978. Estimation of average heterozygosity and genetic distance f r o m a small number of individuals. Genetics 89: 583-590. Nei, M., T. M a r u y a m a and R. Chakraborty. 1975. T h e bottleneck effect and genetic variability in populations. Evol. 29(1): 1-10. N o m u r a , T. 1999. O n the methods for predicting the effective size of populations under selection. H e r e d . 83: 485-489.  89  Norton, B. 1986. Commodity, amenity and morality: the limits of quanitication in valuing biodiversity. Pp. 200-205 in Biodiversity. E.O. Wilson, Ed. National A c a d . Press, W a s h . D.C. Noss, R.F. 1990. C a n w e maintain biological and ecological integrity? C o n s . Biol. 4(3): 241-243. . 2 0 0 1 . Beyond Kyoto: forest management in a time of rapid climate c h a n g e . C o n s . Biol. 15(3): 578-590. Nunney. L. 1999. T h e effective size of a hierarchically structured popuation. Evol. 53(1): 1-10. Ogilvie, R . t . 1990. Distribution and ecology of whitebark pine in western C a n a d a . Pp. 54-60 in Proc. s y m p o s i u m on whitebark pine ecosystems: ecology and m a n a g e m e n t of a high-mountain resource. B o z e m a n , MT., USA. March 2 9 - 3 1 , 1989. Ohta, T. 1982. Linkage disequilibrium with the island model. Genetics. 1 0 1 : 139-155. O h t a , T. a n d M. Kimura. 1973. A model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a finite population. Genet. Res. C a m b . 2 2 : 2 0 1 - 2 0 4 . Oldfield, S. 1984. Ethical issues in valuing biological diversity. Pp. 235-246 in Preservation of species: the value of biological diversity. B.G. Norton, Ed. Princeton Univ. Press, Princeton, N.J. 3 0 5 pp. Parker, P.G., A.A. Snow, M.D. S c h u g , G.C. Booton a n d P.A. Fuerst. 1998. W h a t molecules can tell us about populations: choosing and using a molecular marker. Ecol. 79(2): 361-382. Perkins, D.L. a n d T.W. S w e t n a m . 1996. A dendroecological assessment of whitebark pine in the Sawtooth - Salmon River region, Idaho. C a n . J . For. Res. 2 6 : 2123-2133. Perry, D.A., J.G. Borchers, D.P. Turner, S.V. Gregory, C.R. Perry, R.K. Dixon, S.C. Hart, B. Kauffman, R.P. Neilson and P. Sollins. 1 9 9 1 . Biological feedbacks to climate c h a n g e : terrestrial ecosystems as sources of carbon and nitrogen. Northwest Env. J . 7: 2 0 3 - 2 3 2 . Perry, D.J. a n d B.P. Dancik. 1985. Mating system dynamics of lodgepole pine in Alberta, C a n a d a . Silv. G e n . 35(5-6): 190-195. Perry, D.J. a n d J . Bousquet. 1998. Sequence-tagged-site (STS) markers of arbitrary genes: the utility of black spruce-derived S T S primers in other conifers. Theor. Appl. Genet. 9 7 : 735-743. Peters II, R.L. 1986. T h e effect of global climatic change on natural communties. Pp. 450-461 in Biodiversity. E.O. Wilson, Ed. National A c a d . Press, W a s h . D.C. Peterson, D.L., M.J. Arbaugh, L.J. Robinson and B.R. Derderian. 1990. Growth trends of whitebark pine a n d lodgepole pine in a subalpine Sierra Nevada forest, California, U.S.A. Arctic a n d Alpine Res. 2 2 : 233-243. Petkau, D., L.P. Waits, P.L. Clarkson, L. Craighead and C. Strobeck. 1997. A n empirical evaluation of genetic distance statistics using microsatellite data from bear (Ursidae) populations. Genetics. 147: 1943-1957. Pitel, J.A. a n d B.S.P. W a n g . 1990. Physical and chemical treatments to improve germination of whitebark pine seeds. Pp. 130-133 in Proc. symposium on whitebark pine ecosystems: ecology a n d m a n a g e m e n t of a high-mountain resource. B o z e m a n , MT., USA. March 2 9 - 3 1 , 1989. Politov, D.V. and K.V. Krutovskii. 1994. Allozyme polymorphism, heterozygosity, a n d mating system of stone pines. Pp. 36-42 in Proc. international workshop o n subalpine stone pines a n d their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5 - 1 1 , 1992. 90  Pollock, D.D., A. Bergman and D.B. Goldstein. 1998. Microsatellite behaviour with range constraints: parameter estimation and improved distances for use in phylogenetic reconstruction. Theor. Pop. Biol. 53: 2 5 6 - 2 7 1 . Prentice, I.C., D. Jolly and B I O M E 6000 participants. 2000. Mid-Holocene a n d glacial-maximum vegetation geography of the northern continents and Africa. J . Biogeog. 2 7 : 507-519. Price, R.A., A. Liston and S.H. Strauss. 1998. Phylogeny and systematics of Pinus. Pp. 49-68 in Ecology and Biogeography of Pinus, D.M. Richardson, Ed. Cambridge Univ. Press. Primack, R.B. 1998. Essentials of conservation biology, 2 Mass.  n d  Ed. Sinauer Assoc., Sunderland  Province of British Columbia. 1996. Progress report on initiatives for the conservation of biodiversity in British Columbia. Victoria, B.C. 30 pp. Randall, A. 1986. H m a n preferences, economics, and the preservation of species. Pp. 79-109 in Preservation of species: the value of biological diversity. B.G. Norton, Ed. Princeton Univ. Press, Princeton, N.J. 305 pp. Richards, C M . 2000. Inbreeding depression and genetic rescue in a plant metapopulation. A m . Nat. 155(3): 303-394. Richardson, B.A. 2 0 0 1 . G e n e flow and genetic structure of whitebark pine (Pinus albicaulis): inferences into bird-dispersed seed movement and biogeography. M.Sc. Thesis. Univ. of Idaho, Moscow, ID. 55 pp. Ridgeway, F T . , S . C C o n w a y and M. Maples. 1970. Polymorphic protein detected in brook trout. J . Fish. 4 2 : 1 6 6 - 1 7 1 . Ritland, K. 1989. Correlated matings in the partial selfer, Mimulus  guttatus.  Evol. 4 3 : 8 4 8 - 8 5 9 .  Ritland, K. 1990. A series of F O R T R A N computer programs for estimating plant mating systems. J . Hered. 8 1 : 2 3 5 - 2 3 7 . Ritland, K. and S.K. Jain. 1 9 8 1 : A model for the estimation of outcrossing rate a n d gene frequencies using n independent loci. Hered. 4 7 : 35-52. Ritland, K. and Y.A. El-Kassaby. 1985. T h e nature of inbreeding in a seed orchard of Douglas-fir a s s h o w n by an efficient multilocus model. Theor. Appl. G e n . 7 1 : 375-384. Rogers, D . L , C.I. Millar and R.D. Westfall. 1999. Fine-scale genetic structure of whitebark pine (Pinus albicaulis): associations with watershed and growth f o r m . Evol. 53(1): 74-90. Running, S.W. and R.R. Nemani. 1 9 9 1 . Responses of forests resulting f r o m potential climate c h a n g e . Climate change 19: 349-368. Salwasser, H. 1990. Conserving biological diversity: a perspective o n scope and approaches. For. Ecol. Mgmt. 35: 79-90. Sawicki, J . 2000. Protected areas of British Columbia Act. Schlotterer, C , R. Ritter, B. Harr and G. Brem. 1998. High mutation rate of a long microsatellite allele in Drosophila melanogaster provides evidence for allele-specific mutation rates. Mol. Biol. Evol. 15(10): 1269-1274. Schoettle, A . W . and S.G. Rochelle. 2000. Morphological variation of Pinus flexilis (Pinaceae), a bird-dispersed pine, across a range of elevations. A m . J . Bot. 87(12): 1797-1896. 91  Schuster, W . S . , D.L. Alles and J.B. Mitton. 1989. Gene flow in limber pine: evidence f r o m pollination phenology and genetic differentiation along an elevational transect. A m e r . J . Bot. 76(9): 1395-1403. Schuster, W . S . F . a n d J.B. Mitton. 1 9 9 1 . Relatedness within clusters of a bird-dispersed pine a n d the potential for kin interactions. Hered. 67: 41 -48. Sedjo, R.A. and A . M . Solomon. 1988. Climate and Forests. Chapter 8 in Proc. G r e e n h o u s e w a r m i n g : abatement and adaptation. Eds. Roseburg, N.J., W . E . Easterling III, P.R. Crosson a n d J.Darmstadter. Washington, D.C. June 14-15, 1988. pp. 105-120. Shafer, S.L., P.J. Bartlein and R.S. T h o m p s o n . 2 0 0 1 . Potential changes in the distributions of western North America tree and shrub taxa under future climate scenarios. Ecosystems 4: 200-215. S h a w , D.V., A.L. Kahler a n d R.W. Allard. 1 9 8 1 . A multilocus estimator of mating system parameters in plant populations. P N A S . 78(2): 1298-1302. Simberloff, D. 2000. Global climate change and introduced species in United States forests. Sci. Total Env. 202: 2 5 3 - 2 6 1 . Sokal, R.R. a n d F.J. Rohlf. 1995. Biometry. 3  rd  ed. W . H . Freeman and Co., N.Y., N.Y. 887 pp.  Stettler, R.F. and H.D. Bradshaw. 1994. T h e choice of genetic material for mechanistic studies of adaptation in forest trees. Tree Phys. 14: 781 -796. Stuart-Smith, G.J. 1998. Conservation of whitebark pine in the Canadian Rockies: blister rust a n d population genetics. M.Sc. Thesis. Univ. of Alberta, Edmonton, A B . 125 pp. Swofford, D.L., R.B. Selander and W . C . Black IV. 1997. BIOSYS-2: a computer program for the analysis of allelic variation in genetics. 35 pp. Szmidt, A . E . and X-R. W a n g . 1999. Genetic markers in forest genetics - the tunnel remains dark. For. G e n . and Sustainability 63: 1-18. Tani, N., N. T o m a r u , Y. T s u m u r a , M. Araki and K. O h b a . 1998. Genetic structure within a J a p a n e s e stone pine (Pinus pumila Regel) population on Mt. Aino-Dake in central H o n s h u , J a p a n . J . Plant Res. 1 1 1 : 7-15. T o m b a c k , D.F. 1982. Dispersal of whitebark pine seeds by Clark's nutcracker: a mutualism hypothesis. J . Animal Ecol. 5 1 : 451-467. T o m b a c k , D.F. 1986. Post-fire regeneration of krummholz whitebark pine: a c o n s e q u e n c e of nutcracker seed caching. Madrono 33(2): 100-110. T o m b a c k , D.F. 2 0 0 1 . Clark's nutcracker: agent o f regeneration. Pp. 89-104 in Whitebark pine communities: ecology and restoration. D.F. Tomback, S.F. Arno and R.E. Keane, eds. Island Press, Washington D.C. T o m b a c k , D.F. a n d W . F . Schuster. 1994. Genetic population structure a n d growth f o r m distribution in bird-dispersed pines. Pp. 43-50 in Proc. international workshop o n subalpine stone pines and their environment: the status of our knowledge. St. Moritz, Switzerland. Sep. 5-11,1992. T o m b a c k , D.F. and Y . B . Linhart. 1990. T h e evolution of bird-dispersed pines. Evol. Ecol. 4 : 185219.  92  T o m b a c k , D.F., J.K. Clary, J . Koehler, R.J. Hoff and S.F. Arno. 1995. T h e effects of blister rust on post-fire regeneration of whitebark pine: the Sundance burn of northern Idaho (U.S.A.). C o n s . Biol. 9(3): 654-664. T o m b a c k , D.F., S.F. Arno and R.E. Keane. 2 0 0 1 . T h e compelling case for m a n a g e m e n t intervention. Pp. 3-25 in Whitebark pine communities: ecology and restoration. D.F. T o m b a c k , S.F. Arno a n d R.E. Keane, eds. Island Press, Washington D.C. Tyagi, M., A . M . Kayastha and B. Sinha. 2000. T h e role of peroxidase and polyphenol oxidase isozymes in wheat resistance to Alternaria triticina. Biol. Plant. 43(4): 559-562. United States Environmental Protection Agency. 2000. Impacts of global w a r m i n g : forests. http://www.epa.qov/qlobalwarmina/impacts/forests/index.html United States Environmental Protection Agency. 2001 a. Changing climate. http://www.epa.qob/qlobalwarminq/climate/index.html United States Environmental Protection Agency. 2001 a. Uncertainties. http://www.epa.qob/qlobalwarminq/uncertainties.html Vander Wall, S.B. and R.P. Balda. 1977. Coadaptations of the Clark's nutcracker a n d the pinon pine for efficient seed harvest and dispersal. Ecol. Monog. 4 7 : 8 9 - 1 1 1 . Vida, G. 1994. Global issues of genetic diversity. Pp. 9-21 in V. Loeshcke, J . T o m i u k a n d S.K. Jain, eds. Conservation genetics. Birkhauser, Berlin, Germany. Vujanovic, V., M. St-Arnaud and P.-J. N e u m a n n . 2000. Susceptibility of cones and seeds to fungal infection in a pine (Pinus spp.) collection. For. Path. 30: 305-320. W a d e , M.J. a n d C.J. Goodnight. 1998. Perspective: the theories of Fisher and Wright in the context of metapopulations: w h e n nature does many small experiments. Evol. 52(6): 15371553. W a t s o n , R.T., M.C. Zinyowera and R.H. Moss, eds. 1997. T h e regional impacts of climate c h a n g e : an assessment of vulnerability. Intergovernmental Panel on Climate C h a n g e , G e n e v a , Switzerland. 4 9 pp. W e a v e r , T. and F. Forcella. 1986. Cone production in Pinus albicaulis forests. Pp. 68-76 in Proc. conifer tree seed in the inland mountain west symposium. Missoula, MT. A u g . 5 - 6 , 1 9 8 5 . W e a v e r , T. and J . Jacobs. 1990. Occurrence of multiple stems in whitebark pine. Pp. 156-159 in Proc. s y m p o s i u m on whitebark pine ecosystems: ecology and m a n a g e m e n t of a high-mountain resource. B o z e m a n , MT., USA. March 2 9 - 3 1 , 1 9 8 9 . Williams, C.G., J.L. Hamrick and P.O. Lewis. 1995. Multiple-population versus hierarchical conifer breeding programs: a comparison of genetic diversity levels. Theor. A p p l . Genet. 9 0 : 5 8 4 594. Wilson, B. and J . Stuart-Smith. 2000. Whitebark pine conservation for the Canadian Rocky Mountain National Parks. Draft version. 54 pp. Wright, S. 1 9 3 1 . Evolution in Mendelian populations. Genetics 16: 97-159. Wright, S. 1965. T h e interpretation of population structure by F-statistics with special regard to systems of mating. Evol. 19: 395-420. Yanchuk, A . D . 2 0 0 1 . A quantitative framework for breeding and conservation of forest tree genetic resources in British Columbia. C a n . J . For. Res. 7 1 : 566-576.  93  Yanchuk, A . D . and D.T. Lester. 1996. Setting priorities for conservation of the conifer genetic resources of British Columbia. For. Chron. 72(4): 406-415. Yandell, U.G. 1992. A n allozyme analysis of whitebark pine (Pinus albicaulis Thesis. Univ. of N e v a d a , Reno, NV. 7 3 pp.  Engl, [sic]) M.Sc.  Y e h , F.C., R.-C. Y a n g and T. Boyle. 1999. Popgene Version 3.2: Microsoft W i n d o w - b a s e d freeware for population genetic analysis. 18 pp. Y o u n g , A., D. Boshier and T. Boyle, eds. 2000. Forest conservation genetics - principles and practice. C S I R O Pub., Victoria, Australia. 352 pp. Zeglen, S. 2000. Whitebark pine a n d white pine blister rust in British Columbia. Interim report. B.C. M o F , Vancouver Forest Region. Feb. 2000. 10 pp. plus appendices. . 1999. Whitebark pine and white pine blister rust in British Columbia. Interim report. B.C. M o F , Vancouver Forest Region. June 1999. 8 pp. plus a p p e n d i c e s . '  94  A P P E N D I X I - Table of allele frequencies by population  Table A.1.1. Allele frequencies for 17 populations; names and poulation numbers as in Table 3.1. Pop  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  19  20  22  27  29  24  19  25  17  30  30 ' 17  20  27  30  29  17  Allele N  17 "  MDH1-1 1.000 1.000 1.000 0.981 1.000 0.958 1.000 1.000 1.000 1.000 1.000 0.941 1.000 0.981 0.933 0.983 1.000 MDH1-2 0.000 0.000 0.000 0.019 0.000 0.042 0.000 0.000 0.000 0.000 0.000 0.059 0.000 0.019 0.067 0.017 0.000 N  20  25  21  32  32  27  19  25  20  30  30  17  24  33  29  30  20  MDH2-1 0.850 0.860 0.881 0.703 0.688 0.741 0.737 0.660 0.750 0.683 0.683 0.588 0.646 0.788 0.759 0.750 0.750 MDH2-2 0.150 0.140 0.119 0.297 0.313 0.259 0.263 0.340 0.250 0.317 0.317 0.412 0.354 0.212 0.241 0.250 0.250 N  20  31  28  32  32  26  19  25  25  28  30  17  24  35  27  30  21  MDH3-1 1.000 0.984 1.000 1.000 0.953 1.000 1.000 0.980 1.000 1.000 0.867 1.000 0.979 0.971 1.000 1.000 1.000 MDH3-2 0.000 0.016 0.000 0.000 0.047 0.000 0.000 0.020 0.000 0.000 0.067 0.000 0.000 0.014 0.000 0.000 0.000 MDH3-3 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.067 0.000 0.021 0.014 0.000 0.000 0.000 N  16  13  17  15  21  4  5  25  4  26  30  17  19  20  30  27 ^  8  PGM1-1 0.844 0.808 0.912 0.833 0.548 0.500 0.200 0.540 0.125 0.423 0.750 0.500 0.447 0.575 0.633 0.463 0.500 PGM1-2 0.156 0.192 0.088 0.167 0.452 0.500 0.800 0.460 0.875 0.577 0.250 0.500 0.553 0.425 0.367 0.537 0.500 N  20  24  26  32  28  27  17  25  24  30  30  17  24  34  30  30  18  SKD1-1 0.725 0.875 0.750 0.875 0.839 0.722 0.676 0.660 0.688 0.533 0.683 0.765 0.708 0.603 0.783 0.833 0.889 SKD1-2 0.275 0.125 0.250 0.125 0.161 0.278 0.324 0.340 0.313 0.467 0.317 0.235 0.292 0.397 0.217 0.167 0.111 N  19  21  26  32  30  28  17  25  25  29  30  17  24  34  30  30  18  SKD2-1 0.500 0.619 0.673 0.672 0.483 0.446 0.676 0.620 0.620 0.517 0.583 0.676 0.771 0.632 0.633 0.550 0.722 SKD2-2 0.500 0.381 0.327 0.328 0.517 0.554 0.324 0.380 0.380 0.483 0.417 0.324 0.229 0.368 0.367 0.450 0.278 N  20  2  16  23  16  13  8  25  19  29  30  17  24  35  26  27  21  FDP1-1 0.900 0.500 1.000 1.000 0.938 1.000 1.000 1.000 0.947 1.000 1.000 1.000 1.000 1.000 0.962 0.963 1.000 FDP1-2 0.100 0.500 0.000 0.000 0.063 0.000 0.000 0.000 0.053 0.000 0.000 0.000 0.000 0.000 0.038 0.037 0.000 N  15  17  22  31  22  29  8  25  24  28  30  17  24  35  30  30  19  GDH1-1 0.967 1.000 1.000 1.000 0.955 1.000 1.000 0.840 1.000 0.875 0.833 1.000 0.750 0.957 0.733 0.933 1.000 GDH1-2 0.000 0.000 0.000 0.000 0.045 0.000 0.000 0.160 0.000 0.107 0.000 0.000 0.000 0.000 0.033 0.033 0.000 GDH1-3 0.033 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.018 0.167 0.000 0.250 0.043 0.233 0.033 0.000 N  20  21  18  25  15  29  10  18  26  30  30  17  16  35  30  22  11  LAP1-1 0.900 0.857 0.889 0.600 0.867 0.914 1.000 1.000 0.923 0.833 0.833 0.941 0.938 1.000 1.000 0.932 1.000 LAP1-2 0.100 0.143 0.111 0.400 0.133 0.086 0.000 0.000 0.077 0.167 0.167 0.059 0.063 0.000 0.000 0.068 0.000 N  16  19  10  25  13  28 . 10  18  21  29  30  17  16  34  30  22  6  LAP2-1 1.000 0.842 0.700 0.500 0.538 0.714 1.000 0.639 0.857 0.621 0.750 0.735 0.813 0.721 0.433 0.795 0.833 LAP2-2 0.000 0.158 0.300 0.500 0.462 0.286 0.000 0.361 0.143 0.379 0.250 0.265 0.188 0.279 0.517 0.205 0.167 LAP2-3 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.050 0.000 0.000 N  17  16  18  25  32  26  19  25  20  30  30  17  16  25  30  28  21  IDH1-1 0.676 0.656 0.944 0.720 0.828 0.962 0.947 0.540 0.850 0.950 0.950 0.971 0.938 0.920 0.967 0.875 0.952 IDH1-2 0.324 0.344 0.000 0.080 0.172 0.038 0.053 0.460 0.150 0.050 0.050 0.029 0.063 0.080 0.033 0.089 0.024 IDH1-3 0.000 0.000 0.056 0.200 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.036 0.024 N  15  4  15  11  16  22  8  25  20  30  30  17  12  26  30  30  15  PGI2-1 0.633 0.625 0.633 0.182 0.438 0.636 0.625 0.500 0.725 0.550 0.500 0.471 0.458 0.769 0.550 0.617 0.533 PGI2-2 0.333 0.250 0.200 0.818 0.281 0.227 0.250 0.000 0.200 0.100 0.050 0.088 0.250 0.077 0.167 0.117 0.000 PGI2-3 0.033 0.125 0.167 0.000 0.281 0.136 0.125 0.500 0.075 0.350 0.450 0.441 0.292 0.154 0.283 0.267 0.467  96  Table  A.1.2.  Locus  Pgii Pgi2 Idh  Pgm  6Pg1  6Pg2  Mdh1 Mdh2  Mdh3 Mdh4  Ovule and pollen allele frequencies for Mannning and Baldy; standard deviations in parentheses. Allele Baldy Manning Pollen  Ovule  Pc Men  1  .965 (0.014)  .850 (0.037)  .819 (0.052)  .760  2  .035 (0.014)  .150 (0.037)  .181 (0.052)  .240 (0.053)  1  .852 (0.017)  .833 (0.045)  .917 (0.017)  .900  (0.040)  2  .148 (0.017)  .167 (0.045)  .083 (0.017)  .100  (0.040)  1  .916 (0.017)  .820 (0.041)  .953 (0.009)  .941 (0.020)  2  .002 (0.000)  .016 (0.000)  .002 (0.001)  .020  (0.000)  3  .082 (0.017)  .164 (0.041)  .045  (0.009)  .039  (0.020)  1  .997 (0.000)  .951 (0.020)  .954  (0.024)  .941 (0.029)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  (0.000)  3  .002 (0.000)  .033 (0.020)  .044 (0.024)  .039  (0.029)  1  .279 (0.028)  .164 (0.047)  .232  .216 (0.058)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  (0.000)  3  .720 (0.028)  .820 (0.047)  .766 (0.044)  .765  (0.058)  1  .998 (0.000)  .984 (0.001)  .990 (0.005)  .961 (0.015)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  (0.000)  3  -  .008 (0.005)  .020  (0.015)  1  .998 (0.000)  .984 (0.001)  .998 (0.000)  .980  (0.000)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  (0.000)  1  .766 (0.043)  .574 (0.063)  .775 (0.031)  .686  (0.075)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  (0.000)  3  .232  (0.043)  .410 (0.063)  .223 (0.031)  .294  (0.075)  1  .998 (0.000)  .983 (0.010)  .998 (0.000)  .980  (0.013)  2  .002 (0.000)  .017 (0.010)  .002 (0.000)  .020  (0.013)  1  .725 (0.039)  .656 (0.049)  .732 (0.031)  .667 (0.060)  2  .002 (0.000)  .016 (0.000)  .002 (0.000)  .020  3  .273 (0.039)  .328  .266 (0.031)  .314 (0.060)  (0.050)  97  (0.044)  Ovule (0.053)  (0.000)  A P P E N D I X II -  Buffer recipes  Table A.2.1. Extraction Buffer - modified from Mitton 1979 Item  Quantity  Polyvinylpyrrolidoninone (PVP-40)  2.00 g  Sucrose  2.00 g  EDTA, Na salt  0.04 g  Dithiothreitol (DTT)  0.03 g  Ascorbic acid, Na salt  0.01 g  Bovine albumin  0.02 g  p-NAD  0.01 g  p-NADP  0.01 g.  Pyridoxal-5'-phosphate  0.001 g  B-mercaptoethanol  2 drops  In 20 mL distilled deionized water (ddH 0), add ingredients sequentially while stirring. Add Bmercaptoethanol in fume hood immediately prior to grinding samples. Keep cold, use immediately. Discard after 24 hours. 2  Table A.2.2. Morpholine Electrode Buffer Item  Quantity  Andydrous citric acid  30.74 g  N-3-aminopropyl morpholine  72 mL  Morpholine Gel Buffer Mix a 1:20 dilution of the electrode buffer.  In 4 L of d d H 0 , dissolve citric acid while stirring. Add morpholine; adjust pH to 8.0 with morpholine. 2  Table A.2.3. Tris-Citrate Electrode Buffer  Tris-Citrate Gel Buffer  Item  Quantity  Item  Quantity  Tris-HCI  27.00 g  Tris-HCI  1.4521 g  Anhydrous citric acid  16.52 g  Anhydrous citric acid  0.8646 g  in 1 L d d H 0 , dissolve the ingredients while stirring. Adjust pH to 6.3 with 1 M NaOH.  Dissolve ingredients in 100 mL d d H 0 while stirring, then m'ake a 1:15 dilution to make 1.6 L. Adjust pH to 6.7 with 1 M NaOH.  2  2  Table A.2.4. Ridgeway Electrode Buffer Item  Quantity  Item  Quantity  Boric Acid  11.875 g  TRIZMA base  6.20 g  Lithium Hydroxide "  1.60 g  Citric acid monohydrate  1.50 g  Dissolve ingredients in 1 L d d H 0 while stirring. Adjust pH to 8.3 with LiOH.  Dissolve ingredients in 1 L d d H 0 ; then mix a 9:1 dilution of gehelectrode solutions. Adjust pH to 8.3.  2  2  Gel fixative Mix a 1:5:5 solution of glacial acetic acid:methanol:water, soak gel slices for 1-2 hours or until unstained surfaces appear opaque, at least 30 minutes.  99  A P P E N D I X III -  Locations of all populations sampled  Table A.3.1. List of populations and s a m p l i n g locations Pop#  Location  Area  NTS 1:50,000  Latitude (N)  Elevation  Mapsheet  Longitude (W)  (m)  Smithers 93L714  54°56'25" 127°19'15"  1850  Driftwood Ck 93U15  54°54'20" 126°46'55"  1600  Newcombe Lk 93E/14  53°45'25" 127°12'35"  1630  TusulkoR93C/12  52°32'20" 125°48'40"  1600  TatlaLk92N/15  51°50'45" 124°59'10"  1700  Tchaikazan R 920/4  51°12'00" 123°39'30"  1600  1  Hudson Bay Mtn  Smithers  2  Higgins Creek  Babine Mtns PP  3  Sweeney Lake  Houston  4  Heckman Pass  Tweedsmuir PP •  5  Perkins Peak  Chilcotin  6  Tchaikazan R  Ts'yl-os PP  7  Yalakom R  Lillooet  Big Bar 920/1  51°04'50" 122°27'05"  1900  8  D'arcy  D'arcy  Birkenhead Lk 92J/10  50°31'15" 122°34'35"  1910  9  Van Horlick Ck  Lillooet  Duffy Lk 92J/8  50°16'20" 122°14'45"  2000  10  Whistler Mtn  Whistler  Whistler 92J/2  50°03'45" 122°56'00"  1700  11  Lime Lookout  Clinton  Clinton 92P/4  51°05'25" 121°39'55"  1980  12  Hart's Pass Okanogan (Washington, U.S.A.) National Forest  USGS 1:24,000 Slate Peak  48°42'30" 120°41'00"  2050  Salmo 82F/3  49°05'10" 117°02'30"  1940  N4837.5W12037.5/7.5  13  Kootenay Pass  Stagleap PP  14  Jumbo Pass  Purcell Mts  Duncan Lk 82K/7  50°20'20" 116°38'00"  2060  15  Stanley Glacier  Kootenay NP  Mt Goodsir 82N/1  51°11'10" 116°04'40"  1850  16  Paget Peak  Yoho NP  Lk Louise 82N/8  51°26'50" 116°21'55"  2240  17  Mt Edith Cavell  Jasper NP  Amethyst Lks 83D/9  52°42'00" 118°03'30"  1750  18  Apex Mtn  Hedley  Penticton 82E/5  49°22'40" 119°55'00"  2170  19  Puddingburn Mtn  Cranbrook  St Mary Lk 82F/9  49°34'00" 116°05'35"  2150  20  Galton Pass  Roosville  Inverted Ridge 82G/2  49°00'45" 114°54'30"  1940  21  Morrissey Ridge  Fernie  Flathead Ridge 82G/7  49°27'00" 114°56'10"  2000  22  Line Ck Mine  Sparwood  Tornado Mtn82G/15  49°45'50" 114°50'25"  2100  23  Mt Seven  Golden  Golden 82N/7  51°15'50" 116°51'30"  2150  24  Castle Mtn  Banff NP  Castle Mtn 820/5  51°17'55" 116°56'30"  2200  25  Parker Ridge  Jasper NP  Columbia Icefield 83C/16  52°10'50" 117°04'50"  2200  26  Scout Mtn  Cathedral PP  Ashnola R 92H/1  49°04'40" 120°11'30"  2220  27  Blackwall Peak  Manning PP  Manning Park 92H/2  49°05'35" 120°45'35"  2000  28  Thynne Mtn  Merritt  Tulameen 92H/10  49°42'25" 120°55'50"  1940  29  McBride Mtn  McBride  McBride 93H/8  53°15'00" 120°14'45"  1970  101  Figure A.3.1. Map of sampling locations. Population numbers as in Table A.3.1.  102  A P P E N D I X IV-Zymograms  A.GDH  Figure A.4.1. Zymograms. Zymograms of alleles detected and scored from bud tissue using isozyme analysis. Numbers to the left refer to locus number; numbers atop each banding pattern refer to alleles, or where there are multip le numbers, patterns representing putative combinations of alleles in diploid bud tissue. Not all enzymes or loci depicted here were used in this study due to inconsistent staining, but are included to aid in interpretation in future studies. All loci whi ch appeared with some consistency are depicted on the zymogram to facilitate interpretation by other researchers; the thickness of the line indicates strength of banding across multiple runs. Sod, which had a negative staining pattern, requiring  104  APPENDIX V  - T a b l e s of genetic diversity a n d Wright's F-statistics s u p p l e m e n t a l to the text  Statistics for M a n n i n g a n d Baldy are p r e s e n t e d s e p a r a t e l y as they w e r e a s s a y e d using s e e d s , a n d u s e d different buffers, a slightly different set of loci, h a d a s a m p l e size up to 3 0 t i m e s that of the other populations. Table A.5.1. Summary of genetic parameters by locus; standard errors of the mean in parentheses. M = Manning, B = Baldy, C = the two populations combined, A = alleles per locus (no criterion), H„ = expected heterozygosity, H„ = observed heterozygosity. Locus  A B  M Pgil Pgi2 Idh Pgm 6Pg1 6Pg2 Mdh1 Mdh2 Mdh3 Mdh4  Mean  C  H, B  M  H B  0  C  M  C  2 2 2 0.365 0.255 0.312 0.400 0.233 0.345 2 2 2 0.180 0.278 0.238 0.200 0.333 0.273 3 3 3 0.077 0.182 0.255 0.080 0.233 0.164 3 3 3 0.077 0.064 0.071 0.000 0.067 0.036 3 3 3 0.320 0.326 0.326 0.240 0.367 0.309 3 2 • 3 0.077 0.000 0.036 0.080 0.000 0.036 2 2 2 0.000 0.000 0.000 0.000 0.000 0.000 3 3 3 0.420 0.486 0.467 0.280 0.500 0.400 2 2 2 0.039 0.033 0.036 0.040 0.033 0.036 3 3 3 0.449 0.444 0.451 0.520 0.600 0.564 2.6 2.5 2.6 0.204 0.218 0.212 0.184 0.243 0.216 (0.16) (0.17) (0.16) (0.055) (0.058) (0.055) (0.056) (0.068) (0.060)  Table A.5.2. Summary of Wright's F-statistics; standard errors of the mean in parentheses. F = Wright's fixation index, F = the reduction in H„ of inbred individuals within subpopulations, F the reduction of H„ of inbred individuals over all populations, F = the degree of population subdivision. ls  IT  ST  Locus Pgil Pg\2 Idh Pgm 6Pg1 6Pg2 Mdh1 Mdh2 Mdh3 Mdh4  Mean  M -0.096 -0.111 -0.042 1.000 0.250 -0.042 -  0.333 -0.020 -0.159 0.124 (0.117)  F B -0.176 -0.200 -0.085 -0.034 -0.124 -  C -0.136 -0.156 -0.064 0.483 0.063 -0.042  -  -  -0.029 -0.017 -0.350 -0.127 (0.040)  0.152 -0.019 -0.255 0.003 (0.072)  .  106  F C -0.115 -0.151 0.078 0.495 0.063 -0.023 0.000 0.141 -0.000 -0.245 -0.025 (0.025) IS  F C 0.010 0.003 0.046 -0.028 -0.019 0.026 0.000 0.008 -0.018 -0.014 0.008 (0.002) IT  F  ST  c  -0.104 -0.147 0.121 0.481 0.045 0.004 0.000 0.148 -0.019 -0.263 -0.024 (0.027)  Table A.5.3. Genetic diversity statistics for the other 17 populations combined by locus; standard errors of the mean in parentheses. For H , corrections for small sample size were included where applicable following Levene (1949) which is equivalent to Nei's (1978) unbiased estimate of H„. Locus Sample size F H H Mdh1 804 0.012 0.027 0.541 Mdh2 868 0.355 0.390 0.090 Mdh3 900 0.022 0.035 0.368 Pgm 594 0.599 0.481 -0.246 Skd1 872 0.115 0.387 0.703 Skd2 870 0.481 0.478 -0.005 Fdp 702 0.000' 0.039 1.000 Gdh 812 0.106 0.146 0.274 Lap1 746 0.011 0.179 0.940 Lap2 688 0.102 0.416 0.755 Idh 790 0.127 0.237 0.466 652 0.730 0.581 Pgi2 -0.258 774 0.221 0.283 0.386 Mean (28.4) (0.073) (0.057) (0.124) u  0  107  u  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items