Conservation Genetics 4: 581–593, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.
581
Inbreeding and conservation genetics in whitebark pine J. Krakowski∗ , S.N. Aitken & Y.A. El-Kassaby Department of Forest Sciences, 3041-2424 Main Mall, University of British Columbia, Vancouver, B.C. V6T 1Z4, Canada (∗ Corresponding author: Phone: (604)822-1821; Fax: (604)822-9102; E-mail:
[email protected]) Received 29 August 2002; accepted 14 October 2002
Key words: mating system, Nucifragia columbiana, Pinus albicaulis, population structure
Abstract Whitebark pine (Pinus albicaulis Engelm.) is threatened across its native range by an exotic fungal pathogen introduced within the last century. Mortality has been extensive, and projected potential range shifts based on impending climate change have revealed further pressures to survival and adaptation for this long-lived, high-elevation conifer. Quantifying genetic variation and the mating system of whitebark pine in its northern range provides a basis for effective conservation measures. Isozyme analysis of vegetative bud tissue revealed high expected heterozygosity (0.262), moderate population differentiation (FST = 0.061) and highly significant correlations between observed heterozygosity and geographic variables (R2 = 0.36, latitude; R2 = 0.30 longitude), supporting the hypothesis that this species recolonized its current northern range following glacial retreat from several refugia in the Washington and Oregon Cascades and in the northern Rockies. Mating system analysis based on simultaneous isozyme analyses of embryo and haploid megagametophyte tissues found relatively high levels of consanguineous mating and selfing for a conifer (tm = 0.73) within populations. Avian seed distribution by the Clark’s nutcracker (Nucifragia columbiana Wilson) appears to be the overriding factor influencing genetic patterns: being a mutualistic seed disperser, caches comprised of related seeds develop into clumped stands with strong family substructure. While it is a critical wildlife habitat component, lack of commercial utilization has made in situ adaptation the primary conservation focus. Encouraging regeneration success and nutcracker caching by maintaining natural fire regimes will provide an ecosystem-based conservation solution; however, in the Rocky Mountains between 52◦ N and 47◦ N, disease-resistant individuals should be located and propagated in order to ensure long-term survival of the species in high pathogen hazard areas.
Introduction Whitebark pine (Pinus albicaulis Engelm.) is a keystone species of many high elevation ecosystems in western North America. The seeds are an important food source for animals, and it plays a role in soil stabilization, regulation of snow melt and initiation of high-elevation forest succession (Arno and Hoff 1990). It grows from the subalpine to timberline from central British Columbia and Alberta south to the Sierra Nevadas, from 55◦ N to 37◦ N, along the Cascade and Coast ranges and the Rocky Mountains. There are eastern and western subdivisions, separated at the closest point (in southern British Columbia) by 0.95. Berg and Hamrick (1997) advocate the use of no percentage criterion (i.e. not to use an arbitrary allele frequency cutoff limit for inclusion of loci in summary statistics, commonly employed by genetic data analysis software, e.g., 0.95 or 0.99) when calculating genetic diversity statistics in order to avoid artificially low allelic richness estimates. Population genetic parameters for the two populations selected for mating system analysis are also included, although results were obtained for a different but overlapping set of loci (Table 1).
Yalakom, 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 (Table 1). The highest allelic diversities were 2.6, in Manning, and 2.5 in Mount Baldy from 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, and the latter an outlying population further to the east. Yalakom had the lowest proportion of polymorphic loci (P) at 50%, Edith was the next lowest at 58%. Paget in the northern Rockies and Perkins in the north Coast Mountains, shared the highest value of P at 92%, while Manning had 90% (Table 1).
585 Table 1. Genetic diversity statistics for all populations. Standard errors of the mean in parentheses Population
Pop #
N
A
P
Ho
He
Hudson Higgins Sweeney Heckman Perkins Tchaikazan Yalakom D’arcy Van Horlick Whistler Lime Northern Washington Kootenay Jumbo Stanley Paget Edith Manning Mount Baldy Mean
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
18 (0.6) 18 (2.4) 20 (1.5) 26 (2.0) 24 (2.2) 24 (2.2) 23 (1.6) 24 (0.8) 21 (1.7) 30 (0.4) 30 (0.0) 17 (0.0) 20 (1.3) 31 (1.5) 29 (0.4) 28 (0.9) 16 (1.5) 25 (0.0) 30 (0.0) 23∗ (1.3)∗
1.8 (0.2) 1.9 (0.1) 1.8 (0.2) 1.8 (0.2) 2.0 (0.1) 1.8 (0.2) 1.6 (0.2) 1.8 (0.1) 1.8 (0.2) 1.9 (0.2) 2.0 (0.2) 1.8 (0.2) 1.9 (0.1) 2.0 (0.2) 2.1 (0.2) 2.2 (0.2) 1.7 (0.2) 2.6 (0.2) 2.5 (0.2) 1.9 (0.1)
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)
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)
0.237 (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)
N = mean individual trees analyzed; A = mean alleles per locus, no criterion; P = percentage of polymorphic loci, no criterion; Ho = observed heterozygosity; He = expected heterozygosity. Manning and Mount Baldy statistics based on ten loci. ∗ Not including Manning and Mount Baldy.
Observed heterozygosity (Ho ) averaged 0.213; Hudson, in the northern Coast Mountains had the lowest value (0.119) and the southernmost population sampled, in northern Washington, had the highest (0.289). Other populations in the northwestern portion of the range also had low expected heterozygosity (He ), while those in the south were markedly higher (Table 1, Figure 1). Expected heterozygosity averaged 0.257 across all populations, with the minimum estimate at Yalakom in the southern Coast Mountains (0.194) and the maximum at Perkins (0.312). Populations in the Rocky Mountains did not exhibit clear trends with respect to heterozygosity. Out of 19 populations, He was greater than Ho in three: northern Washington, Paget and Mount Baldy, and the differences between Ho and He in these three were statistically non-significant (p < 0.05). In most cases, He was greater than Ho by at least 50%, and sometimes by well over 100%. The strongest heterozygote deficiencies tended to be at the northwest margin of the species range (with the exception of the Tchaikazan population).
Population structure The mean overall FIS of 0.345 reflects pronounced inbreeding within populations, with wide variability among loci (Table 2). Nine loci had positive values of FIS , many with values > 0.5, indicating significantly lower observed heterozygosity (relative to that expected under random mating) for trees within each subpopulation at these loci (Table 2). The overall mean FST was 0.061, corresponding to Wright’s (1931) category of moderately differentiated populations. Geographic patterns Estimates of expected heterozygosity (He ) were calculated for regional subdivisions, but none of the subdivisions tested differed significantly (95% C.I.). With stratification into northern and southern populations, the southern group (populations 7–14, 18 and 19) averaged 0.254, and the northern group (1–6 and 15–17) averaged 0.260. When stratified into eastern and western regions, the Coast Mountain populations (1–12, 18 and 19) and the Rocky Mountain populations (13–17) both had an average He of 0.257.
586 Table 2. Wright’s F-statistics for 17 populations assessed for genetic diversity∗ ; S.E. is standard error of the mean. Ho and He as in Table 1∗ Allele
FIS
FIT
FST
Ho
He
Mdh1 Mdh2 Mdh3 Pgm Skdh1 Skdh2 Fdp Gdh Lap1 Lap2 Idh Pgi2 Mean (S.E.)
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)
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)
0.004 0.006 0.031 0.113 0.023 0.013 0.071 0.105 0.077 0.067 0.123 0.093 0.061 (0.012)
0.013 0.354 0.020 0.550 0.112 0.470 0.000 0.092 0.009 0.092 0.132 0.706 0.213 (0.027)
0.025 0.385 0.030 0.421 0.372 0.470 0.072 0.115 0.153 0.356 0.213 0.537 0.262 (0.026)
∗ Not including Manning and Mount Baldy.
Populations in the northwestern portion of the range had relatively low He , while those in the south were higher (Table 1). Regional trends were stronger for Ho than for He , although differences were not statistically significant. Average Ho was higher in the southern populations (0.223) than in the north (0.202) and higher in the Rocky Mountains (0.235) than in the Coast Mountains (0.205). While there were north-south differences in the Coast mountains, the numbers of populations sampled within the Rocky Mountains were not sufficient to observe any trends with respect to Ho . He was slightly less variable than Ho . Wright’s inbreeding coefficient F (where F = 1 − Ho /He ) was always positive, indicating systematic homozygote excess compared to expectations. It was highest (0.225) in the northern populations, and lowest (0.084) in the five Rocky Mountain populations. There was a heterozygote deficiency of 0.204 (20%) in the Coast Mountains and 0.124 (12%) in southern populations. Observed heterozygosity was strongly correlated with both latitude and longitude (Figures 2a, b). Correlations with Ho were significant (p < 0.05), and showed the strongest trends (for latitude: R2 = 0.36, p = 0.01; for longitude: R2 = 0.30, p = 0.02). Regressions of latitude and longitude on He were not significant. Observed heterozygosity generally increased towards the south and east, as did the deficiency of heterozygotes as measured by Wright’s inbreeding coefficient (Figures 2c, d).
Nei’s (1972) genetic distance was not highly informative since populations were too genetically similar to accurately gauge relative genetic distances: many pairs of populations had Nei’s distances of 0.000, very few were >0.1, thus Cavalli-Sforza and Edward’s (1967) chord distance provided a higher resolution for examining relationships among populations. The minimum chord distances were between Tchaikazan and northern Washington, which are in the Coast Mountains 370 km apart; and Perkins and Paget, located in different mountain ranges. The maximum chord distance was between Heckman and D’arcy. Mean chord distance among all populations was 0.167. Dendrograms based on the iterative Wagner distance procedure using Cavalli-Sforza and Edwards’ (1967) chord distance revealed only slight correlation between spatial and genetic distances (not shown). Population groupings, especially among mountain ranges, were variable, as were positions and branch lengths of most populations. Physical and chord distances were compared with a nonparametric Mantel test (Jorgensen and Hamrick 1997). Results from 10,000 random permutations were significant (p = 0.05, adjusted for the number of pairwise tests), indicating that physical and genetic distances are correlated in this species, at least in the northern portion of its range.
587
Figure 2. Statistically significant regressions (p < 0.05) of geographical variables with respect to genetic diversity parameters. (a) Longitude on Ho ; (b) latitude on Ho ; (c) longitude on heterozygote deficiency; (d) latitude on heterozygote deficiency.
Table 3. Population-level estimates of tm and ts (multilocus and single-locus outcrossing rates, respectively) by locus; rp is correlation of paternity among progeny; rt is correlation of outcrossing rates among progeny arrays, averaged for all parent trees sampled. Standard errors of the mean in parentheses Locus
Mount Baldy
Manning
Pgi1 Pgi2 Pgm Idh Mdh2 Mdh3 Mdh4 6Pg1 6Pg2∗
0.762 (0.109) 0.888 (0.083) 0.396 (0.282) 0.621 (0.085) 0.646 (0.089) 1.319 (0.952) 0.897 (0.107) 0.913 (0.114) –
0.493 (0.230) 0.777 (0.150) 0.123 (0.057) 0.709 (0.155) 0.758 (0.055) 0.614 (0.385) 0.759 (0.069) 0.684 (0.160) 0.294 (0.237)
Combined SL Combined ML tm − ts rt rp No. of families No. of observations
0.735 (0.048) 0.736 (0.042) 0.001 (0.014) 0.082 (0.052) 0.208 (0.082) 30 853
0.650 (0.061) 0.722 (0.054) 0.068 (0.025) 0.074 (0.046) 0.148 (0.063) 25 750
∗ 6Pg2 locus monomorphic at Mount Baldy.
Mating system At Mount Baldy, mean single-locus (ts ) and multilocus (tm ) outcrossing estimates were nearly identical (0.735 and 0.736, respectively), while for Manning they were 0.650 and 0.722, respectively (Table 3). ts and tm were significantly different in the Manning population (p< 0.05). This difference implies inbreeding is predominantly biparental (i.e., parents are distinct individuals but are closely related to each other); whereas at Mount Baldy inbreeding reflects more self-pollination rather than consanguineous matings (i.e., among relatives). There was a wide range of multilocus outcrossing rates among individual trees (Figures 3a, b). We expected that that the majority of trees would show high outcrossing rates, producting a “j”-like pattern like many other temperate conifers (Perry and Dancik 1985; El-Kassaby et al. 1993), but individual tree outcrossing rates were represented by all classes, with values ranging from self-pollination to complete outcrossing. Furthermore, there was a conspicuous gap in the distribution for both populations in the 0.90–0.99 category (Figures 3a, b). Mount Baldy had slightly higher rt (the correlation between intrapopulation outcrossing rates), although populations were not significantly different (p = 0.3). For Mount Baldy, rp (the likelihood that two randomly
588
Figure 3. Single-tree multilocus outcrossing rate (tm ) distributions for (a) Manning and (b) Mount Baldy.
selected progeny are full sibs) was 0.21, and for Manning 0.15, supporting strongly structured populations comprised of individuals sharing at least one parent or grandparent (Table 3).
Discussion Genetic diversity The somewhat low allelic richness of whitebark pine (relative to other estimates for whitebark pine, Table 4) may reflect repeated founder effects from avian seed caching and subsequent relocation of cached seeds, whereas the high heterozygosity (compared to other studies of whitebark pine and conifers in general, see Table 4 and Hamrick et al. 1992) could result from founding events from multiple populations within the same area (Richardson et al. 2002; Comps et al. 2001). Overall, these findings corroborate the results of both Stuart-Smith (1998) and Yandell (1992), but Rogers et al. (1999), Bruederle et al. (1998) and Jorgensen and Hamrick (1997) found lower He estimates (0.3 for most loci) using neutral markers (e.g., isozymes) is the occurrence of selfing and of crosses between relatives, although Cuguen et al. (1988) also suggest such results may arise from limited intrapopulation gene flow, leading to neighbourhoods within populations and not strictly inbreeding. Geographic patterns and groups The significant but weak correlation observed between physical and genetic distances supports the hypothesis of a stepping-stone repopulation model via founder events from glacial refugia. Very little, if any, cross-Cordilleran migration would have occurred: the closest distance between known populations in the Coast and Rocky Mountain ranges is still much greater than the furthest recorded nutcracker caching distance (22 km, cf. Vander Wall and Balda 1977), even taking
589 into account the potential for relocation of cached seeds by birds and other animals. Genetic diversity patterns found in this study confirm reconstructed biogeographic patterns of postglacial recolonization (Richardson et al. 2002), reflecting recent founder effects. The higher heterozygosity in the south and east suggests that bird-mediated seed dispersal generally progressed northwestwards, and that there were more refugia in the east, or possibly refugia in the Rockies at more northerly latitudes in the Rocky Mountains (but see Widmer and Lexer 2001). There is an overall geographic relationship between blister rust mortality trends (highest in the south and east, decreasing to the north and west) (Stuart-Smith 1998; Campbell and Antos 2000) and Wright’s index of inbreeding, F (Figures 2a-d), with regions with the highest mortality having a greater heterozygote excess, and those where the pathogen is absent or has less impact tending towards heterozygote deficiency. This trend, also found by Stuart-Smith (1998), provides circumstantial support for selection by the pathogen against homozygous genotypes or for increased tolerance of heterozygote genotypes (Bush and Smouse 1992). A causal basis for this apparent correlation between genomic heterozygosity (reflecting inbreeding) and degree of blister rust infection could be verified by investigating this relationship on an individual-tree basis. Mating system Some maternal parents produce seed through pronounced inbreeding, including selfing, while others appear to be primarily outcrossing. The correlation between outcrossing rates of parents and offspring was small, implying within-family variation in this trait (Table 3). If selection, environment and their interaction had similar effects across families and loci, outcrossing rates would be normally distributed; instead, the individual trees appear to vary widely in this characteristic, with some trees being primarily outcrossed and others having high rates of selfing or biparental inbreeding. Outcrossing rates are potentially influenced by many genes that directly and indirectly affect the mating system, leading to a continuous, rather than a discrete, distribution. These genes may impact factors such as male and female fecundity and fertility, preand post-zygotic barriers to fertilization (especially for inbred individuals), and reproductive phenology. In both populations where mating system was studied,
samples had been collected over a wide area of several hectares. The high range of outcrossing values detected among trees reveals that family structuring and individual-tree variation in t are characteristic features. The numerous factors affecting the life history of whitebark pine interact across many spatial scales and exert contradictory influences at different life stages (Bruederle et al. 1998). These factors, which are all affected by selection, include pollen and seed dispersal, and successional stage. The isolated nature of the Mount Baldy population may be reflected in adaptation to higher selfing tolerance levels. As a consequence of the small population size, individuals are more closely related and likely to have more synchronous phenology, which could lead to reduced pollen flow from outside sources. This may lead to a higher tolerance for selfing. The higher paternity correlation (rp ) also reflects a slightly more closely related population in Mount Baldy than in Manning. In other conifers, differences in mating system parameters and selfing tolerance have been found not only among individual seed parents and populations, but also within individuals over time or even among crown strata (El-Kassaby 1984; El-Kassaby et al. 1987, 1993). However, crown strata is unlikely to be a major cause of variation with whitebark pine as cones are only produced at the top of the crown. Mating system differentiation can therefore operate at very fine scales and be exerted by many environmental and genetic factors. Whitebark pine could manifest similar variation in selfing tolerance, both at the individual and population levels. Since it was not possible to determine the paternity of aborted seeds in this study, our estimates may thus be lower than the actual inbreeding rate at the time of fertilization. Coevolution of Nucifragia spp. and stone pines may have resulted in different patterns of mating than in pines with wind-dispersed seeds (Table 4). The caching of groups of related individuals and their subsequent germination and synchronized phenology, combined with highly variable, often low tree densities, increase the opportunities for self-pollination and consanguineous mating. This would account for the relatively high inbreeding found in this study, but not for the wide range of outcrossing estimates (t) for individual families. This wide range of outcrossing rates may reflect variability in the amount of relatedness in bird or mammal seed caches, resulting in varied stand structures with individuals arising from different mating opportunities within the same area.
590 Table 4. Outcrossing data for stone pines (subsection Cembrae) and other pines. Standard errors of the mean in parentheses Taxonomic subsection
Species
tm
ts
Cembrae Cemtrae Cembrae Cembrae Ponderosae Ponderosae Contortae Strobi Cembroides
Pinus albicaulis1 P. cembra2 P. koraiensis2 P. sibirica2 P. ponderosa3 P. jeffreyi4 P. contorta5 P. monticola6 P. maximartinezii7
0.693 (0.055) 0.707 (0.045) 0.936 (0.051) 0.862 (0.054) 0.933 (0.052) 0.911 (0.081) 0.974 (0.016) 0.925 (0.056) 0.816 0.761
0.729 (0.048) 0.686 (0.025) 0.974 (0.058) 0.894 (0.057) 0.960 (0.030) 0.935 (0.021) 0.926 (0.034) 0.977 (0.023)
1 This study; 2 Krutovskii et al. 1995; 3 Mitton et al. 1981; 4 Furnier and Adams 1986; 5 Perry and Dancik 1985; 6 ElKassaby et al. 1987; 7 Ledig et al. 1999.
The relatively low number of generations since glaciation may account for the persistence of families with intermediate outcrossing rates. This large range of outcrossing rates may result from the recolonization of the species’ range in British Columbia’s southern interior by two differentiated populations that are now introgressed, one predominantly outcrossing and the other primarily selfing; high selfing rates could indeed evolve in small refugial populations where purging of recessive lethal or deleterious alleles may have occurred. This could also explain the currently high within-population levels of variation, but will require further evidence before confirmation. Conservation strategy for BC Up to 100% mortality is anticipated for some whitebark pine stands in southeastern British Colombia due to white pine blister rust (Campbell and Antos 2000). Since trees can only expand their range through seed dispersal, whitebark pine’s long generation time, combined with the need to develop cost-effective conservation measures for this non-timber species, makes in situ adaptation to climate change a critical component of any conservation strategy. Targeting genetically diverse and unique populations is the most efficient use of resources. A target effective population size (Ne ) can be a surrogate for the proportion of genetic variability captured in a population since He is approximately proportional to 1/(2Ne ). Widespread, contiguous populations of whitebark pine currently exist in B.C. and adjacent areas
throughout the Coast Mountains and the Northern Rocky Mountains. Although only some populations are protected in parks and reserves, most populations are situated in remote, rugged terrain not subject to immediate human impact, and provide sufficient numbers to meet estimated targets of 5000 census individuals to maintain a sustainable Ne of over 1000 (Yanchuk 2001). Even if the wide range of outcrossing rates found in this study is taken into account, necessitating a slightly higher census population size to achieve the same Ne (e.g., >5000), most populations in these regions should contain enough individuals to fulfil these recommendations without further active measures. Southern Rocky Mountain populations are at a greater immediate risk of mortality from blister rust. Rust-resistant individuals, as opposed to populations, must be targeted for seed collection and subsequent propagation and planting in these high-risk areas, many of which currently have little or no natural regeneration. A key goal of whitebark pine conservation should be to capture a sufficient percentage of disease resistance genes or genotypes (a value of one to five percent was suggested by Hoff et al. 2001). The positive geographic correlation observed between heterozygote deficiency and white pine blister rust intensity suggests, but does not establish, a relationship between inbreeding level and ability to survive in high disease risk areas. Inbreeding depression has not been quantified for this species, but is typically high for conifers (Kormutak and Lindgren 1996). Further investigation on the role of inbreeding in blister rust resistance should be a high priority. It may be possible to enhance the health and survival rate of planted stands by locally augmenting outcrossing rates in areas targeted for seed collection, e.g., through selecting high density whitebark stands with few multistem clusters of trees for seed production, or by reducing some multistem clusters to one vigorous, healthy stem before collecting. A regime of controlled burns and follow-up monitoring has been instituted in selected areas, including Canada’s Rocky Mountain National Parks, to restore whitebark pine habitat and enhance seedling survival (Wilson and Stuart-Smith 2002). Lightningcaused wildfires should not be suppressed in remote areas where human habitation and other activities would not be at risk. Nutcrackers have been frequently observed caching pine seeds in recently burned areas, which provide ideal competition-free mineral soil seedbeds for germination; this would help main-
591 tain more genotypes as more cached seeds would germinate. Where a phenotypically resistant genotype has been located, fire should be limited or suppressed until cones have been collected. This study has not addressed population variation for adaptive traits in this species, a standard prerequisite for decisions on seed movement for reforestation or restoration, as genetic markers are often poor indicators of genetic structure for adaptive traits (McKay and Latta 2002). A large, rangewide genecological study is currently underway that will provide information on the degree of local adaptation in this species (A.D. Bower, University of British Columbia, pers. comm.). Results from the ongoing genecological study, in combination with the FST values estimated in the present study, will aid in developing a deployment strategy for seed collected from putatively resistant phenotypes. Depending on the results of the genecological study, it may be expedient to mitigate the effects of climate change anticipated by facilitating a rapid, northward range shift through planting of seedlings several hundred kilometres north of their current provenance. Determining the appropriate scale for such seed movements and source populations will depend upon patterns of adaptive variation, in conjunction with results from the current study. It was found that while trees growing further north were somewhat less genetically diverse (in terms of Ho ) than those in the south, a substantial amount of genic diversity was captured within individual populations (i.e., high He combined with low to moderate FST ). The short-term impacts of a pre-emptive range shift northward on suboptimally-adapted individuals may be compensated for in the long term by increased survival at the regional level, and by mitigating the lower observed heterozygosity and higher inbreeding found in bird-cached populations as a consequence of founder effects.
Acknowledgements Support for this study was generously provided by Forest Renewal B.C. and the Centre for Forest Gene Conservation. M.G. Meagher and D.G. Edwards collected seed; P. Heppner, C. Stephenson, D. Thorburn and G. Davidson assisted in field collections; S. Zeglen, E.M. Campbell, and D. Kolotelo also provided valuable asssistance. S. Barnes analyzed seeds in the laboratory; Dr. C. Ritland provided labora-
tory support. We thank two anonymous reviewers for their helpful comments.
References Arno SF, Hoff RJ (1990) Pinus albicaulis Engelm. In: Silvics of North America, Vol. 1. Conifers (tech. co-ords. Burns RM, Honkala BH), pp. 268–279. Agriculture Handbook 654, USDA For. Serv., Washington D.C. Baker BH, Amman GD, Trostle GC (1971) Does the mountain pine beetle change hosts in mixed lodgepole and whitebark pine stands? USDA For. Serv. Res. Note, INT-151. Ogden, UT. Berg EE, Hamrick JL (1997) Quantification of genetic diversity at allozyme loci. Can. J. For. Res., 27, 415–424. Bruederle LP, Tomback DF, Kelly KK, Hardwick RC (1998) Population genetic structure in a bird-dispersed pine, Pinus albicaulis (Pinaceae). Can. J. Bot., 76, 83–90. Burdon JJ, Thrall PH, Brown AHD (1999) Resistance and virulence structure in two Linum marginale-Melampsora lini host-pathogen populations with different mating systems. Evol., 53, 704–716. Bush RM, Smouse PE (1992) Evidence for the adaptive significance of allozymes in forest trees. New For., 6, 179–196. Callaway RM (1998) Competition and facilitation on elevation gradients in subalpine forests of the northern Rocky Mountains, USA. Oikos, 82, 561–573. Campbell EM, Antos JA (2000) Distribution and severity of white pine blister rust and mountain pine beetle on whitebark pine in British Columbia. Can. J. For. Res., 30, 1051–1059. Cavalli-Sforza JL, Edwards DW (1967) Phylogenetic analysis: Models and estimation procedures. Evol., 21, 550–570. Clayton JB, Tretiak EM (1972) Differences in allozyme variation in Atlantic herring. J. Fish Res. Board Can., 35, 1167–1170. Comps B, Gömöry D, Letouzey J, Thiébaut B, Petit RJ (2001) Diverging trends between heterozygosity and allelic richness during postglacial colonization in the European beech. Genetics, 157, 389–397. Conkle MT (1971) Isozyme specificity during germination and early growth of knobcone pine. For. Sci., 17, 494–498. Critchfield WB (1986) Hybridization and classification of the white pines (Pinus section Strobus). Taxon, 35, 647–656. Cuguen J, Merzeau D, Thiébaut B (1988) Genetic structure of the European beech stands (Fagus sylvatica L.): F-statistics and importance of mating system characteristics in their evolution. Hered., 60, 91–100. Delcourt PA, Delcourt HR (1998) Paleoecological insights on conservation of biodiversity: A focus on species, ecosystems and landscapes. Ecol. Appl., 8, 921–934. El-Kassaby YA (1984) Variation in the outcrossing rate among crown strata of Sitka spruce. Silvae Gen., 32, 232–237. El-Kassaby YA, Meagher MD, Parkinson J, Portlock FT (1987) Allozyme inheritance, heterozygosity and outcrossing rate among Pinus monticola near Ladysmith, British Columbia. Hered., 58, 173–181. El-Kassaby YA, Meagher MD, Davidson R (1993) Temporal variation in the outcrossing rate in a natural stand of western white pine. Silvae Gen., 42, 131–135. Furnier GR, Adams WT (1986) Mating system in natural populations of Jeffrey pine. Am. J. Bot., 73, 1002–1008. Hamrick JL, Godt M-JW, Sherman-Broyles SL (1992) Factors influencing levels of genetic diversity in woody plant species. New For., 6, 95–124.
592 Hoff RJ, Hagle SK, Krebill RG (1994) Genetic consequences and research challenges of blister rust in whitebark pine forests. In: Proceedings of an International Workshop on Subalpine Stone Pines and Their Environment: The Status of Our Knowledge, pp. 118–126, 21. USDA For. Serv. Tech. Rep. INT-184. Ogden, UT. Hoff RJ, Ferguson DE, McDonald GI, Keane RE (2001) Strategies for managing whitebark pine blister rust. In: Whitebark Pine Communities: Ecology and Restoration (eds. Tomback DF, Arno SF, Keane RE), pp. 346–366. Island Press, Washington D.C. Huntley B (1991) How plants respond to climate change: Migration rates, individualism and the consequences for plant communities. Ann. Bot., 67(Supp. 1), 15–22. Intergovernmental Panel on Climate Change (2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability. Jorgensen SM, Hamrick JL (1997) Biogeography and population genetics of whitebark pine, Pinus albicaulis. Can. J. For. Res., 27, 1574–1585. Keane RE, Arno SF (1993) Rapid decline of whitebark pine in western Montana: Evidence from 20-year remeasurements. W. J. Appl. For., 8, 44–47. Kendall KC, Arno SF (1992) Whitebark pine – an important but endangered wildlife resource. In: Proceedings of a Symposium on Whitebark Pine Ecosystems: Ecology and Management of a High-Mountain Resource, pp. 264–273. USDA For. Serv. Tech. Rep. INT-176, Bozeman, MT. Kormuták A, Lindgren D (1996) Mating system and empty seeds in silver fir (Abies alba Mill.). For. Genet., 3, 231–235. Krutovskii KV, Politov DV, Alkhutov YP (1995) Isozyme study of population genetic structure, mating system and phylogenetic relationships of the five stone pine species (subsection Cembrae, section Strobi, subgenus Strobus). In: Population Genetics and Conservation of Forest Trees (eds. Baradat P, Adams WT, Müller-Starck G), pp. 270–304. SPB Acad. Pub., the Netherlands. Lanner RM (1982) Adaptations of whitebark pine for seed dispersal by Clark’s nutcracker. Can. J. For. Res., 12, 391–402. Leadam CL (1986) Seed dormancy in three Pinus species of the inland mountain west. In: Proceedings: Conifer Tree Seed in the Inland Mountain West Symposium, pp. 117–124. USDA For. Serv. Res. Rep. 230, Missoula, MT. Ledig FT (1988) The conservation of diversity in forest trees: Why and how should genes be conserved? BioScience, 38, 471–479. Ledig FT, Conkle MT, Bermejo-Velásquez B, Eguiluz-Piedra T, Hodgskiss PD, Johnson DR, Dvorak WS (1999) Evidence for an extreme bottleneck in a rare Mexican pinyon: Genetic diversity, disequilibrium, and the mating system in Pinus maximartinezii. Evol., 53, 1–99. Liedloff RT (1999) Mantel Nonparametric Test Calculator V.2.00© http://www.mantel.com Mattson DJ, Reinhart DP (1997) Excavation of red squirrel middens by grizzly bears in the whitebark pine zone. J. Appl. Ecol., 34, 926–940. McCaughey WW, Schmidt WC (2001) Taxonomy, distribution and history. In: Whitebark Pine Communities: Ecology and Restoration (eds. Tomback DF, Arno SF, Keane RE), pp. 29–40. Island Press, Washington D.C. McKay JK, Latta RG (2002) Adaptive population divergence: Markers, QTL and traits. Trends Ecol. Evol., 17, 285–291. Mitton JB, Linhart YH, Sturgeon SW, Farrar JP (1977) Protein polymorphisms detected in mature needle tissue of ponderosa pine. Am. J. Bot., 42, 1154–1157.
Mitton JB, Linhart YB, Davis ML, Sturgeon KB (1981) Estimation of outcrossing in ponderosa pine, Pinus ponderosa Laws., from patterns of segregation of protein polymorphisms and from frequencies of albino seedlings. Silvae Gen., 30, 117– 121. Mitton JK, Kreiser BR, Latta RG (2000) Glacial refugia of limber pine (Pinus flexilis James) inferred from the population structure of mitochondrial DNA. Mol. Ecol., 9, 91–97. Nei M (1972) Genetic distance between populations. Am. Nat., 106, 283–291. Ohta T (1982) Linkage disequilibrium with the island model. Genetics, 101, 139–155. Perkins DL, Swetnam TW (1996) A dendroecological assessment of whitebark pine in the Sawtooth – Salmon River region, Idaho. Can. J. For. Res., 26, 2123–2133. Perry DJ, Dancik BP (1985) Mating system dynamics of lodgepole pine in Alberta, Canada. Silvae Gen., 35, 190–195. Petit RJ, Bahrman N, Baradat PH (1995) Comparison of genetic differentiation in maritime pine (Pinus pinaster Ait.) estimated using isozyme, total protein and terpenic loci. Hered., 75, 382– 389. Pitel JA, Wang BSP (1992) Physical and chemical treatments to improve germination of whitebark pine seed. In: Proceedings of a Symposium on Whitebark pine Ecosystems: Ecology and Management of a High-Mountain Resource, pp. 130–133. USDA For. Serv. Bozeman, MT. Price RA, Liston A, Strauss SH (1998) Phylogeny and systematics of Pinus. In: Ecology and Biogeography of Pinus (ed. Richardson DM), pp. 49–68. Cambridge Univ. Press, Cambridge, MA. Richardson BA, Brunsfeld SJ, Klopfenstein NB (2002) DNA from bird-dispersed seed and wind-disseminated pollen provides insights into postglacial colonization and population genetic structure of whitebark pine (Pinus albicaulis). Mol. Ecol., 11, 215–227. Ridgeway FT, Conway SC, Maples M (1970) Polymorphic protein detected in brook trout. J. Fish., 42, 166–171. Ritland K (1990) A series of FORTRAN computer programs for estimating plant mating systems. J. Hered., 81, 235–237. Rogers DL, Millar CI, Westfall RD (1999) Fine-scale genetic structure of whitebark pine (Pinus albicaulis): Associations with watershed and growth form. Evolution, 53, 74–90. Shafer SL, Bartlein PJ, Thompson RS (2001) Potential changes in the distributions of western North America tree and shrub taxa under future climate scenarios. Ecosystems, 4, 200–215. Stuart-Smith GJ (1998) Conservation of Whitebark Pine in the Canadian Rockies: Blister Rust and Population Genetics. MSc Thesis, University of Alberta, Edmonton, AB. Swofford DL, Selander RB, Black IV WC (1997) BIOSYS-2: A computer program for the analysis of allelic variation in genetics. Tomback DF (1982) Dispersal of whitebark pine seeds by Clark’s nutcracker: A mutualism hypothesis. J. Animal Ecol., 51, 451– 467. Tomback DF, Arno SF, Keane RE (2001) The compelling case for management intervention. In: Whitebark Pine Communities: Ecology and Restoration (eds. Tomback DF, Arno SF Keane RE), pp. 3–25. Island Press, Washington D.C. Vander Wall SB, Balda RP (1977) Coadaptations of the Clark’s nutcracker and the piñon pine for efficient seed harvest and dispersal. Ecol. Monogr., 4, 89–111. Wheeler NC, Guries RP (1982) Population structure, genic diversity, and morphological variation in Pinus contorta Dougl. Can. J. For. Res., 12, 595–606.
593 Widmer A, Lexer C (2001) Glacial refugia: sanctuaries for allelic richness, but not for gene diversity. Trends Ecol. Evol., 16, 267– 269. Wilson B, Stuart-Smith J (2002) Whitebark Pine Conservation for the Canadian Rocky Mountain National Parks. Cordilleran Ecological Research, Winlaw, B.C. Wright S (1931) Evolution in Mendelian populations. Genetics, 16, 97–159. Wright S (1965) The interpretation of population structure by Fstatistics with special regard to systems of mating. Evol., 19, 395–420.
Yanchuk AD (2001) A quantitative framework for breeding and conservation of forest tree genetic resources in British Columbia. Can. J. For. Res., 71, 566–576. Yandell UG (1992) An Allozyme Analysis of Whitebark Pine (Pinus albicaulis Engl.). MSc Thesis, University of Nevada, Reno, NV. Yeh FC, Yang R-C, Boyle T (1999) Popgene Version 3.2: Microsoft Window-Based Freeware for Population Genetic Analysis. Zeglen S (2002) Whitebark pine and white pine blister rust in British Columbia, Canada. Can. J. For. Res., 32, 1265–1274.
