| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Population Biology |
Institute of Forest Genetics, Pacific Southwest Research Station, USDA Forest Service, and Department of Plant Science, Mail Stop 6, University of California, One Shields Avenue, Davis, California 95616 USA
Received for publication January 19, 2005. Accepted for publication August 25, 2005.
ABSTRACT
Brewer spruce (Picea breweriana), a relict of the widespread Arcto-Tertiary forests, is now restricted to a highly fragmented range in the Klamath Region of California and Oregon. Expected heterozygosity for 26 isozyme loci, averaged over 10 populations, was 0.121. More notable than the relatively high level of diversity when compared to other woody endemics was the strong decrease in expected heterozygosity with latitude. Differentiation (FST) was 0.152, higher than values for many north temperate conifers with larger distributions. The number of migrants per generation (Nm) was 1.34 or 2.70, depending on the method of estimation. Inbreeding appeared low; FIS was only 0.003, in agreement with multilocus population outcrossing rates (tm), which were generally well above 0.90. No difference in tm was found between isolated vs. clustered trees. However, the number of seeds per cone was greatest in the densest populations; tm is a measure of effective outcrossing after mortality in the embryonic stage, whereas a reduced number of seeds per cone indicates self pollination. Selfing increased after logging; outcrossing rate before logging was 0.961 and after logging, 0.756. Despite Brewer spruce's narrow, fragmented distribution, the outlook for its conservation was good, with the exception of possible negative effects of logging.
Key Words: fragmentation • genetic structure • heterozygosity • isozymes • relict species • selfing
Brewer spruce (Picea breweriana Wats.) is a relict of the Arcto-Tertiary forest, most species of which are now extinct in western North America. According to the fossil record, Brewer spruce had a wide distribution in the Pliocene and Miocene, at least as far east as Idaho and Nevada, north to central Oregon, and south to central California (Wolfe, 1964
). The fossil species, Sonoma spruce (Picea sonomensis Axelrod), which is synonymous with Brewer spruce, occurred in the Creede Flora in the San Juan Mountains of southwestern Colorado in the Oligocene (Axelrod, 1987
).
The Arcto-Tertiary flora was reduced by increasingly dry climates in the interior West, especially as mountain building accelerated at the close of the Miocene. Cool moist forests shrunk toward the coast and higher elevations (Whittaker, 1961
). Brewer spruce is now endemic to the Klamath Geomorphological Province, which retains forests most nearly equivalent to the western North American Arcto-Tertiary forests (Whittaker, 1960
; Sawyer and Thornburgh, 1977
). Boundaries of the Klamath Province are defined in different ways depending on whether they are based on a strict geomorphological interpretation or weighted by ecological considerations (Fig. 1). The region generally includes the Siskiyou Mountains, Salmon Mountains, Marble Mountains, North Yolla Bolly Mountains, Scott Bar Mountains, and Trinity Alps, an area of about 50 300 km2 (Miles and Gouday, 1997
). The substrate is a complex of Paleozoic and Mesozoic formations distinct from the surrounding younger rocks of the Coast Ranges, Cascade Range, and Sacramento Valley (Whittaker, 1960
; Sawyer and Thornburgh, 1977
).
|
). Regions surrounding the Klamaths were submerged at various periods in their paleohistory, but the Klamaths escaped submergence. Neither were they extensively glaciated nor subject to the volcanism of the Cascade Range to the east (Whittaker, 1961
; Irwin, 1966
). Neighboring floras were decimated while the Klamath Geological Province, relatively stable for at least 100 million years (Myr), became a refuge for many species; i.e., about 3500 plants, of which 280 are endemic (Villa-Lobos, 2003
).
The Klamath Province is a meeting ground for species that are at the southern limit of their range and others that are at their northern limit, as well as species from the Cascades and the Sierra Nevada. The rugged terrain and complex geology and soils contribute to species richness (Whittaker, 1960
). Patches of serpentine soil, in particular, harbor many plants rare outside the Klamath region and may have provided stepping stones for Brewer spruce and the Klamath flora to migrate across the topography in response to climatic alterations (R. H. Waring, Oregon State University, personal communication). Although the regional climate is Mediterranean, a strong moisture and temperature gradient exists from the coast inland (Whittaker, 1960
; Waring, 1969
). Brewer spruce is excluded from the dryer, eastern portion of the Klamath Province because it is adapted to cool temperatures during the growing season and closes its stomata when evaporative demand is high, which puts it at a disadvantage against competitors (Waring et al., 1975
).
Brewer spruce covers a northwest–southeast range of 228 km from Iron Mountain, Oregon, to East Weaver Lake, California, nearly the entire north–south extent of the Klamath Province (Fig. 1). It occurs within 22 km of the Pacific Coast (Waring et al., 1975
), and an isolate near Castle Crags, California (Haddock, 1938
), is about 145 km inland. The northernmost population, at Iron Mountain, may be isolated from other Brewer spruce by as much as 29 km. The species ranges in elevation from 560 to 2300 m a.s.l. (Waring et al., 1975
). It is thin-barked and susceptible to fire (Thornburgh, 1990
), so it often occurs in relatively open stands on rocky and infertile soils, which are less prone to hot fires. Brewer spruce is notable for its long, delicate, pendulous branches, which gives it its other common name, weeping spruce.
Brewer spruce is associated with at least 15 other conifers and 21 broadleaf evergreen trees and shrubs (Waring et al., 1975
). Two of its most common associates are Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) and Shasta red fir (Abies magnifica A. Murr. var. shastensis Lemmon). These and other associates attain greater height and older ages, so Brewer spruce is often found in the subcanopy. Maximum height, diameter at breast height (dbh), and age of Brewer spruce are 54 m, 162 cm, and 900 yr, respectively (Waring et al., 1975
; Cowley, 2000
).
An isolate of Engelmann spruce (Picea engelmannii Parry ex Engelm.) is sympatric with Brewer spruce on the slopes of Russian Peak (Sawyer and Thornburgh, 1969
, 1970
). Another spruce, Sitka spruce [Picea sitchensis (Bong.) Carr.], occurs along the Pacific Coast within 18.5 km of the nearest Brewer spruce but much lower in elevation (Waring et al., 1975
). Brewer spruce is generally found between 560–2300 m a.s.l., while Sitka spruce in the Klamath Region is not found more than 200 m a.s.l. (Hickman, 1993
). In an ecological reconnaissance, no hybrids between Brewer spruce and either Engelmann or Sitka spruce were observed (Sawyer and Thornburgh, 1969
). In fact, repeated attempts at controlled pollination have failed to produce hybrids between Brewer spruce and other spruce species (Gordon, 1986
), and phylogenies based on molecular markers suggest that Brewer spruce may stand alone in the genus with no close relatives (Wellendorf and Simonsen, 1979
; Sigurgeirsson and Szmidt, 1993
; Ledig et al., 2004
; C. S. Campbell, University of Maine, personal communication). The presence of alkaloids from two distinct biosynthetic pathways, one of which is, so far, novel for conifers, underscores the uniqueness of the species (Schneider et al., 1995
).
Within its present range, Brewer spruce occurs singly or in small disjunct populations (Thornburgh, 1990
). Thornburgh (1990)
thought that low density, isolation, the overlapping distribution of male and female strobili within the crown, and their concurrent development encouraged selfing. Certainly, other relict spruces with fragmented ranges and small populations have mixed mating systems characterized by rates of selfing unusually high for conifers (Ledig et al., 1997
, 2000
, 2002
). Selfing or close inbreeding reduces seed production (Franklin, 1970
), reducing population viability in typically outbreeding conifer populations.
Brewer spruce is not protected from logging or disturbance, and projected climate change may adversely affect it and other montane and subalpine species in the Klamath Mountains. The Klamath peaks at only about 2750 m a.s.l., thus limiting the opportunity for montane species to migrate higher in elevation. Therefore, it is important for the conservation of Brewer spruce to know whether it has lost genetic diversity that might restrict its ability to respond to environmental changes, and whether like other rare spruces, it also experiences high levels of selfing, which could lead to reproductive failure.
Areas high in genetic diversity are likely to represent populations best able to cope with change and, therefore, are the best areas to focus conservation efforts. The structure of diversity should inform decisions on whether few or many reserves should be managed and monitored. Mating system estimates indicate whether inbreeding and subsequent depression of fitness may be a problem. Therefore, we undertook a survey to determine the level and pattern of genetic diversity in Brewer spruce and estimate mating system parameters. Because of its relictual status, and based on studies of Mexican spruces that occur in small, fragmented populations (Ledig et al., 1997
, 2000
, 2002
), we hypothesized that Brewer spruce would have low levels of diversity within populations and high levels of differentiation among populations. We further hypothesized that it would have a mixed mating system with substantial selfing. Our sample also provided the opportunity to compare mating systems before and after selective timber harvest and in relatively isolated trees vs. those with many near neighbors.
MATERIALS AND METHODS
USDA Forest Service personnel from the Pacific Southwest Region's North Zone Tree Improvement Center collected cones from nine populations of Brewer spruce, and the staff of the Gold Beach Ranger District, Siskiyou National Forest, collected cones from one other population, Collier Butte (Fig. 1). Most populations were sampled in the 1985 cone year. We intended to sample 20 trees in each population. However, at Iron Mountain, only 10 trees were initially sampled; therefore, a second cone collection was made from 32 trees in 1999. Cones were collected at Baldy Mountain in 1989. At Flattop, 11 trees in the sample were collected in 1981 by the staff of the USDA Forest Service Pacific Northwest Region's Dorena Tree Improvement Center, and the remaining nine trees in the sample were collected in 1985 by the staff of the Pacific Southwest Region's North Zone Tree Improvement Center.
To meet the goal of at least 20 cone-bearing trees, it was often necessary to extend samples over several kilometers. The populations and their characteristics relevant to the genetic analysis are as follows:
Iron Mountain
(Powers Ranger District, Siskiyou National Forest; 42°41' N, 124°09' W, 1190 m a.s.l.)—This, the northernmost population of Brewer spruce, occupies the peak of Iron Mountain, which is about 22 km from the coast (Shea, 1996
). Iron Mountain is actually in the Rogue River Mountains, immediately north of the Rogue River, which is often accepted as the northern boundary of the Klamath Region. The population of Brewer spruce at Iron Mountain is the most compact of the populations sampled. The sampled trees were separated by no more than 100 m and were the canopy dominants. The vegetation on the ridge and the slopes surrounding the summit is floristically rich and includes eight other conifers (Baker, 1956
). The Brewer spruce are small, but fairly old; Shea (1996)
reported increment cores for three trees that ranged in age from over 180 to over 400 yr. The oldest tree aged was only 12 m tall and 43.7 cm dbh.
Flattop
(Galice Ranger District, Siskiyou National Forest; 42°26' N, 123°48' W, 1340 m a.s.l.)—The terrain is relatively level, in contrast to the north slopes on which Brewer spruce usually occurs, and the soils are ultramafic. Cones were collected from 11 trees immediately prior to selective timber harvest and from a different set of nine trees 4 years later after logging. The trees sampled were scattered over at least 2.8 km and 130 m in elevation. Although many Brewer spruce were cut, Brewer spruce probably represented a higher proportion of the stand after logging than before harvest. Prior to logging, the Brewer spruce were associated with larger dominants, true fir (Abies spp.), sugar pine (Pinus lambertiana Dougl.), and Jeffrey pine (Pinus jeffreyi Grev. and Balf.). After logging, the stand was very open and the spruces were fully exposed to sunlight. Fifteen years later, when we revisited the stand, many Brewer spruce were infested with dwarf mistletoe (Arceuthobium campylopodum Engelm.) and were dying or dead. No cones were produced on heavily infected trees.
Collier Butte
(Gold Beach Ranger District, Siskiyou National Forest; 42°22' N, 124°08' W, 1070 m a.s.l.)—The stand was compact and fairly dense, with less than 500 m between the most distant trees in the sample and little change in elevation. Brewer spruce were associated with Douglas-fir, Sadler oak (Quercus sadleriana R. Br.), rhododendron (Rhododendron sp.), and manzanita (Ceanothus sp.).
Little Grayback
(Happy Camp Ranger District, Klamath National Forest; 41°59' N, 123°31' W, 1330 m a.s.l.)—Brewer spruce makes its best development at Little Grayback (Waring et al., 1975
), where it grows under a canopy of Shasta red fir, white fir [Abies concolor (Gord. and Glend.) Hildebr.], and Douglas-fir. The stand is dense, and sample trees spanned only about 500 m and a range in elevation of 60 m. The trees were among the largest sampled, with many over 30 m tall and 75 cm dbh. Reproduction was good.
Poker Flat
(Happy Camp Ranger District, Klamath National Forest; 41°56' N, 123°32' W, 1510 m a.s.l.)—Samples were widely distributed over at least 2.4 km and 180 m in elevation. Some trees were very isolated, with no other Brewer spruce in sight. Much of the area has been selectively cut, leaving the smaller spruce, although one tree was over 30 m tall and about 130 cm dbh.
Doolittle Creek
(Happy Camp Ranger District, Klamath National Forest; 41°50' N, 123°28' W, 1150 m a.s.l.)—Samples were distributed over about 1.2 km and 90 m based on global positioning system (GPS) readings in 2000. The area is floristically rich, and associates include white fir, incense-cedar (Libocedrus decurrens Torr.), Port-Orford-cedar (Chamaecyparis lawsoniana [A. Murr.] Parl.), Douglas-fir, and Pacific yew (Taxus brevifolia Nutt.). The stand has been selectively logged, and dwarf mistletoe is common on the spruces.
Prescott Cabin
(Gasquet Ranger District, Six Rivers National Forest; 41°49' N, 123°44' W, 1260 m a.s.l.)—Samples were widely distributed over about 4.25 km, based on GPS, and 120 m in elevation. Some trees in the sample were quite isolated from other Brewer spruce, and some grew in dense clusters on rocky, north slopes. Reproduction was good.
Baldy Mountain
(Happy Camp Ranger District, Klamath National Forest; 41°48' N, 123°28' W, 1580 m a.s.l.)—Samples were distributed over a total of about 1.5 km. However, 17 of the 20 trees in the sample were within 300 m of each other in a fairly dense stand. Conifer associates included white fir, Shasta red fir, Douglas-fir, incense-cedar, sugar pine, western white pine (Pinus monticola Dougl. ex D. Don), and Pacific yew.
Rock Creek Butte
(Ukonom Ranger District, Klamath National Forest; 41°33' N, 123°39' W, 1300 m a.s.l.)—Samples were distributed in clusters over about 2 km and 275 m in elevation. Some trees exceeded 35 m in height and 85 cm dbh. Reproduction was prolific.
Russian Peak
(Scott River Ranger District, Klamath National Forest; 41°20' N, 122°56' W, 1420 m a.s.l.)—This is the only area where Brewer spruce is sympatric with a congener, Engelmann spruce. The area had been selectively logged but is now protected in the Russian Peak Wilderness Area. The samples were distributed over 550 m of elevation and the most distant trees sampled were nearly 5 km apart.
Except for Flattop, stands occurred on north slopes and rocky soils or outcrops that offer some protection from fire.
Cones were maintained separately by trees within populations. For seven populations (Iron Mountain, Little Grayback, Poker Flat, Doolittle Creek, Prescott Cabin, Baldy Mountain, and Rock Creek Butte) staff of the North Zone Tree Improvement Center counted the number of cones. Once the cones dried and opened, seed was extracted by shaking, then hollow seeds were removed by blowing and full seeds weighed. The total number of seeds from each tree was calculated from the relationship between mass and number in samples that were both weighed and counted, providing a measure of full or viable seeds per cone.
Seeds were stored under refrigeration until shipped to the Institute of Forest Genetics, where they were germinated in petri dishes. When radicles emerged, megagametophytes and embryos were dissected from the seeds, separated, and extracted for isozyme electrophoresis. Isozymes from the 1985 samples were analyzed in 1988; the 1989 and 1999 samples were analyzed in 2001.
In spruces, the nutritive tissue of the seed is a haploid megagametophyte that gives rise to, and is of the same genotype as, the egg. Alleles at a locus can be detected by segregation among megagametophytes from a heterozygote. The genotype of the seed parent can be determined by analyzing a number of megagametophytes. When two different alleles at a locus are detected, the seed parent is unequivocally a heterozygote. When only one allele is detected, the tree is classified as a homozygote, although the possibility remains that it is a heterozygote, but by chance the sampled seeds all carried the same allele. The probability of misclassification decreases with increase in sample size. We assayed at least six megagametophytes per tree. For Baldy Mountain and the 1999 sample from Iron Mountain, we analyzed eight megagametophytes. With a sample of six, there is a probability of 0.03125 of misclassifying a heterozygote as a homozygote. That is, the probability that all six megagametophytes in a sample from a heterozygous tree carry the same allele is 2(
)6 = 0.03125. With eight megagametophytes, the probability of misclassifying a heterozygote as a homozygote is only 0.00781. For estimating allele frequencies, the number of parent trees per population, N, varied from 20 to 32 (or 2N = 40–64 genomes).
We used techniques of starch gel electrophoresis based on the laboratory manual of Conkle et al. (1982)
to assay enzyme systems. In the megagametophytes, we were able to consistently score 26 presumptive loci in 15 enzyme systems in all 10 populations. We interpreted the number of loci and alleles by drawing on the experience gained in our laboratory from studies of allozymes of other conifer species (Conkle, 1981
; Ledig et al., 1997
, 2000
, 2002
, 2004
). Samples of red pine (Pinus resinosa Ait.), which are almost invariably homozygous at all loci, were included as standards on each gel to aid interpretation.
We estimated percentage polymorphic loci, alleles per locus, heterozygosity, and Nei's (1978)
unbiased genetic distance with BIOSYS (Swofford and Selander, 1981
). For small samples such as ours, BIOSYS calculates unbiased heterozygosity (Nei, 1978
). Fixation indices (F) within populations were calculated as the mean deviation of loci from Hardy-Weinberg expectations. We also used BIOSYS to calculate Wright's (1965)
F statistics. BIOSYS calculates FIS and FIT, the fixation indices of trees relative to the population and the meta-population, respectively, as weighted averages across alleles. FST, the proportion of the total genic diversity among populations, is calculated from the relationship: 1 – FIT = (1 – FIS)(1 – FST). Pairwise values of FST were calculated using FSTAT2.9.3.2 (Goudet, 2001
). All inferences apply to the population of mature, cone-bearing trees.
The degree of genetic isolation among populations was estimated by Nm, the number of migrants per generation. Nm was calculated from Wright (1951)
as Nm = (1 – FST)/4FST, and from the number and mean frequency of private alleles, the unique alleles found in only one population (Slatkin, 1985
; Barton and Slatkin, 1986
).
We used the computer program BOTTLENECK (Cornuet and Luikart, 1996
) to determine whether effective population numbers had been restricted in the recent past. We employed the infinite allele model (Kimura and Crow, 1964
) because empirically it tends to fit allozyme data better than alternatives (Luikart and Cornuet, 1998
). The Wilcoxon sign–rank test was preferred to the sign test because the former has higher power and can be used with as few as four polymorphic loci (Piry et al., 1999
). However, we had at least six polymorphic loci available for every population, and in most cases, 10 or more.
The association between geographic and genetic distances was checked using Mantel's (1967)
generalized regression procedure. Latitude and longitude of each population sampled was determined with a GPS. Geographic distance between populations was calculated from latitude and longitude using Kindred's (1997)
distance calculator.
We investigated phylogeographic relationships among the populations with cluster analysis, using the SEQBOOT, NEIGHBOR, and CONSENSE programs in the PHYLIP package (Felsenstein, 1995
). Bootstrapping was performed via random resampling of our data set with the SEQBOOT program to generate 1000 matrices of Nei's (1972)
genetic distances. The goal of bootstrapping was to test the consistency with which our data set supported the phenetic relationships among populations. High bootstrap scores (>75%) suggest strong support for a particular cluster, whereas lower levels of support suggest a lower order of differentiation. The unweighted pair group (UPGMA) method was used to produce phylograms using the NEIGHBOR clustering program, and majority-rule consensus trees were generated from bootstrap trees using the program CONSENSE. The phylogenetic trees were viewed and drawn using TREEVIEW (Page, 1996
).
The outcrossing rate was estimated for all populations except two, Baldy Mountain and Russian Peak. For mating system analysis, we assayed embryos alongside their megagametophytes. Knowing the contribution of the egg (the haploid genotype of the megagametophyte) to the embryo, the pollen contribution can be deduced by subtraction, so small progeny arrays are adequate to estimate the outcrossing rate with little bias. For three populations (Iron Mountain, Collier Butte, and Little Grayback) that represented a range of site conditions, we assayed seeds from 20 trees each and about 30 megagametophyte–embryo pairs per tree (about 600 progeny per population). In four populations in which the sampled trees were distributed over distances of several kilometers (Poker Flat, Doolittle Creek, Prescott Cabin, and Rock Creek Butte), we assayed seeds from about 50 megagametophyte–embryo pairs per tree from each of four trees in each population (about 200 progeny in each population). Two trees in each of these four populations were relatively isolated and two in each had many near neighbors. In addition to population estimates of outcrossing, we tested whether isolated trees had lower rates of outcrossing than trees with conspecific neighbors. Finally, at Flattop we compared the mating system in 11 trees sampled in 1981, before logging, to nine trees sampled in 1985, after logging. About 60 megagametophyte–embryo pairs were assayed for each of the 20 trees from Flattop (about 1200 progeny).
We used Ritland's (1986
, 1989
, 1990a
, b
, 1994
) MLTR and MLTF programs to calculate the single- and multiple-locus estimates of the outcrossing rate, ts and tm, at the population level. MLTF was used to estimate ts and tm for each polymorphic locus because it uses information on the megagametophyte genotype and the estimate of outcrossing becomes much more accurate (Ritland, 1990a
). An exception was 6Pg-2 in Poker Flat; because of the failure of MLTF to return a valid estimate of error, MLTR was used to estimate ts with 1000 bootstraps. Neither MLTF or MLTR returned a valid estimate of ts for Pgi-2 in the sample from Prescott Cabin; MLTF failed to return an estimate of error, and in MLTR, estimates converged on whatever initial estimate was used as input or fluctuated between 0.00 and 1.99. Allele frequency and t were always estimated jointly, and the Newton-Raphson method was used for iteration, except in a very few cases where it was necessary to use the expectation maximization method. MLTR was used with 1000 bootstraps to estimate the mean single-locus outcrossing rate (ts), the correlation of outcrossed paternity (rp) among progeny within families, and family-level estimates of tm.
RESULTS
Of the 26 loci consistently scored, 18 were polymorphic in at least one population (Table 1), and eight (Cat-1, Idh-1, Mdh-3, Mdh-4, Mpi, Skd-1, Skd-2, Tpi-2) were monomorphic in all populations. Many of these monomorphic loci are polymorphic in other spruce species (Ledig et al., 2004
).
|
|
|
|
|
|
|
0.87. Although a value of 0.750 for Rock Creek Butte clearly indicates a mixed mating system, the 95% confidence interval includes 0.830, and in each of these four populations, half of the four progenies sampled represented isolated trees, which might have biased tm downward. The mean single-locus outcrossing rates were often as great as or greater than the multilocus rates and never significantly lower than the multilocus rate, suggesting that little inbreeding, in addition to that contributed by selfing, occurred by crossing among relatives. The correlation of outcrossed paternity, rp, among progeny within a family ranged from low to moderately high, which suggests that some populations sampled a diverse pollen pool, while others were pollinated by a more limited number of fathers (Table 6).
|
No difference in tm was observed between isolated trees vs. those with near neighbors in the Poker Flat, Doolittle Creek, Prescott Cabin, or Rock Creek Butte samples (Table 7). In fact, outcrossing estimates were higher for isolated trees than for trees with immediate neighbors at Poker Flat and Rock Creek Butte, although the difference was not significant.
|
|
Genic diversity
The estimates of genic diversity in Brewer spruce were much higher than average for long-lived, woody endemics and close to average for outcrossing endemic plant species in general. Percentage polymorphic loci (P) in Brewer spruce was about 44%, which is less than the mean for outcrossing endemics (54.4%) but higher than that for woody endemics (26.3%) reviewed by Hamrick and Godt (1996b)
and Hamrick et al. (1992)
. Mean expected heterozygosity (He) in Brewer spruce averaged 0.129 compared to 0.056 for 20 long-lived woody endemics (Hamrick et al., 1992
) and 0.142 for 57 outcrossing endemic plant species (Hamrick and Godt, 1996b
). He in Brewer spruce was about 15% lower than values for gymnosperms in general (0.151) or outcrossing, wind-pollinated trees (0.154) reviewed by Hamrick et al. (1992)
. However, if the genically depauperate Iron Mountain is omitted, mean He for the other nine populations of Brewer spruce is 0.137, and at Russian Peak, He is 0.161.
Compared to other endemic spruces of western North America, Brewer spruce is at least as diverse; for 10 populations of Chihuahua spruce (Picea chihuahuana Martínez), a species endemic to the Sierra Madre Occidental, He averaged only 0.093 (Ledig et al., 1997
); for two populations of Martínez spruce (Picea martínezii T.F. Patterson), a very rare endemic of the Sierra Madre Oriental, He was 0.101 and 0.121 (Ledig et al., 2000
); and for the only three known populations of Mexican spruce (Picea mexicana Martínez; Ledig et al., 2002
), He ranged from 0.117 to 0.130.
Russian Peak is the only area where Brewer spruce is sympatric with any congener. An isolated population of Engelmann spruce is found below and mixed with Brewer spruce at its lowest elevation on Russian Peak. However, hybridization does not seem to be the reason for the high heterozygosity observed in Brewer spruce. No natural hybrids between Brewer and Engelmann spruces have been reported and controlled crosses have failed completely (Gordon, 1986
). The two species are distinct in their isozyme profiles (Ledig et al., 2004
).
Perhaps the most notable point about genic diversity in Brewer spruce was the decrease in heterozygosity with increase in latitude. Similar trends in other species, plant and animal, have been interpreted as the trace of postglacial dispersion northward (Gooch and Glazier, 1986
; Cwynar and MacDonald, 1987
; Moran et al., 1989
; Hamrick and Godt, 1996a
; Ledig, 2000
). In Coulter pine (Pinus coulteri D. Don), the trace is unequivocal, marked by the progressive loss of alleles along a route from the San Jacinto Mountains of southern California to the Diablo Range in the north (Ledig, 2000
). Although a few alleles in the southern populations of Brewer spruce disappear northward, the pattern is irregular and suggests random genetic drift in small populations rather than founder effects. The northern isolates, Iron Mountain and Collier Butte, are homozygous at several loci (Table 1), but this may be attributed to their small size and isolation equally as well as to founder effects arising from dispersal northward during the Holocene or previous interglacials. A general dispersal northward conflicts with the interpretation of Brewer spruce as an Arcto-Tertiary relict that has persisted in the Klamath Region, perhaps since the Miocene (Whittaker, 1961
; Wolfe, 1969
; Sawyer and Thornburgh, 1977
).
A Wahlund effect might be expected in several of the "populations" sampled because the trees were distributed over distances exceeding a kilometer in some cases, and even up to 5 km. This does not seem to be the case; inbreeding coefficients were generally low. One of the largest estimates of F was for Collier Butte, a population in which samples were confined within a relatively small area and unlikely to show a Wahlund effect. Where a strong Wahlund effect might have been expected (i.e., for Poker Flat, Doolittle Creek, Prescott Cabin, Baldy Mountain, and Rock Creek Butte), estimates of F ranged only from –0.007 to 0.099.
Genetic structure
Brewer spruce may be characterized as moderately structured. Even excluding the outlier, Iron Mountain, FST was 0.103 among the other nine populations, which is high for conifers. Only two of 18 studies of north temperate and boreal spruces reported estimates of FST greater than 0.10 (reviewed in Ledig et al., 1997
), and none as high as the estimate of 0.157 when Iron Mountain was included. Only in Chihuahua spruce, with its highly fragmented distribution, was FST larger, 0.248 (Ledig et al., 1997
).
Gene exchange among Brewer spruce populations was low relative to many conifers; the estimate of Nm based on FST, was 1.34. Because estimates of Nm do not reach equilibrium for perhaps 100–1000 generations after gene flow ceases (Slatkin and Barton, 1989
), an Nm of 1.34 is likely to reflect past contact, not the present rate of migration. By comparison, Nm in red spruce (Picea rubens Sarg,) was 20.1 in Canada's Maritime Provinces and 3.0 in Ontario (Rajora et al., 2000
). A high FST and low Nm in Brewer spruce suggest the possibility of differentiation by random genetic drift. This interpretation was further supported by the lack of relationship between genetic distance and geographic distance, which would have been expected if gene flow was extensive because near neighbors are more likely to exchange genes than distant populations. Although some near neighbors clustered on an UPGMA tree calculated from Nei's genetic distance (Fig. 3), the tree revealed little phylogeographic pattern that would suggest isolation by distance through a stepping stone model. The relationship between FST or D and the geographic distance between pairs of populations suggests little gene flow between near neighbors. Despite a genetic structure influenced by drift, a test for recent bottlenecks found no greater heterozygosity than that expected for populations at mutation–drift equilibrium, except for Baldy Mountain.
Mating systems
The mixed mating model for the estimation of outcrossing rate assumes loci are in linkage equilibrium. Our samples were too small to estimate disequilibrium or even the recombination rate between loci. However, single-locus estimates are not affected by linkage disequilibrium, and the means of the single-locus estimates, ts, of outcrossing were similar to estimates of tm. Therefore, linkage probably had little effect on the estimate of tm reported here.
Differences between single-locus and multilocus estimates of outcrossing can be used to infer crossing between relatives that is in addition to inbreeding resulting from self-pollination. Crosses between relatives may occur as a result of limited dispersion of progeny around seed trees followed by pollination between neighbors. In our samples, ts was slightly greater than tm in many cases and never significantly less, so there is no evidence of crossing among relatives.
Neither is there evidence of much selfing from the mating system analyses. Given the spatial structure of Brewer spruce populations, the rate of outcrossing seems, in general, quite high. Brewer spruce typically occurs in mixed forest under larger true firs and Douglas-fir, which would seem to impose a barrier to exchange of pollen. However, outcrossing was as high in mixed conifer stands (e.g., tm = 0.926 ± 0.020 at Little Grayback and 0.961 ± 0.018 at Flattop before logging) as it was in almost pure stands dominated by Brewer spruce (e.g., tm = 0.959 ± 0.028 at Iron Mountain).
All populations of Brewer spruce, even in samples represented in part by isolated trees (Poker Flat, Doolittle Creek, Prescott Cabin, and Rock Creek Butte), had higher rates of outcrossing than other endemic spruces. Two small populations of Chihuahua spruce had outcrossing rates of 0 and 15% (Ledig et al., 1997
); rates in Mexican spruce ranged from 59 to 81% (Ledig et al., 2002
); and rates in Martínez spruce ranged from 40 to 69% (Ledig et al., 2000
). Too little information exists on any of these species to speculate on the reasons for the comparatively high level of outcrossing in Brewer spruce.
Outcrossing rates in Brewer spruce were typical of wide-ranging species that grow in large, uninterrupted populations, where estimates generally range from 84 to 98% (King et al., 1984
; Cheliak et al., 1985
; Shea, 1987
; Kuittinen and Savolainen, 1992
; Chaisurisri et al., 1994
). Likewise, the rare Serbian spruce [Picea omorika (Panc.) Purk.] whose populations occupy, in total, only 60 ha in its native range and which is almost entirely self-fertile has an outcrossing rate of 84% (Kuittinen and Savolainen, 1992
). Exceptions to these high rates of outcrossing are Norway spruce [Picea abies (L.) Karst.] at the northern limits of its range in Finland (Muona et al., 1990
) and white spruce [Picea glauca (Moench) Voss] in Newfoundland (Innes and Ringius, 1990
), in which relatively low levels of outcrossing were observed (74 and 73%, respectively). In red spruce, which has a regional distribution in Canada and the United States, Rajora et al. (2000)
observed outcrossing rates of only 65% in Ontario and 54% in the Maritime provinces. However, these estimates were based on ungerminated seeds, and Rajora et al. (2000)
pointed out that allozyme-based estimates from germinable seeds measure only "tolerable" inbreeding events, which is the case in our study of Brewer spruce.
Even the Brewer spruce trees chosen for analysis because of their isolated condition seemed to have fairly high rates of outcrossing, with an estimated mean tm of 1.06 (100%) for eight isolated trees (Table 7). This suggests either a copious pollen cloud or some mechanism(s) for the prevention of inbreeding, such as self-infertility or incompatibility. Incompatibility, however, is unknown in the Pinaceae.
The equilibrium fixation index, Fe, can be calculated from the outcrossing rate, t, as (Allard et al., 1968
): Fe = (1 – t)/ (1 + t). The relatively low estimates of F were consistent with the observed high rates of outcrossing, especially for the four populations in which estimates of tm were based on samples of 20 trees.
rp is the probability that a randomly chosen pair of embryos from the same seed tree were full-sibs. Values around 0.10 (e.g., Iron Mountain, Flattop, and Rock Creek Butte in Table 6) suggest at least 10 pollen parents in the progeny array if all were equally represented, a fairly diverse pollen pool. Higher values, around 0.50 (e.g., Collier Butte, Little Grayback, Poker Flat, and Prescott Cabin) suggest fewer pollen parents per progeny array, only about two if each was equally represented. A high rp in these populations suggests that most trees sampled a limited pollen pool, composed, perhaps, of a few neighbors. However, standard errors for rp were high, so any conclusions are weak.
Given the unexpectedly high outcrossing rates of isolated trees, the low post-logging outcrossing rate at Flattop was surprising. No heterozygote deficiency was observed among progeny in the pre-logging sample, but the inbreeding coefficient for progeny sampled after logging was 0.161. A lower outcrossing rate might be anticipated in closed stands than in stands opened by logging. The position of Brewer spruces as subdominants scattered below Douglas-firs and other species, as at Little Grayback, might be expected to reduce dispersal of Brewer spruce pollen because air movement beneath canopies is less than that above the canopy. The stems and crowns of Brewer spruce's taller associates in unlogged stands might constitute a barrier to pollen exchange and, therefore, to outcrossing. This does not seem to be the case. Instead, outcrossing was reduced when the stand at Flattop was opened, even though the frequency of Brewer spruce increased relative to its associated species. A reduction in density of Brewer spruce after logging does not seem a likely explanation for the increase in selfing because the relatively isolated trees in other populations had higher rates of outcrossing than did the post-logging Brewer spruces at Flattop.
Dwarf mistletoe might be an explanation. Brewer spruce can survive overstory removal, but evaporative demand increases and results in early stomatal closure and vigor reduction (Waring et al., 1975
). This may stress the trees and favor mistletoe infestation. Cone production in heavily infested trees was noticeably reduced or eliminated entirely. Mistletoe may also reduce the number of pollen strobili and, therefore, the pollen cloud, but this has not been quantified. In any case, the increase in selfing suggests that overstory removal by logging may be detrimental to regeneration of Brewer spruce, and create a problem in maintaining the species.
Seed yields
Seeds per cone should reflect the rate of selfing in the Pinaceae. In most conifers that have been experimentally pollinated, seed yields were considerably reduced by self-pollination relative to cross-pollination (Kraus and Squillace, 1964
; Franklin, 1970
; Coles and Fowler, 1976
; Ledig, 1986
). Trees at Poker Flat, Prescott Cabin, and Rock Creek Butte were sparsely distributed over substantial distances, and the numbers of seeds per cone were much lower than at Iron Mountain, Little Grayback, or Baldy Mountain (Fig. 4). Iron Mountain and Little Grayback had comparatively dense populations of Brewer spruce, and at Baldy Mountain, 17 of 20 trees were within a relatively dense group. The difference in numbers of seeds per cone might suggest more self-pollination within Poker Flat, Prescott Cabin, and Rock Creek Butte than within the three denser populations. The situation at Doolittle Creek, where cone-bearing trees were spread over long distances, similar to the situation at Poker Flat, Prescott Cabin, and Rock Creek Butte, is less obvious. Numbers of seeds per cone at Doolittle Creek were intermediate between those in the relatively dense populations and those populations in which trees were sparsely distributed.
Numbers of seeds per cone were unrelated to a tree's outcrossing rate, either overall (r = –0.04, n = 159) or within Iron Mountain or Little Grayback, populations for which we had outcrossing estimates for 20 and 19 trees, respectively. Estimating outcrossing from viable seeds (i.e., those that germinate) measures the effective rate of inbreeding within the surviving progeny, but not the actual rate of self-pollination in the ovules. It appears that almost no inbreds in Brewer spruce survive embryogeny, which suggests a high genetic load. In the Pinaceae, the genetic load of recessive lethal equivalents is usually high and selfing results in fully developed seed coats in which the embryo has aborted (Sorensen, 1969
; Bramlett and Popham, 1971
; Koski, 1971
; Franklin, 1972
; Bishir and Pepper, 1977
; Park and Fowler, 1984
; Ledig, 1986
). High effective outcrossing rates but low seed yields are consistent with the suggestions from paleontology that Brewer spruce is an ancient species that might be expected to have accumulated a high load of embryonic lethal equivalents over millions of years. Even though only a few inbred progeny survive to the seedling stage and inbreeding coefficients are low, self-pollination is detrimental because it reduces reproductive output by depressing seed yields.
Variance among trees in seed production may also contribute to random genetic drift because it reduces effective population size. For example, only six full seeds were obtained from 52 cones of one relatively isolated tree in the Prescott Cabin population (0.12 seeds per cone), while another tree produced 1938 seeds from 25 cones (77.5 per cone). High variance in number of offspring reduces effective population size (Ne) as Ne = 4N/(2 + 
f2), where f2 is the variance in family size. If variance among trees in the number of seeds per cone reflects variance in number of offspring, effective population sizes could be about one-hundredth of the observed population sizes. The ratio of Ne to N in Brewer spruce populations estimated from the variance in number of seeds per cone was about 0.005 to 0.015, except for Rock Creek Butte, where the variance in seeds per cone was small (see Fig. 4) and Ne/N was 0.1.
Conservation of Brewer spruce
Relatively high genetic diversity and rates of outcrossing are positive factors for the conservation of Brewer spruce. However, some of the most isolated populations of Brewer spruce have limited diversity, most notably Iron Mountain, and low diversity may be a problem under projected climate change scenarios or if other environmental changes threaten. In general, diversity increases southward, suggesting that populations in California are most genetically robust. Although genetic structure, as indicated by FST, is high, it is not as pronounced as in the endemic spruces of México, which suggests that most of the genetic diversity in Brewer spruce could be captured in relatively few populations. In fact, most of the observed structure resulted from inclusion of the northern outliers on Iron Mountain and Collier Butte, and gene flow seems to have provided at least a minimum of connection among the other populations. Outcrossing rates were surprisingly high, especially considering the low rates found in other spruces with fragmented distributions. Outcrossing was high even in isolated trees, which is puzzling and suggests either strong mechanisms for the avoidance of inbreeding or effective local pollen dispersal exceeding that encountered in other spruces of southwestern North America (Jackson, 1994
; Jackson and Smith, 1994
). Pollen flight and reproductive biology of Brewer spruce might be a fruitful area of research.
On one hand, the high rates of outcrossing and low estimates of inbreeding are welcome news for the conservation of Brewer spruce. However, a mating system analysis measures only effective inbreeding. The differences in seed yields among populations suggests extensive self-pollination in populations of sparsely distributed trees, and low seed yields could be a barrier to natural regeneration of Brewer spruce and reduce possibilities for dispersal to new habitat.
The most troubling result for conservation of this living fossil was the impact of logging on outcrossing rate. The increase in selfing in a heavily logged stand may be related to mistletoe infection. In any case, observation indicated that mistletoe reduced or eliminated cone production, which is bound to interfere with regeneration of the species. Further research should be conducted to verify whether opening stands by logging is a threat to Brewer spruce and to what degree low seed yields resulting from selfing is a problem for natural regeneration.
FOOTNOTES
The authors thank C. L. Frank, D. A. Davis, D. Burton, and the Dorena Tree Improvement Center for locating Brewer spruce, collecting cones, and extracting seeds; R. Z. Callaham and C. I. Millar, formerly of the University of California's Wildland Resources Center, for financial assistance during sampling; and J. A. Baldwin, J. O. Sawyer, J. B. St. Clair, R. H. Waring, F. C. Yeh, and two anonymous reviewers for helpful comments on the manuscript. 
2 Author for correspondence (tledig{at}ucdavis.edu
)
LITERATURE CITED
Allard R. W. S. K. Jain P. L. Workman 1968 The genetics of inbreeding populations. Advances in Genetics 14: 55-131
Axelrod D. I. 1987 The Late Oligocene Creede flora, Colorado. University of California Publications in Geological Sciences 130: 1-235
Baker W. H. 1956 Plants of Iron Mountain, Rogue River Range, Oregon. American Midland Naturalist 56: 1-53[CrossRef]
Barton N. H. M. Slatkin 1986 A quasi-equilibrium theory of the distribution of rare alleles in a subdivided population. Heredity 56: 409-415
Bishir J. W. D. Pepper 1977 Estimation of number of embryonic lethal alleles in conifers. I. Self-pollinated seed. Silvae Genetica 26: 50-54[Web of Science]
Bramlett D. L. T. W. Popham 1971 Model relating unsound seed and embryonic lethal alleles in self-pollinated pines. Silvae Genetica 20: 192-193
Chaisurisri K. J. B. Mitton Y. A. El-Kassaby 1994 Variation in the mating system of Sitka spruce (Picea sitchensis): evidence for partial assortative mating. American Journal of Botany 81: 1410-1415[CrossRef][Web of Science]
Cheliak W. M. J. A. Pitel G. Murray 1985 Population structure and mating system of white spruce. Canadian Journal of Forest Research 15: 301-308
Coles J. F. D. P. Fowler 1976 Inbreeding in neighboring trees in two white spruce populations. Silvae Genetica 25: 29-34
Conkle M. T. 1981 Isozyme variation and linkage in six conifer species. In M. T. Conkle [tech. coord.], Proceedings of the Symposium on Isozymes of North American Forest Trees and Forest Insects, 1979, Berkeley, California, 11–17. USDA Forest Service General Technical Report PSW-49. Pacific Southwest Forest and Range Experiment Station, Berkeley, California, USA
Conkle M. T. P. D. Hodgskiss L. B. Nunnally S. C. Hunter 1982 Starch gel electrophoresis of pine seed: a laboratory manual. USDA Forest Service General Technical Report PSW-64. Pacific Southwest Forest and Range Experiment Station, Berkeley, California, USA
Cornuet J. M. G. Luikart 1996 Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144: 2001-2014[Abstract]
Cowley A. 2000 California's biggest trees: a cultural and economic resource. Northern California Society of American Foresters, Porterville, California, USA
Cwynar L. C. G. M. MacDonald 1987 Geographical variation of lodgepole pine in relation to population history. American Naturalist 129: 463-469[CrossRef][Web of Science]
DellaSala D. A. S. B. Reid T. J. Frest J. R. Strittholt D. M. Olson 1999 A global perspective on the biodiversity of the Klamath-Siskiyou Ecoregion. Natural Areas Journal 19: 300-319[Web of Science]
Felsenstein J. 1995 PHYLIP: a phylogeny inference package, version 3.57c. Computer program and documentation available at website, http://evolution.genetics.washington.edu/phylip/software.html
Franklin E. C. 1970 Survey of mutant forms and inbreeding depression in species of the family Pinaceae. USDA Forest Service Research Paper SE-61. Southern Forest and Range Experiment Station, New Orleans, Louisiana, USA
Franklin E. C. 1972 Genetic load in loblolly pine. American Naturalist 106: 262-265[CrossRef][Web of Science]
Gooch J. L. D. S. Glazier 1986 Levels of heterozygosity in the amphipod Gammarus minus in an area affected by Pleistocene glaciation. American Midland Naturalist 116: 57-63
Gordon A. G. 1986 Breeding, genetics and genecological studies in spruce for tree improvement in 1983 and 1984, Sault Ste. Marie, Ontario. In C. W. Yeatman and T. J. B. Boyle [eds.], Proceedings of the Twentieth Meeting of the Canadian Tree Improvement Association, part 1, 112– 116. Quebec City, Quebec, Canada
Griffin J. R. W. B. Critchfield 1972 The distribution of forest trees in California. USDA Forest Service Research Paper PSW-82. Pacific Southwest Forest and Range Experiment Station, Berkeley, California, USA
Goudet J. 2001 FSTAT, a program to estimate and test gene diversities and fixation indices, version 2.9.3. Available at website, http://www.unil.ch/ izea/softwares/fstat.html
Haddock P. G. 1938 Picea breweriana in Shasta County. Madroño 4: 176
Hamrick J. L. M. J. W. Godt 1996a Conservation genetics of endemic plant species. In J. C. Avise and J. L. Hamrick [eds.], Conservation genetics: case histories from nature, 281–304. Chapman and Hall, New York, New York, USA
Hamrick J. L. M. J. W. Godt 1996b Effects of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society of London, B, Biological Sciences 351: 1291-1298[CrossRef]
Hamrick J. L. M. J. W. Godt S. L. Sherman-Broyles 1992 Factors influencing levels of genetic diversity in woody plant species. New Forests 6: 95-124[CrossRef]
Hickman J. C. [ed.] 1993 The Jepson manual: higher plants of California. University of California Press, Berkeley, California, USA
Innes D. J. G. G. Ringius 1990 Mating system and genetic structure of two populations of white spruce (Picea glauca) in eastern Newfoundland. Canadian Journal of Botany 68: 1661-1666[Web of Science]
Irwin W. P. 1966 Geology of Klamath Mountains Province. Geology of northern California. California Division of Mines and Geology Bulletin 190: 19-38
Jackson S. T. 1994 Pollen and spores in Quaternary lake sediments as sensors of vegetation composition: theoretical models and empirical evidence. In A. Traverse [ed.], Sedimentation of organic particles, 253– 286. Cambridge University Press, Cambridge, UK
Jackson S. T. S. J. Smith 1994 Pollen dispersal and representation on an isolated, forested plateau. New Phytologist 128: 181-193[CrossRef][Web of Science]
Kindred D. 1997 "How far is it?" Distance calculator. Available at website, http://www.indo.com/distance/
King J. N. B. P. Dancik N. K. Dhir 1984 Genetic structure and mating system of white spruce (Picea glauca) in a seed production area. Canadian Journal of Forest Research 14: 639-643[CrossRef]
Kimura M. J. F. Crow 1964 The number of alleles that can be maintained in a finite population. Genetics 49: 725-738
Koski W. 1971 Embryonic lethals of Picea abies and Pinus sylvestris. Communicationes Instituti Forestalis Fenniae 75.3. Helsinki, Finland
Kuittinen H. O. Savolainen 1992 Picea omorika is a self-fertile but outcrossing conifer. Heredity 68: 183-187[Web of Science]
Kraus J. F A. E. Squillace 1964 Selfing vs. outcrossing under artificial conditions in Pinus elliottii Engelm. Silvae Genetica 13: 72-76
Ledig F. T. 1986 Heterozygosity, heterosis, and fitness in outcrossing plants. In M. E. Soulé [ed.], Conservation biology: the science of scarcity and diversity, 77–104. Sinauer, Sunderland, Massachusetts, USA
Ledig F. T. 2000 Founder effects and the genetic structure of Coulter pine. Journal of Heredity 91: 307-315
Ledig F. T. V. Jacob-C P. D. Hodgskiss T. Eguiluz-P 1997 Evolution and divergence among populations of a rare Mexican endemic, Chihuahua spruce, following Holocene warming. Evolution 51: 1815-1827[CrossRef][Web of Science]
Ledig F. T. B. Bermejo-V P. D. Hodgskiss D. R. Johnson C. Flores-L V. Jacob-C 2000 Mating system and genic diversity in Martínez spruce, an extremely rare endemic of México's Sierra Madre Oriental: an example of facultative selfing and survival in interglacial refugia. Canadian Journal of Forest Research 30: 1156-1164[CrossRef]
Ledig F. T. P. D. Hodgskiss V. Jacob-C 2002 Genic diversity, the mating system, and conservation of a Mexican subalpine relict, Picea mexicana Martínez. Conservation Genetics 3: 113-122[CrossRef][Web of Science]
Ledig F. T. P. D. Hodgskiss K. V. Krutovskii D. B. Neale T. Eguiluz-P 2004 Relationships among the spruces (Picea, Pinaceae) of southwestern North America. Systematic Botany 29: 275-295[CrossRef][Web of Science]
Little E. L. Jr. 1971 Atlas of United States trees, vol. 1, Conifers and important hardwoods. USDA Miscellaneous Publication no. 1146. USDA Forest Service, Washington, D.C., USA
Luikart G. J.-M. Cornuet 1998 Empirical evaluation of a test for identifying recently bottlenecked populations from allele frequency data. Conservation Biology 12: 228-237[CrossRef][Web of Science]
Mantel N. 1967 The detection of disease clustering and a generalized regression approach. Cancer Research 27: 209-220
Miles S. R. C. B. Gouday [compilers] 1997 Ecological subregions of California: section and subsection descriptions. R5-EM-TP-005. USDA Forest Service, Pacific Southwest Region, San Francisco, California, USA
Moran G. F. J. C. Bell J. W. Turnbull 1989 A cline in genetic diversity in river she-oak Casuarina cunninghamiana. Australian Journal of Botany 37: 169-180[CrossRef]
Muona O. L. Paule A. E. Szmidt K. Kärkkäinen 1990 Mating system analysis in a central and northern European population of Picea abies. Scandinavian Journal of Forest Research 5: 97-102
Nei M. 1972 Genetic distance between populations. American Naturalist 106: 283-292[CrossRef][Web of Science]
Nei M. 1978 Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583-590
Page R. D. M. 1996 TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12: 357-358 Available at website, http://taxonomy.zoology.gla.ac.uk/ rod/treeview.html
Park Y. S. D. P. Fowler 1984 Inbreeding in black spruce (Picea mariana Mill. B.S.P.): self-fertility, genetic load, and performance. Canadian Journal of Forest Research 14: 17-21
Piry S. G. Luikart J.-M. Cornuet 1999 BOTTLENECK: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal of Heredity 90: 502-503
Rajora O. P. A. Mosseler J. E. Major 2000 Indicators of population viability in red spruce, Picea rubens. II. Genetic diversity, population structure, and mating behavior. Canadian Journal of Botany 78: 941-956[CrossRef][Web of Science]
Ritland K. 1986 Joint maximum likelihood estimation of genetic and mating structure using open-pollinated progenies. Biometrics 42: 25-43[CrossRef][Web of Science]
Ritland K. 1989 Correlated matings in the partial selfer Mimulus guttatus. Evolution 43: 848-859[CrossRef][Web of Science]
Ritland K. 1990a Generalized multilocus estimation program MLTF. K. Ritland, Toronto, Ontario, Canada
Ritland K. 1990b A series of FORTRAN computer programs for estimating plant mating systems. Journal of Heredity 81: 235-237[Web of Science]
Ritland K. 1994 Multilocus mating system program MLTR, version 0.9. K. Ritland, Department of Botany, University of Toronto, Toronto, Ontario, Canada
Sawyer J. O. Jr. D. A. Thornburgh 1969 Ecological reconnaissance of relict conifers in the Klamath Region. USDA Forest Service, Pacific Southwest Coop-Aid Grant no. 7, contract no. 21-52
Sawyer J. O. Jr. D. A. Thornburgh 1970 The ecology of relict conifers in the Klamath Region, California. USDA Forest Service, Pacific Southwest Coop-Aid Grant no. 9, contract no. 21-52
Sawyer J. O. D. A. Thornburgh 1977 Montane and subalpine vegetation of the Klamath Mountains. In M. G. Barbour and J. Major [eds.], Terrestrial vegetation of California, 699–732. John Wiley, New York, New York, USA
Schneider M. J. S. Brendze J. A. Montali 1995 Alkaloids of Picea breweriana. Phytochemistry 39: 1387-1390[CrossRef][Web of Science]
Shea D. 1996 Brewer spruce on Iron Mountain. Bulletin of the Native Plant Society of Oregon 29: 5-7
Shea K. L. 1987 Effects of population structure and cone production on outcrossing rates in Engelmann spruce and subalpine fir. Evolution 41: 124-136[CrossRef][Web of Science]
Sigurgeirsson A. A. E. Szmidt 1993 Phylogenetic and biogeographic implications of chloroplast DNA variation in Picea. Nordic Journal of Botany 13: 233-246[Web of Science]
Slatkin M. 1985 Rare alleles as indicators of gene flow. Evolution 39: 53-65[CrossRef][Web of Science]
Slatkin M. N. H. Barton 1989 A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43: 1349-1368[CrossRef][Web of Science]
Sorensen F. 1969 Embryonic genetic load in coastal Douglas-fir, Pseudotsuga menziesii var. menziesii. American Naturalist 103: 389-398[CrossRef][Web of Science]
Swofford D. L. R. B. Selander 1981 BIOSYS 1: a FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72: 281-283
Thornburgh D. 1990 Picea breweriana Wats. Brewer spruce. In R. M. Burns and B. H. Honkala [tech. coords.], Silvics of North America, vol. 1, Conifers, 181–186. Agriculture Handbook no. 654. USDA Forest Service, Washington, D.C., USA
Villa-Lobos J. 2003 Klamath-Siskiyou: jewel of the Pacific Coast. Plant Talk 31: 29-33
Waring R. H. 1969 Forest plants of the Eastern Siskiyous: their environmental and vegetational distribution. Northwest Science 43: 1-17
Waring R. H. W. H. Emmingham S. W. Running 1975 Environmental limits of an endemic spruce, Picea breweriana. Canadian Journal of Botany 53: 1599-1613[Web of Science]
Wellendorf H. V. Simonsen 1979 A chemotaxonomic study in Picea with isoenzymes in the seed endosperm. In D. Rudin [ed.], Proceedings of the Conference on Biochemical Genetics of Forest Trees, 182–193. Report 1, Department of Forest Genetics and Plant Physiology, Swedish University of Agricultural Sciences, Umeå, Sweden
Whittaker R. H. 1960 Vegetation of the Siskiyou Mountains, Oregon and California. Ecological Monographs 30: 279-338[CrossRef]
Whittaker R. H. 1961 Vegetation history of the Pacific Coast states and the "central" significance of the Klamath Region. Madroño 16: 5-23
Wolfe J. A. 1964 Miocene floras from Fingerrock Wash southwestern Nevada. United States Geological Survey Professional Paper no. 454-N
Wolfe J. A. 1969 Neogene floristic and vegetational history of the Pacific Northwest. Madroño 20: 83-110
Wright S. 1951 The genetical structure of populations. Annals of Eugenics 15: 323-354[Web of Science]
Wright S. 1965 The interpretation of population structure by F-statistics with special regard to systems of mating. Evolution 19: 395-420[CrossRef][Web of Science]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
