Direct and indirect estimates of seed versus pollen movement within a population of ponderosa pine.
Until recently, most surveys of natural populations were restricted to biparentally inherited markers (primarily allozymes), which allow inferences only about the total amount of gene flow. The availability of uniparentally inherited organellar DNA markers has opened a new avenue for exploring patterns of gene flow via seed and pollen (McCauley 1994). Maternally inherited markers are dispersed only via seed movement, whereas paternally inherited markers are dispersed by movement of both seed and pollen. Thus the difference between spatial patterns of maternal, paternal, and biparental markers permits inferences to be drawn about the relative movement of seed and pollen within and among populations (Ennos 1994; McCauley 1994). It is expected that the degree of spatial structure of paternally inherited markers will always be less than or equal to that shown by maternally inherited polymorphisms, unless pollen movement is zero, in which case the two are equal (Ennos 1994).
Members of the Pinaceae are particularly well suited to studies of pollen and seed movement because they show opposite uniparental inheritance of their organellar genomes. Mitochondrial DNA (mtDNA) is maternally inherited in pines, whereas chloroplast DNA (cpDNA) is paternally inherited (Wagner 1992). Though exceptions exist, conifers generally show little spatial structure of allozyme allele frequencies either within (Epperson and Allard 1989) or between populations (Hamrick and Godt 1990), and this is generally attributed to extensive gene flow mediated by wind-dispersed pollen. Seeds of wind-dispersed conifers are not expected to move as far as pollen because of their considerably greater mass. Consistent with this expectation, Dong and Wagner (1993 1994) and Latta and Mitton (1997) have demonstrated significantly greater differentiation of mtDNA than cpDNA haplotype frequencies between populations of lodgepole pine and limber pine, respectively. However, no studies have yet described the relative spatial structure of maternal and paternal markers within a population.
In this study, we examine the relative rates of seed and pollen movement within a population of ponderosa pine, Pinus ponderosa Laws (Pinaceae). Our approach combines analysis of the spatial structure seen at maternal, paternal, and biparentally inherited markers with direct estimates of pollen dispersal using genetically marked pollen. The main hypothesis tested is that pollen is the primary agent of gene flow within the population. Thus, we predict that spatial structuring will be much greater for mtDNA than for cpDNA polymorphisms. We also predict that where patches of individuals can be identified that share a common mtDNA haplotype, these will be related through their mother as half-sibs.
We test the hypothesis that the dispersal of genetically marked pollen will show no distance limitation. Dispersal may be quantified in two ways: (1) by the proportion of ovules in recipient trees at a given distance that are fertilized by the source tree; and (2) by the proportion of source tree pollen that fertilizes ovules at a given distance. For clarity, we will use the term "frequency of source tree pollen" to refer to proportion 1 and "dispersal probability" to refer to proportion 2. These two measures have different biological interpretations. Frequency of source tree pollen represents the likely source of pollen fertilizing a given ovule, while dispersal probability represents the likely fate of a pollen grain. We define unlimited dispersal as a constant dispersal probability across distance from the source tree. However, because pollen spreads over a greater area with increasing distance from the source, we predict that the frequency of source tree pollen (proportion 1, above) will be inversely related to the distance from the source tree (frequency = k/d, where k is a constant).
The Study Site
Ponderosa pine is a long-lived conifer widespread in the montane regions of western North America. The study site contains an all-aged pure stand of ponderosa pine occupying approximately 1.5 ha on a dry, south-facing slope. It is located in Boulder Canyon approximately 2 km west of Boulder, Colorado, at an elevation of 1740 m, and is contiguous with the main distribution of ponderosa pine in this region. This population has been the site of extensive long-term research into the ecological genetics of ponderosa pine (e.g., Linhart et al. 1981; Linhart and Mitton 1985). All of the trees within the study area were tagged in 1977, and the location of each tree was mapped using the nearest-neighbor procedure of Rohlf and Archie (1978).
Spatial Structure of Organellar DNA Variation
We collected needle tissue from all 182 surviving tagged trees in 1996. Total genomic DNA was extracted using the CTAB extraction method of Doyle and Doyle (1987). Length variation in the cpDNA was found in the region between the trnG, and trnR genes. This 270-bp region was amplified in a polymerase chain reaction using primers (5[prime]-TCGATTCCCGCTACCCGCT-3[prime] and 5[prime]-TGTCCTATCCATTAGACGAT-3[prime]) kindly provided by Dr. Craig Newton of the Forest Biotechnology Centre at B.C. Research, Inc. Length variation in the mtDNA was found in a 2.2-kb fragment containing the second intron of nad1, which was amplified using the primers of Demesure et al. (1995). Reaction mixtures contained 50 mM KCl, 10 mM Tris, 0.01% gelatin, 4 mM MgCl, 200 [[micro]molar] of each dNTP, 0.5U of Taq polymerase, 200 nM of each primer, and 20 ng of template DNA in a 15-[[micro]liter] reaction volume. We used a "touchdown" amplification profile for the reactions. After an initial three-minute denaturation at 94 [degrees] C, the reaction tubes went through four cycles at each of a descending series of annealing temperatures (63 [degrees] C, 61 [degrees] C, 59 [degrees] C, 57 [degrees] C), and 39 cycles with an annealing temperature of 55 [degrees] C. Denaturation was at 94 [degrees] C (1 min) and extension was at 72 [degrees] C (2 min) in all cycles, with a final 10-minute extension at 72 [degrees] C to end the reaction. The mitochondrial fragment was digested with the restriction enzyme Rsa I to increase resolution of the small insertion/deletion, subjected to electrophoresis on a 1.5% agarose gel, and visualized with ethidium bromide. The cpDNA fragment was subjected to electrophoresis on a 5% polyacrylamide gel and visualized with silver staining.
The spatial structure exhibited by each organellar genome was analyzed using spatial autocorrelation of join count statistics (Sokal and Oden 1978; Epperson 1990). Briefly, this analysis compares all pairs of trees within a specified distance of one another (in this case, [less than] 4m) and asks whether the pairs exhibit the same haplotype more often than expected by chance under a random spatial arrangement. Formulae for the number of such pairs expected (and its sampling variance) under random spatial arrangements are given in Sokal and Oden (1978). This process is then repeated for pairs falling within other distance classes (4-8 m, 8-12 m etc.). Distance classes were set at 4-m intervals since this provided a good compromise between resolution (having distance classes small enough to detect small scale spatial structure), and power (having enough pairs within each distance class). An excess or deficiency of pairs with the same haplotype is expressed as a standard normal deviate (SND) of the expectation, and plotted against distance. If a patchy distribution exists, there will be an excess of pairs with the same haplotype at short distance classes, dropping to a deficiency at larger distances. This is because near neighbors will fall within the same patch, while more distant individuals will likely be drawn from different patches. The reverse pattern implies hyperdispersion. Because this analysis involves multiple statistical tests, which can inflate Type I error, and because we are specifically interested in clustering at small spatial scales, we used only near neighbors ([less than] 4 m apart) to test for the significance of patch structure. We specifically examined spatial autocorrelation of the rare haplotype of each organelle, because these will show patch structure more clearly. Patch structure is less detectable for the common haplotype, because adjacent patches will often both be composed of individuals with the common haplotype.
Allozyme genotypes at 11 polymorphic loci were available for each tree from prior studies (Linhart et al. 1981; Mitton et al. 1981; Linhart and Mitton 1985). We wished to test the hypothesis that patches of individuals sharing the rare mtDNA haplotype are related through their mother as half-sibs and cluster together due to limited seed dispersal. We therefore calculated Queller and Goodnight's (1989) estimate of relatedness for individuals within patches. As in the autocorrelation analysis, we include only those individuals with the rare haplotype, because these patches can be most clearly delimited.
Direct Estimates of Pollen Dispersal
Two trees on the site were heterozygous for unique alleles [ILLUSTRATION FOR FIGURE 1B OMITTED], one at phosphoglucose-isomerase (Pgi), and the other at alcohol dehydrogenase (Adh). To estimate dispersal distances of pollen from these trees, seeds were collected in 1987 from recipient trees, which were chosen so as to sample all parts of the stand. As many seed as could be collected from each tree were analyzed, but equal sample sizes per tree were not possible (sample sizes in Table 1). The embryo of each seed was dissected out and its genotype assayed. Assuming that immigration of the rare alleles (q [less than] 0.004) from outside the population was negligible, any embryo containing the unique allele is most likely to have been sired by pollen from the source tree. This method therefore estimates the distribution of dispersal distances of successful pollen only. To estimate the degree of self-fertilization, we also genotyped seeds from the source tree. Both the embryo and the haploid megagametophyte were genotyped to distinguish maternal and paternal alleles in these seeds.
The number of pollen grains detected from each source tree was tabulated for 10-m distance classes (0-10, 10-20, etc.) from the source tree. The frequency of source tree pollen among all seeds analyzed at a given distance class was estimated as twice the frequency of seeds carrying the unique allele (because there was no evidence of segregation distortion, we assume that only half of the pollen from the heterozygous source trees carries the unique allele). The area over which pollen may disperse increases with distance from the source tree, such that a larger area is included in the 50-60-m distance class than the 0-10-m distance class. To obtain an estimate of dispersal probability, the frequency of source tree pollen at each distance was multiplied (weighted) by the relative area falling within each distance class and normalized to sum to one.
Spatial Structure of Organellar DNA Variation
Two mtDNA haplotypes were detected that differed by an insertion/deletion of approximately 30 bp in the second intron of nad1. The haplotype frequencies were 0.837 for the longer haplotype and 0.163 for the shorter. Similarly, two cpDNA haplotypes were detected in the region between trnR and trnG, differing in length by about 10 bp, with frequencies of 0.615 and 0.385. There was no association between cpDNA and mtDNA haplotypes ([[Chi].sup.2] = 0.016, ns).
The maternally inherited mtDNA haplotypes show a patchy distribution in space in this population [ILLUSTRATION FOR FIGURE 1A OMITTED]. Small groups of individuals carrying the rare mtDNA haplotype can be identified surrounded by individuals bearing the more common haplotype. A few individuals carrying the rare haplotype are not obviously associated with any of these patches, but these are in the minority. Patches were identified by eye, and confirmed by UPGMA clustering of individuals with the rare mtDNA haplotype based upon the physical distance separating them (not shown).
Spatial autocorrelation of the rare mtDNA haplotype is significant at short distance classes ([ILLUSTRATION FOR FIGURE 2 OMITTED]; SND = 1.99, P [less than] (0.05). Thus pairs of trees separated by 4 m or less both have the rare haplotype more often than would be expected if the rare haplotype were randomly distributed in space within the population. As more distantly separated trees are considered, pairs where both trees carry the rare mtDNA haplotype become less frequent than expected under a random spatial distribution, although this deficiency is not quite statistically significant (SND [greater than] -1.96, P = 0.06; [ILLUSTRATION FOR FIGURE 2 OMITTED]). As distance increases, the correlogram shows a fluctuating pattern, as expected for patch-structured populations. The first x-intercept of the correlogram occurs at approximately 6 m. This represents the point at which trees in a pair are equally likely to be in the same patch as to be in adjacent patches, and therefore estimates the radius of patches in the population (Sokal and Oden 1978).
By contrast, the paternally inherited cpDNA polymorphism shows no evidence of patch structure [ILLUSTRATION FOR FIGURE 1B OMITTED]. Both cpDNA haplotypes are distributed roughly randomly across the population. Chloroplast polymorphisms are not significantly autocorrelated [TABULAR DATA FOR TABLE 1 OMITTED] at short distances. Pairs of trees separated by 4 m or less both carry the rare cpDNA haplotype slightly less often than expected by a random spatial arrangement ([ILLUSTRATION FOR FIGURE 2 OMITTED]; SND = -0.67, ns). By contrast, pairs of trees that are more distant from one another tend to carry the same haplotype more often than expected by chance, suggesting that cpDNA haplotypes may be hyperdispersed at least over a short spatial scale.
Individuals sharing the rare mtDNA haplotype within the patches identified in Figure 1a are significantly related (note again that we are considering only those individuals with the rare haplotype). Queller and Goodnight's (1989) estimator of the fraction of genes shared by descent gave r = 0.266 averaged over all eight patches. Jackknife 95% confidence limits were 0.124 [less than] r [less than] 0.408. On average, groups of half-sibs will share one-quarter of their genes through common ancestry. Thus patches appear to represent groups of half-sibs related through their maternal parent.
Direct Estimates of Pollen Dispersal
The frequency of the unique Adh and Pgi alleles in seeds declines as distance from the source tree increases (Table 1, [ILLUSTRATION FOR FIGURE 3 OMITTED]). Thus, the farther a recipient tree is from a source tree, the smaller will be the proportion of its ovules that are pollinated by that source tree. The frequency of source tree pollen is expected to be inversely related to distance from the source tree, even with unlimited pollen movement. The observed pattern of marked pollen dispersal is not significantly different from such an inverse relationship [ILLUSTRATION FOR FIGURE 3 OMITTED]. No relationship between distance and dispersal probability is evident (Table 1). Although confidence limits are naturally quite wide at greater distance classes, pollen seems as likely to fertilize a tree 60 m away as it is to fertilize one at 10 m. In addition, we found no evidence of directionality to pollen movement (not shown). Direct estimates of seed dispersal were not possible. Although several trees on the site are heterozygous for chlorophyll deficiency alleles, only nine albino seedlings have been observed. Of these, seven fell within 4 m of a tree known to carry the recessive allele, while the maximum distance between an albino seedling and a carrier of the allele was 23 m.
Taken together, the data presented above strongly support the hypothesis that seed movement is restricted, while pollen movement is not. Mitochondrial DNA polymorphisms show a patchy spatial arrangement, consistent with limited seed movement. Moreover, the patches of individuals defined by common haplotype are related as half-sibs (r = 0.266), indicating that the patches represent matrilineal groups. This measure represents an average over all patches. Visual inspection of the allozyme genotypes and ages of trees within patches indicates that some pairs of individuals within a patch are probably more closely related (e.g., mother-daughter; r = 0.5), since one is at least 50 years older than the other, and allozyme genotypes are consistent with this relationship. Other patches do not contain an individual old enough to be the mother of the other trees in that patch, so these trees may be more distantly related.
By contrast, both direct and indirect estimates indicate that pollen movement is essentially unrestricted within the stand. Pollen is as likely to travel across the stand as it is to fertilize neighboring trees (Table 1). Moreover, cpDNA polymorphisms, which are inherited through the pollen, show no evidence of patch structure within the population. The lack of association between mtDNA and cpDNA haplotypes further strengthens the inference of high pollen movement, since such a lack of association is expected with random mating (Schnabel and Asmussen 1989). Dong and Wagner (1994) reported a similar lack of mtDNA-cpDNA association within eight lodgepole pine and eight jack pine populations, indicating that extensive movement of pollen among trees creates panmixia within these populations, despite limited seed movement.
At least two other studies have examined the spatial distribution of organellar DNA within populations. In both cases, the maternally inherited cpDNA of angiosperm species was examined. McCauley et al. (1996) examined the spatial distribution of cpDNA polymorphism within a population of Silene alba, and Tarayre (1996) examined differentiation of cpDNA haplotype frequencies among subpopulations of Thymus vulgaris. Both studies found strong spatial substructuring of cpDNA haplotype frequencies within populations. Moreover cpDNA showed much greater spatial structure than did biparentally inherited allozyme allele frequencies. Although no paternally inherited markers are available in these species, the results suggest that pollen is the primary agent of gene movement.
Between-population differentiation of organellar haplotype frequencies has been more extensively studied than within-population structure. In general, for both gymnosperms and angiosperms, maternally inherited markers show greater differentiation among populations than do allozymes (Dong and Wagner 1993; Strauss et al. 1993; McCauley 1994). Where paternally inherited markers are available, these generally show very little differentiation among populations (Dong and Wagner 1994; Latta and Mitton 1997). These results are generally consistent with the hypothesis that pollen is the primary agent of gene flow between populations as well as within them. Two exceptions are worth mentioning. Milligan (1991) found little differentiation of maternally inherited cpDNA between populations of Trifolium repens, suggesting high seed movement, which he attributed to human transport of this agricultural species. By contrast, Hong et al. (1994) found high differentiation of both cpDNA and mtDNA among populations of Bishop pine (Pinus muricata). This result suggests limited movement of both seed and pollen, which is in strong contrast to the much lower level of allele frequency differentiation exhibited by allozymes in that species (Millar et al. 1988).
Although the confidence limits are wide at high distance classes, there is no evidence that dispersal probability of the pollen declines with distance (Table 1). Thus pollen released by a tree is just as likely to fertilize one of the (few) nearby trees as one of the (many) trees further away. Conversely, a maternal tree receives as much pollen from its near neighbors (i.e., from the small group of trees falling within 10 m of the recipient) as from more distant trees (i.e., the larger group of trees falling between, say, 50 m and 60 m from the recipient). However, the total amount of pollen received from near neighbors is contributed by fewer trees, so that each individual near neighbor contributes more pollen than each of the (relatively many) more distant trees. Thus the frequency of pollen from an individual source tree declines with distance from the source.
Pollen seems less likely to self-fertilize than to fertilize an ovule at any other distance. Only 4-7% of pollen is likely to self-fertilize (Table 1), concordant with a prior estimate of the selfing rate in this stand from progeny arrays (s = 0.04; Mitton et al. 1981). Thus, on average, only 4% of ovules are self-fertilized. Intriguingly, however, the selfing rate of ovules on the two source trees is much higher than this average (Table 1). While the selfing rate of seed and pollen must be equal across the population, individual trees may have different selfing rates of their ovules and pollen. It is possible, for example, that the source trees export more pollen than they receive from other trees in the stand. High selfing rates for individual trees in otherwise outcrossing populations have been reported for several species (Mitton 1992).
Attempts to measure pollen movement have yielded diverse results. Studies using a linear array of traps to measure pollen dispersal have usually given low estimates of pollen dispersal distances for wind-dispersed (e.g., Bateman 1947; Colwell 1951; Wolfenberger 1959; Boshier et al. 1995) as well as water-dispersed species (Ruckleshaus 1996). By contrast, studies using paternity analysis have given larger estimates of pollen dispersal distance, particularly within populations (Devlin et al. 1988; Campbell 1991; Dow and Ashley 1996; Stacy et al. 1996; Schuster and Mitton, unpubl., but see Meagher 1986). There are two possible explanations for this pattern. A decrease of marked pollen density with distance from a source is expected to follow an inverse relationship, as pollen spreads out from the source. Such an inverse relationship can give the impression of a leptokurtic distribution of pollen dispersal distances. Adjustment for the greater area falling at greater distances from the source drastically changes the interpretation of the result, such that dispersal probability shows no relationship with distance (Table 1).
The second possibility is that post-pollination events may cause the distribution of successful pollen to differ from that based upon measurements of pollen movement (Levin 1981; Campbell 1991). The correlogram for cpDNA suggests that chloroplast polymorphisms are spatially hyperdispersed [ILLUSTRATION FOR FIGURE 2 OMITTED]. While the deficiency of near neighbor pairs carrying the same cpDNA haplotype is not significant, the pattern across the first four classes is qualitatively opposite to that shown by mtDNA. One possible mechanism that could create a hyperdispersed spatial arrangement is the selective elimination of offspring from near-neighbor matings through inbreeding depression. This certainly appears to occur for self-fertilizations. Source tree pollen was less likely to be found in a selfed seed than in seed of almost any other distance class (Table 1). Because seeds were surveyed at maturity, there was ample time for selective elimination of selfed embryos. Such elimination prior to seed maturity is known to occur in other species of conifer (e.g., Cheliak et al. 1985). Since near neighbors tend to be members of the same family, offspring of matings over a short distance may also be inbred (Coles and Fowler 1976). However, there is no evidence that pollen is less likely to sire seed of near neighbors than of more distant trees (Table 1), which would be necessary to give a hyperdispersed spatial arrangement of cpDNA. Potentially, some elimination of biparentally inbred offspring takes place between seed set (the life stage used in the direct study) and adult trees (which were surveyed for cpDNA and mtDNA haplotypes). However, for the moment, hyperdispersion of paternally inherited cpDNA haplotypes must remain an intriguing possibility to be investigated further.
We gratefully acknowledge the work of numerous students and collaborators of YBL, who helped to collect the data on allozyme genotypes and spatial locations that laid the foundation for this study. C. Newton of British Columbia Research Inc., Vancouver, British Columbia, Canada, kindly provided the cpDNA primers, and K. Goodnight and D. Queller provided the computer program to estimate relatedness. J. Mitton, T. Ranker, M. Grant, and two anonymous reviewers provided many helpful comments on an earlier draft of this paper. Financial support for this project was provided by National Science Foundation grant DEB 7816798, USDA grant 95-37101-1638 to YBL, and by an NSERC Postgraduate Fellowship and National Science Foundation Doctoral Dissertation Improvement grant to RGL.
BATEMAN, A. J. 1947. Contamination of seed crops. II. Wind pollination. J. Genet. 49:235-246.
BOSHIER, D. H., M. R. CHASE, AND D. S. BAWA. 1995. Population genetics of Cordia alliodora (Boraginaceae), a neotropical tree. 3. Gene flow, neighborhood and population substructure. Am. J. Bot. 82:484-490.
CAMPBELL, D. R. 1991. Comparing pollen dispersal and gene flow in a natural population. Evolution 455:1965-1968.
CHELIAK, W. M., B. P. DANCIK, K. MORGAN, F. C. YEH, AND C. STROBECK. 1985. Temporal variation of the mating system in a natural population of jack pine. Genetics 109:569-584.
COLES, J. E, AND D. P. FOWLER. 1976. Inbreeding in neighboring trees in two white spruce populations. Silvae Genet. 25:29-34.
COLWELL, R. N. 1951. The use of radioactive isotypes in determining spore distribution patterns. Am. J. Bot. 38:511-523.
DEMESURE, B., N. SODZI, AND R. J. PETIT. 1995. A set of universal primers for amplification of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants. Mol. Ecol. 4:129-131.
DEVLIN, B., K. ROEDER, AND N. C. ELLSTRAND. 1988. Fractional paternity assignment: theoretical development and comparison to other methods. Theor. Appl. Genet. 76:369-380.
DONG, J., AND D. B. WAGNER. 1993. Taxonomic and population differentiation of mitochondrial diversity in Pinus banksiana and Pinus contorta. Theor. Appl. Genet. 86:573-578.
-----. 1994. Paternally inherited chloroplast polymorphism in Pinus: estimation of diversity and population subdivision and tests of disequilibrium with a maternally inherited mitochondrial polymorphism. Genetics 136:1187-1194.
DOW, B. D., AND M. V. ASHLEY. 1996. Microsatellite analysis of seed dispersal and parentage of saplings in bur oak, Quercus macrocarpa. Mol. Ecol. 5:615-627.
DOYLE, J. J., AND J. C. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19:11-15.
ENNOS, R. A. 1994. Estimating the relative rates of pollen and seed migration among plant populations. Heredity 72:250-259.
EPPERSON, B. K. 1990. Spatial patterns of genetic variation within plant populations. Pp. 229-253 in A. H. D. Brown, M. T. Clegg, A. L. Kahler, and B. S. Wier, eds. Plant population genetics, breeding and genetic resources. Sinauer, Sunderland, MA.
EPPERSON, B. K., AND R. W. ALLARD. 1989. Spatial autocorrelation analysis of the distribution of genotypes within populations of lodgepole pine. Genetics 121:840-858.
HAMRICK, J. L., AND M. J. W. GODT. 1990. Allozyme diversity in plant species. Pp. 43-63 in A. H. D. Brown, M. T. Clegg, A. L. Kahler, and B. S. Wier, eds. Plant population genetics, breeding and genetic resources. Sinauer, Sunderland, MA.
HAMRICK, J. L., AND J. D. NASON. 1996. Consequences of gene flow in plants. Pp. 203-236 in O. E. Rhodes Jr., R. K. Chesser, and M. H. Smith, eds. Population dynamics in ecological space and time. Univ. of Chicago Press, Chicago.
HONG, Y. P., V. D. HIPKINS, AND S. H. STRAUSS. 1994. Chloroplast DNA diversity among trees, populations and species in the California closed cone pines (Pinus radiata, P. muricata and P. attenuata). Genetics 135:1187-1196.
LATTA, R. G., AND J. B. MITTON. 1997. A comparison of population structure across four classes of gene marker in limber pine. Genetics 146:1153-1163.
LEVIN, D. A. 1981. Dispersal versus gene flow in plants. Ann. Mo. Bot. Gard. 68:233-253.
LEVIN, D. A., AND H. W. KERSTER. 1974. Gene flow in seed plants. Evol. Biol. 7:139-220.
LINHART, Y. B., AND J. B. MITTON. 1985. Relationships among reproduction, growth rate, and protein heterozygosity in ponderosa pine. Am. J. Bot. 72:181-184.
LINHART, Y, B., J. B. MITTON, K. B. STURGEON, AND M. L. DAVIS. 1981. Genetic variation in space and time in a population of ponderosa pine. Heredity 46:407-426.
MCCAULEY, D. E. 1994. Contrasting the distribution of chloroplast DNA and allozyme polymorphism among local populations of Silene alba: implications for studies of gene flow in plants. Proc. Nat. Acad. Sci. USA 91:8127-8131.
MCCAULEY, D. E., J. E. STEVENS, P. A. PERONI, AND J. A. RAVEILL. 1996. The spatial distribution of chloroplast DNA and allozyme polymorphisms within a population of Silene alba (Caryophyllaceae). Am. J. Bot. 83:727-731.
MEAGHER, T. R. 1986. Analysis of paternity within a population of Chamelirium luteum. I. Identification of most likely male parents. Am. Nat. 127:199-215.
MILLAR, C. I., S. H. STRAUSS, M. T. CONKLE, AND R. D. WESTFALL. 1988. Allozyme differentiation and biosystematics of the Californian closed cone pines (Pinus subsect. Oocarpae). Syst. Bot. 13:351-370.
MILLIGAN, B. G. 1991. Chloroplast DNA diversity within and among populations of Trifolium pratense. Curr. Genet. 19:411-416.
MITTON, J. B. 1992. The dynamic mating systems of conifers. New For. 6:197-216.
MITTON, J. B., Y. B. LINHART, M. L. DAVIS, AND K. B. STURGEON. 1981. Estimation of outcrossing in ponderosa pine, Pinus ponderosa, Laws, from patterns of segregation in protein polymorphisms and from frequencies of albino seedlings. Silvae Genet. 30:117-121.
QUELLER, D.C., AND K. E GOODNIGHT. 1989. Estimating relatedness using genetic markers. Evolution 43:258-275.
ROHLF, E J., AND J. W. ARCHIE. 1978. Least squares mapping using interpoint distances. Ecology. 59:126-132.
RUCKELSHAUS, M. H. 1996. Estimation of genetic neighborhood parameters from pollen and seed dispersal in the marine angiosperm Zostera marina L. Evolution 50:856-864.
SCHNABEL, A., AND M. A. ASMUSSEN. 1989. Definition and properties of disequilibria within nuclear-mitochondrial-chloroplast and other nuclear-dicytoplasmic systems. Genetics 123:199-215.
SOKAL, R. R., AND N. L. ODEN. 1978. Spatial autocorrelation in biology. 1. Methodology. Biol. J. Linn. Soc. 10:199-228.
STACY, E. A., J. L. HAMRICK, J. D. NASON, S. P. HUBBELL, R. B. FOSTER, AND R. CONDIT. 1996. Pollen dispersal in low-density populations of three Neotropical tree species. Am. Nat. 148:275-298.
STRAUSS, S. H., Y. E HONG, AND V. D. HIPKINS. 1993. High levels of population differentiation of mitochondrial DNA haplotypes in Pinus radiata, muricata and attenuata. Theor. Appl. Genet. 86:605-611.
TARAYRE, M. 1996. Structure genetique des populations et systeme de reproduction chez une espece gynodioique, Thymus vulgaris L. Ph.D. diss. Universite de Montpellier, Montpellier, France.
WAGNER, D. B. 1992. Nuclear, chloroplast and mitochondrial DNA polymorphisms as biochemical markers in population genetic analyses of forest trees. New For. 6:373-390.
WASER, N.M. 1993. Population structure, optimal outbreeding and assortative mating in angiosperms. Pp. 173-199 in N. Wilmsen-Thornhill, ed. The natural history of inbreeding and outbreeding: theoretical and empirical perspectives. Univ. of Chicago Press, Chicago.
WOLFENBERGER, D. O. 1959. Dispersion of small organisms. Incidences of viruses and pollen, dispersion of fungus spores and insects. Lloydia 322:1-106.
WRIGHT, S. 1951. The genetical structure of populations. Ann. Eugen. 15:323-354.
ZAR, J. H. 1984. Biostatistical analysis. 2d ed. Prentice Hall, Englewood Cliffs, NJ.
|Printer friendly Cite/link Email Feedback|
|Author:||Latta, Robert G.; Linhart, Yan B.; Fleck, David; Elliot, Michael|
|Date:||Feb 1, 1998|
|Previous Article:||Fitness consequences of maternal and nonmaternal components of inbreeding in the gynodiodecious Phacelia dubia.|
|Next Article:||Evidence from the fossil record of an antipredatory exaptation: conchiolin layers in corbulid bivalves.|