Journal of Heredity Advance Access originally published online on July 4, 2007
Journal of Heredity 2007 98(4):367-372; doi:10.1093/jhered/esm040
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brief Communications |
Fine-Scale Spatial Genetic Structure within Continuous and Fragmented Populations of Trillium camschatcense
From the Graduate School of Environmental Earth Science, Hokkaido University, Sapporo 060-0810, Japan (Yamagishi and Ohara); and the Department of Biological Sciences, Tokyo Metropolitan University, Hachioji 192-0397, Japan (Tomimatsu)
Address correspondence to M. Ohara at the address above, or e-mail: ohara{at}ees.hokudai.ac.jp.
Spatial genetic structure (SGS) within populations was analyzed for the ling-lived understory perennial herb Trillium camschatcense using allozyme loci. We used Sp statistics to compare SGS between 2 life-history stages, juveniles (J) and reproductives (R), as well as between 2 populations, continuous and fragmented, with different habitat conditions. In the continuous population, significant SGS was detected in both stages but the extent was greatly reduced with the progress of the stage (J, Sp = 0.0475; R, Sp = 0.0053). We inferred that limited seed dispersal and subsequent random loss of individuals from the family patches are responsible for the J and R stage structures, respectively. The fragmented population differed in the patterns of SGS; significant structure was detected in the R stage, but not in the J stage (J, Sp = 0.0021; R, Sp = 0.0165) despite significant positive inbreeding coefficients (J, FIS = 0.251). The observed differences in the J-stage structures between populations may be explained by habitat fragmentation effects because reduced recruitment in the fragmented population prevents the development of maternal sibling cohort. Such comparative analysis between populations and life-history stages can be useful to understand the different underlying causes of SGS.
Spatial genetic structure (SGS) within plant populations is determined by many processes, including gene flow and local selection at different life-history stages (Hedrick 1986; Ennos 1994; Latta et al. 1998; Kalisz et al. 2001). In the case of long-lived perennial plant species, SGS of juveniles primarily reflects their parental structure and the patterns and distances of both pollen and seed dispersal. This structure can be modified in subsequent life-history stages by postdispersal selection or mortality of individuals. Therefore, analyses of SGS over the life cycle can infer the roles of these demographic and genetic processes in the formation and maintenance of the within-population genetic structure (Kalisz et al. 2001; Chung et al. 2003). For example, some studies have found that significant structure in juveniles is greatly reduced in reproductive adults, suggesting that limited dispersal and the subsequent random loss of individuals from the family patches are responsible for the juvenile and adult structures, respectively (e.g., Epperson and Alvarez-Buylla 1997; Parker et al. 2001; Chung et al. 2003). Conversely, the opposite trend was also observed in some species, and the increasing SGS has been attributed to local selection during recruitment or historical events (Tonsor et al. 1993; Ueno et al. 2000; Kalisz et al. 2001).
Moreover, the extent of SGS may vary among populations under different habitat conditions, and such interpopulation variation can also improve our understanding of the relationship between environmental factors and life-history factors. In particular, habitat alteration due to human activities affects many demographic and genetic processes including the breeding system (e.g., Rajimann et al. 1994) and recruitment (e.g., Tomimatsu and Ohara 2002; Ward and Johnson 2005), which are likely reflected in the patterns of genetic structuring. Several studies have revealed habitat-associated variations of SGS, for example, between fragmented and continuous populations of Acer saccharum (Young and Merriam 1994) and between old-growth and logged forests of Pinus strobus (Epperson and Chung 2001); however, few studies have examined the differences in SGS of different life-history stages both among and within populations (but see Epperson and Chung 2001; Cruse-Sanders and Hamrick 2004).
In this study, we analyzed SGS of 2 life-history stages in 2 populations of the long-lived understory herb, Trillium camschatcense Ker Gawler (Trilliaceae or Melanthiaceae). First, we compared SGS between juveniles and reproductives in a large, continuous population. Although the species has an outcrossing mating system in our populations (Ohara et al. 1996), the dispersal distance of seeds by ants has been reported to be short, only 0.60 m on the average (Ohara and Higashi 1987). Thus, we expect that the limited seed dispersal ability in our species develops significant SGS in juveniles. Juveniles are subsequently subject to density-dependent mortality and many are thus excluded from the family patches (Ohara and Kawano 1986), so that SGS in juveniles is expected to decrease or even become undetected in reproductives. Second, we compared the patterns of SGS in a large population with those of a small, fragmented population. In the Tokachi plain of eastern Hokkaido, Japan, T. camschatcense has experienced extensive habitat loss because of recent agricultural development since the 1880s and now is often found in the understory of small and isolated forests. Our previous studies (Tomimatsu and Ohara 2002, 2003a) using 14 populations with different sizes have demonstrated that smaller populations showed lower seed production and seedling recruitment and sometimes higher inbreeding coefficients (i.e., biparental inbreeding). Other things being equal, the reduced recruitment in small populations may prevent the development of maternal sibling cohorts and thus reduce SGS in juveniles, whereas biparental inbreeding is expected to intensify SGS. To examine these expectations, we used spatial autocorrelation analysis with new Sp statistics that has recently been proposed to make a comparative analysis of genetic structure (reviewed by Vekemans and Hardy 2004). Although previous studies used estimators of genetic relatedness (Moran's I or kinship coefficients) or the intercept of spatial correlograms to analyze their variations (e.g., Stehlik and Holderegger 2000; Diniz-Filho and Telles 2002), these estimators strongly depend on the sampling schemes adopted. By contrast, under isolation by distance in 2-dimensional space, the Sp statistics primarily depends on the rate of decrease of pairwise kinship coefficients between individuals with the logarithm of the spatial distance separating them.
 |
Materials and Methods |
|---|
 Top Materials and Methods ResultsDiscussionFundingReferences |
|---|
Study Species and Populations
Trillium camschatcense is a diploid (2n = 10), nonclonal perennial herb that commonly occurs in the understory of broad-leaved deciduous forests of Hokkaido, Japan. Vegetative individuals of T. camschatcense comprise 2 morphologically distinct forms: 1-leaf and 3-leaves stages. It takes more than 10 years for seedlings to become flowering plants. Reproductives have one or several flowers, and seed production results from obligatory outcrossing by insect pollination in our study area (Ohara et al. 1996). After reaching flowering, the plants can live for at least 25 years with continuous flowering every year (Ohara et al. 2006). The flowers are visited by a wide range of insects, primarily beetles (e.g., Nitidulidae and Melandryidae) and flies (e.g., Bibionidae and Scathophagidae; Tomimatsu and Ohara 2003b), and the fruits contain
80 seeds on average (Tomimatsu and Ohara 2002). As each seed has a soft, juicy elaiosome, the seeds are attractive to ants, which consequently contribute to seed dispersal (Ohara and Higashi 1987).
This study was conducted in 2 populations, Hiroo and Kiyokawa, in the Tokachi plain of eastern Hokkaido, Japan. In this study, we selected 2 distinct populations of 14 populations used in our previous studies (Tomimatsu and Ohara 2002, 2003a, 2004), one large continuous (Hiroo) and one small fragmented (Kiyokawa). The Hiroo population (42°19'N, 143°20'E; hereafter "continuous") grows in a large, continuous forest (300 000 m2), whereas the Kiyokawa population (42°45'N, 143°7'E; hereafter "fragmented") grows in a small, fragmented forest (7700 m2), fully surrounded by agricultural fields. The estimated habitat size and the number of reproductive plants are 50 000 m2 and 126 000 for the continuous and 7700 m2 and 10 200 for the fragmented population, respectively. Our previous study demonstrated that the fragmented population showed a significant inbreeding coefficient (FIS = 0.129), possibly due to biparental inbreeding (Tomimatsu and Ohara 2003a). Considering the size of habitats and the development history of the region, habitat conditions must be much more altered in the fragmented population. Thus, when interpreting the data, we assume that environmental conditions that gave rise to SGS in the reproductive and juvenile stages are similar in the continuous population, but they are quite different in the fragmented population owing to the fragmentation events.
Leaf Sampling and Allozyme Electrophoresis
In this study, we sampled individuals in 2 life-history stages, reproductives, that is flowering individuals (hereafter "R"), and juveniles, that is seedlings and one-leaf individuals (hereafter "J"). We did not sample 3-leaves nonflowering individuals because not enough number of them were found in the plots. In each population, leaf materials were collected in 2 ways. First, to examine the level of genetic diversity and inbreeding coefficient (FIS), leaves were randomly sampled from >40 individuals in each life-history stage (R, n = 41 and J, n = 94 for the continuous population; R, n = 52 and J, n = 94 for the fragmented population). Second, to analyze SGS, we established a single plot (2 x 10 m for the continuous and 2 x 16 m for the fragmented; Figure 1) in the center of each population and mapped all individuals of the R and J stages within the plots. For each stage, we collected leaf samples from plants located nearest all intersections of 2 x 2 cm grids within the quadrants unless there were no plants within 1 cm of the intersection point. Consequently, we sampled 151 of R and 754 of J plants in the continuous and 154 of R and 98 of J plants in the fragmented populations. Multilocus genotypes of all sampled individuals were determined by allozyme electrophoresis as described below. The density of R and J plants within the study plots was 7.6 and 71.9 m–2 in the continuous and 3.1 and 7.8 m–2 in the fragmented population, respectively.
|
Leaf materials were transported on ice to the laboratory and kept at –80 °C until electrophoresis. Approximately 70 mg of leaf tissue from R plants was homogenized in 0.9 ml of extraction buffer made up of 0.1 M Tris–HCl (pH 8.0), 0.2 g/ml glycerol, 63 mg/ml Tween 80, 8 mM dithiothreitol, 0.50% (v/v) ß-mercaptoethanol, 0.40% (w/v) ß-nicotiamide adenine dinucleotide, 0.45% (w/v) ß-nicotiamide adenine dinucleotide phosphate, 0.3% (w/v) bovine serum albumin, and 7% (w/v) polyvinylpolypyrrolidone (modified from Shiraishi 1988). As the leaves of J plants were small, we homogenized
2 mg of leaf tissue in 0.25 ml of extraction buffer. After the homogenates were centrifuged (15 000 rpm for 15 min at 4 °C), 10 µl of the resulting supernatant was used for vertical polyacrylamide gel electrophoresis. We tested 14 enzymes first, and then examined 6 enzyme systems that consistently showed clear and interpretable banding patterns: aspartate aminotransferase (AAT, 2 loci [EC 2.6.1.1
[EC]
]), leucine aminopeptidase (LAP, 1 locus [EC 3.4.11.1
[EC]
]), glutamate dehydrogenase (GDH, 1 locus [EC 1.4.1.2
[EC]
]), malate dehydrogenase (MDH, 1 locus [EC 1.1.1.37
[EC]
]), sorbitol dehydrogenase (SODH, 1 locus [EC 1.1.1.14
[EC]
]), and alcohol dehydrogenase (ADH, 2 loci [EC 1.1.1.1
[EC]
]). On the basis of the banding patterns, we assumed Mendelian inheritance of all loci.
Data Analysis
Standard measures of genetic diversity, including the percentage of polymorphic loci (P), the mean number of alleles per locus (A) corrected for differences in sample size (i.e., rarefied; Leberg 2002), the observed heterozygosity (Ho), and the expected heterozygosity (He) were calculated for each life-history stage in each population. Inbreeding coefficient (FIS) was calculated as estimates of Weir and Cockerham (1984) using the computer program FSTAT version 2.9.3.2
[EC]
(Goudet 2002), and the levels of significance were obtained by randomization-based procedures (for details, see Goudet 2002). To investigate fine-scale SGS, we conducted spatial autocorrelation analyses with kinship coefficients (Loiselle et al. 1995) using SPAGeDi version 1.2 (Hardy and Vekemans 2002). Mean multilocus kinship coefficients (F(j)) were computed for the following distance classes: 0.2, 0.3, 0.4, 0.5, 0.75, 1, 2, 4, and 8 (upper-bound distance in meters) and were plotted against the logarithm of the geographic distance (j). Standard errors for the kinship coefficients were estimated using a jackknife procedure over the loci. The linear regression slope (b) can be a good estimator of the extent of SGS (Vekemans and Hardy 2004). We tested the significance of b against the null hypothesis H0: b = 0 (i.e., the overall absence of SGS) by comparing the observed values with those obtained after 1000 random permutations of individuals among positions. We also quantified SGS by Sp statistics, calculated as –b/(1 – F(1)) where F(1) is the mean kinship coefficient in the first distance class (see Vekemans and Hardy 2004 for details). The Sp statistics could be considered robust in our sampling scheme because F(j) tended to decrease linearly with ln(j).
|
Results |
|---|
|
TopMaterials and MethodsResultsDiscussionFundingReferences |
|---|
Genetic Diversity and Levels of Inbreeding
Seven of the 8 loci resolved were polymorphic in at least one of the 2 populations examined: 6 (AAT [2 loci], ADH, SODH, LAP, GDH) and 4 (AAT [2 loci], ADH, LAP) loci were polymorphic in the continuous and fragmented populations, respectively; MDH was monomorphic. In both stages, the continuous population had higher genetic diversity (P, A, Ho, and He) than the fragmented population (Table 1). For example, when the differences in sample sizes were adjusted, the mean number of alleles per locus (A) was 1.875 for R and 2.227 for J stages in the continuous population and 1.620 for R and 1.484 for J stages in the fragmented population. In the continuous population, inbreeding coefficients were nearly equal to zero (R, FIS = 0.072; J, FIS = 0.079). In contrast, the fragmented population showed high inbreeding coefficients in both stages (R, FIS = 0.189; J, FIS = 0.251). Randomization tests detected a significant deficit of heterozygosity relative to Hardy–Weinberg expectations only in the J stage.
|
Spatial Genetic Structure
The slopes (b) of linear regressions between the kinship coefficient and the logarithm of the geographic distance were found to be significantly negative (P < 0.05) in all but the J stage in the fragmented population (Table 1; Figure 2). In the continuous population, the Sp statistics showed that the J stage was much more genetically structured than the R stage (J, Sp = 0.0475; R, Sp = 0.0053). In contrast, the fragmented population exhibited significant SGS at the R stage (Sp = 0.0165), but not in the J stage (Sp = 0.0021). When the strength of SGS of each life-history stage was compared between populations, the R stage in the fragmented population was slightly more structured than that in the continuous population (continuous, Sp = 0.0053; fragmented, Sp = 0.0165). In contrast, the J stage exhibited about 20-fold stronger structure in the continuous population than in the fragmented population (continuous, Sp = 0.0475; fragmented, Sp = 0.0021).
|
|
Discussion |
|---|
|
TopMaterials and MethodsResultsDiscussionFundingReferences |
|---|
This study indicates how fine-scale SGS of T. camschatcense differs across 2 life-history stages and 2 populations under contrasting habitat conditions. In addition, the fragmented population exhibited lower allelic diversity and higher inbreeding coefficients than the continuous population (Table 1). These results are consistent with our previous study (Tomimatsu and Ohara 2003a). The fragmented population would have experienced a stochastic loss of alleles at the time of fragmentation and biparental inbreeding due to localized pollen transfer.
In the continuous population, as expected, we found significant SGS in juveniles (Table 1; Figure 2A). This structure could develop as a result of limited pollen and seed dispersal (Wright 1946; Slatkin 1985). Although the distance of pollen dispersal would not be far, the value of inbreeding coefficient (FIS) indicates that mating within this population was essentially random (Table 1). Therefore, the results suggest that restricted seed dispersal is mainly responsible for SGS in juveniles. In fact, seeds that fell on the ground were rarely transported by ants even after 72 h and many remained around the maternal plants (dispersal frequency 11.7%; Yamagishi H, unpublished data). By contrast, the reproductive stage showed a weak structure relative to the juvenile stage (Table 1; Figure 2A). The reduction pattern of SGS with the progress of the stage could occur when juveniles experience genetically random mortality that occurs in a density-dependent manner. Several previous studies also demonstrated the same pattern, which was attributed to random mortality during recruitment (Hamrick et al. 1993; Epperson and Alvarez-Buylla 1997; Chung et al. 2003). It should be noted, however, that the opposite trend was observed in a related species of the same genus, Trillium grandiflorum (Kalisz et al. 2001); significant SGS was detected in reproductive but not in juvenile stages, and it was suggested that the increase in SGS between the stages could arise from local selection resulting from microenvironmental conditions or from specific historical events.
The fragmented population differs in the extent and pattern of SGS from the continuous population; namely, juveniles did not show significant SGS despite relatively high levels of inbreeding coefficients (Table 1; Figure 2B). Given our previous knowledge on the effects of habitat fragmentation, juvenile stage SGS could either increase due to biparental inbreeding or decrease due to reduced recruitment. Our results suggest that the low recruitment may be responsible for the absence of SGS. In this population, seedling recruitment is much more restricted than in the continuous population probably because of reduced seed production, microclimatic edge effects, and biparental inbreeding (Tomimatsu and Ohara 2002, 2004). Actually, the frequency of juveniles was only 2.5 times that of reproductives in the fragmented population (cf., 9.5 times that of the continuous population; Figure 1). This suggests that only a few maternal siblings exist among juveniles, and thus the low recruitment could explain the absence of genetic structure. Interestingly, a related species, T. grandiflorum, also did not show significant SGS at the juvenile stage (Kalisz et al. 2001). The number of seeds produced per reproductive plant is small in T. grandiflorum (16 seeds; Kalisz et al. 1999), so that the resulting low recruitment, in conjunction with a moderate level of seed dispersal, would have a homogenizing effect on the genetic structure (Kalisz et al. 2001). Although the difference between Sp statistics was not tested in our study, the reproductive stage SGS in the fragmented population was slightly stronger with lower adult density than that in the continuous population (Table 1; Figure 2). This result is consistent with theoretical and empirical evidence that density is a major determinant of SGS because it affects the strength of local genetic drift and the level of gene flow distance (Heywood 1991; Vekemans and Hardy 2004). We are not sure, however, whether this difference in SGS is significant because F(1) in these populations are similar (Table 1) and F(j) in the short distance classes had large variance in the fragmented population (Figure 2B).
Some limitations and caveats should be considered before we draw any conclusions from this study. First, we examined only a single plot within each population. Although we carefully established the plots to reflect typical environmental conditions and demographic structures of populations, 2 or more replicates are needed to draw more corroborative evidence. Second, the accuracy of Sp statistics critically depends on F(j) in the first several distance classes, particularly F(1) (Vekemans and Hardy 2004). For example, F(j) in the first 4 distance classes had large standard errors in the continuous population (Figure 2A) and this could partly be caused by how we defined distance classes. We recalculated F(1) and Sp statistics under different sets of distance classes, but these calculations gave essentially the same interpretation of our data (results not shown). Finally, in the fragmented population, we do not infer the causes of creating the strong pattern of structure observed in the reproductives from the pattern in the juveniles. Considering the long life-span of T. camschatcense (>50 years) and the population history, many reproductive individuals may still have survived fragmentation. Therefore, the existent structure in this reproductive population was likely to develop under habitat conditions before or immediately after habitat fragmentation, whereas the structure in the juveniles was established recently, long after the fragmentation events.
In summary, under a high level of recruitment, SGS appears at the juvenile stage probably because of limited seed dispersal around maternal plants, and subsequent random loss of individuals from family patches is a plausible explanation for the loss of SGS from reproductive to juvenile stages. However, under low recruitment due to habitat fragmentation, juveniles do not show a significant genetic structure possibly because only few maternal siblings exist. Although replicates of plots or populations may be needed to confirm our conclusions, such comparative analysis between populations as well as between life-history stages can be useful to understand the different potential causes of the SGS.
|
Funding |
|---|
|
TopMaterials and MethodsResultsDiscussionFundingReferences |
|---|
The Akiyama Memorial Foundation; the Foundation for Earth Environment; the Ministry of Education, Culture, Sports, Science; and Technology for the 21st Century Center of Excellence Program (E-01) and the Japan Society for the Promotion of Science for Scientific Research (Nos. 16370007 and 18405010).
|
Acknowledgments |
|---|
We are grateful to K. Nemoto for assistance in the field and Hiroo town office for offering the use of their facility. We also thank M. T. Kimura and G. Kudo for helpful advice.
|
Footnotes |
|---|
Corresponding Editor: David Wagner
Received May 6, 2006
Accepted April 2, 2007
|
References |
|---|
|
TopMaterials and MethodsResultsDiscussionFundingReferences |
|---|
-
Chung MY, Epperson BK, Chung MG. Genetic structure of age classes in Camellia japonica (Theaceae). Evolution (2003) 57:62–73.[ISI][Medline]
Cruse-Sanders JM, Hamrick JL. Spatial and genetic structure within populations of wild American ginseng (Panax quinquefolius L. Araliaceae). J Hered (2004) 95:309–321.
Diniz-Filho JA, Telles MPC. Spatial autocorrelation analysis and the identification of operational units for conservation in continuous populations. Conserv Biol (2002) 16:924–935.[CrossRef]
Ennos RA. Estimating the relative rates of pollen and seed migration among plant populations. Heredity (1994) 72:250–259.[ISI]
Epperson BK, Alvarez-Buylla ER. Limited seed dispersal and genetic structure in life stages of Cecropia obtufifolia. Evolution (1997) 51:275–282.[CrossRef][ISI]
Epperson BK, Chung MG. Spatial genetic structure of allozyme polymorphisms within populations of Pinus strobus (Pinaceae). Am J Bot (2001) 88:1006–1010.
Goudet J. FSTAT, a program to estimate and test gene diversities and fixation indices, version 2.9.3.2 [Internet][cited 2005 Oct 1] (2002) Available from: http://www.unil.ch/izea/softwares/fstat.html.
Hamrick JL, Murawski D, Nason JD. The influence of seed dispersal mechanisms on the genetic structure of tropical tree populations. Vegetatio (1993) 107–108:281–297.
Hardy OJ, Vekemans X. SPAGeDi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Mol Ecol Notes (2002) 2:618–620.[CrossRef][ISI]
Hedrick PW. Genetic polymorphism in heterogeneous environments: a decade later. Annu Rev Ecol Syst (1986) 17:235–255.
Heywood JS. Spatial analysis of genetic variation in plant populations. Annu Rev Ecol Syst (1991) 22:335–355.[CrossRef][ISI]
Kalisz S, Hamzawa FM, Tonsor SJ, Thiede DA, Voigt S. Ant-mediated seed dispersal alters pattern of relatedness in a population of Trillium grandiflorum. Ecology (1999) 80:2620–2634.[CrossRef][ISI]
Kalisz S, Nason JD, Hanzawa FM, Tonsor SJ. Spatial population genetic structure in Trillium grandiflorum: the roles of dispersal, mating, history, and selection. Evolution (2001) 55:1560–1568.[CrossRef][ISI][Medline]
Latta RG, Linhart YB, Fleck D, Elliot M. Direct and indirect estimates of seed versus pollen movement within a population of ponderosa pine. Evolution (1998) 52:61–67.[CrossRef][ISI]
Leberg PL. Estimating allelic richness: effects of sample size and bottlenecks. Mol Ecol (2002) 11:2445–2449.[CrossRef][Medline]
Loiselle BA, Sork VL, Nason JD, Graham C. Spatial genetic structure of a tropical understory shrub. Am J Bot (1995) 82:1420–1425.[CrossRef][ISI]
Ohara M, Higashi S. Interference by ground beetles with the dispersal by ants of seeds of Trillium species (Liliaceae). J Ecol (1987) 75:1091–1098.[CrossRef][ISI]
Ohara M, Kawano S. Life history studies on the genus Trillium (Liliaceae) IV. Stage class structures and spatial distribution of four Japanese species. Plant Species Biol (1986) 1:147–161.[CrossRef]
Ohara M, Takeda H, Ohno Y, Shimamoto Y. Variations in the breeding systems and the population genetic structure of Trillium kamtschaticum (Liliaceae). Heredity (1996) 76:476–484.[ISI]
Ohara M, Tomimatsu H, Takada T, Kawano S. Importance of life history studies for conservation of fragmented populations: a case study of the understory herb, Trillium camschatcense. Plant Species Biol (2006) 21:1–12.[CrossRef]
Parker KC, Hamrick JL, Parker AJ, Nason JD. Fine-scale genetic structure in Pinus clausa (Pinaceae) populations: effects of disturbance history. Heredity (2001) 87:99–113.[CrossRef][ISI][Medline]
Rajimann LEL, van Leeuwen NC, Kersten R, Oostermeijer JGB, den Nijs HCM, Menken SBJ. Genetic variation and outcrossing rate in relation to population size in Gentiana pneumonanthe L. Conserv Biol (1994) 8:1014–1026.[CrossRef]
Shiraishi S. Inheritance of isozyme variations in Japanese black pine, Pinus thunbergii Parl. Silvae Genet (1988) 37:93–100.
Slatkin M. Gene flow in natural populations. Annu Rev Ecol Syst (1985) 16:393–430.[ISI]
Stehlik I, Holderegger R. Spatial genetic structure and clonal diversity of Anemone nemorosa in late successional deciduous woodlands of Central Europe. J Ecol (2000) 88:424–435.[CrossRef]
Tomimatsu H, Ohara M. Effects of forest fragmentation on seed production of the understory herb Trillium camschatcense. Conserv Biol (2002) 16:1277–1285.[CrossRef]
Tomimatsu H, Ohara M. Genetic diversity and local population structure of fragmented populations of Trillium camschatcense (Trilliaceae). Biol Conserv (2003a) 109:249–258.[CrossRef]
Tomimatsu H, Ohara M. Floral visitors of Trillium camschatcense (Trilliaceae) in fragmented forests. Plant Species Biol (2003b) 18:123–127.[CrossRef]
Tomimatsu H, Ohara M. Edge effects on recruitment of Trillium camschatcense in small forest fragments. Biol Conserv (2004) 117:509–519.[CrossRef]
Tonsor SJ, Kalisz S, Fisher J. A life-history based study of population structure: seed bank to adults in Plantago lanceolata. Evolution (1993) 47:833–843.[CrossRef][ISI]
Ueno S, Tomaru N, Yoshimaru H, Manabe T, Yamamoto S. Genetic structure of Camellia japonica L. in an old-growth evergreen forest, Tsushima, Japan. Mol Ecol (2000) 9:647–656.[CrossRef][Medline]
Vekemans X, Hardy OJ. New insights from fine-scale spatial genetic structure analyses in plant populations. Mol Ecol (2004) 13:921–935.[CrossRef][Medline]
Ward M, Johnson SD. Pollen limitation and demographic structure in small fragmented populations of Brunsivigia radulosa (Amaryllidaceae). Oikos (2005) 108:253–262.[CrossRef][ISI]
Weir BS, Cockerham CC. Estimating F-statistics for analysis of population structure. Evolution (1984) 43:413–418.
Wright S. Isolation by distance under diverse systems of mating. Genetics (1946) 31:39–59.
Young AG, Merriam HG. Effects of forest fragmentation on the spatial genetic structure of Acer saccharum Marsh. (sugar maple) populations. Heredity (1994) 72:201–208.[ISI]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||