PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of annbotAboutAuthor GuidelinesEditorial BoardAnnals of Botany
 
Ann Bot. 2009 April; 103(6): 885–899.
Published online 2009 February 10. doi:  10.1093/aob/mcp018
PMCID: PMC2707893

Effect of altitude on the genetic structure of an Alpine grass, Poa hiemata

Abstract

Background and Aims

The persistence of plants inhabiting restricted alpine areas under climate change will depend upon many factors including levels of genetic variation in adaptive traits, population structure, and breeding system.

Methods

Using microsatellite markers, the genetic structure of populations of a relatively common alpine grass, Poa hiemata, is examined across three altitudinal gradients within the restricted Australian alpine zone where this species has previously been shown to exhibit local adaptation across a narrow altitudinal gradient.

Key Results

Genetic variation across six microsatellite markers revealed genetic structuring along altitudinal transects, and a reduction in genetic variation at high and low altitude extremes relative to sites central within transects. There was less genetic variation among transect sites compared with altitudinal gradients within transects, even though distances among transects were relatively larger. Central sites within transects were less differentiated than those at extremes.

Conclusions

These patterns suggest higher rates of gene flow among sites at similar altitudes than along transects, a process that could assist altitudinal adaptation. Patterns of spatial autocorrelation and isolation by distance changed with altitude and may reflect altered patterns of dispersal via pollen and/or seed. There was evidence for selfing and clonality in neighbouring plants. Levels of gene flow along transects were insufficient to prevent adaptive changes in morphological traits, given previously measured levels of selection.

Key words: Poa hiemata, genetic structure, altitudinal gradient, microsatellite, gene flow, climate change

INTRODUCTION

Plants have the potential to respond to climate change through plastic changes, adaptation and/or migration; however, the ability of many species to track optimal environments under the current rate of climate change is largely unknown (Jump and Penuelas, 2005). Adaptation and plastic responses are particularly important in plants occupying areas that are already geographically restricted such as alpine areas (Beniston, 2003). There is some evidence for shifts in the distribution of species inhabiting alpine ecosystems in response to global warming (Korner, 1999; Pauli et al., 2007). Re-visitation studies show an increase in vascular plant species richness on the summits of mountains, and mean (over 10-year periods) up-slope migration rates of 1–4 m were estimated in alpine-nival flora in the Alps (Korner, 1999). Other studies have documented the upward encroachment of subalpine woodland species into alpine herbfields (Grabherr et al., 1994; Lloyd and Fastie, 2003). This suggests that plants in alpine zones will need to adapt to changing climate conditions and competition with encroaching woodland species in order to persist.

Levels of gene flow between populations within mountainous regions appear to be highly variable. Restricted gene flow (Nem < 1) has been found between altitudinal populations of bromeliads (Alcantarea imperialis and A. geniculata; Barbara et al., 2007) whereas in pines (Pinus oocarpa) high rates of gene flow (Nem = 227) have been found (Saenz-Romero and Tapia-Olivares, 2003). High levels of gene flow might assist in adaptation through the spread of favourable alleles and promotion of high levels of genetic variation in marginal populations. However, high levels of gene flow in marginal populations can also limit adaptation because gene swamping prevents the maintenance of locally adapted individuals (Antonovics, 1976; Hoffmann and Parsons, 1991; Kirkpatrick and Barton, 1997).

As alpine habitats are typically heterogenous at small spatial scales with respect to many ecological factors (Korner, 1999), local patterns of genetic variation could be complex (Bennet, 1970; McGraw and Antonovics, 1983; Crawford and Abbott, 1994; Galen et al., 1997). Indeed, several studies on mountain plants have found significant genetic differentiation and structuring between populations (e.g. Aradhya et al., 1993; Wen and Hsiao, 2001; Reisch et al., 2003a, b; Hirao and Kudo, 2004). Diversity may change with altitude including higher levels of diversity within central (mid-altitude) compared with lower levels in peripheral (low and high altitude margins) populations (e.g. Isik and Kara, 1997; Saenz-Romero and Tapia-Olivares, 2003; Ohsawa et al., 2007). If diversity levels are generally higher mid-altitude than at extremes, gene swamping might prevent adaptation in peripheral alpine populations where selection is likely to be most intense under climate change.

The ability of an organism to persist and/or migrate in response to varying conditions depends upon reproductive strategies as well as gene flow and genetic variation for traits under selection. Arctic and alpine tundra communities contain a large number of plant species that can reproduce clonally (Billings and Mooney, 1968; Bliss, 1971). Clonality enhances individual longevity and maintains well-adapted genotypes; however, sexual reproduction allows evolution and adaptation to changing environmental conditions as well as the colonization of new areas (Jackson et al., 1985; Korner, 1999). Mixed reproductive systems may be advantageous for plants in alpine environments that are variable (Korner, 1999) and clonality might be favoured at mid-altitudes where the colonization of new areas is less important than at margins.

In this paper, genetic diversity, structure and gene flow are assessed in a restricted alpine grass, Poa hiemata, across three replicated altitudinal transects. This species has been described in the literature as having the potential for a mixed reproductive system (Vickery, 1970), but there is no direct evidence for different reproductive modes in the field. Although pollination and seed dispersal mechanisms have not been directly tested in Poa hiemata, this species is likely to be wind and insect pollinated (Inouye and Pyke, 1988; Friedman and Harder, 2004) with seeds dispersed by wind, rain and birds as is common for many alpine forbs and grasses (Chambers, 1995; Tackenberg and Stocklin, 2008). Poa hiemata shows localized adaptation and differentiation in fitness-related traits across its altitudinal range (Byars et al., 2007). Gene flow along transects might restrict adaptive differentiation, although it might be enhanced if there is gene flow across sites found at similar altitudes. The following questions are addressed. (a) To what extent are sites along an altitudinal gradient genetically separated from each other and from isolated populations of Poa hiemata? How large are genetic differences among altitudes when compared with differences among transects across slopes, and how do these patterns compare with local adaptation in Poa hiemata? (b) Do range margins have relatively lower levels of genetic variability compared with central populations? (c) What breeding system(s) (sexual/clonal) is/are present in field Poa hiemata? Does the presence of clonality increase at peripheral populations (i.e. potentially more stressful conditions) and/or in populations that are isolated? (d) Is there directional gene flow? (e) Can gene flow constrain the development of adaptive differences between high and low altitude sites given previously observed levels of selection? Answers to these questions are linked to patterns of adaptive differentiation in Poa hiemata for quantitative traits.

MATERIALS AND METHODS

Species and sample collection

Species and sites

Poa hiemata (soft snow-grass) is widespread within alpine and subalpine zones of Australia above 1500 m, and is common on the Bogong High Plains (Alpine National Park) in North-eastern Victoria and also the Kosciuszko Alpine National Park in New South Wales (Vickery, 1970). The Australian alpine zone is particularly restricted in terms of high-altitude retreat compared with other such areas around the world, as there are no areas above the alpine zone (Brereton et al., 1995; Hughes, 2003). Plants flower only once per season, are bisexual and exhibit panicle inflorescences, hermaphrodite spikelets and vegetative growth through rhizomatous shoots (Vickery, 1970). The genus of Poa is large with over 200 species in temperate-to-cold climates; 34 native Australian species are recognized (Vickery, 1970).

The altitudinal sites used in this study are described in detail in Byars et al. (2007) and plotted in Fig. 1. Briefly, three east-facing altitudinal transects around 1 km long were placed along three mountains within the Bogong High Plains, Alpine National Park, Australia. Transects were selected for their comparability (similar in species composition, aspect and altitudinal range of approx. 1700–1880 m a.s.l. – typical of high alpine ranges in Australia). The transect sites were at Mount Cope (MC; high altitude, 147°16′E 36°55′S, 1847 m, to low altitude, 147°17′E 36°55′S, 1700 m), Mount Nelse Central (MNC; high altitude, 147°20′E 36°50′S, 1880 m to low altitude, 147°20′E 36°49′S, 1780 m) and New Country Spur (NCS; high altitude, 147°21′E 36°50′S, 1877 m to low altitude, 147°21′E 36°49′S, 1746 m). Sampling sites of P. hiemata along each of the transects represented continuously distributed populations; beyond both the high and low altitude extremes of the transects they became fragmented as range margins were approached. One environmental variable that has been previously quantified along the transects is temperature; averages from data loggers (at 100-m intervals) along each transect taken between November and April 2005/2006 indicated mean temperatures at the low altitude sites were consistently 1°C (or more) higher compared with the cooler high altitude sites measured over the snow-free growing season (Byars et al., 2007). Many other variables that were not quantified are likely to vary along such altitudinal transects including depth and duration of snow, wind regime, soil conditions and many others (Körner, 1999). Also, two fragmented populations were included near the New Country transect (designated New Countryisolated; 147°21′E 36°49′S, 1700 m) and Nelse Central transect (designated Nelse Centralisolated; 147°21′E 36°50′S, 1660 m). These fragments were below the tree-line but represented open areas where P. hiemata were found. Two other populations were sampled from Mount Feathertop (MF; 147°7′E 36°53′S, 1760 m) and Mount Bogong (MB; 147°17′E 36°44′S, 1920 m; Fig. 1).

Fig. 1.
Location of the three altitudinal transects (New Country Spur, Mount Nelse Central and Mount Cope), isolated populations [New Country isolated (NCSisolated); Nelse Central isolated (MNCisolated)], Mount Feathertop and Mount Bogong within the Alpine National ...

Population sampling

Leaf tissue was collected from a minimum of 20 individual plants randomly sampled within 50-m2 sample plots in December 2006. Three or four young leaves per plant were removed to 2-mL Eppendorf tubes and placed in a cooled cryogenic container in the field before transport, longer-term storage (at –80°C) and DNA extraction. This was repeated every 100 m from high to low altitude sites along each of the 1-km-long altitudinal transects (New Country, Nelse Central and Cope; Fig. 1) giving a total of 11 sample sites (33 in total) or 220 individuals per transect (660 in total across all transects). Also 20 individuals each from the two isolated (Nelse Centralisolated, New Countryisolated) and the two sites from other peaks (Fig. 1) were randomly sampled. Field sampling was also used to assess the density of P. hiemata (per m2) using 30 randomly assigned 1 m2 quadrats within a 50-m2 area at each site along all transects.

Genetic analysis

DNA extraction and microsatellite analysis

DNA was isolated from frozen plant material by a CTAB (cetyl trimethyl ammonium bromide) method (Rogers and Bendich, 1994) using approx. 20–30 mg (wet weight) of leaf material. After re-suspension in 50 µL TE buffer (10 mm Tris–HCl, pH 8·0; 1 mm EDTA, pH 8·0), the concentration was estimated spectrophotometrically (GeneQuant Pro) at 260 nm.

An initial screen of potential candidate microsatellite loci was undertaken to identify markers with a high reproducibility. Three of the six microsatellite markers used in this study were assayed for the first time in Poa hiemata by cross-species amplification from the related species Poa alpina (Maurer et al., 2005), two were obtained from a study in Lolium and Poa species (Kindiger, 2006) and one from Poa pratensis (Albertini et al., 2003) (Table 1). As primers were not designed specifically for use in Poa hiemata, initial gradient polymerase chain reaction (PCR) analyses were performed on the Mastercycler® Gradient (ranging from 50–65°C) from Eppendorf to determine optimal annealing temperatures and MgCl2 concentrations. Optimization also helped remove the presence of non-specific binding in initial PCR reactions with these primers.

Table 1.
Polymorphic microsatellites adapted from various sources successfully used with Poa hiemata

All populations were genotyped for six microsatellite loci in PCR reactions with 10-μL volumes containing 1 µL of genomic DNA (approx. 10–20 ng), 1 µL 10× PCR amplification buffer (Bioline), 0·3 µL 50 mm MgCl2 (1·5 mm optimal for all loci; Bioline), 0·4 µL 1·5 mm dNTPs (New England BioLabs), 0·5 µL 100× purified bovine serum albumin (New England Biolabs), 0·1 µL of Immolase™ DNA polymerase (Bioline), 0·3 µL 10 mm forward primer end-labelled with [γ-33P]ATP (prepared following the protocol in Sambrook et al., 1989), 0·1 µL 10 mm unlabelled forward primer, 0·12 µL 10 mm reverse primer and 6·18 µL Milli-Q H2O. PCR cycling conditions were the same for all loci (except for annealing temperature), with initial denaturation at 95°C (7 min), followed by 30 cycles of 95°C (45 s), primer annealing (48–60°C; see Table 1) (45 s), and extension at 72°C (60 s) with a final extension step of 72°C for 8 min. To test the reproducibility of allelic profiles and minimize errors associated with PCR and scoring of alleles, 30 individuals were chosen at random and re-amplified at all six microsatellite loci and scored as above. Re-amplified samples matched the original genotypes obtained for all loci. Amplified fragments were separated through a 5% denaturing polyacrylamide gel [per gel, 216 g urea (8m final concentration), 57 mL 40% acrylamide (19 : 1 ratio of acrylamide to bisacrylamide, Bio-Rad®), 45 mL 10× TBE, 177 mL Milli-Q H20; gel size, 32 × 39 cm] at 65 W for 2–4 h and exposed for 24–72 h (depending on levels of radioactivity) to autoradiography film (Fuji medical imaging film). Alleles were also sized with a λgt11 ladder (fmol DNA Cycle Sequencing System, Promega). Approximately ten samples were repeated from one gel run to the next (co-loaded as an allelic ladder) to allow another standardization of allele size.

Statistical analysis

Genetic variation

To estimate overall genetic variation, the following measures of genetic diversity were calculated per site and transect using Fstat version 2·9·3 (Goudet, 1999) after adjusting repeating identical multilocus genotypes (IMGs) (see section below for calculation) to one sample: allelic richness (A, mean number of alleles per locus based on the minimal sample size), observed heterozygosity (HO) and expected heterozygosity (HE) for each locus and site. Departures from Hardy–Weinberg (HW) equilibrium were tested by Fisher's exact tests (Genepop version 3·4; Raymond and Rousset, 1995). The presence of linkage disequilibrium was tested for all pairs of markers within each site using the log-likelihood ratio G statistic (with 3300 permutations) implemented in Fstat. All tests involving multiple comparisons were corrected at the table-wide α = 0·05 level with the Dunn–Šidák method (Sokal and Rohlf, 1995).

Spatial analyses

Mantel tests were performed to investigate the correlation between matrices of genetic and geographic (estimated from GPS co-ordinates in ArcView™, ESRI®) distances with two methods. First, the correlation was computed at the site level whereby genetic similarity based on the overall fixation indices (FST)/(1 – FST) calculated between sites within and across transects (dependent variable) was compared with matrices of logarithmically transformed geographic distances (independent variable) in Fstat. The significance of inter-site relationships was evaluated with 2000 randomizations in each case. Secondly, the correlation was computed at the individual level, comparing tri-matrices of genetic and geographic distances between samples along transects using Mantel tests in GenAlEx version 6 (Peakall and Smouse, 2006). Both tests were performed with and without repeated cases of clonality to determine clone influence on spatial genetic structure. Genotypic data from sites within transects were also analysed using spatial autocorrelation analyses in GenAlEx. The autocorrelation coefficient (r) provides a measure of genetic similarity between pairs of individuals whose geographic separation falls within the specified distance class. A distance class of 30 m for all analyses was chose to ensure that the first distance class contained only within-territory comparisons. Tests for significance were also performed using random permutation and bootstrap estimates of r, with the number of iterations set to 1000 for each.

Population structure and differentiation

The fixation index (FST; Weir and Cockerham, 1984) was estimated among pairs of altitudinal sites within transects and between isolated sites to those within transects relating to the same mountain using Fstat. The statistical significance of FST estimates was tested by 1000 random permutations of individuals across sites with Genetix version 4·03 (Belkhir et al., 2004). Estimates per transect were also obtained. To partition the genetic variance among transects, among sites within transects, and within sites, an analysis of molecular variance (AMOVA, Excoffier et al., 1992) was conducted in Arlequin (Schneider et al., 2000). Because most of the differentiation was between the altitudinal extremes of each transect, each high and low altitude site from each transect was also compared in an AMOVA. As RST estimates gave very similar results to FST estimates in terms of the magnitude of differences between sites and transects, only the latter are presented. The geographical and altitudinal distribution of genetic variability was further investigated using factorial correspondence analysis as implemented in Genetix, and the first two underlying factors that explained the majority of the variation within the multilocus genotypes were plotted. For comparison, the Bayesian clustering method implemented in Structure we run (Pritchard et al., 2000) that clusters individuals using Markov chain Monte Carlo (MCMC) methods. For determining the most appropriate number of clusters (K), the admixture model was applied with ten runs for each K value from 1 to 15, each run comprising a burn-in period of 30 000 iterations followed by 100 000 iterations.

Migration versus adaptive divergence

As gene flow estimates from Wright's model [FST = 1/(1 + 4Nm)] have often been criticized for underlying assumptions and conditions that do not typify many natural systems (Hutchison and Templeton, 1999), two alternative approaches were carried out. Migration was calculated within transects (between altitudinal sample sites) and between transects with Slatkin's private allele method (NemSlatkin) (Slatkin, 1985) and the MCMC-based maximum likelihood method (NemBeerli) (Beerli and Felsenstein, 1999, 2001) as implemented in the programs Genepop and Migrate, respectively. Effective population size was also calculated as a function of the mutation rate (Θ = 4Neμ) in Migrate. These were then converted to Ne, assuming a mutation rate of 10–4 per locus per generation found in other related species in the family Poaceae (Thuillet et al., 2002; Vigouroux et al., 2002). For simulations in Migrate, ten short and three long chains were run, with short sampling increments of 20 for each. The number of discarded trees per chain (‘burn-in’) was set to 10000. Metropolis-coupled MCMC (‘MCMCMC’) or ‘heating’ was applied for auxiliary searches with more permissive acceptance criteria, to act as ‘scouts’ for the main analysis (Kuhner, 2006). The search was run with four chains at different temperatures (1·0, 1·2, 1·5, and 3·0) with an adaptive heating scheme that manipulated these temperatures according to their swapping success.

To test if adaptive divergence between high and low altitude sites (separately within each transect) might be constrained by levels of gene flow, a quantitative model from eqn 7 in Hendry et al. (2001) was applied: D*ms = Dθ[(G)/(G(1−m) + (ω2 + P)m)], where D*ms is the mean trait value after mixing and selection. Parameter definitions (Dθ, G, P, ω – in the simulations ω was substituted for βavggrad, which is the selection gradient acting on the trait) are described in detail in Hendry et al. (2001) and were estimated from Byars et al. (2007). The rate of migration (m) was obtained directly from Migrate and Slatkin's rare allele method by dividing Nem by estimates of Ne (Waples, 1991, in Hendry et al., 2001). The equilibrium difference in adaptive trait value between high and low populations was estimated when mixing occurs before selection (more appropriate for sessile organisms) and hence the degree to which adaptive divergence might be constrained by gene flow.

Comparing altitudes

To examine potential changes in population genetic parameters [A, HE, clonality, ŝ(ML), FIS, Ne] and plant density with altitude, ANOVAs on the dataset combined across transects were undertaken, and terms for the transects included as well as linear or polynomial terms for altitudinal position within transects. By using altitudinal position within transects (rather than altitude itself), differences were controlled between transects with respect to the altitudes at which plants were sampled. This allowed the relative importance to be examined of how position (centre versus range margins) within transects affects genetic parameters. Alternatively, how transects differed for these estimates by using altitude as a covariate was also examined. Post hoc Tukey b tests were undertaken to examine which sites differed significantly for genetic diversity estimates. All analyses were carried out with SPSS for Windows version 14

Clonality and selfing estimates

To test for the presence of clonality, banding patterns were examined to assess the number of different and identical multilocus genotypes in each site with Genclone 1·0 (Arnaud-Haond and Belkhir, 2007). The proportion of IMGs was calculated as G/N, where G is the number of identical genotypes and N is the number of samples per site. When the same genotype was detected more than once in a sample site, the probability (Psex) that the samples actually originated from distinct reproductive events (i.e. from separate genets) was estimated from the binomial expression (Tibayrenc et al., 1990; Parks and Werth, 1993). The clonal sub-range (i.e. the maximum distance in metres between two identical genotypes belonging to the same clone) (Harada et al., 1997; Alberto et al., 2005) was estimated for each site. All estimates of clonality and probabilities were calculated with Genclone 1·0. For most of the other tests described, an adjusted data set was used with repeating IMGs reduced to a single representative.

To obtain estimates of selfing, a method based on maximum likelihood of the whole distribution [ŝ(ML)] was applied (David et al., 2007), which potentially avoids bias in the presence of incorrectly scoring microsatellites due to null alleles or short-allele dominance that may occur when estimates of selfing are based on inbreeding coefficients (FIS) (Zouros and Foltz, 1984; Bonin et al., 2004; Pompanon et al., 2005; David et al., 2007). However, inbreeding coefficients (FIS; Weir and Cockerham, 1984) were also calculated due to their wide use in (and for comparison to) previous studies. Both estimates were calculated per site (20 individuals) to assess variability and per transect (220 individuals) for more accurate estimates of inbreeding.

RESULTS

Genetic variation

A total of 99 alleles at six microsatellite loci was detected across 740 individuals at all sites. Mean number of alleles across all loci (mean, 16·5; range, 7–37) demonstrated a high level of polymorphism. Allelic richness (A) and both expected and observed heterozygosities (HE, HO) were variable across sites within transects (A, 4·5–8·3; HO, 0·45–0·89; HE, 0·515–0·769; see Appendix) but were similar in the different transects (Nelse Central, etc). HE estimates for the isolated sites (Nelse Centralisolated, New Countryisolated) were at the lower end of the range compared with values at other sites (Appendix).

Significant departures from HW equilibrium across loci (reflecting heterozygote deficiency) were detected in 10 of 37 (or 27·03%) sites after corrections for multiple comparisons. Of the 222 locus per site tests for HW equilibrium, 55 (or 24·77%) were significant after corrections for multiple comparisons including eight at locus CA1D4, 19 at locus CA1F4, nine at M16-B, two at AE5 and 17 at GAC1. As no locus displayed consistent departures from HW equilibrium across all sites, it is likely that these instances reflect occasional departures from random mating rather than the presence of null alleles. No consistent linkage disequilibrium was found between any loci across sites, with only nine out of 555 pair-wise tests (across all sampled sites) significant after correction for multiple comparisons.

Altitudinal variation

When data were examined across transects in an ANCOVA with mountain as a fixed factor, highly significant differences were found between altitudinal position and both allelic richness (A, F1,33 = 8·72, P = 0·006) and expected heterozygosity (HE, F1,33 = 15·16, P < 0·001) when a polynomial expression was used. The linear expression was also significant for both, but less so (A, F1,33 = 4·64, P = 0·040; HE, F1,33 = 9·93, P = 0·004). Plants central within transects tended to have higher diversity when compared with plants found at range margins, and plants at high altitude sites tended to have relatively lower diversity overall (Fig. 2A, B). When altitude was used as the covariate, differences in the intercept of the altitudinal relationships were present for A (F1,33 = 8·66, P = 0·007) and HE (F1,33 = 31·36, P < 0·001; Fig. 2C, D). No significant relationships were found between estimates of clonality, inbreeding, effective population size or plant density with altitudinal position within transects.

Fig. 2.
Combined (A, B) and separate (C, D) polynomial regressions of expected heterozygosity (HE) and allelic richness (A) for all transects (NCS, New Country Spur; MNC, Mount Nelse Central; MC, Mount Cope) against altitudinal position within transects and altitude ...

There was a significant difference between transects for clonality (F2,33 = 3·87, P = 0·033) and inbreeding (FIS, F2,33 = 20·59, P < 0·001). This reflected slightly higher levels of clonality and negative inbreeding coefficients at lower altitudes at Cope transect compared with lower yet positive values, respectively, at Nelse Central transect (Appendix). For inbreeding (FIS) and clonality there was also a difference in the slope of the altitudinal relationships as reflected by a significant interaction between altitude and transect (clonality, F2,33 = 4·01, P = 0·030; FIS, F2,33 = 20·47, P < 0·001).

Spatial analyses

Mantel tests for correlation between genetic and geographic distances were significant between sites (New Country, r2 = 0·664, P < 0·001; Nelse Central, r2 = 0·559, P < 0·001; Cope, r2 = 0. 589, P < 0·001) and individuals (New Country, r2 = 0·124, P < 0·001; Nelse Central, r2 = 0·139, P < 0·001; Cope, r2 = 0·104, P=< 0·001) suggesting an isolation by distance effect across all transects. These values remained significant when these tests were run with repeated clonal replicates removed.

Spatial autocorrelation analyses (with repeated clonal replicates removed) of each transect revealed significantly positive r-values ranging from 30 m to 120 m distance classes for both New Country and Nelse Central transects and from 30 m to 210 m distance classes for Cope transect (Fig. 3).

Fig. 3.
Correlograms of the genetic correlation coefficient (r) as a function of distance along New Country, Nelse Central and Cope transects. The permuted 95% confidence interval (dashed lines) and the bootstrapped 95% confidence error bars are also shown. Distance ...

Genetic structure

Highly significant levels of pair-wise differentiation (FST) were found between sites 1 km apart down to sites only 100 m apart (Table 2) within all transects. Local differentiation was most apparent at Cope transect where all 100 m pair-wise values were significant. Highest differentiation occurred at sites near margins of transects compared to those centrally situated (Table 2) and this became obvious when all pair-wise values were averaged for each particular site within transects and compared against altitudes (Fig. 4).

Fig. 4.
Polynomial regressions between genetic distance (pairwise FST) and altitude for each transect. For each sample site, FST values are based on an average of all pair-wise site differences between that site and all other sites within each transect separately. ...
Table 2.
Matrix of pairwise differentiation (FST; lower diagonal) and migration (Nem; upper diagonal) based on Slatkin's (1985) private allele method for transects of Poa hiemata between sites at (A) New Country, (B) Nelse Central and (C) Cope

Comparisons between each of the two isolated populations and altitudinal sites within transects from the same mountain revealed FST values ranging between 0·016 and 0·138 for Nelse Centralisolated (P < 0·01 for 6 of 11 comparisons) and FST values between 0·006 and 0·103 for New Countryisolated (P < 0·01 for 5 of 11 comparisons), indicating moderate genetic differentiation. Consistent with this pattern of altitudinal genetic differentiation, the spatial analyses in Genetix and Structure both showed clear graphical separation between high- and low-altitude sites within transects, and less clear separation between sites located centrally within transects (results not shown).

The results of the AMOVA analysis with data from all sites within transects (Table 3A) showed highly significant differentiation at all levels tested (all P-values < 0·001). There was significant variation among sites within transects (4·45%) and less (yet significant) variation among transects (1·73%), with the remaining variation attributable to the within-site component. When only sites at range margins at each transect were considered, the amount of variation among transects became very small and non-significant and the value among sites within transects increased noticeably (13·63%) (Table 3B).

Table 3.
Analysis of molecular variance (FST based) comparing genetic variation between (A) all altitudinal sites across transects (33 sites total), and (B) high and low altitude sites across transects (six sites total) for Poa hiemata

Migration versus adaptive divergence

Estimates of number of migrants from both methods (NemSlatkin, NemBeerli) showed similar changes in magnitude when sites within transects and between transects were compared. Generally higher levels were found between transects (NemSlatkin 5·01–9·41; NemBeerli 3·10–7·41) than between altitudes within transects (NemSlatkin 0·005–2·57; NemBeerli 0·72–3·57) (Fig. 5). Estimates of gene flow between sites at transect margins (high, low) were often smaller than values to and from central sites, suggesting less-restricted gene flow centrally within transects, although these values were also small (Fig. 5). The majority of migration estimates between altitudes (17/27 based on NemSlatkin; Table 2) were greater than 1. Estimates of Ne (calculated from Θ) were similar across altitudes (apart from NCSH) and when averaged per transect (NCS, 1855; MNC, 1458; MC, 1804; Appendix). By combining estimated rates of gene flow (m) from this study with selection coefficients from Byars et al. (2007) into the quantitative model in Hendry et al. (2001), a clear shift in trait values was found after a generation of mixing and selection at high and low altitudes (leaf length, D*ms ranging 0·94–2·49 cm; plant circumference, D*ms ranging 1·28–2·13 cm across all transects).

Fig. 5.
Migration estimates for Poa hiemata between altitudinal sites within transects and between transects based on (A) maximum likelihood estimates (NemBeerli; Beerli and Felsenstein, 1999) and (B) rare alleles method (NemSlatkin; Slatkin, 1985). Key: MCH, ...

Clonality and selfing

The frequency of IMGs (also referred to as clones) ranged from 0% to 40% when sites were compared within transects and was high at the two isolated sites (Nelse Centralisolated, 40%; New Countryisolated, 25%), but clones were absent at Bogong or Feathertop sample sites (Appendix). The chances of obtaining the same multi-locus genotype by sexual recombination were considered small (all P < 0·01). The same clones were never found at different sample sites along transects. Instead, distances between the same clones were small and varied from 4 m to 10 m within the 50-m2 sample plots.

The majority (34/37 sample sites) of selfing ŝ(ML) estimates were small and non-significant (Appendix). Estimates per transect were low [ŝ(ML) – New Country, 0·039; Nelse Central, 0·018; Cope, 0·066] yet significant at Cope transect (P < 0·01). Estimates of FIS suggested a higher incidence of inbreeding, with significance found at 10 of 37 sites (Appendix). FIS estimates per transect (New Country, 0·159; Nelse Central, 0·199; Cope, 0·119) were all significant (all P-values < 0·01). Reducing cases of clonality to a single representative had little effect on size or significance of inbreeding estimates.

DISCUSSION

Genetic variability

Estimates of genetic diversity based on microsatellites between sample sites were generally high in Poa hiemata, and similar to levels found in other studies using various molecular markers on plants from mountainous regions (e.g. Sun and Salomon, 2003; Woodhead et al., 2005; Heuertz et al., 2006; Barbara et al., 2007; Ohsawa et al., 2007; Schneller and Liebst, 2007). RAPD markers indicated a lower level of diversity (HO ranging from 0·133 to 0·199) for a closely related species, Poa fawcettiae, that shares the same habitat as P. hiemata (James et al., 1997). This could reflect differences in resolution of molecular marker systems and/or life history differences between these species. In sites that were isolated from the larger more continuous sites of P. hiemata, higher estimates of clonality and lower diversity levels were obtained, suggesting the presence of potentially negative effects of colonization and fragmentation on diversity in these isolated sites.

Patterns of genetic variation in P. hiemata were associated with altitudinal positioning in the three transects. Higher genetic diversity was found at mid-altitudes or at sites central within transects compared with lower levels in peripheral (high and low altitudes) sites. Similar patterns of diversity with altitude have been found in other studies on alpine plants along altitudinal gradients (e.g. Young et al., 1993; Schaal and Leverich, 1996; Isik and Kara, 1997; Lammi et al., 1999; Jump et al., 2003; Saenz-Romero and Tapia-Olivares, 2003) and may occur through loss of diversity in peripheral populations as they are more geographically isolated from the central part of the range (Brown, 1984), due to reduced gene flow and genetic drift (Kapralov, 2004; Ohsawa et al., 2007). Alternatively, loss of diversity may occur through natural selection if loci are linked to markers under selection (Amos and Harwood, 1998), although this would only be expected to influence some microsatellite markers. In the recent study by Byars et al. (2007) on the quantitative genetic variation in Poa hiemata along these same altitudinal gradients, local adaptation to high and low altitude was detected leading to differentiation in fitness-related traits.

Reproductive strategies

Findings from this study support the mixed mating system previously inferred (Vickery, 1970) for P. hiemata. This type of mating system may be typical of alpine grasses in the Australian Alps as it was also suggested possible for P. fawcettiae based on a molecular (RAPD) marker study (James et al., 1997). Evidence consistent with a predominantly outcrossing mating system in P. hiemata includes the fact that most genetic diversity was found within sites (93·82% of the total genetic variance), and expected heterozygosity was high, typical of diploid outcrossers (Loveless and Hamrick, 1984). Estimates of genetic differentiation (global FST range 0·030–0·061) and gene flow (Nem was generally >1) fell within the range of values expected in outcrossing plants or plants with mixed mating systems as described in the literature (Morjan and Rieseberg, 2004). Most of the sites studied here were also in HW equilibrium. Evidence for alternative strategies of reproduction comes from the presence of relatively low levels of inbreeding yet clonal replication at some of the sample sites.

The clonality found in the present study matches with the general finding that a high proportion of alpine plant species may have the potential for clonal replication (Stocklin and Baumler, 1996; Klimes et al., 1997). The distance between clones in P. hiemata was never greater than 10 m and clones were never found across sampled sites, suggesting clonal patches are restricted and their occurrence fairly sporadic. Most estimates of inbreeding were relatively low, yet significant values were found with both methods applied even when clonal replicates were removed, suggesting occasional mating between relatives. Shifts in reproductive strategies with altitude have been documented in some alpine forbs (Young et al., 2002) and other plants inhabiting alpine areas (Douglas, 1981; Bauert, 1996); however, there was no consistent (across transects) association of clonality or inbreeding with altitude in the present study. Differences in reproductive strategies may be related to other conditions apart from altitude or position within a continuous population. More reliable estimates of selfing could help elucidate this (i.e. progeny arrays; Ritland and Jain, 1981).

Genetic structure and differentiation

The microsatellite data indicate significant genetic differentiation (FST) between altitudinal sites of P. hiemata, in the three transects. Similar levels of differentiation have been found between altitudes in other studies (e.g. Murawski and Hamrick, 1990; Hirao and Kudo, 2004). Levels of pair-wise differentiation were higher in sites at range margins compared with sites central within transects, suggesting limited gene flow across different altitudes. Consistent with this result, Ohsawa et al. (2007) found higher levels of differentiation among peripheral populations of Quercus crispula and Gapare et al. (2005) found high gene flow among core populations of Picea stichensis and lower values among peripheral populations along altitudinal gradients.

Genetic differences in P. hiemata among sites may have developed through limited gene flow. Some other studies have found that isolation-by-distance can occur over short distances in alpine plants. For instance, Hirao and Kudo (2004) found significant correlation between geographic and genetic distance in an alpine plant across a snowmelt gradient <1 km long. Most seeds of arctic–alpine plants (including those of P. hiemata) lack wings and this is likely to limit dispersal. In an analysis of the arctic–alpine flora of Fennoscandia, 70% of a total of 251 species were classified as having no adaptations to long-distance dispersal, 23% as wind-dispersed, and 2% as animal-dispersed (Dahl, 1963). However, long-distance dispersal in plants appears to be possible and even common (Alsos et al., 2007) in some alpine regions. Significant isolation-by-distance and spatial autocorrelations suggest limited long-distance dispersal capabilities of P. hiemata; however, migration was at levels (Nem > 1 in the majority of cases) considered sufficient to prevent divergence resulting from genetic drift (Wright, 1931; Slatkin, 1987) and both plant density and effective population size estimates were similar across altitudes. This suggests that limited geographic gene flow and differences in population size are not the only mechanisms responsible for genetic structuring in P. hiemata. Using a quantitative genetic model, it was found that rates of altitudinal gene flow estimated in this study were likely to be insufficient to counter the strong selection detected on high and low altitude phenotypes of P. hiemata (leading to adaptive genetic divergence) in a recent study (Byars et al., 2007).

In the AMOVA analysis, there was more variation between altitudinal sites within transects than among transects (4·4% versus 1·7%, respectively), and when only the high and low altitudes were compared in an AMOVA these differences became strikingly different (13% vs. 0%, respectively). This reflects larger genetic structuring between high and low altitudes than between transects, even though the distances between transects were larger. This pattern suggests higher gene flow among high (and/or low) altitude sites from different transects than along transects (Fig. 5). This type of pattern in alpine plants has been found in some other studies: Schneller and Liebst (2007) found more variation between altitudes within regions than between sample regions in the fern Athyrium filix-femina. Isolated sites displayed some separation and differentiation based on clustering and pair-wise FST tests, although not as large as some values between sites within transects. In some cases, fragmentation may not completely eliminate gene flow (Gunter et al., 2000) and the same could be true in P. hiemata.

One reason why gene flow might be high across sites at similar altitudes is that there could be phenological separation along altitudinal gradients. Alpine plants often exhibit differences in the timing of flowering depending on altitude (Levesque et al., 1997; Korner, 1999; Ohsawa et al., 2007). The timing of reproduction with altitude has not been investigated in Poa hiemata, although there are substantial differences in growth form between plants at different ends of the gradient we examined. Prevailing winds on the Bogong High Plains are in a north-westerly direction (Williams, 1987) and could contribute to horizontal pollen transfer at least in one direction across the east-facing altitudinal transects. In addition, insect pollinators that track flowering with altitude may allow more opportunities for horizontal (compared with altitudinal) gene flow.

Significance for adaptive responses

The microsatellite data indicate that Poa hiemata exhibits a mixed reproductive system with genetic structuring of neutral variation across altitudes that suggests restricted gene flow in the recent past. Low gene flow can increase the potential for adaptive differentiation to variable altitudinal conditions, and a previous quantitative field study in populations of P. hiemata along these same altitudinal gradients indicated localized adaptation in peripheral populations (Byars et al., 2007). Further adaptive differentiation may occur in response to climate change, although, as populations become increasingly isolated and fragmented, adaptation might be limited because favoured mutations are not exchanged between populations. Also, gene flow from populations central within current distributions could limit adaptation in peripheral populations (Kirkpatrick and Barton, 1997), particularly as different traits are selected along altitudinal gradients.

However, patterns of gene flow found in this study are likely to enhance rates of adaptation to altitudinal gradients. Because gene flow across different sites at a similar altitude appears to be high, alleles favoured at high or low altitudes will be able to spread to sites where they have a high fitness. In a previous study on altitudinal adaptation, grasses had a high fitness at a similar altitude from where they originated, regardless of the transect where they were planted (Byars et al., 2007). Thus alleles from one high (or low) altitude site would be favoured if they were moved to another site at a similar altitude.

ACKNOWLEDGEMENTS

We thank Parks Victoria for vehicle access to sites and other advice, S. Hadden for flora collection permits and M. Blackett, A. Weeks, R. Hallas, P. Mitrovski and N. Endersby for advice in the laboratory. This work was supported by the Australian Research Council (ARC) through their Linkage program as well as the Department of Natural Resources and Environment, Parks Victoria, E.S Link, Mount Hotham Alpine Resort Management, Howmans Gap Alpine Centre, Falls Creek Alpine Resort Management. S.G.B. was supported by the David Hay Memorial Fund from the University of Melbourne and A.A.H. by a Federation Fellowship from the Australian Research Council.

APPENDIX

Genetic diversity and selfing estimates within populations of Poa hiemata. Key to variables: Altitude, m a.s.l.; n, sample size; A, average number of alleles per locus; HO, observed heterozygosity; HE, expected heterozygosity; ŝ(ML), maximum-likelihood estimate of selfing; FIS, within-population coefficient of inbreeding; IMGs, identical multilocus genotypes; Ne, effective population size. Bold values indicate significant (P < 0·05) parameters. Population labels at transects relate to position of altitudinal population from high (H) to low (L) altitude (H, 10, 9, 8, 7, 6, 5, 4, 3, 2, L) for each transect (NCS = New Country Spur; MNC = Mount Nelse Central; MC = Mount Cope) including outlying populations (MF = Mount Feathertop; MB = Mount Bogong; MNCisolated = Nelse Centralisolated; NCSisolated = New Country isolated).

PopulationAltitudenAHOHEŝ(ML)FISIMGsNe
NCSH1880205·50·6470·5980·004–0·08224146
NCS101875207·00·7010·6680·001–0·04921665
NCS91867207·60·5990·6540·0040·08441770
NCS81853207·30·6580·6590·0020·00201705
NCS71836208·00·6320·6540·0020·03421524
NCS61819208·30·5460·6480·0060·15721408
NCS51789207·60·5930·6620·0030·10441818
NCS41784206·80·5920·7130·0680·17001635
NCS31771207·00·5010·6590·2840·24001669
NCS21764206·80·5230·6620·0020·21021359
NCSL1746208·00·5420·6540·0850·17161707
MNCH1880204·50·4560·5150·1720·11501289
MNC101878205·80·5210·6610·0030·21201311
MNC91874206·20·4980·6570·0030·24201560
MNC81868207·00·5310·6230·0020·14821392
MNC71860206·70·5740·5860·020·02001537
MNC61839208·20·6400·6670·0040·04062011
MNC51827207·70·5790·6600·0020·12321719
MNC41812206·80·5210·6170·0050·15641599
MNC31798207·00·5640·6590·2320·14401141
MNC21790206·30·5500·6140·2480·10401133
MNCL1780207·00·5920·6640·0060·10801349
MCH1847205·00·4610·5640·0040·18361913
MC101810206·20·5210·6720·0070·22582008
MC91780205·20·5170·6710·1990·23001703
MC81762206·70·6480·664<0·0010·02441603
MC71750205·70·5930·7070·0010·16141795
MC61746207·80·6320·6760·0010·06521953
MC51732206·50·7880·7640·001–0·03121883
MC41725206·70·7250·7690·0210·05701899
MC31712206·30·7750·7330·002–0·05701386
MC21704204·50·7170·6900·143–0·03901460
MCL1700207·20·8950·7090·001–0·26222239
MF1760206·70·6010·6590·0320·0880
MB1920206·80·6610·7150·0130·0760
MNCisolated1660205·80·5400·6110·0090·1168
NCSisolated1680207·70·5560·6410·0030·1335

LITERATURE CITED

  • Albertini E, Porceddu A, Marconi G, Barcaccia G, Pallottini L, Falcinelli M. Microsatellite-AFLP for genetic mapping of complex polyploids. Genome. 2003;46:824–832. [PubMed]
  • Alberto F, Gouveia L, Arnaud-Haond S, Perez-Llorens JL, Duarte CM, Serrao CA. Within-population spatial genetic structure, neighbourhood size and clonal subrange in the seagrass Cymodocea nodosa. Molecular Ecology. 2005;14:2669–2681. [PubMed]
  • Alsos I, Eidesen P, Ehrich D, et al. Frequent long-distance plant colonisation in the changing arctic. Science. 2007;316:1606–1609. [PubMed]
  • Amos W, Harwood J. Factors affecting levels of genetic diversity in natural populations. Philosophical Transactions of The Royal Society of London Series B - Biological Sciences. 1998;353:177–186. [PMC free article] [PubMed]
  • Antonovics J. Nature of limits to natural selection. Annals of the Missouri Botanic Garden. 1976;63:224–247.
  • Aradhya KM, Muellerdombois D, Ranker TA. Genetic structure and differentiation in Metrosideros polymorpha (Myrtaceae) along altitudinal gradients in Maui, Hawaii. Genetical Research. 1993;61:159–170.
  • Arnaud-Haond S, Belkhir K. GENCLONE: a computer program to analyse genotypic data, test for clonality and describe spatial clonal organisation. Molecular Ecology Notes. 2007;7:15–17.
  • Barbara T, Martinelli G, Fay MF, Mayo SJ, Lexer C. Population differentiation and species cohesion in two closely related plants adapted to neotropical high-altitude ‘inselbergs’, Alcantarea imperialis and Alcantarea geniculata (Bromeliaceae) Molecular Ecology. 2007;16:1981–1992. [PubMed]
  • Bauert M. Genetic diversity and ecotypic differentiation in arctic and alpine populations of Polygonum viviparum. Arctic and Alpine Research. 1996;28:190–195.
  • Beerli P, Felsenstein J. Maximum-likelihood estimation of migration rates and effective population numbers in two populations using a coalescent approach. Genetics. 1999;152:763–773. [PubMed]
  • Beerli P, Felsenstein J. Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences of the USA. 2001;98:4563–4568. [PubMed]
  • Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F. GENETIX 4·03, logiciel sous Windows pour la genetique des populations. France: Laboratoire Genome et Populations, Universite de Montpellier II, Montpellier; 2004.
  • Beniston M. Climatic change in mountain regions: a review of possible impacts. Climatic Change. 2003;59:5–31.
  • Bennet E. Adaptation in wild and cultivated plant populations. In: Frankel HO, Bennett E, editors. Genetic resources in plants: their exploration and conservation. Philadelphia: FA Davis; 1970. pp. 115–129.
  • Billings WD, Mooney HA. Ecology of arctic and alpine plants. Biological Reviews of the Cambridge Philosophical Society. 1968;43:481–501.
  • Bliss L. Arctic and alpine life cycles. Annual Review of Ecology and Systematics. 1971;2:405–439.
  • Bonin A, Bellemain E, Eidesen PB, Pompanon F, Brochmann C, Taberlet P. How to track and assess genotyping errors in population genetics studies. Molecular Ecology. 2004;13:3261–3273. [PubMed]
  • Brereton R, Bennett S, Mansergh I. Enhanced greenhouse climate-change and its potential effect on selected fauna of SouthEastern Australia – a trend analysis. Biological Conservation. 1995;72:339–354.
  • Brown JH. On the relationship between abundance and distribution of species. American Naturalist. 1984;124:255–279.
  • Byars SG, Papst W, Hoffmann AA. Local adaptation and cogradient selection in the alpine plant, Poa hiemata, along a narrow altitudinal gradient. Evolution. 2007;61:2925–2941. [PubMed]
  • Chambers J. Disturbance, life history strategies, and seed fates in alpine herbfield communities. American Journal of Botany. 1995;82:421–433.
  • Crawford RMM, Abbott RJ. Pre-adaptation of arctic plants to climate change. Botanica Acta. 1994;107:271–278.
  • Dahl E. Plant migrations across the North Atlantic Ocean and their importance for the palaeogeography of the region. In: Löve D, Löve A, editors. North Atlantic biota and their history. Oxford: Pergamon Press; 1963. pp. 173–188.
  • David P, Pujol B, Viard F, Castella V, Goudet J. Reliable selfing rate estimates from imperfect population genetic data. Molecular Ecology. 2007;16:2474–2487. [PubMed]
  • Douglas D. The balance between vegetative and sexual reproduction of Mimulus primuloides (Scrophulariaceae) at different altitudes in California. Journal of Ecology. 1981;69:295–310.
  • Excoffier L, Smouse PE, Quattro JM. Analysis of molecular variance inferred from metric distances among DNA haplotypes – application to human mitochondrial DNA restriction data. Genetics. 1992;131:479–491. [PubMed]
  • Friedman J, Harder LD. Inflorescence architecture and wind pollination in six grass species. Functional Ecology. 2004;18:851–860.
  • Galen C, Stanton M, Shore J, Sherry R. Source-sink dynamics and the effect of an environmental gradient on gene flow and genetic substructure of the alpine buttercup, Ranunculus adoneus. Opera Botanica. 1997;132:179–188.
  • Gapare WJ, Aitken SN, Ritland CE. Genetic diversity of core and peripheral Sitka spruce (Picea sitchensis) populations: implications for conservation of widespread species. Biological Conservation. 2005;123:113–123.
  • Goudet J. FSTAT version 2·8, updated from Goudet (1995). FSTAT: a computer program to calculate F-statistics. Journal of Heredity. 1999;86:485–486.
  • Grabherr G, Gottfried M, Pauli H. Climate effects on mountain plants. Nature. 1994;369:448–448. [PubMed]
  • Gunter LE, Tuskan GA, Gunderson CA, Norby RJ. Genetic variation and spatial structure in sugar maple (Acer saccharum Marsh.) and implications for predicted global-scale environmental change. Global Change Biology. 2000;6:335–344.
  • Harada Y, Kawano S, Iwasa Y. Probability of clonal identity: inferring the relative success of sexual versus clonal reproduction from spatial genetic patterns. Journal of Ecology. 1997;85:591–600.
  • Hendry AP, Day T, Taylor E. B. Population mixing and the adaptive divergence of quantitative traits in discrete populations: a theoretical framework for empirical tests. Evolution. 2001;55:459–466. [PubMed]
  • Heuertz M, De Paoli E, Kallman T, et al. Multilocus patterns of nucleotide diversity, linkage disequilibrium and demographic history of Norway spruce (Picea abies) Genetics. 2006;174:2095–2105. [PubMed]
  • Hirao AS, Kudo G. Landscape genetics of alpine-snowbed plants: comparisons along geographic and snowmelt gradients. Heredity. 2004;93:290–298. [PubMed]
  • Hoffmann A, Parsons P. Evolutionary genetics and environmental stress. Oxford: Oxford University Press; 1991.
  • Hughes L. Climate change and Australia: trends, projections and impacts. Austral Ecology. 2003;28:423–443.
  • Hutchison DW, Templeton AR. Correlation of pairwise genetic and geographic distance measures: inferring the relative influences of gene flow and drift on the distribution of genetic variability. Evolution. 1999;53:1898–1914.
  • Inouye DW, Pyke GH. Pollination biology in the Snowy Mountains of Australia: comparisons with montane Colorado, USA. Australian Journal of Ecology. 1988;13:191–210.
  • Isik K, Kara N. Altitudinal variation in Pinus brutia and its implication in genetic conservation and seed transfers in southern Turkey. Silvae Genetica. 1997;46:113–120.
  • Jackson J, Buss L, Cook R. Population biology and evolution of clonal organisms. New Haven, CT: Yale University Press; 1985.
  • James E, Isenegger D, Papst W. A study of the genetic variation of selected populations of Poa fawcettiae in the Mount Kosciusko area. 1997 Report to the National Parks and Wildlife Service of New South Wales.
  • Jump AS, Penuelas J. Running to stand still: adaptation and the response of plants to rapid climate change. Ecology Letters. 2005;8:1010–1020.
  • Jump AS, Woodward FI, Burke T. Cirsium species show disparity in patterns of genetic variation at their range-edge, despite similar patterns of reproduction and isolation. New Phytologist. 2003;160:359–370.
  • Kapralov MV. Genotypic variation in populations of the clonal plant Saxifraga cernua in the central and peripheral regions of the species range. Russian Journal of Ecology. 2004;35:413–416.
  • Kindiger B. Cross-species amplification of Lolium microsatellites in Poa ssp. Grassland Science. 2006;52:105–115.
  • Kirkpatrick M, Barton NH. Evolution of a species' range. American Naturalist. 1997;150:1–23. [PubMed]
  • Klimes L, Klimesova J, Hendriks R, Groenendal Jv. Clonal plant architecture: a comparative analysis of form and function. In: De Kroon H, van Groenendael J, editors. The ecology and evolution of clonal plants. Leiden: Backhuys; 1997. pp. 1–29.
  • Korner C. Alpine plant life: functional plant ecology of high mountain ecosystems. Berlin: Springer; 1999.
  • Kuhner MK. LAMARC 2·0: maximum likelihood and Bayesian estimation of population parameters. Bioinformatics. 2006;22:768–770. [PubMed]
  • Lammi A, Siikamaki P, Mustajarvi K. Genetic diversity, population size, and fitness in central and peripheral populations of a rare plant Lychnis viscaria. Conservation Biology. 1999;13:1069–1078.
  • Levesque E, Henry GH. R, Svoboda J. Phenological and growth responses of Papaver radicatum along altitudinal gradients in the Canadian high arctic. Global Change Biology. 1997;3:125–145.
  • Lloyd AH, Fastie CL. Recent changes in treeline forest distribution and structure in interior Alaska. Ecoscience. 2003;10:176–185.
  • Loveless MD, Hamrick JL. Ecological determinants of genetic structure in plant populations. Annual Review of Ecology and Systematics. 1984;15:65–95.
  • McGraw JB, Antonovics J. Experimental ecology of Dryas octopetala ecotypes. I. Ecotypic differentiation and life cycle stages of selection. Journal of Ecology. 1983;71:879–897.
  • Maurer K, Gautschi B, Weyand A, Stocklin J, Fischer M. Isolation and characterisation of microsatellite DNA markers in the grass Poa alpina. Molecular Ecology Notes. 2005;5:719–720.
  • Morjan CL, Rieseberg LH. How species evolve collectively: implications of gene flow and selection for the spread of advantageous alleles. Molecular Ecology. 2004;13:1341–1356. [PMC free article] [PubMed]
  • Murawski DA, Hamrick JL. Local genetic and clonal structure in the tropical terrestrial bromeliad. Aechmea magdalenae. American Journal of Botany. 1990;77:1201–1208.
  • Ohsawa T, Tsuda Y, Saito Y, Sawada H, Ide Y. Altitudinal genetic diversity and differentiation of Quercus crispula in the Chichibu mountains, central Japan. International Journal of Plant Sciences. 2007;168:333–340.
  • Parks JC, Werth CR. A study of spatial features of clones in a population of bracken fern, Pteridium aquilinum (Dennstaedtiaceae) American Journal of Botany. 1993;80:537–544.
  • Pauli H, Gottfried M, Reier K, Klettner C, Grabherr G. Signals of range expansions and contractions of vascular plants in the high Alps: observations (1994–2004) at the GLORIA*master site Schrankogel, Tyrol, Austria. Global Change Biology. 2007;13:147–156.
  • Peakall R, Smouse PE. GenAlEx 6: genetic analysis in Excel: population genetic software for teaching and research. Molecular Ecology Notes. 2006;6:288–295. [PMC free article] [PubMed]
  • Pompanon F, Bonin A, Bellemain E, Taberlet P. Genotyping errors: causes, consequences and solutions. Nature Reviews Genetics. 2005;6:847–859. [PubMed]
  • Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–959. [PubMed]
  • Raymond M, Rousset F. Genepop version 1·2 upgraded to version 3·4. Genepop: population genetics software for exact tests and ecumenicism. Journal of Heredity. 1995;86:248–249.
  • Reisch C, Poschlod P, Wingender R. Genetic differentiation among populations of Sesleria albicans (Poaceae) from ecologically different habitats in central Europe. Heredity. 2003;a 91:519–527. [PubMed]
  • Reisch C, Poschlod P, Wingender R. Genetic variation of Saxifraga paniculata (Saxifragaceae): molecular evidence for glacial relict endemism in central Europe. Biological Journal of the Linnean Society. 2003;b 80:11–21.
  • Ritland K, Jain S. A model for the estimation of outcrossing rate and gene frequencies using n independent loci. Heredity. 1981;47:35–52.
  • Rogers S, Bendich A. Extraction of total cellular DNA from plants, algae and fungi. In: Gelvin SB, Schilperoot RA, editors. Plant molecular biology manual. Dordrecht: Kluwer Academic Press; 1994. pp. 1–8.
  • Sambrook J, Fritsch E, Maniatis T. Molecular cloning: a laboratory manual. 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989.
  • Saenz-Romero C, Tapia-Olivares BL. Pinus oocarpa isoenzymatic variation along an altitudinal gradient in Michoacan, Mexico. Silvae Genetica. 2003;52:237–240.
  • Schaal B, Leverich W. Molecular variation in isolated plant populations. Plant Species Biology. 1996:33–40.
  • Schneider S, Roessli D, Excoffier L. Switzerland: Genetics and Biometry Laboratory, University of Geneva; 2000. ARLEQUIN, version 2·001: a software for population genetics and data analysis.
  • Schneller J, Liebst B. Patterns of variation of a common fern (Athyrium filix-femina; Woodsiaceae): population structure along and between altitudinal gradients. American Journal of Botany. 2007;94:965–971. [PubMed]
  • Slatkin M. Gene flow in natural populations. Annual Review of Ecology and Systematics. 1985;16:393–430.
  • Slatkin M. Gene flow and the geographic structure of natural populations. Science. 1987;236:787–792. [PubMed]
  • Sokal R, Rohlf F. Biometry: the principles and practice of statistics in biological research. New York, NY: W.H. Freeman; 1995.
  • Stocklin J, Baumler E. Seed rain, seedling establishment and clonal growth strategies on a glacier foreland. Journal of Vegetation Science. 1996;7:45–56.
  • Sun G, Salomon B. Microsatellite variability and heterozygote deficiency in the arctic-alpine Alaskan wheatgrass (Elymus alaskanus) complex. Genome. 2003;46:729–737. [PubMed]
  • Tackenberg O, Stocklin J. Wind dispersal of alpine plant species: a comparison with lowland species. Journal of Vegetation Science. 2008;19:109–118.
  • Thuillet A, Bru D, David J, et al. Direct estimation of mutation rate for 10 microsatellite loci in durum wheat, Triticum turgidum (L.) Thell. ssp durum. Molecular Biology and Evolution. 2002;19:122–125. [PubMed]
  • Tibayrenc M, Kjellberg F, Ayala FJ. A clonal theory of parasitic protozoa – the population structures of Entamoeba, Giardia, Leishmania, Naegleria, Plasmodium, Trichomonas, and Trypanosoma and their medical and taxonomical consequences. Proceedings of the National Academy of Sciences of the USA. 1990;87:2414–2418. [PubMed]
  • Vickery JW. A taxonomic study of the genus Poa in Australia. Contributions of the N.S.W National Herbarium. 1970;4:145–243.
  • Vigouroux Y, Jaqueth J, Matsuoka Y, et al. Rate and pattern of mutation at microsatellite loci in maize. Molecular Biology and Evolution. 2002;19:1251–1260. [PubMed]
  • Weir BS, Cockerham CC. Estimating F-statistics for the analysis of population structure. Evolution. 1984;38:1358–1370.
  • Wen CS, Hsiao JY. Altitudinal genetic differentiation and diversity of Taiwan lily (Lilium longiflorum var. formosanum; Liliaceae) using RAPD markers and morphological characters. International Journal of Plant Sciences. 2001;162:287–295.
  • Williams R. Patterns of air temperature and accumulation of snow in subalpine heathlands and grasslands on the Bogong High Plains, Victoria. Australian Journal of Ecology. 1987;12:153–163.
  • Woodhead M, Russell J, Squirrell J, et al. Comparative analysis of population genetic structure in Athyrium distentifolium (Pteridophyta) using AFLPs and SSRs from anonymous and transcribed gene regions. Molecular Ecology. 2005;14:1681–1695. [PubMed]
  • Wright S. Evolution in Mendelian populations. Genetics. 1931;16:97–159. [PubMed]
  • Young AG, Warwick SI, Merriam HG. Genetic variation and structure at three spatial scales for Acer saccharum (sugar maple) in Canada and the implications for conservation. Canadian Journal of Forest Research – Revue Canadienne De Recherche Forestiere. 1993;23:2568–2578.
  • Young AG, Hill JH, Murray BG, Peakall R. Breeding system, genetic diversity and clonal structure in the subalpine forb Rutidosis leiolepis (Asteraceae) Biological Conservation. 2002;106:71–78.
  • Zouros E, Foltz D. W. Possible explanations of heterozygote deficiency in bivalve molluscs. Malacologia. 1984;25:583–591.

Articles from Annals of Botany are provided here courtesy of Oxford University Press