Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2013 Feb 12;111(4):611–621. doi: 10.1093/aob/mct021

Extensive contemporary pollen-mediated gene flow in two herb species, Ranunculus bulbosus and Trifolium montanum, along an altitudinal gradient in a meadow landscape

Philippe Matter 1,*, Chris J Kettle 1, Jaboury Ghazoul 1, Andrea R Pluess 1
PMCID: PMC3605955  PMID: 23408831

Abstract

Background and Aims

Genetic connectivity between plant populations allows for exchange and dispersal of adaptive genes, which can facilitate plant population persistence particularly in rapidly changing environments.

Methods

Patterns of historic gene flow, flowering phenology and contemporary pollen flow were investigated in two common herbs, Ranunculus bulbosus and Trifolium montanum, along an altitudinal gradient of 1200–1800 m a.s.l. over a distance of 1 km among five alpine meadows in Switzerland.

Key Results

Historic gene flow was extensive, as revealed by Fst values of 0·01 and 0·007 in R. bulbosus and T. montanum, respectively, by similar levels of allelic richness among meadows and by the grouping of all individuals into one genetic cluster. Our data suggest contemporary pollen flow is not limited across altitudes in either species but is more pronounced in T. montanum, as indicated by the differential decay of among-sibships correlated paternity with increasing spatial distance. Flowering phenology among meadows was not a barrier to pollen flow in T. montanum, as the large overlap between meadow pairs was consistent with the extensive pollen flow. The smaller flowering overlap among R. bulbosus meadows might explain the slightly more limited pollen flow detected.

Conclusions

High levels of pollen flow among altitudes in both R. bulbosus and T. montanum should facilitate exchange of genes which may enhance adaptive responses to rapid climate change.

Keywords: Elevation, flowering phenology, gene flow, herbs, managed meadows, microsatellites, pollen flow, pollen-pool analysis, Ranunculus bulbosus, Trifolium montanum

INTRODUCTION

Gene flow connects populations, maintains genetic diversity, prevents inbreeding and facilitates the spread of adaptive genes across a species range (Slatkin, 1985, 1987; Ellstrand and Elam, 1993). In plants, the extent and magnitude of gene flow is determined by a variety of pollen and seed dispersal mechanisms. The demography of the species, the mating system, spatial distances and the presence of physical barriers may greatly influence patterns of gene flow (Manel et al., 2003). The genetic composition of populations is influenced by colonization, bottlenecks and gene flow, whose effects can accumulate over multiple generations as well as through selection processes.

Plant species in mountain habitats are thought to be especially sensitive to rapid climatic changes, and upward shifts of plant species distributions are already well documented (Grabherr et al., 1994; Walther et al., 2002; Frei et al., 2010). Climate change might also modify the underlying drivers of gene flow by acting on flowering time and pollinator abundance and behaviour. At the population scale, adaptation to the direct and indirect effects of climate change is likely to be facilitated by population genetic diversity. Thus, understanding how plant populations are connected by gene flow, particularly across altitudinal gradients in mountain regions, will improve our understanding of how they will respond to future environmental change.

Mountain environments are characterized by strong gradients of temperature, exposition, UV radiation and duration of snow cover. These have a major influence on plant metabolism and physiology and, by extension, on their ability to grow and reproduce (Körner, 2003). Such ecologically relevant gradients are important in shaping patterns of local adaptation in plant species (Byars et al., 2007; Gonzalo-Turpin and Hazard, 2009), and, for example, thermal gradients across altitudes [the adiabatic lapse rate of approx. 0·5 °C per 100 m elevation (Körner, 2003)] can cause differential flowering of populations at different altitudes (Dittmar and Elling, 2006; Larcher, 2006; Ziello et al., 2009). Empirical studies show that differences in flowering periods can influence patterns of genetic structure by constraining pollen dispersal among populations (Stanton and Galen, 1997; Gerber et al., 2004; Kitamoto et al., 2006; Hirao and Kudo, 2008). However, many of these studies focus on species whose flowering time is greatly influenced by the time of snow melt, which does not directly relate to altitudinal difference but rather to differences in micro-topography and exposure (Scherrer and Körner, 2011). It thus remains unclear how phenological differences driven by altitude alone act as a barrier to gene flow within plant species.

In the European Alps, temperatures are expected to increase by about 3 K over the next 100 years (IPCC, 2007). The impacts of these changes for onset of first flowering have been well documented (Walther et al., 2002); however, the implications for flowering overlap among populations at different altitudes remain unexplored. Because pollen-mediated gene flow is often more efficient than seed-mediated gene flow in terms of dispersal distance (Ennos, 1994), it is particularly important to study patterns of contemporary pollen flow in plants occupying altitudinal gradients, especially in the context of land use change.

Studying patterns of contemporary pollen dispersal in species at the landscape scale can be challenging due to the high abundance of conspecifics. Many herbaceous species are aggregated in large populations and, consequently, most empirical studies of contemporary pollen dispersal have focused on relatively small spatial scales within populations (Miyazaki and Isagi, 2000; Hardy et al., 2004b; Scheepens et al., 2012), while landscape-scale studies focused on species with local low-abundance of conspecifics (Buehler et al., 2012). This greatly limits our understanding of patterns of gene flow along altitudinal gradients and over larger scales. For a long time it has been thought that pollen-induced contemporary gene flow only occurs over small distances in herbaceous plants (Hardy et al., 2004b; Pluess and Stocklin, 2004; Gonzales et al., 2006; Ishihama et al., 2006; Kitamoto et al., 2006). More recently, advances in molecular techniques (paternity and pollen-pool analyses instead of pollen analogues and direct observation of pollinators) have enabled us to study these processes at larger scales. For example, a recent study on Arabis alpina using exhaustive sampling at the landscape scale (up to 2 km distance and about 350 m elevation difference) by Buehler et al. (2012) found that the vast majority of pollen dispersal events were over short distances, but dispersal events can occur beyond 1 km.

In this paper, we investigate patterns of contemporary gene flow by pollen-dispersal and historic gene flow in the two common semi-dry grassland perennial plant species Ranunculus bulbosus and Trifolium montanum, along an altitudinal gradient located in Grindelwald, Switzerland. These species have similar distributional ranges encompassing colline and sub-alpine belts but have different pollination syndromes and thus they provide a useful study system for exploring contemporary gene flow across altitude; an issue of relevance with respect to likely plant population responses to climate change and consequent elevation shifts. In a recent study of historic gene flow of both these species in the Swiss mountains, Hahn et al. (2012) showed that altitude appeared not to influence genetic diversity, and showed little genetic differentiation at the regional scale. However, the patterns of contemporary gene flow in these species across altitude at landscape scales remain unresolved.

We sampled individuals and whole seed-families of R. bulbosus and T. montanum at five meadows along a 1200–1800 m a.s.l. altitudinal gradient stretching over approx. 1 km distance, and genotyped them at seven and 11 neutral nuclear microsatellite markers, respectively. This enabled us to explore patterns of historic gene flow and contemporary pollen dispersal over a landscape scale relevant to predicted climate change. A 600-m elevation range was chosen because it corresponds to the expected 3 K in temperature increase in the next 100 years (IPCC, 2007). Using population genetics, pollen-pool analysis and ecological surveys of the flowering phenology within this study system we address the following three questions: (1) Do patterns of genetic differentiation and genetic diversity differ across altitude? (2) Does contemporary pollen dispersal link meadows along an altitudinal gradient in R. bulbosus and T. montanum? (3) Do flowering phenology, demography and mating system influence pollen flow across altitude?

MATERIALS AND METHODS

Plant material

Study species

Ranunculus bulbosus L. (Ranunculaceae) and Trifolium montanum L. (Leguminosae) are both diploid, perennial herbs that occur mainly on dry calcareous grasslands. They are self-compatible species, yet selfing in T. montanum was observed less frequently than in R. bulbosus (Schleuning et al., 2009; Matter et al., 2013). The two species differ in their flower morphology and pollination syndromes. Ranunculus bulbosus flower heads are made of a single flower with a rotate corolla containing on average 40 carpels. Trifolium montanum inflorescences consist of up to 150 zygomorphic flowers each containing one ovule. Ranunculus bulbosus is visited by a wide array of generalist pollinators including Diptera (Syrphidae, Muscidae, Anthomyiidae), Coleoptera, small bees (Halictidae), honey-bees (Apis mellifera; Harper, 1957) and rarely bumble-bees (Steinbach and Gottsberger, 1994). Trifolium montanum is primarily visited by Apidae species such as the honey-bee and bumble-bees (Pettersson and Sjodin, 2000; Schleuning et al., 2009).

Study site and sampling

In summer 2009, five meadows located along a single mountain slope at around 1200, 1350, 1500, 1650 and 1800 m a.s.l. in Grindelwald (Switzerland) were sampled for whole seed families (SF; leaf material and seeds from the same individual) of R. bulbosus and T. montanum (see Fig. 1). The sampling area within each meadow was based upon the abundance of the plants within the focal altitudinal range (±40 m elevation difference). At the lowest elevation the area was clearly delineated by forest boundaries; at the other elevations a shrub layer formed some boundaries, but where meadows were more continuous, our sampling areas were defined by our chosen altitudinal range. Directly following collection, seeds of SF were sown in a 1:1 potting soil:sand mix and germinated in the greenhouse. Fresh in-situ sampled maternal leaves as well as fresh seedling leaves were dried in silica-gel for subsequent DNA extraction. Each SF was mapped with a triangular mapping technique including reference points. Measurements were converted into spatial co-ordinates with the Excel-Program Aequometer (www.aequometer.de).

Fig. 1.

Fig. 1.

Open in a new tab

Location of the sampled seed families and additional adults (as indicated in the key). The meadows at the bottom of the picture lie at approx. 1200 m a.s.l., and the ones at the top lie at approx. 1800 m a.s.l.

In R. bulbosus, 75 SF (15 per meadow) and in T. montanum 85 SF (17 per meadow) with ten offspring each were genotyped. For the assessment of overall genetic variation at the parental level, a further 170 R. bulbosus and 149 T. montanum individuals, equally distributed among meadows, were sampled, mapped and genotyped (hereafter called ‘additional adults’).

Census density and overall numbers of individuals

Since counting open flower heads in the field is far more time efficient than counting individuals, we estimated the density of individuals (i.e. individuals m−2) based on (a) the maximum number of open flower heads (OFH) yielded by their exhaustive count in each meadow at 13 time points in summer 2010 (see preceding paragraph for how the meadow area was delineated and the next section for the OFH count), (b) the mean number of OFH per individual (estimated based on flower counts in 2–190 and 72–141 individuals per meadow in R. bulbosus and T. montanum, respectively), and (c) the area of the sampled meadows, which was subsequently measured from aerial pictures of SwissTopo® using ImageJ (http://rsbweb.nih.gov/ij/index.html) and corrected for the slope (Table S1). The resulting estimations of densities at the time point with the maximal numbers of OFH were then used to extrapolate the ‘population’ sizes of R. bulbosus and T. montanum along the whole altitudinal transect to highlight the abundance of the focal species in the study area and variation among meadows. Thereby, all areas up to approx. 3 km distance of the sampled meadows and covered by calcareous grasslands containing the species of interests, as indicated in the database of dry grassland sites of national importance in Switzerland (Eggenberg et al., 2001), were considered.

Quantifying flowering overlap along the altitudinal gradient

We quantified the intensity of flowering overlap along the altitudinal gradient among meadows to investigate its relationship with pollen flow. We counted open flower heads (OFH) weekly from 28 May 2010 to 20 August 2010 (13 time points) in each meadow. Per census, OFH was measured by walking parallel transects spaced by 8–10 m each and counting OFH present within 4–5 m right and left of the transect. The degree of flowering overlap was thus quantified by estimating the number of simultaneously opened flower-heads (NSO-FH) between meadows, which has the advantage of taking into account potential differences in the species abundance. Previous studies have focused solely on overlapping flowering period which ignores this important demographic parameter (Robledo-Arnuncio et al., 2006). We determined curves of number of OFH over time and calculated the NSO-FH of each pair of meadows defined by the area of overlapping curves divided by the number of days during which both meadows were simultaneously flowering (see Supplementary Data Fig. S1). As a result, NSO-FH represents the average number of flower heads in each meadow, which were open during the whole flowering period overlap.

Laboratory methods and genetic dataset

DNA was extracted from approx. 10 mg dried leaf tissue with DNeasy 96 Plant Kits (Qiagen, Hombrechtikon Switzerland). Samples were genotyped using seven (R. bulbosus: Rb204, Rb206, Rb302, Rb306, B127, B129 and B145) or 11 (T. montanum: ats002, ats006, ats029, ats032, Tm10, Tm12, Tm13, Tm16, Tm17, Tm21 and Tm24) nuclear microsatellite markers as described in the two Primer Notes by Matter et al. (2012a, b).

In R. bulbosus, repeatability of the genotyping was assessed based on 52 (markers Rb204, Rb206, Rb302 and Rb306), 77 (B129 and B145) and 139 individuals (B127), and proved to range between 94·2 % and 100 % (mean = 97·4 %). Individuals with at least six (out of seven) loci typed were included in the analysis, resulting in a final dataset of 924 samples [72 SF with 6–10 (median = 10) individuals each, N = 686; 72 maternal plants and 166 additional adults]. Missing data across the dataset were 3·7 %.

In T. montanum, repeatability was assessed on 18–41 individuals per marker. The overall repeatability rate was 99·7 % (97·6 % for Tm17 and 100 % for all other markers). Individuals with at least nine (out of 11) loci typed were included in the analyses resulting in a final dataset of 1031 samples [82 SF with 8–10 (median = 10) individuals each, N = 808; 82 maternal plants and 141 additional adults]. Missing data across the dataset were 1·5 %.

Statistical analyses

Patterns of genetic differentiation among meadows

Since the presence of null alleles in some of the markers has previously been shown (Matter et al., 2012a, b), the degree of inbreeding was estimated with INEst which simultaneously estimates the null allele frequencies and the inbreeding coefficient (Chybicki and Burczyk, 2009), using the Population Inbreeding Model; standard errors were obtained with the jack-knifing procedure accounting for inbreeding. The genetic diversity and the genetic differentiation were estimated with FSTAT (Goudet, 2001). A spatial clustering analysis was done with BAPS 5·3 (Corander et al., 2007) using 25 runs allowing for a varying number of clusters K (max. K = 20) and two spatial co-ordinates. Spatial genetic structure (SGS) was investigated with SPAGEDIv1·3 (Hardy and Vekemans, 2002), yielding correlograms of Loiselle's kinship coefficients (Loiselle et al., 1995) against spatial distance between individuals, calculated based on three spatial co-ordinates. For correlograms across all five meadows, 20 distance classes were set containing about the same number of pairs each; for correlograms on each meadow separately, ten such distances classes were set. Confidence intervals of the estimated kinship coefficients were calculated with jack-knifing over loci; and confidence intervals of the expectation of no spatial genetic pattern were based on 999 permutations of individual locations among all individuals.

Indirect estimates of contemporary pollen flow

Pollen dispersal parameters were estimated using the program KINDIST (Robledo-Arnuncio et al., 2006; Robledo-Arnuncio et al., 2007), which implements the pollen-pool approach. The among-sibships correlated paternities estimated after Loiselle et al. (1995) were used to fit an exponential power function (Austerlitz et al., 2004; Robledo-Arnuncio and Gil, 2005; Klein et al., 2008; Oddou-Muratorio et al., 2010; Field et al., 2011) with a non-linear least squares model. A key assumption made by the program is that among-sibship correlated paternities decay with increasing distance. We tested this with a Mantel test in the R-package ‘Vegan’ (Oksanen et al., 2012) using Spearman's rank correlation coefficient rho (ρ) and 9999 permutations on the kinship matrix. A further assumption of the program to calibrate kinship coefficients is that, beyond a given threshold distance, pollen pools are unrelated, indicated by a stabilization of the among-sibships correlated paternity estimates (plotted against distance) at a slightly negative value. We tested several threshold distances (ranging between 1 m and the maximum distance between two maternal plants) to find the best-fitting model, indicated by the smallest least-squares residuals (LSR; as in Field et al., 2011). Based on the best fitting kernel, the average pollen dispersal distance (δ) was estimated. To test the robustness of the kernel fitting we repeated one to two analyses per dataset or sub-dataset. Note that KINDIST handles two spatial co-ordinates only and uses the allele frequency of the maternal plants to estimate the correlated paternities.

Pollen pool differentiation (global φft) and, subsequently, number of effective pollen donors (Nep) were calculated with TWOGENER (Austerlitz and Smouse, 2001, 2002). The global φft is a measure of correlated paternity among all sibships within a given dataset and is calculated via an AMOVA based on allele frequencies of the maternal plants and additional adults, as suggested by Robledo-Arnuncio et al. (2007). To estimate Nep from the global φft, the latter has to be corrected for parental inbreeding (FP) and selfing (s) (Austerlitz and Smouse, 2001; Burczyk and Koralewski, 2005). Given FP, φft transforms to φ′ft = φft/(1 + FP); and given s as (1 – tm) with tm being the multilocus outcrossing rate, φ′ft transforms to φ″ft = (2φ′ft – s2)/[2(1 – s)2]. Nep was calculated as 1/(2φ″ft). FP and tm were estimated with MLTR (Ritland, 2002) using the expectation-maximization method, pollen gene frequencies inferred from each seed-family and inference of parentage with parents chosen at random.

RESULTS

Census density and overall population size

Census density differed among meadows and species (Fig. 2 and Supplementary Data Table S1): the density of R. bulbosus individuals tended to decrease with altitude (Pearson's correlation coefficient of –0·843; P = 0·0729), while there was no correlation in T. montanum (–0·286, P = 0·640). Given the census density, the average numbers of open flower heads per individual and the area containing the focal species along the altitudinal transect under study (respectively, 26 and 49 ha for R. bulbosus and T. montanum, according to Eggenberg et al., 2001; see Materials and Methods), the overall numbers of individuals (‘population’ size) was about 29 000 individuals in R. bulbosus and 81 000 individuals in T. montanum.

Fig. 2.

Fig. 2.

Open in a new tab

Census plant density of Ranunculus bulbosus and Trifolium montanum in function of altitude. The dashed lines indicate the average census density.

Inbreeding, genetic diversity and genetic differentiation

Both R. bulbosus and T. montanum showed no significant inbreeding coefficients, even when computed separately for each meadow. Allelic richness in each meadow (corrected for sample size) ranged between 4·30 and 4·55 in R. bulbosus (minimum sample size = 27) and 4·25 and 4·39 in T. montanum (minimum sample size = 37) and did not significantly differ among meadows (1000 permutations, two-sided test; P = 0·555 in R. bulbosus and P = 0·746 in T. montanum). Differentiation among the five meadows was very low in both species: Fst = 0·0099 [95 %CI = (0·006; 0·016)] in R. bulbosus and 0·0068 [95 %CI = (0·003; 0·011)] in T. montanum.

There was no evidence for spatial clustering: all individuals of the five meadows got assigned to a single genetic cluster in both species. Moreover, SGS across all five meadows was very weak: only the two first distance classes (i.e. up to 38 m) showed a kinship coefficient significantly different from zero in both species (Supplementary Data Fig. S2a). Due to these indications of SGS at small distances, i.e. within meadows, we repeated the analyses at the meadow level. There was a lack of SGS in all but the upper-most R. bulbosus meadow (‘Ran1800’), where individuals in the shortest distance class were significantly related (Supplementary Data Fig. S2b).

Indirect estimates of contemporary pollen flow

Pollen-pool approach and dispersal kernel fitting

Across all five meadows, the among-sibship correlated paternity estimates showed a weak decay with increasing distance [ρ = –0·086 (P = 5·0e-04) in R. bulbosus and ρ = –0·048 (P = 0·005) in T. montanum; Fig. 3]. At the meadow scale, patterns were slightly different. In the upper three R. bulbosus meadows ‘Ran1500’, ‘Ran1650’ and ‘Ran1800’ and in the T. montanum meadow ‘Trif1650’, the among-sibships correlated paternity estimates (tended to) decrease stronger with distance than across all meadows: ρ = –0·191 (P = 0·033) in ‘Ran1500’; ρ = –0·215 (P = 0·023) in ‘Ran1650’; ρ = –0·18 (P = 0·075) in ‘Ran1800’; and ρ = –0·139 (P = 0·112) in ‘Trif1650’ (Fig. 3). In the two lower R. bulbosus meadows and in all other T. montanum's meadows there were no negative correlations found.

Fig. 3.

Fig. 3.

Open in a new tab

Among-sibships correlated paternity estimated in Ranunculus bulbosus and Trifolium montanum against distance: (A) the five meadows pooled; (B) within each meadow separately. The correlations (ρ) and associated P-values are indicated, calculated with a Mantel test using Spearman's rank correlation coefficient and 9999 permutations on the kinship matrix.

Given that Robledo-Arnuncio et al. (2007) recommends not to run KINDIST when ρ > –0·1 to avoid inflated biases, we estimated the parameters of an Exponential-Power dispersal kernel in the five following datasets only: the single meadows ‘Ran1500’, ‘Ran1650’, ‘Ran1800’ and ‘Trif1650’ (although the correlation in the two latter meadows was not significant as, for example, in de-Lucas et al., 2008), and also all R. bulbosus' meadows pooled (since ρ was close to –0·1 and highly significant).

The analyses across all meadows as well as within the meadows ‘Ran1500’, ‘Ran1650’ and ‘Trif1650’ reached the best fit (i.e. lowest LSR) when the threshold distance equaled the maximum inter-seed-family distance (Supplementary Data Table S2). The resulting kernels had shape parameters below one, indicating fat tails (Austerlitz et al., 2004) and their average pollen dispersal distances (δ) were 141 m in R. bulbosus across all meadows, 297 m within the meadow ‘Ran1500’, 225 m within ‘Ran1650’ and 3·5 km within ‘Trif1650’ (Table 1).

Table 1.

Exponential-power kernel fit on four Ranunculus bulbosus samples and one Trifolium montanum sample

Dataset NSF Noffs a b δ (m)
All R. bulbosus meadows 72 686 0·000033 0·1724 140·72
Ran1500 15 141 0·000011 0·1589 296·53
Ran1650 15 143 0·000003 0·1522 225·43
Ran1800 15 143 0·0001 0·1843 109·46
Trif1650 14 136 0·000003 0·1380 3490·57
Open in a new tab

The kernel was fitted with KINDIST (Robledo-Arnuncio et al., 2007). The best-fitting kernel estimates are indicated.

NSF, Number of seed-families; Noffs, total number of offspring; a and b, scale and shape parameters of the dispersal kernel (for details, see Austerlitz et al., 2004); δ, average pollen dispersal distance.

Results of ‘Trif1650’ are questionable since the same parameter estimates were obtained regardless of the threshold distance used, excepting the maximum inter-seed-family distance. In the upper meadow ‘Ran1800’, LSR indicated a best fit when the threshold distance was set to 45 m, which is smaller than the maximum inter-seed-family distance of 57 m. However, parameter estimates were similar for six out of ten different threshold distances used in ‘Ran1800’ but changed drastically with threshold distances of 44, 45 and 46 m, resulting in a pollen dispersal distance of 250, 109 and 7·5 m, respectively. As the latter estimate of pollen dispersal would result in a spatial genetic structure, which was absent here, the other two values might be better estimates.

The seven kernel fittings which have been replicated (see Supplementary Data Table S2) yielded exactly the same parameter estimations and LSRs.

In summary, as among-sibship correlated paternity decayed less across all meadows in T. montanum compared with R. bulbosus and decayed more often with distance in the latter species, our results indicate higher rates of contemporary gene flow in T. montanum than R. bulbosus.

Correlated paternity and effective number of pollen donors

The correlated paternity (φ″ft) across all meadows was higher in R. bulbosus (0·129) than in T. montanum (0·042; Table 2). Within individual meadows it was also consistently higher in R. bulbosus (range: 0·088–0·155) than in T. montanum (range: 0·017–0·064). The effective number of pollen donors Nep was higher in T. montanum than in R. bulbosus, both at the whole-sample scale (12·0 vs. 3·9) and at the meadow scale (7·8–29·3 vs. 3·2–5·7). φ″ft and Nep did not correlate with altitude and neither did φ″ft with plant density (all correlation tests with Spearman's ranks tests, details not shown).

Table 2.

Correlated paternity, mating system parameters and effective number of pollen donors in five meadows of Ranunculus bulbosus and Trifolium montanum along an altitudinal gradient

Dataset Global φft FP φ′ft s φ″ft Nep (individuals)
R. bulbosus
All meadows pooled 0·128 0·011 0·127 0·010 0·129 3·9
Ran1200 0·097 0·000 0·097 0·007 0·099 5·1
Ran1350 0·124 0·000 0·124 0·000 0·124 4·0
Ran1500 0·086 0·000 0·086 0·011 0·088 5·7
Ran1650 0·141 0·034 0·136 0·013 0·140 3·6
Ran1800 0·157 0·054 0·149 0·020 0·155 3·2
T. montanum
All meadows pooled 0·043 0·038 0·041 0·002 0·042 12·0
Trif1200 0·066 0·041 0·064 0·000 0·064 7·8
Trif1350 0·045 0·025 0·044 0·000 0·044 11·3
Trif1500 0·029 0·031 0·028 0·005 0·029 17·4
Trif1650 0·042 0·065 0·039 0·000 0·039 12·7
Trif1800 0·017 0·021 0·017 0·006 0·017 29·3
Open in a new tab

Global φft, pollen-pool differentiation parameter (calculated with TwoGener); φ′ft and φ″ft, pollen-pool differentiation parameter corrected for parental inbreeding and selfing (see text); FP, inbreeding coefficient of the parental generation; s, selfing rate; Nep, number of effective pollen donors.

Flowering overlap among altitudes

The flowering peak of two consecutive meadows was delayed by <1 week in R. bulbosus and at maximum by 2 weeks in T. montanum. The overlapping flowering period between the highest and the lowest meadow, was 3 weeks for R. bulbosus and 6 weeks for T. montanum (see Fig. 4). The pairwise number of simultaneously opened flower-heads (NSO-FH) between meadow pairs, a potential indicator of pollen flow, was higher in all pairs of T. montanum meadows (144–2355, median = 372, n = 10 pairs) compared with R. bulbosus (17–411, median = 40; see Table 3). This measure of pairwise flowering overlap did not correlate with the pairwise altitudinal difference [Mantel test using Spearman's ρ and 9999 permutations: ρ = –0·152 (P = 0·335) in R. bulbosus and 0·1758 (P = 0·764) in T. montanum].

Fig. 4.

Fig. 4.

Open in a new tab

Numbers of open flower-heads in Ranunculus bulbosus and Trifolium montanum in five meadows at 1200, 1350, 1500, 1650 and 1800 m a.s.l. at 13 weekly census points between 25 May and 20 August 2010.

Table 3.

Pairwise numbers of simultaneously opened flower-heads (NSO-FH) of five Ranunculus bulbosus (lower diagonal) and five Trifolium montanum (upper diagonal) meadows

M1200 M1350 M1500 M1650 M1800
M1200 – 2345 460 183 841
M1350 411 – 407 172 944
M1500 169 162 – 145 337
M1650 41 42 40 – 144
M1800 20 20 18 17 –
Open in a new tab

Each meadow is referred to as ‘M’ with its approximate altitude.

DISCUSSION

This study provides new insights into pollen-mediated contemporary gene flow in abundant mountain herbs across an altitudinal gradient. Based on neutral genetic markers, our evaluation of contemporary pollen dispersal and genetic differentiation among five meadows (Fst = 0·01 in R. bulbosus and 0·007 in T. montanum) indicate extensive contemporary and historic gene flow. Our results suggest that plant density and degree of flowering period overlap among meadows could be important drivers for shaping the patterns of pollen dispersal.. Pollen flow appears to be slightly greater in T. montanum than R. bulbosus. The tail of the pollen dispersal curve extended well beyond the scale of our study area, supporting the idea that contemporary gene flow by pollen is extensive in these species in mountain meadow systems. Below we discuss the processes which may lead to the observed patterns and their implications for maintenance of genetic diversity and adaptive variation in mountain plants facing rapid and unprecedented climate change.

Do patterns of genetic differentiation and genetic diversity differ across altitude?

Patterns of historic gene flow are important because these may underlie patterns of local adaptation in plant species (Gonzalo-Turpin and Hazard, 2009). The absence of any genetic clustering, the very low Fst values and similar levels of genetic diversity among meadows reflect weak genetic differentiation at our study site. This is consistent with the findings of Hahn et al. (2012) who studied the same two species at 1200 and 1800 m a.s.l., i.e. similar upper and lower altitudes, but over a much larger spatial scale and using different molecular markers. It is interesting to note that, although both studies observed little differentiation, contrary to our study Hahn et al. (2012) observed greater differentiation in T. montanum (Fst = 0·118) than in R. bulbosus (Fst = 0·071). This difference can be explained by several factors, including much greater sampling scale (about 200 km), fewer individuals per site sampled (n = 20), and different markers (AFLP vs. microsatellites). In contrast, in Arabis alpina sampled over very similar spatial scales to our study, very strong genetic structuring was observed (Buehler et al., 2012). We attribute these differences to the predominantly selfing mating system of Arabis alpina.

Does contemporary pollen dispersal link meadows along an altitudinal gradient in R. bulbosus and T. montanum?

Recent advances in population genetic statistics and modelling have greatly enhanced our capacity to indirectly evaluate pollen flow, i.e. without exhaustive sampling of populations (Robledo-Arnuncio et al., 2006). However, such approaches do have limitations. First, they require adequate genetic structuring over the scale of the investigation (Robledo-Arnuncio et al., 2006; Rong et al., 2010). Second, over-interpretation of the shape of the dispersal kernel derived from such data should be avoided (Robledo-Arnuncio et al., 2006). Without baseline studies it remains difficult to determine what the appropriate scale of sampling is. Our study provides a baseline analysis and demonstrates that the scales of contemporary gene flow are greater than the scale of sampling in T. montanum and R. bulbosus (1 km horizontal and 600 m of altitudinal difference). This point is particularly evident by the Least-Squares Residuals of the pollen pool analyses, which in most datasets indicate that the best fit of the dispersal kernels is reached at the maximum sample distance. In the seven datasets which were not suited for pollen-pool analysis, and, especially in T. montanum, the lack of a prominent structuring in the pollen pool is likely a consequence of extremely low differentiation of the parental generation. Indeed, since individuals are genetically similar over the whole sampling area, the pollen produced by paternal plants will also be similar across potential pollen donors. Yet, combining information on mating system, flower morphology and phenology, we can infer some additional important factors shaping pollen dispersal.

Do flowering phenology, demography and mating system influence pollen flow across altitude?

Flowering phenology

The overall phenology patterns illustrated in Fig. 4 are consistent with the lack of genetic differences as shown by KINDIST in T. montanum. Its greater population size and the greater proportional flowering overlap (6 weeks of simultaneous flowering for the highest and lowest meadow) among meadows appears to contribute to the more extensive pollen flow. In contrast, in R. bulbosus, flowering overlap across all meadows is clearly much smaller (maximum 3 weeks of simultaneous flowering between the highest and the lowest meadow). Indeed, average pollen dispersal distance proved to be smaller than 300 m in R. bulbosus, which corresponds to a dispersal over three consecutive meadows. Although patterns of genetic differentiation may be influenced by seed dispersal, the tail of the pollen dispersal kernel and average dispersal distances, our results imply that dispersal between the highest and the lowest meadow is more limited, although not excluded, in R. bulbosus compared with T. montanum. This is further supported by the consistently larger number of simultaneously open flower-heads (NSO-FH) per meadow pair in T. montanum compared with R. bulbosus, suggesting that the probability of pollen flow in T. montanum is greater than in R. bulbosus. The fact that NSO-FH does not correlate with pairwise altitudinal difference among meadows, which was unexpected, is attributable to the large differences in number of individuals per meadow. This is evident in R. bulbosus, in which the largest meadow contains over 150 times more OFH than the smallest meadow which is the uppermost location under study.

Gerber et al. (2004) studied patches of Ranunculus alpestris in Switzerland and found that flowering phenology delays induced by time of snow-melt only weakly affected genetic differentiation of sub-populations along the gradient, and that distance was more important for structuring populations. In contrast, Hirao and Kudo (2004) found that genetic distance was correlated with phenological distance in Veronica stelleri and Gentiana nipponica of northern Japan. In the alpine buttercup Ranunculus adoneus in Colorado, USA, Stanton et al. (1997) found that populations showed slight but significant genetic differentiation among the different snow-melt zones. As both pollen and seed dispersal are relevant to the observed patterns of genetic structure, our results suggest that, between altitudinal extremes, seed dispersal may be more important when phenology constrains pollen-mediated gene flow.

Flower morphology and population demography

Additionally to phenology, flower morphology and population density also contribute to determining the diversity of the pollen pool sampled by seed families (Hardy et al., 2004a). Global correlated paternity (φft) was higher in R. bulbosus (0·128) than in T. montanum (0·043) and the effective number of pollen donors (Nep) was lower in R. bulbosus (Nep = 3·2–5·7) than in T. montanum (Nep = 7·2–29·3). These interlinked metrics reflect how these species ‘sample’ the background pollen pool and are influenced by co-dispersion and limited mate availability (Hardy et al., 2004a). Hardy et al. (2004a) report values of effective number of pollen donors ranging from Nep = 1·2–8·8 in eight herb species from five different families. The values reported are comparable to those found for R. bulbosus, but much smaller than those found for T. montanum. One explanation for this difference may be the floral morphology of the species. Ranunculus bulbosus has a single corolla circling many stigmas; in contrast the inflorescence of T. montanum is composed of 100–150 corollas each containing only one stigma. This is consistent with the eight species mentioned above, where the by-far highest Nep value is reported for Centaurea solstitialis, whose inflorescences are also composed of several long and tubular corollas. The probability that several stigmas receive pollen from the same individual through foraging insects is thus greater in R. bulbosus; a pattern already described by Matter et al. (2013). Additionally, limited mate availability could result from the lower density of conspecifics with increasing altitude in R. bulbosus and lower flowering overlap (see preceding paragraph) between meadows of different altitudes. In contrast larger population densities, greater overlap in flowering and floral morphology facilitated a reduced correlated paternity in T. montanum.

Mating system

Generally pollen dispersal distance is thought to be negatively correlated with the selfing rate in plants (Ennos, 1994; Chauvet et al., 2004). In Arabis alpina 84 % selfing rates were detected using maximum likelihood-based paternity analysis while the direct estimate of pollen dispersal showed only a tiny proportion of mating events occurring over distances >1000 m (Buehler et al., 2012). In our study using the same maximum likelihood approach, we observed 25 % selfing in R. bulbosus and 4·3 % in T. montanum (results not shown). While we treat these estimates with caution (due to the limited power to assign paternity in field conditions with many un-sampled potential donors), the slightly lower pollen dispersal in R. bulbosus compared with T. montanum based on this paternity analysis and on the pollen-pool approach is consistent with the idea that selfing may reduce the frequency of long-distance dispersal.

Implications of pollen dispersal among altitudes for mountain perennial herbs in the face of climate change

The implications of climate change for species and habitat conservation is the subject of global concern (Theurillat and Guisan, 2001; Thomas et al., 2004; Hannah et al., 2005). The ability of species to survive these changes at their present location will depend on their adaptive potential. Our findings suggest that populations occurring across an altitudinal gradient have the capacity to exchange genes by pollen dispersal. This ensures population connectivity, large effective population sizes and reduced probability of inbreeding. Extant adaptive variation should thus be maintained and transferable among altitudes in response to climate change.

Limited local adaptation is often attributed to high levels of gene flow and because only very few dispersal events are required to spread advantageous alleles (Morjan and Rieseberg, 2004). However, in perennial herb species patterns of local adaptation appear to be more commonly observed over large scales but seem to be rarer at small spatial scales (Clausen et al., 1940; Snaydon and Davies, 1976; Scheepens et al., 2010). Some studies also show local adaptation at small scales specifically associated with altitude, e.g. in Poa hiemata (Byars et al., 2007) and in Festuca eskia (Gonzalo-Turpin and Hazard, 2009). In such species extensive gene flow might facilitate the exchange of advantageous genes to higher elevations which may enhance population resilience to climatic change.

Perhaps of more immediate threat to mountain species is rapid land use change. Abandonment of traditional management in dry grassland meadows has been reported over the last 60 years across Switzerland (Stöcklin et al., 2007). Such habitat fragmentation is predicted to reduce genetic connectivity. However, our estimates of contemporary pollen dispersal in T. montanum and R. bulbosus suggest that recent land use change has little consequence for contemporary gene flow for some herbaceous semi-dry grassland species. In conclusion, these common semi-grassland herb species are likely to be highly resilient to anthropogenic habitat change provided that sufficient meadow habitat is maintained.

SUPPLEMENTARY DATA

Supplemenatry data are available online at www.aob.oxfordjournals.org and consist of the following. Table S1: characteristics of plant numbers and densities in Ranunculus bulbosus and Trifolium montanum in five meadows along an altitudinal transect and estimated at 13 weekly census days over the growing season 2010. Table S2: exponential-power Kernel fit on four Ranunculus bulbosus and one Trifolium montanum samples with KINDIST. Fig. S1: estimating the number of simultaneously opened flower-heads between two given meadows. Fig. S2: spatial genetic structure in Ranunculus bulbosus and Trifolium montanum.

Supplementary Data
supp_111_4_611__index.html (1.1KB, html)

ACKNOWLEDGEMENTS

We thank Maja Frei, Sarah Frischknecht, Marina Beck, Charlotte Klank, David Emmerth, Sarah Burg, Aline Finger, Kristina Herz and Dominique Buehler for help and/or technical support in the field and greenhouses. Special thanks go to the local farmers, namely Families Wyss, Rubi and Feuz in Burglauenen, Grindelwald. We are also grateful to Kirsti Määttänen and Marianne Leuzinger for help with development of the T. montanum microsatellite markers. Fragment length measurements were done in the Genetic Diversity Centre of ETH Zurich. This work is part of P. Matter's PhD project. This work was supported by the Swiss National Fonds for Research (grant SNF 3100A0-116277 to J.G. and A.R.P.), by S-ENETH and by CCES.

LITERATURE CITED

  1. Austerlitz F, Smouse PE. Two-generation analysis of pollen flow across a landscape. III. Impact of adult population structure. Genetical Research. 2001;78:271–280. doi: 10.1017/s0016672301005341. [DOI] [PubMed] [Google Scholar]
  2. Austerlitz F, Smouse PE. Two-generation analysis of pollen flow across a landscape. IV. Estimating the dispersal parameter. Genetics. 2002;161:355–363. doi: 10.1093/genetics/161.1.355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Austerlitz F, Dick CW, Dutech C, et al. Using genetic markers to estimate the pollen dispersal curve. Molecular Ecology. 2004;13:937–954. doi: 10.1111/j.1365-294x.2004.02100.x. [DOI] [PubMed] [Google Scholar]
  4. Buehler D, Graf R, Holderegger R, Gugerli F. Contemporary gene flow and mating system of Arabis alpina in a Central European alpine landscape. Annals of Botany. 2012;109:1359–1367. doi: 10.1093/aob/mcs066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Burczyk J, Koralewski TE. Parentage versus two-generation analyses for estimating pollen-mediated gene flow in plant populations. Molecular Ecology. 2005;14:2525–2537. doi: 10.1111/j.1365-294X.2005.02593.x. [DOI] [PubMed] [Google Scholar]
  6. 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. doi: 10.1111/j.1558-5646.2007.00248.x. [DOI] [PubMed] [Google Scholar]
  7. Chauvet S, van der Velde M, Imbert E, et al. Past and current gene flow in the selfing, wind-dispersed species Mycelis muralis in western Europe. Molecular Ecology. 2004;13:2871–2871. doi: 10.1111/j.1365-294X.2004.02166.x. [DOI] [PubMed] [Google Scholar]
  8. Chybicki IJ, Burczyk J. Simultaneous estimation of null alleles and inbreeding coefficients. Journal of Heredity. 2009;100:106–113. doi: 10.1093/jhered/esn088. [DOI] [PubMed] [Google Scholar]
  9. Clausen J, Keck DD, Hiesey WM. Experimental studies on the nature of species. I. Effect of varied environments on western North American plants. Carnegie Institute of Washington Publicaton No. 520. 1940 [Google Scholar]
  10. Corander J, Gyllenberg M, Koski T. Random partition models and exchangeability for Bayesian identification of population structure. Bulletin of Mathematical Biology. 2007;69:797–815. doi: 10.1007/s11538-006-9161-1. [DOI] [PubMed] [Google Scholar]
  11. Dittmar C, Elling W. Phenological phases of common beech (Fagus sylvatica L.) and their dependence on region and altitude in Southern Germany. European Journal of Forest Research. 2006;125:181–188. [Google Scholar]
  12. Eggenberg S, Dalang T, Dipner M, Mayer C. Technischer Bericht. Bern: 2001. Kartierung und Bewertung der Trockenwiesen und -weiden von nationaler Bedeutung. [Google Scholar]
  13. Ellstrand NC, Elam DR. Population genetic consequences of small population size: implications for plant conservation. Annual Review of Ecology and Systematics. 1993;24:217–242. [Google Scholar]
  14. Ennos RA. Estimating the relative rates of pollen and seed migration among plant populations. Heredity. 1994;72:250–259. [Google Scholar]
  15. Field DL, Ayre DJ, Whelan RJ, Young AG. The importance of pre-mating barriers and the local demographic context for contemporary mating patterns in hybrid zones of Eucalyptus aggregata and Eucalyptus rubida. Molecular Ecology. 2011;20:2367–2379. doi: 10.1111/j.1365-294X.2011.05054.x. [DOI] [PubMed] [Google Scholar]
  16. Frei E, Bodin J, Walther GR. Plant species' range shifts in mountainous areas: all uphill from here? Botanica Helvetica. 2010;120:117–128. [Google Scholar]
  17. Gerber JD, Baltisberger M, Leuchtmann A. Effects of a snowmelt gradient on the population structure of Ranunculus alpestris (Ranunculaceae) Botanica Helvetica. 2004;114:67–78. [Google Scholar]
  18. Gonzales E, Hamrick JL, Smouse PE, Dyer RJ. Pollen-mediated gene dispersal within continuous and fragmented populations of a forest understorey species, Trillium cuneatum. Molecular Ecology. 2006;15:2047–2058. doi: 10.1111/j.1365-294X.2006.02913.x. [DOI] [PubMed] [Google Scholar]
  19. Gonzalo-Turpin H, Hazard L. Local adaptation occurs along altitudinal gradient despite the existence of gene flow in the alpine plant species Festuca eskia. Journal of Ecology. 2009;97:742–751. [Google Scholar]
  20. Goudet J. FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3) 2001 Updated from Goudet (1995). 2.9.3. http://www2.unil.ch/popgen/softwares/fstat.htm . [Google Scholar]
  21. Grabherr G, Gottfried M, Pauli H. Climate effects on mountain plants. Nature. 1994;369:448. doi: 10.1038/369448a0. [DOI] [PubMed] [Google Scholar]
  22. Hahn T, Kettle CJ, Ghazoul J, Frei ER, Matter P, Pluess AR. Patterns of genetic variation across altitude in three plant species of semi-dry grasslands. Plos One. 2012;7:pe41608. doi: 10.1371/journal.pone.0041608. http://dx.doi.org/10.1371/journal.pone.0041608 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hannah L, Midgley G, Hughes G, Bomhard B. The view from the Cape: extinction risk, protected areas, and climate change. Bioscience. 2005;55:231–242. [Google Scholar]
  24. Hardy OJ, Vekemans X. SPAGEDi: a versatile computer program to analyse spatial genetic structure at the individual or population levels. Molecular Ecology Notes. 2002;2:618–620. [Google Scholar]
  25. Hardy OJ, Gonzalez-Martinez SC, Colas B, Freville H, Mignot A, Olivieri I. Fine-scale genetic structure and gene dispersal in Centaurea corymbosa (Asteraceae). II. Correlated paternity within and among sibships. Genetics. 2004;168:1601–1614. doi: 10.1534/genetics.104.027714. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hardy OJ, Gonzalez-Martinez SC, Freville H, et al. Fine-scale genetic structure and gene dispersal in Centaurea corymbosa (Asteraceae). I. Pattern of pollen dispersal. Journal of Evolutionary Biology. 2004b;17:795–806. doi: 10.1111/j.1420-9101.2004.00713.x. [DOI] [PubMed] [Google Scholar]
  27. Harper JL. Ranunculus acris L. (R. acer auct. plur.) Journal of Ecology. 1957;45:289–342. [Google Scholar]
  28. Hirao AS, Kudo G. Landscape genetics of alpine-snowbed plants: comparisons along geographic and snowmelt gradients. Heredity. 2004;93:290–298. doi: 10.1038/sj.hdy.6800503. [DOI] [PubMed] [Google Scholar]
  29. Hirao AS, Kudo G. The effect of segregation of flowering time on fine-scale spatial genetic structure in an alpine-snowbed herb Primula cuneifolia. Heredity. 2008;100:424–430. doi: 10.1038/hdy.2008.1. [DOI] [PubMed] [Google Scholar]
  30. IPCC. Climate Change 2007: the physical science basis. New York, NY: Cambridge University Press; 2007. [Google Scholar]
  31. Ishihama F, Ueno S, Tsumura Y, Washitani I. Effects of density and floral morph on pollen flow and seed reproduction of an endangered heterostylous herb, Primula sieboldii. Journal of Ecology. 2006;94:846–855. [Google Scholar]
  32. Kitamoto N, Ueno S, Takenaka A, Tsumura Y, Washitani I, Ohsawa R. Effect of flowering phenology on pollen flow distance and the consequences for spatial genetic structure within a population of Primula sieboldii (Primulaceae) American Journal of Botany. 2006;93:226–233. doi: 10.3732/ajb.93.2.226. [DOI] [PubMed] [Google Scholar]
  33. Klein EK, Desassis N, Oddou-Muratorio S. Pollen flow in the wildservice tree, Sorbus torminalis (L.) Crantz. IV. Whole interindividual variance of male fecundity estimated jointly with the dispersal kernel. Molecular Ecology. 2008;17:3323–3336. doi: 10.1111/j.1365-294X.2008.03809.x. [DOI] [PubMed] [Google Scholar]
  34. Körner C. Alpine plant life: functional plant ecology of high mountain ecosystems. Heidelberg: Springer; 2003. [Google Scholar]
  35. Larcher W. Altitudinal variation in flowering time of lilac (Syringa vulgaris L.) in the Alps in relation to temperatures. Oesterreichische Akademie der Wissenschaften Mathematisch-Naturwissenschaftliche Klasse Sitzungsberichte Abt I. 2006;212:3–18. [Google Scholar]
  36. Loiselle BA, Sork VL, Nason J, Graham C. Spatial genetic structure of a tropical understory shrub, Psychotria officinalis (Rubiaceae) American Journal of Botany. 1995;82:1420–1425. [Google Scholar]
  37. Manel S, Schwartz MK, Luikart G, Taberlet P. Landscape genetics: combining landscape ecology and population genetics. Trends in Ecology & Evolution. 2003;18:189–197. [Google Scholar]
  38. Matter P, Määttänen K, Kettle CJ, Ghazoul J, Pluess AR. Eleven microsatellite markers for the mountain clover Trifolium montanum (Fabaceae) American Journal of Botany – Primer Notes & Protocols in the the Plant Sciences. 2012a:e447–e449. doi: 10.3732/ajb.1200102. http://dx.doi.org/10.3732/ajb.1200102 . [DOI] [PubMed] [Google Scholar]
  39. Matter P, Pluess AR, Ghazoul J, Kettle CJ. Eight microsatellite markers for the bulbous buttercup Ranunculus bulbosus (Ranunculaceae) American Journal of Botany – Primer Notes & Protocols in the the Plant Sciences. 2012b:e399–e401. doi: 10.3732/ajb.1200101. http://dx.doi.org/10.3732/ajb.1200101 . [DOI] [PubMed] [Google Scholar]
  40. Matter P, Kettle CJ, Ghazoul J, Hahn T, Pluess AR. Evaluating contemporary pollen dispersal in two common grassland species Ranunculus bulbosus L. (Ranunculaceae) and Trifolium montanum L. (Fabaceae) using an experimental approach. Plant Biology. 2013 doi: 10.1111/j.1438-8677.2012.00667.x. in press. http://dx.doi.org/10.1111/j.1438-8677.2012.00667.x . [DOI] [PubMed] [Google Scholar]
  41. Miyazaki Y, Isagi Y. Pollen flow and the intrapopulation genetic structure of Heloniopsis orientalis on the forest floor as determined using microsatellite markers. Theoretical and Applied Genetics. 2000;101:718–723. [Google Scholar]
  42. 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. doi: 10.1111/j.1365-294X.2004.02164.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Oddou-Muratorio S, Bontemps A, Klein EK, Chybicki I, Vendramin GG, Suyama Y. Comparison of direct and indirect genetic methods for estimating seed and pollen dispersal in Fagus sylvatica and Fagus crenata. Forest Ecology and Management. 2010;259:2151–2159. [Google Scholar]
  44. Oksanen J, Blanchet G, Kindt R, et al. vegan, a Community Ecology Package. 2012 R-package version 2.0–4. http://vegan.r-forge.r-project.org/ , http://cran.r-project.org . [Google Scholar]
  45. Pettersson MW, Sjodin E. Effects of experimental plant density reductions on plant choice and foraging behaviour of bees (Hymenoptera: Apoidea) Acta Agriculturae Scandinavica Section B – Soil and Plant Science. 2000;50:40–46. [Google Scholar]
  46. Pluess AR, Stocklin J. Population genetic diversity of the clonal plant Geum reptans (Rosaceae) in the Swiss Alps. American Journal of Botany. 2004;91:2013–2021. doi: 10.3732/ajb.91.12.2013. [DOI] [PubMed] [Google Scholar]
  47. Ritland K. Extensions of models for the estimation of mating systems using n independent loci. Heredity. 2002;88:221–228. doi: 10.1038/sj.hdy.6800029. [DOI] [PubMed] [Google Scholar]
  48. Robledo-Arnuncio JJ, Gil L. Patterns of pollen dispersal in a small population of Pinus sylvestris L. revealed by total-exclusion paternity analysis. Heredity. 2005;94:13–22. doi: 10.1038/sj.hdy.6800542. [DOI] [PubMed] [Google Scholar]
  49. Robledo-Arnuncio JJ, Austerlitz F, Smouse PE. A new method of estimating the pollen dispersal curve independently of effective density. Genetics. 2006;173:1033–1045. doi: 10.1534/genetics.105.052035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Robledo-Arnuncio JJ, Austerlitz F, Smouse PE. POLDISP: a software package for indirect estimation of contemporary pollen dispersal. Molecular Ecology Notes. 2007;7:763–766. [Google Scholar]
  51. Rong J, Janson S, Umehara M, Ono M, Vrieling K. Historical and contemporary gene dispersal in wild carrot (Daucus carota ssp. carota) populations. Annals of Botany. 2010;106:285–2. doi: 10.1093/aob/mcq108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Scheepens JF, Frei ES, Stocklin J. Genotypic and environmental variation in specific leaf area in a widespread Alpine plant after transplantation to different altitudes. Oecologia. 2010;164:141–150. doi: 10.1007/s00442-010-1650-0. [DOI] [PubMed] [Google Scholar]
  53. Scheepens JF, Frei ES, Armbruster GF, Stocklin J. Pollen dispersal and gene flow within and into a population of the alpine monocarpic plant Campanula thyrsoides. Annals of Botany. 2012;110:1479–1488. doi: 10.1093/aob/mcs131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Scherrer D, Körner C. Topographically controlled thermal-habitat differentiation buffers alpine plant diversity against climate warming. Journal of Biogeography. 2011;38:406–416. [Google Scholar]
  55. Schleuning M, Niggemann M, Becker U, Matthies D. Negative effects of habitat degradation and fragmentation on the declining grassland plant Trifolium montanum. Basic and Applied Ecology. 2009;10:61–69. [Google Scholar]
  56. Slatkin M. Gene flow in natural populations. Annual Review of Ecology and Systematics. 1985;16:393–430. [Google Scholar]
  57. Slatkin M. Gene flow and the geographic structure of natural populations. Science. 1987;236:787–792. doi: 10.1126/science.3576198. [DOI] [PubMed] [Google Scholar]
  58. Snaydon RW, Davies MS. Rapid population differentiation in a mosaic environment. 4. populations of Anthoxanthum odoratum at sharp boundaries. Heredity. 1976;37:9–25. [Google Scholar]
  59. Stanton ML, Galen C. Life on the edge: adaptation versus environmentally mediated gene flow in the snow buttercup, Ranunculus adoneus. The American Naturalist. 1997;150:143–178. doi: 10.1086/286061. [DOI] [PubMed] [Google Scholar]
  60. Stanton ML, Galen C, Shore J. Population structure along a steep environmental gradient: consequences of flowering time and habitat variation in the snow buttercup, Ranunculus adoneus. Evolution. 1997;51:79–94. doi: 10.1111/j.1558-5646.1997.tb02390.x. [DOI] [PubMed] [Google Scholar]
  61. Steinbach K, Gottsberger G. Phenology and pollination biology of 5 Ranunculus species in Giessen, Central Germany. Phyton – Annales Rei Botanicae. 1994;34:203–218. [Google Scholar]
  62. Stöcklin J, Bosshard A, Klaus G, Rudmann-Maurer K, Fischer M. Landnutzung und biologische Vielfalt der Alpen – Fakten, Perspektiven, Empfehlungen. Thematische Synthese zum Forschungsschwerpunkt II, «Land- und Forstwirtschaft im alpinen Lebensraum». 2007 Swiss National Science Foundation. [Google Scholar]
  63. Theurillat JP, Guisan A. Potential impact of climate change on vegetation in the European Alps: a review. Climatic Change. 2001;50:77–109. [Google Scholar]
  64. Thomas CD, Cameron A, Green RE, et al. Extinction risk from climate change. Nature. 2004;427:145–148. doi: 10.1038/nature02121. [DOI] [PubMed] [Google Scholar]
  65. Walther GR, Post E, Convey P, et al. Ecological responses to recent climate change. Nature. 2002;416:389–395. doi: 10.1038/416389a. [DOI] [PubMed] [Google Scholar]
  66. Ziello C, Estrella N, Kostova M, Koch E, Menzel A. Influence of altitude on phenology of selected plant species in the Alpine region (1971–2000) Climate Research. 2009;39:227–234. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_111_4_611__index.html (1.1KB, html)
supp_mct021_mct021supp.pdf (372.5KB, pdf)

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

RESOURCES