Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Inter-Simple Sequence Repeat Data Reveals High Genetic Diversity in Wild Populations of the Narrowly Distributed Endemic Lilium regale in the Minjiang River Valley of China

  • Zhu-hua Wu,

    Affiliations College of Landscape Architecture of Nanjing Forestry University, Nanjing, China, Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, College of Forestry, Nanjing Forestry University, Nanjing, China, Forest and Evolutionary Genomics Laboratory and Centre for Structural and Functional Genomics, Biology Department, Concordia University, Montreal, Quebec, Canada, Quebec Centre for Biodiversity Sciences, Montreal, Quebec, Canada

  • Jisen Shi ,

    juliawu9@gmail.com

    Affiliation Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, College of Forestry, Nanjing Forestry University, Nanjing, China

  • Meng-li Xi,

    Affiliation Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, College of Forestry, Nanjing Forestry University, Nanjing, China

  • Fu-xing Jiang,

    Affiliation Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, College of Forestry, Nanjing Forestry University, Nanjing, China

  • Ming-wen Deng,

    Affiliation Key Laboratory of Forest Genetics and Biotechnology, Ministry of Education, College of Forestry, Nanjing Forestry University, Nanjing, China

  • Selvadurai Dayanandan

    Affiliations Forest and Evolutionary Genomics Laboratory and Centre for Structural and Functional Genomics, Biology Department, Concordia University, Montreal, Quebec, Canada, Quebec Centre for Biodiversity Sciences, Montreal, Quebec, Canada

Inter-Simple Sequence Repeat Data Reveals High Genetic Diversity in Wild Populations of the Narrowly Distributed Endemic Lilium regale in the Minjiang River Valley of China

  • Zhu-hua Wu, 
  • Jisen Shi, 
  • Meng-li Xi, 
  • Fu-xing Jiang, 
  • Ming-wen Deng, 
  • Selvadurai Dayanandan
PLOS
x

Abstract

Lilium regale E.H. Wilson is endemic to a narrow geographic area in the Minjiang River valley in southwestern China, and is considered an important germplasm for breeding commercially valuable lily varieties, due to its vigorous growth, resistance to diseases and tolerance for low moisture. We analyzed the genetic diversity of eight populations of L. regale sampled across the entire natural distribution range of the species using Inter-Simple Sequence Repeat markers. The genetic diversity (expected heterozygosity= 0.3356) was higher than those reported for other narrowly distributed endemic plants. The levels of inbreeding (Fst = 0.1897) were low, and most of the genetic variability was found to be within (80.91%) than amongpopulations (19.09%). An indirect estimate of historical levels of gene flow (Nm =1.0678) indicated high levels of gene flow among populations. The eight analyzed populations clustered into three genetically distinct groups. Based on these results, we recommend conservation of large populations representing these three genetically distinct groups.

Introduction

Lilies (Lilium spp.) are perennial herbaceous monocots of the family Liliaceae and are commercially important as ornamental plants throughout the world. In addition, lily bulbs are regularly consumed as food in Asia [1], and some species, including L. longiflorum Thunb., L.browniiF.E.Br. ex Miellez var. viridulum Baker, L. pensylvanicum Ker Gawl. And L. pumilum DC, have been used traditionally in China as sedative, anti-inflammatory and antitussive agents(cough suppressants)[2,3]. The crude extract “Baihe” is prepared from the bulbs of Lilium spp. and is regularly used to treat lung ailments in China [4].

The native range of Lilium is centered around China with almost half of the species being distributed within China [5].Lilium regale Wilson is endemic to southwestern China, with the distribution limited to an area of approximately170 km2 in the valley region within the upper reach of the Minjiang River [6,7].L. regale blooms annually from May through June, with conspicuous white trumpet-shaped flowers. Since the introduction of L. regale to North America during the period from 1907 to 1911 by Wilson [8],due to its vigorous growth and resistance to diseases[9,10,11], it has been widely used as a parent in breeding programs and this has led to the production of many hybrid cultivars [12],including some resistant to the devastating wilt disease of lilies [8].

The native habitats of L. regale are subject to severe anthropogenic and natural disturbances including land use changes, landslides and earthquakes [13], which threaten the survival of the species. As L. regale is a commercially important endemic species restricted to a narrow geographic area in which severe environmental disturbances are likely to occur, conservation and management plans to protect wild populations are urgently needed. The long-term survival of the species depends on the maintenance of sufficient genetic variability to respond and adapt to new selective pressures under changing environmental conditions [14,15,16]. As such, the assessment of genetic variability serves as the first step for evaluating the long-term survival potential of a given species and developing suitable conservation programs [17,18,19].

In general, species with a narrow range of distribution are considered to have lower levels of genetic diversity than their more widespread counterparts, as they undergo genetic drift, inbreeding and low levels of gene flow resulting from historically small and fragmented populations [20]. Although some studies support this view[21,22,23,24], others have revealed moderate to high levels of genetic diversity in some endemic species with narrow distributional ranges [16,25,26]. Life history traits such as the mating system, seed and pollen dispersal strategies, and mode of reproduction also influence genetic diversity within and among populations[19,20,27].

To date, the genetic diversity of only six Lilium species has been assessed using molecular techniques[28,29,30,31,32,33]. Random Amplified Polymorphic DNA markers have been used to assess the genetic variation in Turk's-cap lily (L. martugon L.), one of the most common lily species in Europe [28], the Taiwan lily (L. longiflorum Thunb. var. formosanum Baker)distributed in wide altitudinal ranges from the lowlands to high mountains of Taiwan[29], and L. brownii in southern China[30]. A chloroplast DNA and Inter-Simple Sequence Repeat (ISSR) marker-based study evaluated the genetic diversity of L.maculatum var. bukosanense, the Miyamasukashi-yuri lily, which is endemic to Japan[31].Simple Sequence Repeat (SSR)-based studies have focused on characterizing the genetic diversity and structure of L. philadelphicum in the midwestern United States [32]and of L. japonicum Thunb. var. Abeanum(Honda) Kitam, one of the rarest plants in Japan [33].

In the present study, we used Inter Simple Sequence Repeat (ISSR) data to analyze genetic diversity of L. regale populations across its entire natural range. ISSR markers represent mostly noncoding regions of the genome flanked by SSRs. Several genetic studies of natural populations using ISSR markers have demonstrated their high variability, reproducibility [16,34,35,36] and relative ease of use in population-level studies in a variety of organisms [36,37,38] including plants [15,22,39,40,41]. ISSR markers are particularly useful for population genetic studies of rare species for which no previously characterized molecular markers are available [21,22,40,41]. We analyzed eight populations of L. regale distributed in the Minjiang River valley of China using ISSR markers to assess the levels and distribution of genetic diversity within and among natural populations.

Materials and Methods

Sample collection and sampling sites

The Minjiang River valley is relatively dry compared to humid surroundings in southwestern China, making this a unique habitat for plants. L.Regale is not an endangered or protected species and therefore no specific permission is currently required for academic studies. Although L. regale individuals are abundant in the area, they are not prominent because of their small size relative to the dominant vegetation, which consists of several shrub species including Caryopteris incana,Ceratostigma minus,Lespedeza Formosa and Artemisia annua[7].Therefore, we collected samples during the flowering season in May and June, when L. regale plants can be easily detected.

We first surveyed for L.regale within its established distribution region [5,7,8] by direct searching and consulting with local forest managers and farmers. We determined the approximate current distribution (Fig. 1), which comprised three major rivers(Minjiang, Heishui, and Zagunao)within the Minjiang valley. Then, based on the potential impact of the watershed on genetic diversity and accessibility of populations, eight sites were chosen for sampling, including five sites along both banks of the Minjiang River, one site in the Heishui tributary valley, and two sites in the Zagunao tributary valley(Table 1, Fig. 1).

thumbnail
Fig 1. Locations of L. regale sampling sites along the Minjiang River valley.

Population names are listed in Table 1.

https://doi.org/10.1371/journal.pone.0118831.g001

thumbnail
Table 1. Locations and sample sizes of 8 populations of L. regale used in the present study.

https://doi.org/10.1371/journal.pone.0118831.t001

Leaf samples from19–23 individuals [42]located at 2-m intervals along a transect in each site were collected, placed in sealed plastic bags containing silica gel to prevent degradation of DNA, transported to the laboratory, and stored at -80°C until extraction of the genomic DNA.

DNA extraction and ISSR amplification

The leaf tissue (100mg) from each sampled individual was ground to powder in liquid nitrogen, and total genomic DNA was extracted following the CTAB-based procedure of Armaleo et al.[43]. The extracted DNA was amplified by PCR using 100 ISSR primers obtained from the University of British Columbia and 35 Lilium-specific ISSR primers from Masumi Yamagishi[44],which were synthesized by SBSGenetechCo., Ltd.Only 10 primers produced clear and reproducible bands and were selected for subsequent genotyping (Table 2).

thumbnail
Table 2. The ISSR primer sequences, number of bands observed and assessment of banding patterns.

https://doi.org/10.1371/journal.pone.0118831.t002

The genotyping of PCR reactions was performed in 20-μl reaction volumes containing 20 ngtemplate DNA,0.4mMdNTPs, 1.25 mM MgCl2,0.6μM of each primer, 0.5UTaq-polymerase(Takara),2μl of 10×PCR buffer and double-distilled water. The thermal cycling steps consisted of an initial denaturation at 94°C for 5 minutes, followed by 45 cycles of denaturation at 94°C for 40 seconds, annealing at 52°C for 45 seconds and extension at 72°C for 90 seconds, followed by a final extension at 72°C for 8 minutes. PCR amplification was performed in a PTC-200 DNA thermocycler (MJ Research, Boston, MA). The amplification products were separated on 1.5% agarose gels containing 0.5 mg/ml ethidium bromide in 1× Tris-borate-EDTA buffer. The gels were examined under ultraviolet light and photographed using a Kodak EDAS 200 digital documentation system. All samples were analyzed at least twice with each primer to assess the reproducibility of ISSR-based genotyping.

Data analysis

The ISSR banding pattern of each individual was scored as the presence or absence of a given band to construct a binary data matrix (1 and 0) consisting of 172 individuals. The informativeness of markers and diversity was evaluated using marker index (MI) [45], Shannon’s index (H′) [46] and the polymorphic information content(PIC) [47,48]. The PIC of dominant bi-allelic data was estimated by the formula PIC = 1-pi2-qi2, where p is the frequency of visual alleles and q is the frequency of null alleles (Hardy-Weinberg equilibrium was assumed, where q = (1band frequency)/2 and p = 1-q).MI was calculated as MI = PIC × number of polymorphic loci. Shannon’s index (H′) was calculated by the formula H′ = -pilnpi. We calculated standard genetic parameters, including the average effective number of alleles (Ne), Nei’s gene diversity (h)[49], Shannon’s information index (I)[46]and Nei’s unbiased measures of genetic distance [49] and gene flow [50],using the software program POPGENE version 1.31 [51].We also estimated the expected heterozygosity(He), observed heterozygosity(Ho), inbreeding coefficient (Fis)and linkage disequilibrium across sampled loci and performed an analysis of molecular variance (AMOVA) to examine the partitioning of genetic variability within and among populations using the Arlequin software (version 3.5)[52]. The unweighted pair group method with arithmetic mean (UPGMA)dendrogram based on the genetic distances[53] was constructed using the software program TFPGA(version1.3). The principal component analysis of pairwise genetic distance between individuals was performed using GenAlEx version 6.4 [54].

We calculated the Bayesian estimate of panmictic heterozygosity (hs)as a measure of genetic diversity and θI(the Bayesian analog of Wright’s pairwise population distribution, Fst) and θB(the Bayesian analog of Nei’s Gst) as measures of population differentiation using the Hickory program (version 1.0)[55], which uses Markov chain Monte Carlo simulations to produce posterior distributions based on the data[55,56,57,58].Several runs of the analyses were performed with default sampling parameters (burn in = 50,000; sample = 250,000; thin = 50) to ensure consistency of results [59].The gene flow (Nm) was estimated using the formula Nm = (1—Fst) / 4Fst[60].Bayesian model-based clustering analysis was used for determining the optimal number of genetic clusters among populations using the software program STRUCTURE 2.3.3 [61], which partitions individuals into numbers of clusters (K) based on the multi-locus genotypic data. The admixture model and correlated allele frequencies were used for each run with a burn-in period of 10,000 and 100,000 Markov chain Monte Carlo replications. The optimal K value, which indicates the number of genetically distinct clusters in the data, was determined from 10 replicate runs for each value of K[62]. The value of ΔK was based on the change in the log probability of the data between successive K values. Structure Harvester version 6.0 [63] was used to calculate parameters described by Evanno et al. [62].We confirmed that five independent runs consistently produced the same values.

We performed the Mantel test [64] in the Isolation By Distance program version 1.52 [65]to examine correlations of pairwise genetic distances of populations [49] and Fst with a geographic distance matrix.

Results

ISSR polymorphism

The 10 selected primers generated a total of 180 unambiguous and reproducible bands with sizes ranging from 220 to 2000 bp, of which 176 (97.7%) were polymorphic. The number of bands per primer ranged between 10 (primer 3A53) and 22 (primer 3A21), with an average of 18 bands per primer (Table 2). Among the 32,400 possible combinations, only 745 pairs (2.3%) showed significant linkage disequilibrium between loci (P ≤0.05). This is within the range expected by random chance, and therefore all 180 bands were considered independent loci. To determine PIC values for each ISSR primer, we analyzed the mean PIC values for all loci of each ISSR primer. We detected a high mean PIC value for ISSR primer3A39 (0.2783) and a low mean PIC value for ISSR primer 3A42 (0.1597) (Table 2). The average PIC per primer was 0.2155. The highest MI was observed with ISSR primer 3A39 (5.566) and the lowest with ISSR primer 3A53 (2.0988). There was a positive correlation between the MI and PIC values (r2 = 0.909, P<0.05).The highest H′ was observed with primer 3A07(0.3858) and the lowest with 3A42 (0.2779).

Genetic variability within populations

The average observed number of alleles per population ranged from 1.672 in YMF to 1.811 in TF. FHB,WZT and SDG showed relatively high levels of observed allelic diversity, whereas populations SEG, XV and SP showed low levels of observed allelic diversity(Table 3). The mean effective number of alleles per population ranged from 1.2047 in WZT to 1.3211 in FHB. The YMF and XV populations showed relatively high levels of effective alleles, and the SEG and SP populations showed relatively low levels of effective allelic diversity. A similar pattern was observed in h and I, with the highest values of h(0.2018) and I(0.3180) in FHB and lowest values (h = 0.1419 and I = 0.2413) in WZT. The SDG, XV and YMF populations showed moderate values of h and I, and the values were lower for the TF,SEG and SP populations. The overall average values of h and I were 0.1989 and 0.3339, respectively. The values of He ranged from 0.2850 in WZT to 0.3473 in FHB. The YMF, XV and SDG populations showed relatively high values of He, TF showed moderate values and SP and SEG showed relatively low values. The highest value of Ho occurred in the FHB population, followed by YMF and TF. The lowest value of Ho was found in SEG. The Bayesian estimates of the panmictic heterozygosity[32]ranged from 0.2192 in SP to 0.2648 in FHB. Overall, all measures of genetic diversity revealed the highest diversity in FHB, whereas SDG,XV,YMF and TF showed moderate levels of genetic diversity, and SEG,WZT and SP showed low levels of genetic diversity.

thumbnail
Table 3. Measures of genetic diversity and inbreeding coefficients in 8 populations of L. regale.

https://doi.org/10.1371/journal.pone.0118831.t003

The Hardy-Weinberg equilibrium test based on Markov chain Monte Carlo iterations revealed significant departures from Hardy-Weinberg equilibrium at five and six loci in populations XV and SEG, respectively. Similar departures from Hardy-Weinberg equilibrium were observed at one to three loci in the remaining populations, suggesting a varying degree of inbreeding or selection in populations.

Genetic differentiation among populations

Overall, the population differentiation was moderate to high (Fst = 0.1909, P = 0.00). The results of the AMOVA indicated that most (80.91%) genetic variation was within populations, with a moderate amount (19.09%) among populations.

The inbreeding coefficient value (Table 4) was highest in the SEG population (0.1613) followed by FHB(0.1600) and SDG (0.1570). The inbreeding coefficient value of the WZT population was the lowest (0.0361). The average θI was0.1897 (Model f = 0,s.d. = 0.0097) and the average θB was 0.1849 (Full model, s.d. = 0.086). The θB value was similar to, but slightly higher than, the Gstvalue (0.1619) estimated using the POPGENE software, while θI was similar to the Fst value (0.1909) calculated with the Arlequin software. The gene flow estimate (Nm) was 1.0678 and suggests a moderate level of migration among populations. The pairwise Nei’s genetic distances (DS) among populations ranged from 0.0201 between SDG in the Minjiang River valley and TF in the Zagunao River valley to 0.0636 between the WZT and SP populations. The pairwise Fst values ranged from 0.1024 between SDG and TF to 0.2581 between XV and WZT (Table 5).

thumbnail
Table 4. The results of the Bayesian model based genetic differentiation and inbreeding coefficients in natural populations of L. regale.

https://doi.org/10.1371/journal.pone.0118831.t004

thumbnail
Table 5. Nei’s unbiased (1978) measures of genetic distance (below the diagonal) and Pairwise Fst (above the diagonal) observed in 8 populations of L. regale.

https://doi.org/10.1371/journal.pone.0118831.t005

There were no significant correlations between the geographic and genetic distances tested (DS: r = -0.2495, P ≤ 0.1130; Fst: r = -0.2744, P ≤ 0.1040) or population differentiation parameters (DS: P ≤ -0.2841, Fst:P≤ 0.8960).

The UPGMA clustering based on DS values grouped populations into three distinct groups (Fig. 2). Group-A in the UPGMA tree consisted only of WZT. Group-B consisted of the SDG,SEG,XV,TF and SP populations, and Group-Cconsisted of YMF and FHB.

thumbnail
Fig 2. UPGMA dendrogram based on DS values of eight L.regale populations.

https://doi.org/10.1371/journal.pone.0118831.g002

The percentages of variation explained by the first three axes in principal component analysis(PCA) were 58.37%,29.80% and10.02%.Theprincipal component analysis also clustered the eight populations into three distinct groups, but these were different from the three groups identified by UPGMA. In PCA, Group i consisted only of population TF, Group ii consisted of FHB, XV and SP, and Group iii consisted of YMF, SEG,SDG and WZT(Fig. 3).The Bayesian-based analysis of the population structure using the STRUCTURE software showed the highest log likelihood of change in K at three(Fig. 4), suggesting the existence of three genetically distinguishable groups. Individuals were then assigned to clusters based on posterior probability estimates of inclusion into each group. Individuals with a posterior probability ≥0.8 were considered confidently assigned to a single cluster. Those with split probabilities were assigned into two or more clusters and considered admixed individuals. The assignment of individuals into three clusters (Fig. 5) revealed that Group-I (blue) contained21.5% of the total individuals, with the majority being from the YMF and FHB populations. Another 20.3% of individuals, all but one (from SDG) of which were from the SEG and WZT populations, formed Group-II (red). Group-III (green) consisted of 30.8% of the individuals, the majority of which were from the XV and SP populations, with a few individuals from SDG and TF. The remaining 27.3% of individuals, mainly from the SDG and TF populations, showed high levels of mixed ancestry. Within this set, 46.8% of individuals clustered with all three groups, 29.8% clustered with Group-II and Group-III, 8.5% clustered with Group-I and Group-III, and 6.4% clustered with Group-I and Group-II. This indicates that the SEG and WZT populations, which are geographically separate, share the same ancestry. Similarly, the SDG and TF populations also showed mixed ancestry.

thumbnail
Fig 4. The relationship between ΔK and Kas revealed by Structure Harvester.

The highest peak was at K = 3.

https://doi.org/10.1371/journal.pone.0118831.g004

thumbnail
Fig 5. Bayesian model-based clustering (STRUCTURE)[60] of L. regale individuals (optimal value of K = 3).

. YMF,2.FHB,3.SDG,4.SEG,5.XV,6.TF,7.WZT,8.SP. Blue: Group-I; red: Group-II; green: Group-III.

https://doi.org/10.1371/journal.pone.0118831.g005

Discussion

ISSR Primers

The PIC of the ISSR primers ranged from 0.1597 to 0.2783, and the MI ranged from 2.0988 to 5.5660,which is lower than the SSR primers for this species[66].Therefore, the universal primers may not be particularly suitable to L.regale. Although the details of the mating system of L.regale are not well known, plants grown ex situin conservation sites have been observed to produce seeds through self-pollination[67].The species is generally considered to have a mixed mating system, and reproduces both sexually and asexually through subterranean bulbs. The seeds of L. regale are winged, semi-circular, very thin and light (~5mg) [68]. Both pollen and seeds show adaptations for wind dispersal. The genetic diversity and population differentiation parameters which we obtained are in agreement with these features of L. regale.

Genetic diversity

The evolutionary potential of a species largely depends on its level of genetic variability, and, for plants, the overall level of genetic variability in a population is often correlated with the geographical distribution range of the species[19,27,69].Compared with other narrowly distributed species of the genus Lilium, the genetic diversity of L. regale(Ho = 0.2750,He = 0.3139, h = 0.1989 and I = 0.3339) was moderately high. Higher genetic diversity in L. regale is likely to be attributable to its potential mixed mating system and long-range pollen and seed dispersal. The genetic diversity of L. regale is higher than the average genetic diversity reported for rare and common plant species at the population (rare: He = 0.113, Ho = 0.100; common: He = 0.150, Ho = 0.139) and species (rare: He = 0.142; common: He = 0.199) levels [70]. Overall, the genetic diversity measures that we found for L. regale were higher than those reported for narrowly endemic plant taxa [20,70,71].

Mating system

The mating system can greatly influence genetic diversity both within and among populations. In general, most of the genetic diversity in self-pollinated plants is distributed among populations, whereas in outcrossed plants, most of the genetic diversity is distributed within populations[20]. The values for inbreeding and population genetic differentiation of L. regale observed in the present study (Fis = 0.1353, Fst = 0.1897, Gst = 0.1619) fall roughly in between those reported for selfed and outcrossed plants, further suggesting a mixed mating system in L. regale. Similarly, the self-pollination of L. regale [56] under exsitu conditions indicates that L. regale is self-compatible. However, the observed Gst value (0.1619) of L. regale was considerably lower than the Gst value (0.510) reported for generally self-pollinated plant species [20]. Furthermore, the strong fragrance of L.regale flowers suggests that they may attract animals as pollinators. In addition, strong and frequent wind in the Minjiang River valley [13]may facilitate pollination and seed dispersal. The trumpet-shaped flowers, large stamens, large numbers of highly viable pollen grains, and large, sticky stigma of L.regale also suggest wind pollination. Thus, it is possible that a combination of animal and wind pollination mechanisms function in L. regale. This is further supported by the fact that the Gst value of L.regale falls between those of typically wind-pollinated (Gst = 0.100) and typically animal-pollinated (Gst = 0.216) plants [20].

Plants with narrow distribution ranges are prone to high levels of inbreeding, leading to loss of genetic variation and differentiation of populations[72,73,74]. Although L. regale is an endemic species with a limited distribution range, the levels of inbreeding in L. regale were lower than the values reported for rare (Fis = 0.175) and common (Fis = 0.184) plant species [70], and measures of genetic structure were also lower than those reported for rare (Fst = 0.212) and common (Fst = 0.198) species. The inbreeding value in L. regale was also lower than the range reported for the congeneric species L. japonicum Thunb. var. abeanum(Honda) Kitam (Fis = 0.042–0.385)[33]. Geneflow among populations decreases inbreeding, and the relatively high gene flow value of L. regale (Nm = 1.0678) that we observed may account for the low levels of inbreeding. The light, thin seeds of L. regale with wind velocity of 1.4–4m/s[75]may be dispersed to considerable distances [76]. However, it is unknown whether L. regale seeds and pollen commonly disperse across the rivers.

Although most species of the genus Lilium grow in the forest understory in shaded and moist conditions, L. regale grows in dry habitats in high sunlight in the Minjiang River valley. The high level of outcrossing due to wind-mediated pollination and seed dispersal may increase the genetic diversity in L.regale, allowing it to escape inbreeding depression and to respond to selection pressure under these harsh environmental conditions. Thus, the endemic and narrowly distributed L.regale maintains relatively high genetic diversity and gene flow through its adaptations for the environmental conditions in the Minjiang River valley.

Genetic structure

The results of the AMOVA indicated that most of the variation in L.regale was within populations rather than among populations. The UPGMA, principal component and STRUCTURE analyses clustered the populations into three groups. However, the three groups determined by the three approaches consisted of different populations, with SEG,SDG,WZT and TF being the most differentially represented, which might be are result of the mixed ancestry of these populations. None of the clusters corresponded to the geographical distribution of populations, except for YMF and FHB.

The among-population genetic variation of L. regale was lower than the corresponding value of the self-incompatible species L. brownii (24.67%) [30]but was higher than the inter-population variation (5.09%) in the self-pollinating Taiwan lily, L. longiflorum Thunb. var. formosanum Baker [29],again suggesting a mixed mating system in L. regale.

Asexual and sexual reproduction

Typically, clonal propagation is expected to give rise to higher levels of linkage disequilibrium, differentiation between populations, and heterozygosity than Hardy-Weinberg expectations for sexual reproduction [77,78]. The L.regale populations showed approximately2.3% linkage disequilibrium, and several of the populations showed significant departure from Hardy-Weinberg predictions, suggesting some clonal reproduction. The high number of ramets per genet will result in a large numbers of viable pollen and seeds, increasing the chances of successful establishment and survival of progeny derived from that genet. Moreover, for species with both asexual and sexual reproduction, efficient clonal propagation may result in a reduction in sexual reproduction[79,80]. As a consequence, a mixture of both clonal spread and sexual reproduction may be beneficial in a heterogeneous environment [81,82]. Therefore, the combination of sexual and clonal reproduction may contribute to the maintenance of high genetic diversity in L. regale.

Clonal plants may improve fitness in stressful environments by increasing resource acquisition through the development of a large rooting system, which may involve the physical integration of ramets[83], and reduction in resource expenditure for seed production and a concomitant increase in the proportion allocated to vegetative propagation[84]. Under these circumstances, low genetic diversity is expected in stressful environments [85]. In contrast, clonal plants may also increase investment in seed production under unfavorable conditions to respond to harsh environments [86,87]. Thus, high genetic diversity as a result of repeated seedling recruitment through sexual reproduction can also be expected under stressful conditions [88]. However, we did not encounter high levels of clonal reproduction in L.regale, suggesting that the high genetic diversity in this species is a consequence of recruitment of individuals through sexual reproduction.

Conclusions

Despite being an endemic plant species confined to a relatively small geographical region in southwest China, L. regale shows relatively high genetic diversity, mostly within populations, and low levels of inbreeding compared with other narrowly distributed endemic plants. Strong winds prevalent in the river valley and a possible mixed mating system in this species may play important roles in maintaining a high level of gene flow among populations through pollen and seed dispersal. The fact that there were three genetically distinct groups of populations of L.regale should be taken into consideration in planning conservation and management programs for this species.

Conservation implications

Lilium regale may serve as an important germplasm for the improvement of many lily varieties, and conservation strategies to protect native populations and prevent genetic erosion are urgently needed. For example, because L.regale grows in relatively dry environments, it likely carries alleles that confer tolerance to drought conditions that are predicted to become widespread under climate change scenarios[89].

Several measures must be implemented for the conservation and protection of L. regale populations. First, measures should be taken to minimize anthropogenic impact on native populations of L. regale. The natural habitat of L. regale is subject to heavy grazing and farming activities. The thin, dry soil of the habitat is prone to erosion leading to the loss of many populations and reducing the sizes of remaining populations[89]. Second, conservation plans should be implemented to preserve the maximum amount of genetic diversity [90]. We recommend the protection of large populations representing the three genetically distinct groups, such as the FHB, SEG and XV populations. Because physical barriers such as topographically dissected terrain and water bodies may hinder gene flow, several populations strategically located to promote gene flow should also be maintained. For instance, the location of the FHB population is central to maintaining gene flow across the river to increase network connectivity [76]. Third, genetic diversity parameters should be routinely monitored to capture long-term changes in the genetic diversity and implement mitigation measures in a timely fashion. Thus, the results of the present study may serve as the baseline data for future monitoring purposes. Finally, conservation of L. regale should focus on maintaining local adaptation. Thus, populations representing diverse habitats that include altitudinal or soil moisture extremes should be conserved.

Further analyses using more robust and more variable markers such as microsatellites or whole genome sequence-based single-nucleotide polymorphisms should be conducted to estimate the levels of self-pollination, consanguineous mating and outcrossing, and genome-wide associations with specific traits of ecological significance. Mating system studies combined with analyses of seed dispersal, germination and viability are needed to test some of the hypotheses resulting from the present study. Capitalizing on recent advances in conservation genomics techniques, including sequencing of whole genomes [91], many population genetic parameters such as genetic diversity, introgression and gene flow need to be studied at a higher precision for effective planning and implementation of conservation programs.

Acknowledgments

The authors thank Qianli Liu, CuijunYuan,WeihongXu, Cheng Zou and JunfangRen for fieldwork assistance and sample collection. We thank BaharulChoudhury for help with the analyses and Edith Pounden for editorial assistance.

Author Contributions

Conceived and designed the experiments: ZW JS MX. Performed the experiments: ZW FJ MD. Analyzed the data: ZW SD. Contributed reagents/materials/analysis tools: ZW JS MX SD. Wrote the paper: ZW SD.

References

  1. 1. Yang Y, Wang Z. Some properties of polyphenol oxidase from lily. Int J Food Sci Technol. 2008; 43: 102–107.
  2. 2. Munafo JP, Ramanathan A, Jiminez LS, Gianfagna TJ. Isolation and structural determination of steroidal glycosides from the bulbs of Easter lily (Lilium longiflorum Thunb.). J Agric Food Chem.2010;58: 8806–8813. pmid:20681669
  3. 3. Jin L, Zhang Y, Yan L, Guo Y, Niu L. Phenolic compounds and antioxidant activity of bulb extracts of six Lilium species native to China. Molecules. 2012;17: 9361–9378. pmid:22864243
  4. 4. Liu DJ, Huang YJ. Lilies. Beijing, China: Traditional Chinese Medicine Press;(2001.
  5. 5. Comber HF. A new classification of the genus Lilium. Lily Yearbook, Royal Horticultural Society.1949;13: 86–105.
  6. 6. Xu X-L, Ma K-M, Fu B-J, Song C-J, Liu W. Relationships between vegetation and soil and topography in a dry warm river valley, SW China. Catena.2008;75: 138–145.
  7. 7. Wu Z-H, Jiang Fuxing, Shi Jisen, Xi Mengli, Xu Weihong, Hu Fengrong, et al. Habitat and Genetic Diversity of Lilium regale. Journal of Northeast Forestry University.2007; 35: 40–43.
  8. 8. Spongberg S. A Reunion of Trees: The Discovery of Exotic Plants and Their Introduction into North American and European Landscapes. Cambridge, MA: Harvard University Press;1990.
  9. 9. Niimi Y, Nakano M, Maki K-I. Production of interspecific hybrids between Lilium regale and L. rubellum via ovule culture. Journal of the Japanese Society for Horticultural Science.1996;64: 919–925.
  10. 10. Fukai S, Tsuji K. Interspecific hybrids between Lilium × formolongi and some Asian trumpet species. Journal of the Japanese Society for Horticultural Science.2004; 73: 447–452.
  11. 11. Suzuki H, Tsuruhara A, Tezuka T. Regulations of the C2H4-forming system and the H2O2-scavenging system by heat treatment associated with self incompatibility in lily. Sexual Plant Reproduction.2001; 13: 201–208.
  12. 12. van Creij MGM, Kerckhoffs DMFJ, van Tuyl JM. Application of four pollination techniques and of hormone treatment for bypassing interspecific crossing barriers in Lilium L. Acta Horticulturae. 1998;508: 267–276.
  13. 13. Liu S, Fu B-J, Ma K, Liu G. Effects of vegetation types and landscape features on soil properties at the plateau in the upper reaches of Minjiang River. Yingyong Shengtai Xuebao.2004; 15: 26–30. pmid:15139181
  14. 14. Vrijenhoek RC. Genetic diversity and fitness in small populations. In: Loeschcke V, Tomiuk J, Jain SK, editors. Conserv Genet. Basel, Switzerland: Birkhäuser; 1994. pp. 37–53.
  15. 15. Torres-Díaz C, Ruiz E, González F, Fuentes G, Cavieres LA. Genetic diversity in Nothofagus alessandrii (Fagaceae), an endangered endemic tree species of the Coastal Maulino Forest of Central Chile. Ann Bot.2007; 100: 75–82. pmid:17513870
  16. 16. Szczecińska M, Sawicki J, Wąsowicz K, Hołdyński C. Genetic variation of the relict and endangered population of Chamaedaphne calyculata (Ericaceae) in Poland. Dendrobiology. 2009;62: 23–33.
  17. 17. Barrett SCH, Kohn JR. The genetic and evolutionary consequences of small population size in plants: Implications for conservation. In: Falk D, Holsinger KE, editors. Genetics and Conservation of Rare Plants. New York, NY: Oxford University Press; 1991.pp. 3–30.
  18. 18. Frankham R. Conservation genetics. Annu Rev Genet. 1995;29: 305–327. pmid:8825477
  19. 19. Hamrick JL, Godt MJ. Conservation genetics of endemic plant species. In: Avise JC, Hamrick JL, editors. Conservation Genetics: Case Histories From Nature. London: Chapman & Hall; 1996.pp. 281–304.
  20. 20. Hamrick JL. Isozymes and analyses of genetic structure of plant populations. In: Soltis DE, Soltis PS, editors. Isozymes in Plant Biology. Portland, OR: Dioscorides Press;1989. pp. 87–105.
  21. 21. Gibson JP, Rice SA, Stucke CM. Comparison of population genetic diversity between a rare, narrowly distributed species and a common, widespread species of Alnus (Betulaceae). Am J Bot. 2008;95: 588–596. pmid:21632385
  22. 22. Shao JW, Chen WL, Peng YQ, Zhu GP, Zhang XP. Genetic diversity within and among populations of the endangered and endemic species Primula merrilliana in China. Biochem Syst Ecol.2009; 37: 699–706.
  23. 23. Ayres DR, Ryan FJ. Genetic diversity and structure of the narrow endemic Wyethia reticulata and its congener W. bolanderi (Asteraceae) using RAPD and allozyme techniques. Am J Bot.1999;86: 344–353. pmid:10077497
  24. 24. Wolf AT, Howe RW, Hamrick JL. Genetic diversity and population structure of the serpentine endemic Calystegia collina (Convolvulaceae) in northern California. Am J Bot. 2000; 87: 1138–1146. pmid:10947998
  25. 25. Maki M, Horie S, Yokoyama J. Comparison of genetic diversity between narrowly endemic shrub Menziesia goyozanensis and its widespread congener M. pentandra (Ericaceae). Conserv Genet.2002;3: 421–425.
  26. 26. González-Astorga J, Castillo-Campos G. Genetic variability of the narrow endemic tree Antirhea aromatica Castillo-Campos & Lorence, (Rubiaceae, Guettardeae) in a tropical forest of Mexico. Ann Bot. 2004;93: 521–528. pmid:15056561
  27. 27. Hamrick JL, Godt MJ. Effects of life history traits on genetic diversity in plant species. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences.1996; 351: 1291–1298.
  28. 28. Persson HA, Lundquist K, Nybom H. RAPD analysis of genetic variation within and among populations of Turk's-cap lily (Lilium martagon L.). Hereditas. 1998;128: 213–220.
  29. 29. Wen CS, Hsaio JY. Altitudinal genetic differentiation and diversity of Taiwan lily (Lilium longiflorum var. formosanum; Liliaceae) using RAPD markers and morphological characters. Int J Plant Sci.2001;162: 287–295.
  30. 30. Huang YF, Yang MX, Zhang H, Zhuang XY, Wu XH, Xie W, et al. Genetic diversity and genetic structure analysis of the natural populations of Lilium brownii from Guangdong, China. Biochem Genet.2009; 47: 503–510. pmid:19543969
  31. 31. Arzate-Fernández AM, Miwa M, Shimada S, Yonekura T, Ogawa K. Genetic diversity of Miyamasukashi-yuri (Lilium maculatum Thunb. var. bukosanense), an endemic and endangered species at Mount Buko, Saitama, Japan. Plant Species Biol. 2005; 20: 57–65.
  32. 32. Horning ME, Maloney SC, Webster MS. Isolation and characterization of variable microsatellite loci in Lilium philadelphicum (Liliaceae). Mol Ecol Notes.2003;3: 412–413.
  33. 33. Kawase D, Hayashi K, Takeuchi Y, Yumoto T. Population genetic structure of Lilium japonicum and serpentine plant L. japonicum var. abeanum by using developed microsatellite markers. Plant Biosystems. 2010;144: 29–37.
  34. 34. Souframanien J, Gopalakrishna T. A comparative analysis of genetic diversity in blackgram genotypes using RAPD and ISSR markers. Theor Appl Genet.2004;109: 1687–1693. pmid:15368042
  35. 35. Prasad ASH, Neeraja CN, Sarla N, Siddiq EA. Comparative evaluation of SSR and ISSR markers for polymorphism in landraces and elite varieties of rice (Oryza sativa L.). Journal of Plant Biochemistry and Biotechnology.2005; 14: 49–51.
  36. 36. Bublyk OM, Andreev IO, Kalendar RN, Spiridonova KV, Kunakh VA. Efficiency of different PCR-based marker systems for assessment of Iris pumila genetic diversity. Biologia.2013; 68: 613–620.
  37. 37. Luan S, Chiang TY, Gong X. High genetic diversity vs. low genetic differentiation in Nouelia insignis (Asteraceae), a narrowly distributed and endemic species in China, revealed by ISSR fingerprinting. Ann Bot. 2006;98: 583–589. pmid:16807255
  38. 38. Culley TM, Wolfe AD. Population genetic structure of the cleistogamous plant species Viola pubescens Aiton (Violaceae), as indicated by allozyme and ISSR molecular markers. Heredity.2001;86: 545–556. pmid:11554971
  39. 39. Ding G, Zhang D, Yu Y, Zhao L, Zhang B. Analysis of genetic variability and population structure of the endemic medicinal Limonium sinense using molecular markers. Gene.2013; 520: 189–193. pmid:23506829
  40. 40. Zeng X, Michalski SG, Fischer M, Durka W. Species diversity and population density affect genetic structure and gene dispersal in a subtropical understory shrub. Journal of Plant Ecology.2012;5: 270–278.
  41. 41. Hedrén M. Species delimitation and the partitioning of genetic diversity–an example from the Carex flava complex (Cyperaceae). Biodiversity & Conservation. 2004;13: 293–316.
  42. 42. Marshall DR, Brown AHD. Optimum sampling strategies in genetic conservation. In: Frankel OH, Hawkes JG, editors. Crop Genetic Resources for Today and Tomorrow. London: Cambridge University Press;1975.pp. 53–80.
  43. 43. Armaleo D, Clerc P. A rapid and inexpensive method for the purification of DNA from lichens and their symbionts. The Lichenologist. 1995;27: 207–213.
  44. 44. Yamagishi M, Abe H, Nakano M, Nakatsuka A. PCR-based molecular markers in Asiatic hybrid lily. Scientia Horticulturae.2002; 96: 225–234.
  45. 45. Powell W, Morgante M, Andre C, Hanafey M, Vogel J, Tingey S, et al. The comparison of RFLP, RAPD, AFLP and SSR (microsatellite) markers for germplasm analysis. Mol Breed.1996;2: 225–238.
  46. 46. Shannon CE, Weaver W. The Mathematical Theory of Communication. Champaign, IL: University of Illinois Press; 1949.
  47. 47. Grativol C, da Fonesca Lira-Medeiros C, Hemerly AS, Ferreira PCG. High efficiency and reliability of inter-simple sequence repeats (ISSR) markers for evaluation of genetic diversity in Brazilian cultivated Jatropha curcas L. accessions. Mol Biol Rep.2011; 38: 4245–4256. pmid:21113672
  48. 48. Tatikonda L, Wani SP, Kannan S, Beerelli N, Sreedevi TK, Hoisington DA, et al. AFLP-based molecular characterization of an elite germplasm collection of Jatropha curcas L., a biofuel plant. Plant Sci. 2009; 176: 505–513.
  49. 49. Nei M. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics.1978; 89: 583–590. pmid:17248844
  50. 50. Slatkin M. Rare alleles as indicators of gene flow. Evolution.1985;39: 53–65.
  51. 51. Yeh FC, Yang R, Boyle T, Ye Z-H, Mao JX. POPGENE, the user friendly shareware for population genetic analysis. 1.31 ed. Edmonton, AB: Molecular Biology and Biotechnology Centre, University of Alberta; 1997.
  52. 52. Excoffier L, Lischer HE. Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. MolEcol Resources.2010;10: 564–567. pmid:21565059
  53. 53. Nei M. Genetic distance between populations. Am Nat.1972;106: 283–292.
  54. 54. Peakall ROD, Smouse PE. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol Ecol Notes.2006; 6: 288–295.
  55. 55. Holsinger KE, Lewis PO. HICKORY: A package for analysis of population genetic data V1. 0. Storrs, CT: Department of Ecology and Evolutionary Biology, University of Connecticut;(2003.
  56. 56. Tian H, Kang M, Liu Y, Ye Q, Yao X. High genetic diversity in remnant natural populations of Myricaria laxiflora, a species once considered to be extinct in the wild. Aquat Bot. 2012;103: 48–53.
  57. 57. Kothera L, Richards CM, Carney SE. Genetic diversity and structure in the rare Colorado endemic plant Physaria bellii Mulligan (Brassicaceae). Conserv Genet. 2007;8: 1043–1050.
  58. 58. Holsinger KE, Lewis PO, Dey DK. A Bayesian approach to inferring population structure from dominant markers. Mol Ecol. 2002; 11: 1157–1164. pmid:12074723
  59. 59. Tero N, Aspi J, Siikamäki P, Jäkäläniemi A, Tuomi J. Genetic structure and gene flow in a metapopulation of an endangered plant species, Silene tatarica. Mol Ecol. 2003;12: 2073–2085. pmid:12859630
  60. 60. Wright S. The genetical structure of populations. Annals of Eugenics. 1949;15: 323–354.
  61. 61. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155: 945–959. pmid:10835412
  62. 62. Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software STRUCTURE: A simulation study. Mol Ecol. 2005; 14: 2611–2620. pmid:15969739
  63. 63. Earl DA. STRUCTURE HARVESTER: A website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources. 2012;4: 359–361.
  64. 64. Mantel N. The detection of disease clustering and a generalized regression approach. Cancer Res.1967;27: 209–220. pmid:6018555
  65. 65. Bohonak AJ. IBD (isolation by distance): A program for analyses of isolation by distance. J Hered.2002;93: 153–154. pmid:12140277
  66. 66. Yuan S, Ge L, Liu C, Ming J. The development of EST-SSR markers in Lilium regale and their cross-amplification in related species. Euphytica. 2013; 189: 393–419.
  67. 67. Liu XL, Li WY, Jia GX. Cross compatibility of 11 species of wild lilies. Journal of Beijing Forestry University.2012; 34: 85–90.
  68. 68. Wang HX, Wang WF. Study on morphology diversity in Lilium regale Wilson. Journal of Anhui Agricultural Sciences. 2008;36: 6762–6763.
  69. 69. Savolainen O, Karhu A. Assessment of biodiversity with molecular tools in forest trees. In: Jain SM, Minocha SC, editors. Molecular Biology of Woody Plants. Dordecht, The Netherlands: Kluwer Academic;2000.
  70. 70. Cole CT. Genetic variation in rare and common plants. Annual Review of Ecology, Evolution, and Systematics. 2003;34: 213–237.
  71. 71. Hamrick JL, Godt MJ, Sherman-Broyles SL. Factors influencing levels of genetic diversity in woody plant species. In: Adams WT, Strauss SH, Copes DL, Griffin AR, editors. Population Genetics of Forest Trees. Dordecht, The Netherlands: Kluwer Academic;1992. pp. 95–124.
  72. 72. Spigler RB, Hamrick JL, Chang SM. Increased inbreeding but not homozygosity in small populations of Sabatia angularis (Gentianaceae). Plant Syst Evol.2010; 284: 131–140.
  73. 73. Reed DH, Briscoe DA, Frankham R. Inbreeding and extinction: the effect of environmental stress and lineage. Conserv Genet. 2002;3: 301–307.
  74. 74. Marchelli P, Baier C, Mengel C, Ziegenhagen B, Gallo LA. Biogeographic history of the threatened species Araucaria araucana (Molina) K. Koch and implications for conservation: a case study with organelle DNA markers. Conserv Genet. 2010;11: 951–963.
  75. 75. Zhang YP, Duan ZX, Dou JX. Comparison of climate characteristics between a dry-warm valley in upper reaches of Min River and a dry-hot valley of Yuanjiang River. Resources and Environment in the Yangtze Basin. 2005;14: 76–82.
  76. 76. Neel MC, Ellstrand NC. Patterns of allozyme diversity in the threatened plant Erigeron parishii (Asteraceae). Am J Bot.2001;88: 810–818. pmid:11353706
  77. 77. Delmotte F, Leterme N, Gauthier JP, Rispe C, Simon JC. Genetic architecture of sexual and asexual populations of the aphid Rhopalosiphum padi based on allozyme and microsatellite markers. Mol Ecol.2002;11: 711–723. pmid:11972759
  78. 78. Halkett F, Simon JC, Balloux F. Tackling the population genetics of clonal and partially clonal organisms. Trends Ecol Evol.2005; 20: 194–201. pmid:16701368
  79. 79. Ricardo CT, Corrado CA, Mandujano MC, Molina-Freaner F. Reproductive consequences of clonal growth in Stenocereus eruca, a rare clonal cactus of the Sonoran desert. Evol Ecol. 2006; 20: 131–142.
  80. 80. Honnay O, Jacquemyn H. A meta-analysis of the relation between mating system, growth form and genotypic diversity in clonal plant species. Evol Ecol. 2008;22: 299–312.
  81. 81. Gaudeul M, Stenøien HK, Ågren J. Landscape structure, clonal propagation, and genetic diversity in Scandinavian populations of Arabidopsis lyrata (Brassicaceae). Am J Bot.2007; 94: 1146–1155. pmid:21636482
  82. 82. Gonzales E, Hamrick JL, Smouse PE. Comparison of clonal diversity in mountain and piedmont populations of Trillium cuneatum (Melanthiaceae-Trilliaceae), a forest understory species. Am J Bot. 2008;95: 1254–1261. pmid:21632330
  83. 83. Watson MA. Resource storage and the expression of clonal plant life histories. Evol Ecol. 2008;22: 471–475.
  84. 84. Thompson FL, Eckert CG. Trade-offs between sexual and clonal reproduction in an aquatic plant: experimental manipulations vs. phenotypic correlations. J Evol Biol. 2004;17: 581–592. pmid:15149401
  85. 85. Lachapelle J, Bell G. Evolutionary rescue of sexual and asexual populations in a deteriorating environment. Evolution.2012; 66: 3508–3518. pmid:23106714
  86. 86. Gardner SN, Mangel M. Modeling investments in seeds, clonal offspring, and translocation in a clonal plant. Ecology.1999; 80: 1202–1220.
  87. 87. van Kleunen M, Fischer M. Adaptive evolution of plastic foraging responses in a clonal plant. Ecology.2001; 82: 3309–3319.
  88. 88. Jacquemyn H, Brys R, Honnay O, Hermy M, Roldán-Ruiz I. Local forest environment largely affects below-ground growth, clonal diversity and fine-scale spatial genetic structure in the temperate deciduous forest herb Paris quadrifolia. Mol Ecol. 2005;14: 4479–4488. pmid:16313608
  89. 89. Zhao Y, Wang C, Wang S, Tibig LV. Impacts of present and future climate variability on agriculture and forestry in the humid and sub-humid tropics. Clim Change. 2005; 70: 73–116. pmid:15870023
  90. 90. Mattner J, Zawko G, Rossetto M, Krauss SL, Dixon KW, Sivasithamparam K, et al. Conservation genetics and implications for restoration of Hemigenia exilis (Lamiaceae), a serpentine endemic from Western Australia. Biol Conserv. 2002;107: 37–45.
  91. 91. Primmer CR. From conservation genetics to conservation genomics. Ann N Y Acad Sci. 2009;1162: 357–368. pmid:19432656