Plant Systematics and Evolution

, Volume 298, Issue 8, pp 1483–1491

Genetic diversity assessment and ex situ conservation strategy of the endangered Dendrobium officinale (Orchidaceae) using new trinucleotide microsatellite markers

Authors

  • Beiwei Hou
    • College of Life SciencesNanjing Normal University
  • Min Tian
    • College of Life SciencesNanjing Normal University
  • Jing Luo
    • College of Life SciencesNanjing Normal University
  • Yuan Ji
    • College of Life SciencesNanjing Normal University
  • Qingyun Xue
    • College of Life SciencesNanjing Normal University
    • College of Life SciencesNanjing Normal University
Original Article

DOI: 10.1007/s00606-012-0651-3

Cite this article as:
Hou, B., Tian, M., Luo, J. et al. Plant Syst Evol (2012) 298: 1483. doi:10.1007/s00606-012-0651-3

Abstract

Dendrobium officinale (Orchidaceae) is an endangered plant species with important medicinal value. To evaluate the effectiveness of ex situ collection of D. officinale genetic diversity, we developed 15 polymorphic trinucleotide microsatellite loci of D. officinale to examine the genetic diversity and structure of three D. officinale germplasm collections comprising 120 individuals from its germplasm collection base and their respective wild populations consisting of 62 individuals from three provinces in China. The three germplasm collections showed reductions in gene diversity and average number of alleles per locus, but an increase in average number of rare alleles (frequency ≤ 0.05) per locus in comparison to their wild populations. However, the differences in gene diversity between the germplasm collections and wild populations were not statistically significant. The analysis using STRUCTURE revealed evident differences in genetic composition between each germplasm collection and its wild population, probably because the D. officinale individuals with distinct genotypes in each wild population were unevenly selected for establishing its germplasm collection. For conservation management plans, we propose that D. officinale individuals with rare alleles need to be conserved with top priority, and those individuals with the most common alleles also should be concerned. The 15 new microsatellite loci may be used as a powerful tool for further evaluation and conservation of the genetic diversity of D. officinale germplasm resources.

Keywords

Dendrobium officinaleGenetic diversityTrinucleotide microsatelliteEx situ collection

Introduction

Traditional Chinese Medicine (TCM) is an essential part of the health care system in several Asian countries, also becoming increasingly popular worldwide (Jiang et al. 2010). A famous TCM called “Tiepi Fengdou” is made of the stems of Dendrobiumofficinale Kimura et Migo, which is ranked “first of the nine kinds of supernatural Chinese medicinal herbs” (Wang and Xie 2004) and listed in The Chinese Pharmacopoeia (The State Pharmacopoeia Commission of P. R. China 2010). Dendrobium officinale has excellent medicinal merits, such as nourishing the kidneys, clearing away "heat-evil," benefiting the stomach, promoting the production of body fluids, resisting cancer, and prolonging life. Dendrobium officinale is commonly adnascent on cliffs or tree trunks covered with humus and moss at altitudes between 800 and 1,500 m, indicating that D. officinale has special requirements for its habitats, so it is susceptible to habitat deterioration. The marvelous therapeutic effects of D. officinale bring a great market demand for it, thus leading to the over-exploitation of its germplasm resources. Consequently, D. officinale has become an endangered species.

For protecting the germplasm resources of D. officinale with maximum diversity as well as maintaining its sustainable industrial development, seed breeding and tissue culture technologies have been rapidly developed and applied. In some provinces in China, such as Zhejiang and Yunnan Provinces, D. officinale germplasm collection bases have been established (Wu and Si 2010) to develop ex situ germplasm collections, construct a seed bank, and promote Good Agricultural Practice (GAP) for this plant.

According to a recent study by Xu et al. (2008), D. officinale germplasm from Yunnan Province and the Yandang area of Zhejiang Province gives rise to plants with moderate plant size, robust stems, and a relatively high polysaccharide content. D. officinale germplasm from Fujian Province and the Yandang area of Zhejiang Province displays the same plant morphology. Based on our investigations on most of the D. officinale germplasm collection bases in China, the D. officinale germplasm used for large-scale production of “Tiepi Fengdou” originated from Fujian Province, Yunnan Province, and the Yandang area of Zhejiang Province.

Amplified fragment length polymorphism (AFLP), sequence-related amplified polymorphism (SRAP), inter-simple sequence repeats (ISSR), and random amplified polymorphic DNA (RAPD) markers have been applied to studies of the genetic diversity and conservation of D. officinale (Ding et al. 2008, 2009; Li et al. 2008). In addition, microsatellites, which are also referred to as simple sequence repeats (SSRs), have been employed to assess the genetic diversity and structure of D. officinale (unpublished data). Recently, microsatellite markers have also been used to evaluate the effectiveness of in situ and ex situ conservation of the germplasm of some plants, such as Cynoglossum officinale, bean (Phaseolus vulgaris) and Jala maize (Zea mays) (Negri and Tiranti 2010; Gomez et al. 2005; Rice et al. 2006; Ensslin et al. 2011). Critical evaluation of living conservation collections with population genetic analysis can directly inform the ex situ conservation strategy (Namoff et al. 2010). The sustainable conservation of germplasm resources provide the ideal basis for its commercial use. However, the effectiveness of ex situ collection of D. officinale in a germplasm collection base has not been documented.

In this study, with magnetic beads containing trinucleotide repeats (TNRs) (AAG)10, 15 polymorphic microsatellite loci of D. officinale were isolated from its microsatellite-enriched genomic library. These 15 microsatellite loci were utilized to analyze the genetic diversity and structure of three D. officinale germplasm collections established in a D. officinale germplasm collection base and their original wild populations from Zhejiang Province, Fujian Province, and Yunnan Province, China. The aims of this study were to calculate allele frequencies and assess the genetic diversity within and among these germplasm collections and wild populations, to determine the genetic relationships among the three germplasm collections and their wild populations for assessing the effectiveness of ex situ collection of D. officinale diversity, and to infer a strategy for integrating the conservation of its genetic diversity with the maintenance of its sustainable industrial development.

Materials and methods

Plant materials and total DNA extraction

A total of 62 wild D. officinale individuals were collected between 2006 and 2009 in Yandang in Zhejiang Province, Wuyi and Shunchang in Fujian Province, and Guangnan and Wenshan in Yunnan Province, representing the three wild populations designated ZJ-W (20), FJ-W (21), and YN-W (21), respectively (Table 2). In the D. officinale germplasm collection base located in Yandang of Zhejiang Province, 120 individuals were randomly sampled in 40 individuals per collection from the three established D. officinale germplasm collections designated ZJ-G, FJ-G, and YN-G, originating from ZJ-W, FJ-W, and YN-W, respectively (Table 2). These collections had been cultivated in the above germplasm collection base for 5 years from 2004 to 2009 and used for large-scale production of “Tiepi Fengdou.” New D. officinale germplasm was also collected in the germplasm collection base every year. The genomic DNA of each sample was extracted from its dry or fresh leaves using the standard CTAB (cetyltrimethyl ammonium bromide) method (Doyle and Doyle 1987) with a minor modification: the lysis buffer was added with polyvinylpyrrolidone-40 at a final concentration of 2 % to eliminate polyphenols.

Microsatellite isolation and primer design

A microsatellite-enriched genomic library of D. officinale was constructed using a sample of the wild population YN-W according to the protocols as previously described (Xie et al. 2010; Fan et al. 2009). The biotinylated oligonucleotide probe (AAG)10 was used to isolate microsatellites from the enriched library; the isolated microsatellites were sequenced at the Beijing Genomics Institute (Beijing, China). By means of the primer analysis software OLIGO 7 (Rychlik 2007), primer pairs were designed on the basis of the upstream and downstream flanking sequences of the microsatellite repeat motifs.

PCR amplification of microsatellite loci

Polymerase chain reaction (PCR) amplification was carried out in 10 μl of PCR reaction mixture containing approximately 40 ng template DNA, 1× PCR buffer containing 50 mM KCl and 10 mM Tris-HCl (TaKaRa), 2 mM MgCl2, 200 μM dNTP (TaKaRa), 0.2 μM of each primer, and 0.5 U of Taq DNA polymerase (TaKaRa). PCR amplification was performed in a Peltier Thermal Cycler PTC-200 (BIO-RAD) according to the following program: initial denaturation at 94 °C for 5 min; subsequent 30 cycles of denaturation at 94 °C for 1 min; annealing at the optimal temperature (see Table 1) for 1 min; extension at 72 °C for 2 min; a final extension at 72 °C for 10 min. Subsequently, amplification products were subjected to 8 % polyacrylamide gel electrophoresis (PAGE) with a standard size marker loaded on the first lane of the gel at 180 V for about 2.0 h after pre-electrophoresis for 20 min. Then, the gel was stained according to the silver staining method (Xu et al. 2002).
Table 1

Genetic characteristics and amplification information of 15 trinucleotide microsatellite loci of Dendrobium officinale

Locus

Forward primer sequence (5′–3′)

Reverse primer sequence (5′–3′)

Repeat motif

TA (°C)

Expected size (bp)

Size range (bp)

AT

HO

HE

PIC value

BWH001

ATCAATGGCAACGAAGAT

TCCATTCACGCAATACAG

(CTT)10

50

268

253–295

7

0.786

0.801

0.771

BWH002

GGCAGAACTCTAGCGTCAC

TATCATCCGTCCATCCGAC

(CTT)10

52

281

269–302

6

0.476

0.609

0.692

BWH003

TTTGGGTGGTATTGGTTG

AGTGGAGGTATCGCTTGG

(CTT)12

52

137

119–161

7

0.614

0.723

0.734

BWH004

AGCTACCTCTCGTAGGCGTT

CTTGGACACCACTTGTCAG

(CTT)11

58

158

149–169

3

0.159

0.195

0.183

BWH005

CTTACAGCACGGAGGACA

CCAACTCGTTGACGGAGAT

(CTT)11

54

384

366–411

8

0.628

0.8462*

0.858

BWH006

CTGTCCTCCTGACCACCT

TCTGATAAGCGAGAAACCTA

(CTT)8

51

114

108–123

4

0.356

0.389

0.366

BWH007

GCACTCTTTCGCCATCCG

AAGCCAACTTTCAGACAATG

(CTT)13

58

235

220–259

7

0.476

0.759

0.73

BWH008

TCGCTGTCACATAGGAGTTG

CCATGCCATTCCATAAATCTA

(CTT)12

52

214

196–232

9

0.752

0.853

0.794

BWH009

AATCGAAGAGCGAGGTAA

ACAGAATAGAGTGGAATAAAC

(CTT)8

50

132

126–153

5

0.595

0.638

0.576

BWH010

TGACTGAGGTGCAGAGGTTT

GATGATGAGTATGAAGAGCC

(CTT)9

55

148

136–160

4

0.386

0.429

0.368

BWH011

TTCTCGGAAGTTTACTGTT

AAGAGTAACGGTATGGAG

(CTT)10

50

247

235–280

10

0.881

0.877

0.865

BWH012

CCCTCTTGCGATTCTTCT

AATGGTGGGAAACTTGGA

(CTT)11

52

201

186–225

9

0.571

0.812

0.789

BWH013

CAAGGAAGGTGGGCTGTC

AGATGTTGGAACGGAGGG

(CTT)8

55

333

327–354

5

0.535

0.641

0.634

BWH014

TTCTCCTCCTCTGCTCAC

CACCCATACATTCAACAACT

(CTT)8

50

360

354–375

6

0.456

0.623

0.531

BWH015

TTGGGTGGTATTGGTTGC

TGGAGGTATCGCTTGGAG

(CTT)12

54

136

121–157

7

0.762

0.8421*

0.822

TA annealing temperature, AT total number of alleles, HO observed heterozygosity, HE expected heterozygosity, PIC polymorphism information content

* Significant Hardy–Weinberg disequilibrium

Statistical analysis

The expected heterozygosity, observed heterozygosity, and polymorphic information content (PIC) of each tested locus in a population were calculated using CERVUS version 3.0.3 (Kalinowski et al. 2007). HP-Rare 1.1 (Kalinowski 2005) was utilized to calculate average numbers of alleles and average allelic richness. Allele frequencies at each tested locus were calculated according to the number of bands resulting from the PAGE.

The DetSel (Vitalis et al. 2001) test was used to detect outlier loci in spatial comparisons between wild population and the germplasm collection. Outlier loci could be subject to selective effects and play important roles in changing patterns of genetic variation and differentiation in regeneration cycles (Negri and Tiranti 2010).

Arlequin 3.1 (Excoffier et al. 2005) was used to calculate the observed heterozygosity and gene diversity of each wild population and germplasm collection. The GENEPOP 3.4 package (Raymond and Rousset 1995) was employed to perform the exact test for Hardy-Weinberg equilibrium by microsatellite loci (test multi-population) and by wild populations or germplasm collections (test multi-locus) using the Markov chain method with 1,000 iterations and considering the heterozygote deficit as the alternative hypothesis.

With Arlequin 3.1 (Excoffier et al. 2005), FST values of pairwise comparisons were calculated for all pairs of sampling localities to investigate the differentiation of wild populations and germplasm collections. The model-based program STRUCTURE (Falush et al. 2003) was used to determine the genetic structure of three germplasm collections and their original wild populations by performing a simulation analysis involving all of the 182 D. officinale samples. The methods Evanno’s ΔK and ad hoc ln Pr(X|K) were employed to determine a reasonable range of K values. The parameter ln Pr(X|K) was calculated for K = 1–6 with three independent runs for each K, using the following settings: MCMC iteration number = 1,200,000; Burn-in = 500,000.

Results

Development and characterization of microsatellite loci

With the biotinylated oligonucleotide probe (AAG)10, 92 positive colonies containing microsatellite loci were isolated from the constructed microsatellite-enriched genomic library of D. officinale; 28 of them harbored the sequences with length and quality appropriate for primer design. Accordingly, 28 primer pairs were designed; 15 of them resulted in clear amplification products—15 SSR loci with non-ambiguous genotypes and evident polymorphism, which were designated as BWH001 to BWH015 with GenBank accession nos. HQ905560-HQ905574, respectively (Table 1).

All of the 62 individuals from the three wild populations ZJ-W, FJ-W, and YN-W were genotyped at the 15 SSR loci, whose genetic characteristics are displayed in Table 1. For each of the 15 loci, the number of alleles ranged from 3 to 10 with a mean number of 6.13; the observed heterozygosity varied from 0.159 to 0.881 with a mean value of 0.553; and the expected heterozygosity ranged from 0.195 to 0.877 with a mean value of 0.662. Among the 15 loci, only two (BWH005 and BWH015) revealed significant deviations from Hardy-Weinberg equilibrium. In addition, the 15 loci exhibited higher PIC values ranging from 0.183 to 0.865 with a mean value of 0.641. These loci were subsequently utilized as molecular markers to assess the genetic diversity and genetic structure of the three D. officinale germplasm collections and their wild populations.

The DetSel spatial pairwise comparisons between wild population and germplasm collection showed that locus BWH004 was putatively under selective effects in all comparisons. Other loci were detected as outliers by DetSel tests, but, since they only occurred in one or two pairwise comparison, they were not considered to be reliable outliers (Negri and Tiranti 2010). Locus BWH004 as an outlier SSR may be considered a candidate for genes under selection, and 14 other SSR loci were all not under selective effect.

Genetic diversity among and within germplasm collections

The three germplasm collections ZJ-G, FJ-G, and YN-G differed in several genetic parameters, including allelic richness and gene diversity (Table 2). Among them, the germplasm collection YN-G showed the highest levels of allelic diversity (6.27), allelic richness (5.87), observed heterozygosity (0.56055), and gene diversity (0.65708). The average gene diversity of the three germplasm collections was 0.637, indicating that these collections had a high level of genetic diversity. Moreover, none of the three germplasm collections revealed significant deviations from Hardy-Weinberg equilibrium. Based on the previously established criteria (Wright 1978), a low level of genetic differentiation existed among the three germplasm collections, as indicated by the corresponding FST value of 0.0959 (data not shown).
Table 2

Genetic variation in three wild populations and three germplasm collections of Dendrobium officinale at 15 microsatellite loci

Code

Sampling locality

N

AT

AT ≤ 0.05

AD

AR

HO

HE

ZJ-W

Yandang of Zhejiang province

20

85

10

5.67

5.07

0.53206

0.68185

FJ-W

Wuyi and Shunchang of Fujian province

21

94

15

6.27

5.36

0.46821

0.72457

YN-W

Guangnan and Wenshan of Yunnan province

21

95

6

6.33

5.72

0.54113

0.71775

ZJ-G

Germplasm collection from ZJ-W

40

80

17

5.33

4.98

0.51976

0.60987

FJ-G

Germplasm collection from FJ-W

40

91

23

6.07

5.66

0.4743

0.64326

YN-G

Germplasm collection from YN-W

40

94

18

6.27

5.87

0.56055

0.65708

The number of common alleles (AT > 0.05) can be calculated by subtracting the value of AT ≤ 0.05 from the value of AT

N the number of individuals, AT the total number of alleles, AT ≤ 0.05 the number of rare alleles, AD allelic diversity, AR average allelic richness, HO observed heterozygosity, and HE gene diversity

Genetic diversity between germplasm collections and wild populations

Among the three D. officinale wild populations, the population YN-W displayed the highest values of allelic diversity (6.33) and richness (5.72), whereas the population ZJ-W demonstrated the lowest values in these two parameters (Table 2). The three germplasm collections differed in the level of allelic diversity in the same order: YN-G > FJ-G > ZJ-G (Table 2). However, all three wild populations had higher allelic diversity than their respective germplasm collections (Table 2), indicating that some alleles of the wild populations were lost in their respective germplasm collections. As shown in Table 2, five, three, and one allele(s) present in the wild populations ZJ-W, FJ-W, and YN-W, respectively, were absent in their respective germplasm collections. However, the germplasm collections FJ-G and YN-G had higher allelic richness than their respective wild populations, whereas the germplasm collection ZJ-G showed lower allelic richness than its wild population (Table 2).

According to calculated allele frequencies, the numbers of rare alleles (frequency < 0.05) within wild populations ZJ-W, FJ-W, and YN-W were 10 (11.8 %), 15 (16.0 %), and 6 (6.3 %), respectively; the counterparts of their respective germplasm collections were 17 (21.3 %), 23 (25.3 %), and 18 (19.1 %), respectively, of which the proportions were calculated without considering the lost alleles in these collections (Table 2). These results indicated that the germplasm collections ZJ-G, FJ-G, and YN-G had more rare alleles than their respective wild populations. However, the germplasm collections ZJ-G, FJ-G, and YN-G contained fewer common alleles (frequency > 0.05) than their respective wild populations (Table 2).

The gene diversity of the three wild populations ranged from 0.68185 to 0.72457, with a mean value of 0.708, which was higher than the counterpart of the corresponding germplasm collections (Table 2). However, statistically significant differences in the observed heterozygosity and gene diversity were not observed between each germplasm collection and its wild population. For each germplasm collection and wild population, the observed heterozygosity was less than the expected heterozygosity, which is consistent with the common observation for microsatellites (Nybom 2004). In addition, none of the three wild populations showed significant deviations from Hardy-Weinberg equilibrium. According to the previously established criteria (Wright 1978), there was almost no genetic differentiation among the three wild populations as indicated by the corresponding FST value of 0.0283 (data not shown).

Genetic structure of germplasm collections and wild populations

The pairwise estimates of the genetic differentiation for all pairs of the analyzed D. officinale germplasm collections and wild populations were indicated by FST values (Table 3). The results were in accordance with those described above for the genetic differentiation among germplasm collections or wild populations alone. A statistically significant level of genetic differentiation was detected between every two germplasm collections, but neither between every two wild populations, nor between any germplasm collection and its original wild population. However, the genetic differentiation between the germplasm collection ZJ-G and a wild population from a different province (either FJ-W or YN-W) was statistically significant, as was that between the germplasm collection YN-G and the wild population ZJ-W.
Table 3

Pairwise FST values for Dendrobium officinale wild populations and germplasm collections

 

ZJ-W

FJ-W

YN-W

ZJ-G

FJ-G

ZJ-W

     

FJ-W

0.0471

    

YN-W

0.0356

0.0028

   

ZJ-G

0.0018

0.1192*

0.0985*

  

FJ-G

0.0405

0.0008

0.0329

0.0986*

 

YN-G

0.0675*

0.0280

−0.0125

0.1287*

0.0603*

* Significant values based on permutation tests following a sequential Bonferroni correction

With STRUCTURE, the genetic structure of all tested germplasm collections and wild populations was determined via a simulation analysis. The results indicated that the genetic structure of three germplasm collections exhibited a higher level of differentiation than that of three wild populations (Fig. 1), which was consistent with the results of the analysis based on FST values. The value of LnP(D) gradually increased as the value of K was elevated from 1 to 6, but showed an evident knee at K = 2. Based on the values of LnP(D) and Evanno’s ΔK for K = 2, the genetic structure of each wild population or germplasm collection consisted of two clusters, the green and the red (Fig. 1). However, the differences in the genetic composition of the two clusters among three germplasm collections were more evident than those among three wild populations. Specifically, the genotypes of the germplasm collection ZJ-G were almost equally assigned to the two clusters, whereas those of the collections FJ-G and YN-G were mainly assigned to the green and the red cluster, respectively (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs00606-012-0651-3/MediaObjects/606_2012_651_Fig1_HTML.gif
Fig. 1

Genetic structure of three wild populations and three germplasm collections of Dendrobium officinale for K = 2 and K = 3 chain

The genetic structure of the three germplasm collections and three wild populations for K = 3 chain demonstrated marked differences in genetic composition between each germplasm collection and its original wild population (Fig. 1). According to the standard of assigning a genotype to a specific cluster with a probability >90 %, there were 15.0 and 40.0 % of the genotypes of the population ZJ-W and its collection ZJ-G, respectively, assigned to the green cluster; 9.5 and 35 % of those of the population FJ-W and its collection FJ-G, respectively, were assigned to the blue cluster; and 28.6 and 45 % of those of the population YN-W and its collection YN-G, respectively, were assigned to the red cluster (Fig. 1).

Discussion

Microsatellites comprising 1–6 bp long units are present abundantly and ubiquitously in all life forms, but their genomic distribution is non-random (Li et al. 2002). Much evidence has demonstrated that SSRs located in different positions of a gene play important roles in determining protein function, genetic development, and regulation of gene activity (Lawson and Zhang 2006; Li et al. 2004). In Arabidopsis thaliana and rice (Oryza sativa), SSRs in general were more favored in upstream regions of genes, but TNRs were the most common repeats found in the coding regions compared to other types of SSRs (Lawson and Zhang 2006). In addition, compared to dinucleotide microsatellites, trinucleotide microsatellites are easier to genotype and thus often the markers of choice for human genetic analysis (Gastier et al. 1995). Accordingly, triplet SSRs were developed in the present study, although dinucleotide repeats SSRs, such as CT and GT repeats, were also isolated from the constructed microsatellite-enriched genomic library of D. officinale. However, in a variety of plant species, the TNRs present in the coding regions show bias to some specific nucleotide motif subclasses. The subclass of AAG repeats is the most abundant TNRs in Arabidopsis, while CCG and AGG repeats are the most common TNRs in cereal species (Li et al. 2004). In this study, AAG repeats were chosen as enrichment probes, resulting in isolation of the SSRs consisting of CTT repeats from D. officinale species.

For the 15 polymorphic SSRs developed in this study, the number of alleles per locus ranged from 3 to 10, and the PIC value per locus varied from 0.183 to 0.865. These genetic parameters of the 15 markers are at similar levels to those of the SSR markers of D. officinale available in molecular ecology resource database 1 or previously reported (Xie et al. 2010). The development of these 15 new markers leads to an expansion of the small SSR marker bank of D. officinale. It is noticeable that all of the 15 new SSR markers were appropriate to analyze the genetic diversity of D. officinale in this study, so they may be used as a powerful tool for studying the population genetics and conservation genetics of D. officinale. At present, the SSR marker bank of D. officinale is still small. Therefore, it is necessary to isolate more trinucleotide microsatellite loci of D. officinale; in this aspect, the development of the microsatellite loci consisting of CCG and AGG repeats should take place prior to that comprising other subclasses of TNRs.

Dendrobium officinale has been listed as an endangered species by the Chinese government since 1987 (Fu 1992), and its current distributions are highly fragmented and discontinuous. Effective measures to protect the wild resources of D. officinale include its in situ management and ex situ conservation. With respect to the latter, a few D. officinale germplasm collection bases have been established in China for collecting its germplasm resources, constructing its seed bank, and growing D. officinale plants with different germplasms on a large scale for meeting the market demand. However, the ex situ conservation of D. officinale raises the serious question of whether the established germplasm collections can fully maintain the genetic diversity of their original wild populations. The reduction in gene diversity and the loss of low frequency alleles have been commonly reported in ex situ collections compared to in situ conservation (Gomez et al. 2005; Negri and Tiranti 2010; Rucińska and Puchalski 2011). In this research, compared to three studied wild populations of D. officinale, their respective germplasm collections declined in gene diversity and average number of alleles per locus, but increased in average number of rare alleles per locus (Table 2), indicating that some modifications to the genetic composition of the three wild populations occurred in their respective germplasm collections. Consistently, the analysis using STRUCTURE showed evident differences in genetic composition between each germplasm collection and its wild population (Fig. 1), which was probably attributable to the fact that initially the D. officinale individuals with distinct genotypes in each wild population were unevenly selected for establishing the germplasm collection. however, unconscious selection in garden populations also led to genetic changes in evidence (Ensslin et al. 2011). Hence, for improving the effectiveness of ex situ collection of D. officinale diversity, it is crucial to evenly collect the individuals with distinct genetic makeups from their wild populations for transplantation into and multiplication in a D. officinale germplasm collection base to develop the germplasm collections.

For all rare plant species, Farnsworth et al. (2006) suggested three simple decision matrices for prioritizing collection of rare plant species for ex situ conservation. For one species, knowledge of the genetic diversity and structure of D. officinale is essential for the management and conservation of its germplasm resources. Rare alleles are important in conservation biology, and collection designs oriented to sampling rare alleles provide the management of genetic diversity with adequate tools with which to reinforce declining populations or aid the survival of reintroduced plants (Caujapé-Castells and Pedrola-Monfort 2004). To avoid a loss of rare alleles, allele fixation, and an increase in homozygosity, conservation of D. officinale individuals with rare alleles should take place prior to that for common alleles. For integrating conservation of the genetic diversity of D. officinale with maintenance of its sustainable industrial development, we propose that the D. officinale individuals with rare alleles need to be preserved with top priority. Those individuals with the most common alleles should also be conserved at the same time. Thus, it is increasingly important to develop coherent, systematic strategies for targeting plant populations to maximize the capture of genetic diversity and potentially adaptive alleles (Farnsworth et al. 2006). In this study, the observed differences in gene diversity among the three D. officinale germplasm collections and their wild populations did yet not reach a statistically significant level, so it is high time to conserve its germplasm resources in established germplasm collection bases. However, it is also urgent to implement in situ management of the remaining wild resources of D. officinale by detecting the genetic diversity and structure of all wild populations.

In conclusion, we developed 15 new trinucleotide microsatellite loci of D. officinale with a high rate of polymorphism to examine the genetic diversity and structure of three D. officinale germplasm collections and their wild populations. Compared with their respective wild populations, three germplasm collections showed reductions in the gene diversity and average number of alleles per locus, but an increase in the average number of rare alleles per locus. The analysis using STRUCTURE also displayed marked differences in genetic composition between each D. officinale germplasm collection and its wild population. The above differences between the germplasm collections and wild populations probably occurred because, initially, the D. officinale individuals with distinct genotypes in each wild population were unevenly selected for establishing the germplasm collection. Therefore, even sampling of the D. officinale individuals with distinct genotypes from a wild population for transplantation into a D. officinale germplasm collection base is essential for improving the effectiveness of ex situ collection of D. officinale genetic diversity. For conservation management plans, we should consider how genetic information can inform conservation management. For conservation of other D. officinale germplasm collections, the status of genetic diversity should also be evaluated in the germplasm collection base. We provided a basic guideline in which the D. officinale individuals with rare alleles deserve to be protected with first priority, and those individuals with the most common alleles should also be concerned. The 15 trinucleotide microsatellites developed in the present study can be used as an important tool for further assessing and conserving the genetic diversity of D. officinale germplasm resources.

Acknowledgments

We are grateful to Shu-Miaw Chaw for many helpful comments on previous drafts. This work was financed by the National Natural Science Foundation of China (no. 30870234) and the Natural Science Foundation of Jiangsu Province (no. BK2008431).

Copyright information

© Springer-Verlag 2012