刘焕 欧阳天林 田承清
（江西省林业科技实验中心，江西 信丰 341600）
【摘要】随着《中国生物多样性保护战略和行动指南（2010-2030）》的贯彻实施，生物多样性监测与评价工作将在全国范围陆续开展。进化生态学作为阐述生物多样性演化规律和机理的基础性学科，其数量研究方法在20世纪70年代后得到了迅速的发展。本文从三个层面系统性总结、筛选了进化生态学在植物生态领域的主流研究方法，其中在生态系统层面，群落演替的主成分分析和聚类分析方法、群落的可恢复性、可持续性、变异性、抗干扰性、边缘效应等主题被筛选为主要分析方法；在种群层面，种间关联指数、相关系数、分离指数、生态位宽度指数、生态位重叠指数等概念可以全面阐释植物种群的演替规律；在遗传层面，哈迪-温伯格平衡度的检测、等位基因频率、多态位点百分数、平均位点的等位基因数、平均位点的预期杂合度、Nei氏遗传分化系数、Nei氏遗传一致度、遗传距离、聚类分析、遗传贡献率等方法在分子进化分析中的应用相对广泛。
教育期刊网 http://www.jyqkw.com
关键词 生物多样性评价；生物多样性监测；进化生态学
Review of Evolutionary Ecology Study and Its Application on Biodiversity Monitoring and Assessment
LIU Huan OUYANG Tianlin TIAN Cheng-qing
（Jiangxi Provincial Forest Science and Technology Experiment Cente， Xinfeng Jiangxi 341600，China）
【Abstract】After Chinese Biodiversity Conservation Strategy and Action Planning (2010-2030) is implemented in China, biodiversity monitoring and assessment projects are increasing steadily in national wide. The statistical methods of evolutionary ecology study have been developed quickly since 1970s, which provides the theory underlying the interpretation of biodiversity evolution in ecosystem. This article summarizes the evolutionary ecology methods which have been relatively broadly applied on botanical species from three layers: for ecosystem diversity, the principle component index (PCI) and cluster analysis for community succession analysis, ecosystem resilience, sustainability, variance, resistance capacity and edge effects are identified as the main analysis methods; for species diversity, the conceptions of inter-specific association, rank correlation coefficient, segregation index, coefficient of niche breadth and coefficient of niche overlap can fully interpret the succession of plant populations in ecosystem; for genetic diversity, the methods including Hardy-Weinberg equilibrium, allele frequency, percentage of polymorphic loci, mean number of alleles per locus, mean expected heterozygosity per locus, Nei’ coefficient of gene differentiation, Nei’ genetic identity, genetic distance and cluster analysis, genetic contribution rate have been identified as main methods for analysis of molecular evolution.
【Key words】Biodiversity assessment；Biodiversity monitoring；Evolutionary ecology
0 Introduction
According to the Chinese Biodiversity Conservation Strategy and Action Planning (2010-2030), there are three thorny issues threatening biodiversity conservation in national wide: degradation of ecosystem function in some area; deterioration of endangered species; continuous loss of genetic resources. The methods of evolutionary ecology study from three layers (ecosystem, species, genetics) provides substantial theory explaining these threats so that conservation strategies can be worked out properly.
After Environmental Standard for the Assessment of Regional Biodiversity (HJ623-2011) is implemented in China, multivariate methods of evolutionary ecology study become essential to classify the basic units for biodiversity assessment at both ecosystem layer (classification of communities) and genetic layer (classification of sub-populations).
After Environmental Standard on Classifying the Categories of Genetic Resources (HJ 626-2011) comes into force in China, the methods of evolutionary ecology provide the theoretical basis not only for understanding the evolutionary process of endangered species, but also becomes compulsory for ranking genetic resources (or endangered species) between CategoryⅠand categoryⅡ.
This review article systematically summarizes the main themes of evolutionary ecology study of plant species from three layers, with discussion of selecting suitable methods for biodiversity monitoring and assessment work.
1 Ecosystem Diversity
1.1 Cluster Analysis and Principal Component Analysis (PCA)
According to the Technical Guideline for Ecological Assessment, the significance of dominant plant species is calculated by a combination of density, frequency and dominance, which becomes the basis of cluster analysis or PCA for community classification[1], which becomes the essential units for biodiversity assessment at ecosystem layer. Bu et al.,(2005) adopted both fuzzy cluster analysis and principal component analysis (PCA) methods to classify 13 sampling plots into 5 communities, which included 15 botanical species located in loess hilly region. Both methods led to similar conclusions in terms of community classification. According to the restoration duration required by each community, the temporal succession of 5 plant communities was identified as: Artemisia scoparia community-Leymus scalinus community-Stipa bungeana community-Artemisia gmelinii community-Hippophae rhamnoides community [2].
Anwar et al.,(2009) selected multivariate methods of cluster analysis and principal component analysis to understand corticolous lichen species composition and community structure characteristics in the forest ecosystem of Southern Mounffiins of Urumqi, China. There were thirty nine corticolous lichen species found, which were classified into 5 orders, 13 families and 26 genera. According to the multivariate analysis, three types of communities were classified, including community Lecanora hageni(Ach.)Ach. + Physcia stellaris(L.)Nyl. + L.saligna(Schrad.)Zahlbr; community Physcia aipolia(Humb.)Furm. + Ph.dimidiata(Arn.)Nyl + Cladonia pyxidata(L.)Hoffm; and community Xanthoria fallax (Hepp) Arnold + X.elegans(Link.)Th.Fr, whose structures were significantly influenced by altitude and tree type [3].
The composition and community structure of dominant species were analyzed by Cai et al., (2007) on the basis of multivariate methods of both principal component analysis and cluster analysis with the survey data of phytoplankton in spring, summer, autumn and winter from 1998 to 1999 in the West Guangdong Waters. According to the cluster analysis, phytoplankton species were classified into 2 communities in each season of spring, summer and autumn, with one inshore group and one offshore group, whereas the differentiation of species community was not significant in winter time. The seasonal succession of dominant species was Skeletonema costatum, Navicula subminuscula, Thalassionema nitzschioides, and Thalassiosira subtilis in spring, summer, autumn and winter respectively. However, the freshwater species, Oscillatoria sp. became the dominant species in summer as well [4].
Wang & Peng adopted both species similarity analysis (including coefficient of community, percentage of similarity and coefficient of similarity) and cluster analysis methods to classify plant communities and examine the environmental gradient effects on community succession in Dinghu Mountain, which indicated that Cryptocarya chinensis communities varied with different altitude gradient. Ten plant communities were compared and contrasted, revealing the mutual effects and evolutionary patterns among these communities [5].
1.2 Ecosystem Resilience
Ecological resilience is the capacity of disturbed ecosystem restored into its primitive conditions[6]. Zhang et al., (2013) assessed the ecosystem resilience quantitatively by using social-ecological system (SES) model in Northern Highlands of Yuzhong County, and resulted in the conclusion that the resilience of ecosystem was determined by both drought stress and ecosystem sensitivity to drought condition [7].
To order to assess community resilience and restoration success, Renaud et al., (2013) developed two indices including Community Structure Integrity Index measuring the proportion of species diversity for the reference community in comparison to the restored or degraded community, as well as the Higher Abundance Index assessing the proportion of the species abundance which was higher than the reference community. Three examples were illustrated for the application of two indices, including fictitious communities; A recently restored (2 years) Mediterranean temporary wetland (Camargue in France) for the assessment of restoration efficiency; and a recently disturbed pseudo-steppe plant community (La Crau area in France) assessing the natural community resilience, which demonstrated that these two indices were not only able to assess the static value of ecosystem function, but also to analyze the temporal and spatial dynamics of ecosystem evolution [8]. Nevertheless, compared with Zhang et al., (2013) model, social disturbance was not integrated into Renaud et al., (2013) model.
Additionally, 5 succession phases of the restoration of degraded ecosystem in Jinyun Mountain were investigated by Li et al.,(2007), including Shrubby grass land, Masson Pine early stage, Masson Pine late stage, Coniferous broad-leaved mixed forest and Evergreen broad --- leaved forest stage. Under the same climate conditions, criteria of species diversity, light absorption, community temperature, cumulate cover of arbor and community pole temperature became the main indicators for the succession of ecosystem restoration. However, among these indicators, both cumulate cover of arbor and community pole temperature were identified to be the best two indicators, and the other indicators were advised as the minor ones for consideration [9].
1.3 Ecosystem Sustainability
Ecosystem sustainability is the potential or manifested ability for ecosystem to perpetually sustain its interior composition, structure and function so that ecosystem is able to develop and evolve healthily [6]. Hu Dan (1997) presented methodology for assessment of ecosystem sustainability on the basis of identifying and evaluating ecosystem components, structure and function, which was consisted of 12 items and more than 30 variables, indicating the dynamics of sustainable ecosystem[10]. However, social factors were not considered in this methodology. In comparison, Yu et al., (2007) developed a quantitative index system for the assessment of eco-tourism sustainability in TianMuShan Natural Reserve, which included 25 criteria selected from three aspects: Environment, Society-Culture and Economics. On the basis of this method, a case study in Tianmushan Nature Reserve was introduced to demonstrate sustainability assessment in ecosystem [11].
1.4 Ecosystem Resistance
Ecosystem resistance is the ability of ecosystem to boycott the external disturbance and sustain its primitive conditions[6]. Hou et al., (2012) pointed out that the criteria of assessing eco-resistance were consisted of decomposition rate of ground combustibles, increase of ground combustibles, spontaneous combustion caused by lightning, indigenous pest, invasive pests and occurrence of pest[12]. However, quantitative method (such as the weight of each criterion) was not presented in this research. In comparison, Guo et al., (2012) presented the criteria for the assessment of eco-resistance which were consisted of the degree of pest invasion (or disease infection) and the fire incidence, with a weight of 0.6891 and 0.3109 respectively [13].
1.5 Ecosystem Variance
Ecosystem variance is divided into spatial heterogeneity and functional heterogeneity, which reflects the complex or variance of species distribution pattern and community structure influenced by available resources and environmental conditions [6]. Liu et al., (2010) adopted β Sorenson index to investigate the variability of plant communities of grass land in Ordos, Inner Mongolia of China, which was restored from grazing land. The relations between restoration duration and variability of plant communities was deduced in this research: compared with stabilized sand (25～30 a), higher variability existed in semi- mobile sand (restoration duration:5～10 a) and semi- stabilized sand (restoration duration:15～20 a). β Sorenson index for plant communities with dominant species Artemisia ordosica or Hedysarum laeve (restoration duration:5～20 a) was approximately 1.2, while the variability index of Artemisia ordosica (restoration duration: 30a) sand was twice than that of Hedysarum laeve (restoration duration: 30a), and faster growth rate was reported in Artemisia ordosica (restoration duration: 30a) sand [14].
Zhang et al.,(1988) analyzed the succession of pioneer meadow communities in abandoned farmland located in the high land of Gansu Province South. Heterogeneity index of H1 was deduced in this study, with value ranging from 0 to 1. Two meadow communities were investigated, with H1 heterogeneity indices of 0.11 and 0.15 respectively, which revealed relatively low heterogeneity between them[15].
1.6 Edge Effect
Edge effect typically exists in the ecotone between different plant communities, which is caused by the mutual interactions between different plant species from various communities, leading to characteristics in terms of species composition, configuration and function differed from the original communities [6]. Wang & Peng (1986) quantified the edge effects of plant communities in DingHuShan Nature Reserve by a model, with discussion of both positive and negative effects of community edges [16].
Eugenie et al., (2001) quantified the edge effects on plant communities caused by 6 recent clearcut edges adjacent to Pinus banksiana and Pinus resinosa plantations in the Great Lakes region. 10 sampling plots were randomly placed at 19 distances along a 240 transect which spanned from clearcut, across the edge, into the forest interior, with an estimation of percentage cover of each understory plant species. Species richness was significantly higher in Pinus banksiana lines than Pinus resinosa lines, with 18 and 2 unique species respectively. Species with clear preference for the clearcut, edge habitats or interior were respectively reflected by depth-of-edge influence, with composition gradient examined by the Detrended correspondence analysis (DCA) of distance sampled on the basis of species richness. Finally a synthesis model was designed to calculate the plant species distributions across forest/clearcut edges [17].
2 Species Diversity
2.1 Inter-specific Association, Rank Correlation Coefficient, Segregation Index
Inter-specific association is the mutual association between different species in terms of spatial distribution patterns in various habitats, which is divided into the competition relationship defined by segregation index (negative correlation), as well as interdependence relation calculated by rank correlation coefficient (positive correlation) [6].
On the basis of 25 sampling plots, 375 quadrats and 150 transect lines, Zhang et al., (2013) adopted eight indices of Diffusion Coefficient (C), Negative Binomial Parameters (K), Average Crowed Degree (m＊), Index of Clumping (I), Index of Patchiness (PI), Green index (GI), Cassie index (CA), Moristia index (Ｉδ ) and Variance of Percentages (VP) to analyze the spatial distribution patterns and overall correlation between dominant plant species in Gansu Donghuang xihu Desert Wetland ecosystems. The results revealed that significant positive correlation existed between dominant species populations in shrub layer and tree layer, whereas significant negative correlation was reported between dominant species in tree-shrub-grass layer and grass layer. Further more, the 2×2Contigency Table of Chi-square statistics, Association Coefficient (AC), Percentage of Co-occurence (PC) and other methods were conducted additionally to analyze the correlation significance and intensity between dominant species, leading to the results that correlation between dominant species was not significant in most cases and logarithm with significantly negative correlation was more than positive one, which indicated various requirements of habitat and resources for different species [18].
Yan et al., (2009) adopted Contingency Table and Spearman Rank coefficient to analyze the inter-specific association and inter-specific covariance between Artemisia annua and its associated plant species in the natural fostering base from 2006 to 2007. The results showed that flooding disturbance led to insignificant effects on inter-specific association, but significant effects on inter-specific covariance. However, flooding effect on inter-specific covariance varied between different species pairs, indicating that inter-specific covariance of paired species was depended on both environmental conditions and ecological characters, which became more sensitive to environmental disturbance than inter-specific association [19].
Wang et al., (2014) applied statistical methods of 2×2 contingency table V ratio, X2 (Yate’ s correction), Ochiai Index (OI), Dice Index (DI), Point Correlation Coefficient (PCC), Jaccard index (JI), Association Coefficient (AC) and Spearman correlation coefficient to analyze the inter-specific association between epiphytic plant species in ancient cultivated tea plantation. For the 127 tea trees measured at individual scale, significant inter-specific association was reported, whereas insignificant association was found among 31 plots measured at plot scale. Indices of both Association Coefficient (AC) and Spearman correlation coefficient well indicated the inter-specific association between epiphytic species in consistence with X2 test, which revealed positive association between Bulbophyllum sp. and Drynaria propinqua, Davallia cylindrica and Liparis elliptica, Dendrobium capillipes and Lysionotus petelotii，as well as negative association between Bulbophyllum ambrosia and Dendrobium capillipes, Bulbophyllum ambrosia and Lysionotus petelotii, Bulbophyllum nigrescens and Dendrobium chrysanthum, Ascocentrum ampullaceum and Peperomia tetraphylla [20].
2.2 Coefficient of Niche Breadth and Coefficient of Niche Overlap
Niche breadth is the total available resources which can be utilized by a species (or other biological unit), and niche overlap is the competition phenomenon that two or more species with similar niche breadth compete for the limited resources in the common space for survival [6].
Field study were conducted by Chen et al.,(2014) to analyze the niche breadth and overlap of 12 plant species on 70 forest plots in Bawangling National Nature Reserve, presenting the descending order of niche breadth for 12 species: Aquilaria sinensis, Nephelium topengii, Camellia sinensis var. assamica, Alseodaphne hainanensis, Keteleeria hainanensis, Podocarpus imbricatus, Firmiana hainanensis, Parakmeria lotungensis, Cephalotaxus mannii, Michelia hedyosperma, Ixonanthes reticulata, Dacrydium pierrei. The results revealed that the niche breadth of a species was determined by its range of spatial distribution; in most cases, higher niche overlap value was usually found between species with broader niche breadth, except Michelia hedyosperma and Firmiana hainanensis species of narrow niche breadth; the low niche breadth of Michelia hedyosperma and Ixonanthes reticulate species partially led to smaller populations, which was advised to give the priority for conservation [21].
Both niche breadth and niche overlap of 10 shrub species and 11 herb species were examined by Gao et al., (2014) under a mixed forest consisted of Picea crassifolia and Betula platyphylla in high hill regions in Datong County, Qinghai Province. The results indicated broader niche breadth for species Potentilla fruticosa and Salix cupularis in shrub layer, as well as species Polygonum viviparum and Fragaria orientalis in herb layer. Higher niche overlap was found usually between populations with broader niche breadth. Nevertheless, some populations with narrow niche breadth also showed high niche overlap. The niche overlap between different species of a genera tended to be smaller, which would be attributed to their evolution and succession [22].
Statistical methods of Variance ratio, χ2-test based on a 2×2 contingency table and the test of association indices (Jaccard, Dice and Ochiai) were selected by Yu et al.,(2012) to examine the inter-specific association of 22 Pyrola decorata communities in Taibai Mountain. Results reported that only 5 paired species showed significant positive association (P<0.05), with 2 paired species showing highly significant positive association (P<0.01), whereas insignificant association was reported between the rest species pairs. For Jaccard index analysis, 84.42% of total species pairs were under 0.25 value of Jaccard index, and 12.31 % of total species pairs ranged from 0.25 to 0.50, while only 3.26% of total species pairs were over 0.50. These results revealed weak inter-specific association between investigated communities which tended to be independent [23].
3 Genetic Diversity
3.1 Hardy-Weinberg Equilibrium
Hardy-Weinberg equilibrium is the principle for the parental generation and their offspring to assess the degree of equilibrium between observed genotypic frequencies and allele frequencies in sexual reproduction process[6]. Both Hardy-Weinberg equilibrium and population structure of 283 Hevea brasiliensis Wickham germplasm were examined and analyzed by Fang et al., (2013), with 25 EST-SSRs loci detected. According to the results, 13 of total 25 EST-SSRs loci deviated Hardy-Weinberg equilibrium. The 283 Hevea brasiliensis Wickham germplasm were divided into 4 groups, and the amount of each group was 155, 110, 61 and 22 respectively. 20 locus combinations (6.67%) were significant linkage disequilibrium (P<0.05), and 5 of them were significant linkage disequilibrium at P<0.01 level [24].
3.2 Genetic Diversity
There are a number of conceptions to quantify genetic diversity, mainly including allele frequency, percentage of polymorphic loci, mean number of alleles per locus, mean expected heterozygosity per locus, Nei’ coefficient of gene differentiation, Nei’ genetic identity.
90 accessions were chosen by Xu et al., (1999) from total 22637 accessions in the National Genebank of soybean species, with selection criteria of nine agronomic traits, including disease resistance to SCN race No.3 and SCN race No.4, rust, SMV, and tolerance to cold, drought, salt, 100 seed weight and protein content. Five maximum and five minimum accessions in the Genebank were selected for comparison for each trait. The genetic diversity of 90 (G. max) soybean and one wild soybean (G. soja) accession were assessed by both agronomic trait analysis and microsatellite DNA or SSR markers. In total twelve pairs of SSR primers were applied and 83 alleles were detected with an average of 6.9 alleles per locus. Simple matching similarity coefficients between each pairs of genotypes were analyzed and clustered by Unweighted Paired Group Method Using Arithmetic Averages (UPGMA), revealing that soybean germplasms could be identified by SSR technique. However, the cluster analysis based on agronomic traits was not identical to SSR markers [25].
The genetic diversity of 38 Paulownia fortunei provenances, with 15 individuals per provenance, was deduced by Li et al., (2011) with technique of inter-simple sequence repeats (ISSR). In total 95 amplified DNA fragments were detected by 9 primers leading to clear and unique polymorphic bands, which were screened from 100 ISSR primers. There were 88 polymorphic loci among 95 amplified DNA fragments, resulting in the percentage of polymorphic loci (PPL) of 92.63%. The PPL at species level ranged from 32.63% (Fuzhou, Jiangxi) to 56.84% (WuZhou Guangxi and Jiu Jiang, Jiangxi) with the mean percentage of 47.16%. The mean values of effective number of alleles (Ne), Nei´s gene diversity index (H) and Shannon´s Information index (I) between different provenances were calculated as 1.3910, 0.2424 and 0.3765 respectively, indicating abundant genetic diversity between them. The Coefficient of Gene Differentiation (GGst) of provenances was 0.3539, and the genetic variation between provenances accounted for 35.39% of total genetic variation, revealing that genetic variation between different individuals of each provenance was higher. Genetic Identity of provenances varied from 0.39 to 0.82, showing the relatively broad genetic basis and abundant genetic variation among provenances. According to Genetic Identity, the provenances of Kaili, Guizhou, and Liuzhou, Guangxi showed closest relationship with Genetic Identity of 0.82, whereas longer genetic distance was reported between Hengyang (Hunan) and Zhuji (Zhejiang) populations, and between Hengyang (Hunan) and Zhenning County (Guizhou) populations, with Genetic Identity of 0.39. In total 38 provenances were classified into 3 groups by UPGMA cluster analysis, with little correlation between genetic distance and geographic distance among those provenances [26].
Genetic diversity of wild soybean population in the region of Beijing China was evaluated by Yan et al., (2008) with 40 primer pairs. In total ten populations were sampled with 28-30 individuals per population. 526 alleles were detected with a mean value of 13.15 per locus. The average value of Expected Heterozygosity per locus (He) and Observed Heterozygosity per locus (Ho) were 0.369 and 1.29% respectively for the wild soybean populations, and the mean Shannon index (I) was 0.658. The mean value of between-population genetic diversity (Hs) and within-population genetic diversity (DST) were 0.446 and 0.362 respectively. The average Coefficient of Gene Differentiation for loci (GGst) between populations was estimated as 0.544. Center-Western ecotype showed more abundant genetic diversity than the Northern and Eastern ecotypes, geographic heterozygosity was found in the genetic divergence patterns of natural populations between the Taihang and the Yanshan mountains. The genetic diversity of drought-tolerant population was poor, indicating the potential value of tolerance gene (s) for breeding [27].
Genetic diversity of totally 13 Cannabis populations from different origins was deduced by Hu et al., (2012) using POPGENE 3.2 Software. AFLP results indicated that the most abundant genetic diversity was found in Yunnan population, with Percentage of Polymorphic Loci (PPL) of 88.82%, Nei´s total genetic diversity (He) of 0.3011, and Shannon Index (I) of 0.4571; and followed by the Heilongjiang population with Percentage of Polymorphic Loci (PPL) of 75.66%, Nei´s total genetic diversity (He) of 0.2572, and Shannon Index (I) of 0.3897. The PPL, Ht and Hs of 13 Cannabis populations was 92.11%, 0.3837 and 0.1640 respectively. Coefficient of genetic differentiation between populations (GGst) was 0.5725, revealing that genetic variation between populations accounted for 57.25% of the total genetic variation, and the other 42.75% of total genetic variation was attributed to the genetic variation between individuals within population. Both genetic distance and genetic identity of Cannabis were calculated on the basis of Nei´s (1978) method, for further analysis of genetic differentiation among populations. Genetic identity among populations ranged from 0.6556 to 0.9258, with the highest value of 0.9258 between Guangxi population and Sichuan population. The genetic identity between Yunnan population and Guizhou population, Yunnan population and Sichuan population were 0.9196 and 0.9173 respectively, while the lowest genetic identity was found between Gansu and Shanxi populations. These findings became the scientific evidence for identification of Cannabis seed and provided the indicators for breeding and evolutionary analysis [28].
The genetic diversity of 120 individuals from six natural populations of Abies chensiensis was analyzed by Li et al., (2012) on the basis of 10 simple sequence repeat markers. The genetic diversity, genetic structure and changes in gene flow between different populations were analyzed, revealing 149 alleles in 10 microsatellite loci with a value of 14.9 as the average number of alleles per locus (A). The effective number of alleles per locus (Ne), the mean expected heterozygosity (He), the mean observed heterozygosities per locus (Ho), the Shannon diversity index (I), the proportion of genetic differentiation among populations (FST), and gene flow between the populations were 7.7, 0.841, 0.243, 2.13, 6.7% and 3.45, respectively. Insignificant correlation was found between genetic distance and geographic distance (r=0.4906, P>0.05). The relatively low genetic diversity was reported in the 6 natural populations of A.chensiensis, and inner-population genetic variation accounted for the majority of total genetic variation [29].
However, it is worthwhile mentioning that the analysis of genetic diversity is significantly influenced by sampling size. For example, the genetic integrity of Sorghum bicolor L. Moench. was studied by Xu et al.,(2012) adopting SSRs technique, as one of the most commonly used markers for the assessment of genetic diversity, population structure studies and marker-assisted selection. In total ten groups of sorghum with different sample sizes (including 10, 20, 30, 40, 50, 60, 70, 80, 90, 100 individuals per group) were selected randomly and 25 polymorphic microsatellite primers were conducted for the assessment of genetic diversity indices (the average number of alleles, effective number of alleles, Shannon Index, Observed Heterozygosity, Expected Heterozygosity, Percentage of Polymorphic loci and the Frequency of Rare Alleles). According to the correlation between genetic diversity indices and sample sizes, the number of alleles, effective number of alleles, Shannon index increased correspondingly to the increase of sample sizes, with the peak increase rate at a sample size of 40 individuals. Consequently, the sample size of 40 individuals accounted for 98.5% of total numbers of alleles, 99.1 % of total effective numbers of alleles and 98.5% of total Shannon indexes among 100 individuals, indicating the 40 individuals as optimal sample size for the SSRs technique in gorghum integrity assessment [30].
3.3 Genetic Variation
Genetic variation assessment mainly adopts the conception of genetic distance and evolutionarily significant unit (ESU), usually deduced by cluster analysis, PCA, or evolutionary tree analysis. However, both cytological and DNA molecular markers are able to achieve this.
The karyotype of characteristics and evolutionary relationships among the traditional Chinese medicine Sophora flavescens from four different origins was investigated by Duan et al., (2014). The karyotypes and chromosome numbers of Sophora flavescens were calculated by using root-tip squashing method and clustered by the karyotype resemblance-near coefficient, which linked all the genetic materials.
The chromosome numbers of Sophora flavescens from Chifeng Inner Mongolia， Changzhi Shaanxi, Meixian Shaanxi and Chengdu Sichuan all were 18 and belonged to 1 A type, with karyotype formulas of 2n = 2x = 18 = 18m(2SAT), 2n = 2x = 18 = 14m(1SAT) + 4sm(1SAT), 2n = 2x = 18 = 16m(2SAT) + 2sm and 2n = 2x = 18 = 18m(2SAT) respectively. The karyotype asymmetry index of Sophora flavescens from Chifeng Inner Mongolia, Changzhi Shaanxi, Meixian Shaanxi and Chengdu Sichuan were 56.32%, 57.88%, 59.41 % and 54. 32%, respectively. According to Karyotype clustering analysis, the closest genetic relationship was reported between S. flavescens from Chengdu and Chifeng, with the highest karyotype resemblance-near coefficient of 0.9929, and their evolution distance was 0.0072. In comparison, the farthest genetic relationship was found between S. flavescens from Chengdu and Meixian, with the lowest karyotype resemblance-near coefficient of 0.9533, and their evolution distance was 0.0478. Karyotype of Sophora flavescents from Chengdu was the most primitive among them, followed by those from Chifeng, Changzhi and Meixian. The conclusion of this study provided cytological information for germplasms identification, and became the basis of genetic variation and genetic relationship analysis of Sophora flavescens[31].
To explore the genetic distance in evolutionary process among 6 Bupleurum medical plants, including B.longeradiatum, B.smityii, B. longicaule var. amplexicaule, B. scorzonerifolium, B. chinense, B. falcatum, karyotype parameters identification was adopted by Song et al.,(2012), which used the cluster analysis of karyotype resemblance-near coefficient and evolutionary distance, based on the calculation of the relative length, arm ratio, centromere index. The highest karyotype resemblance-near coefficient (0.9920) and smallest evolutionary distance (De = 0.0080) existed between B. scorzonerifolium and B. chinense, revealing the closest relationship between them. In comparison, the minimum karyotype resemblance-near coefficient (0.4794) and the maximum evolutionary distance (De = 0.7352) was reported between B. smityii and B. falcatum [32].
In Luo et al., (2006) study, 200 two-line combinations were matched by mating 5 photo/ thermal-sensitive genic male-sterile lines and 40 varieties. The genetic distance (GD) between 5 sterile lines and 40 varieties was examined by SSR markers, with the discussion between genetic distance and heterosis. The correlation of genetic distance varied with yield per F1 plant, heterobeltiosis of F1 yield, effective panicles, panicle length spikelets per panicle, density of spikelet setting, seed setting rate, and 1000 grain weight, due to various gene materials or different range of genetic distance. When the genetic distance between Tianfeng S and its paternal varieties ranged from 0.6286 to 2.5257, the correlation of genetic distance with yield per F1 plant or its heterobeltiosis appeared to be significant at P<0.05 level; As the genetic distance between Peiai 64S and paternal varieties ranged from 0.8247 to 1.5315, their correlation between genetic distance and yield per F1 plant was significant at P<0.05 level; furthermore, for all parents of two-line combinations with genetic distance ranged from 0. 5333 to 1.5, the correlation between heterobeltiosis of yield per F1 plant and genetic distance appeared to be significant at P < 0. 05 level; the correlation of yield per F1 plant with genetic distance was significant at P < 0. 05 level, as the genetic distance ranged from 0.5333 to 1.0; the significance of correlation between yield per F1 plant and genetic distance was at P < 0. 01 level, when genetic distance ranged at three layers: between 1.0 and 1.5; 0.5333 and 1.5; 0.5333 and 2.5257. This genetic distance analysis indicated the appropriate range for mating combinations of hybrid rice [33].
An endemic species of Sinomanglietia glauca, which is unique in Yichun in Jiangxi Province and Yongshun of Hunan Province in Central China, has been listed in Category I of the National Key Protected Wild Plants in 1999 (as asynonym of Manglietia decidua). Xiong et al.,(2014) study covered all of four populations of S. glauca, which had been identified so far, and the genetic diversity and genetic variation was investigated by nuclear microsatellite markers. According to the results, S. glauca showed relatively low genetic diversity with the average number of alleles (A) of 2.604 and the mean expected heterozygosity (HE) of 0.423, but presented significant genetic variation with high genetic differentiation FST of 0.425. Cluster analysis by STRUCTURE and Principal Coordinated Analysis indicated that Jiangxi and Hunan populations were classified into two independent groups. Only one natural breeding population was identified in Jiangxi, while two were found in Hunan, with significant genetic variation. The heterozygosity was found to be excessive significantly, which might be caused by allelic frequencies differed between male and female parents occasionally in a small population. The results indicated that S. glauca would experience bottleneck(s) in recent evolution history, which led to reduction of population size, loss of genetic diversity and strong population differentiation. The genetic diversity study resulted in the advices that S.glauca should be classified as three conservation units according to their evolutionary units: Jiangxi unit and Hunan unit, and the Hunan populations could be further divided into two sub-management units (YPC and LJC) [34].
3.4 Genetic Contribution Rate
Genetic contribution rate was firstly proposed by Petit et al., (1998). For the standardization of the allelic richness results across populations, the technique of rarefaction is established to facilitate assessment of the expected number of different alleles among equal-sized samples derived from different populations, which is divided into two components: the first is relevant to the degree of population diversity and the second is related to its divergence from the other populations [35].
4 Conclusion
As discussed above, the multivariate methods of evolutionary ecology study become essential to classify the communities and sub-populations at ecosystem layer and at genetic layer respectively for biodiversity assessment work. Due to three thorny issues threatening biodiversity defined by Chinese Biodiversity Conservation Strategy and Action Planning (2010-2030), biodiversity monitoring projects need to be implemented at three layers, and application of 3S technology on biodiversity monitoring with high-resolution remote sensing imagines is advised by Liu et al., (2014) [36], e.g. investigation of the distribution change of dominant plant species over ten years in a national park by using object-oriented classification of Quickbird remote sensing imagines, and then the temporal and spatial dynamics of biodiversity evolution at both ecosystem layers and species layers should be discussed on the basis of evolutionary ecology study. Additionally, biodiversity monitoring projects should be conducted according to the Technical Guidelines for Biodiversity Monitoring --- Terrestrial Vascular Plant (HJ 710.1-2014).
For genetic layer, a combination of cytological markers and DNA molecular markers is advised by Liu et al., (2014) for classification of sub-populations [37], mainly due to the consideration of saving the cost and reliability of differentiation methods. Nevertheless, it is worthwhile mentioning that the conclusion drawn by multivariate cluster analysis between cytological markers and DNA molecular markers would not be consistent, possibly due to gene recombination and gene mutation. Consequently, the multivariate cluster analysis for sub-population classification would be more reliable on the basis of DNA molecular markers. The software of computing both polygenetic (gene by gene analysis) and phylogenomic (the whole genome comparison) methods is suggested by Ahmed (2009) [38].
According to the Environmental Standard on Classifying the Categories of Genetic Resources (HJ 626-2011) in China, there are three kinds of DNA molecular methods pointed out for ranking genetic resources (or endangered species) between categoryⅠand categoryⅡ, including assessment of genetic diversity, evolutionarily significant unit (ESU), or genetic contribution rate, which have been substantially discussed above. However, it is worthwhile noting that any one of these three methods is acceptable for environmental engineers to conduct this environmental standard, although there is debate between these methods in terms of selection priority, such as Chen et al., (2002) [39].
According to the Chinese Biodiversity Conservation Strategy and Action Planning (2010-2030), prediction of climate change effects on biodiversity conservation is significant, and the application of CTMs model on prediction of climate change effects on biodiversity is advised by Liu Huan (2014) [40]. However, the knowledge of evolutionary ecology study derived from the biodiversity monitoring projects in the past may be required for this prediction work.
【References】
［1］环境保护部环境工程评估中心．全国环境影响评价工程师职业资格考试考点要点分析[M]．中国环境出版社，2008．
［2］卜耀军，等．模糊聚类和排序在植被演替研究中的综合应用[J]．2005.
［3］艾尼瓦尔·吐米尔,热衣木·马木提,阿不都拉·阿巴斯．乌鲁木齐南部山区森林生态系统树生地衣群落结构[J]．2009.
［4］蔡文贵,等.粤西海域浮游植物群落结构特征的多元分析与评价[J].2007.
［5］王伯荪,彭少麟.鼎湖山森林群落分析[J].中山大学学报,1985.
［6］陶玲,任珺.进化生态学的数量研究方法[M].北京:中国林业出版社,2004.
［7］张向龙,等.基于恢复力定量测度的社会 生态系统适应性循环研究:以榆中县北部山区为例[J].2013.
［8］Jaunatre, R., et al., New synthetic indicators to assess community resilience and restoration success[J].2013.
［9］李艳霞,等.退化生态系统恢复指示性指标的研究[J].2007.
［10］胡聃.生态系统可持续性的一个测度框架[J].1997.
［11］于玲,王祖良,李俊清.自然保护区生态旅游可持续性评价：以浙江天目山 自然保护区为例[J].2007.
［12］侯海潮,等.森林健康经营理论应用研究:以塞罕坝机械林场为例[J].河北林果研究,2012,27(1).
［13］郭峰，等．华北土石山区典型天然次生林生态系统健康评价研究[J].水土保持研究,2012,19(4).
［14］刘硕,贺康宁,王晓江.鄂尔多斯沙地不同退牧年限植物群落多样性及变异性研究[J].西北植物学报,2010,30(3).
［15］张大勇,王刚,赵松岭.甘南亚高山草甸弃耕地植物群落演替的数量研究Ⅰ.演替先锋群落的特征分析[J].中国草地,1988(06):14-19.
［16］王伯荪,彭少麟.鼎湖山森林群落分析:边缘效应[J].中山大学学报,1986(4).
［17］Euskirchen, E.S., J. Chen and R. Bi, Effects of edges on plant communities in a managed landscape in northern Wisconsin[J]. Forest Ecology And Management, 2001(148).
［18］张瑾,等.甘肃敦煌西湖荒漠湿地生态系统优势植物种群分布格局及种间关联性[J].中国沙漠,2013,33(2).
［19］闫志刚,等.黄花蒿野生群落的种间关系及其对水淹干扰响应[J].广西植物, 2009,29(6).
［20］王青,等.景迈-芒景古茶园茶树群落寄附生植物种间联结研究[J].西部林业科 学,2014,43(3).
［21］陈玉凯,等.海南岛霸王岭国家重点保护植物的生态位研究[J].植物生态学报,2014,38(6).
［22］高二鹏,等.青海大通脑山区青海云杉+白桦混交林主要种群的生态位特征[J].中国水土保持科学,2014,12(3).
［23］余贝贝,等.太白山雅美鹿蹄草群落植物种间关联性[J].东北林业大学学报, 2012,40(11).
［24］方家林,等.橡胶树魏克汉种质群体结构分析[J].基因组学与应用生物学, 2013,32(5).
［25］许占友,等.利用 SSR 标记鉴定大豆种质[J].中国农业科学,1999,32.
［26］李芳东,等.白花泡桐种源遗传多样性的ISSR分析[J].中南林业科技大学学报,2011(07):1-7.
［27］严茂粉,李向华,王克晶.北京地区野生大豆种群SSR标记的遗传多样性评价[J].植物生态学报,2008,32(4).
［28］胡尊红,等.大麻品种遗传多样性的 AFLP 分析[J].植物遗传资源学报,2012, 13(4).
［29］李为民李思锋,黎斌.利用SSR分子标记分析秦岭冷杉自然居群的遗传多样性[J].植物学报,2012,47(4).
［30］许玉凤,等.高粱微卫星分析中遗传完整性样本量的确定[J].华北农学报, 2012,27(3).
［31］段永红,等.不同产地苦参核型及似近系数聚类分析[J].中国药学杂志,2014(14):1194-1199.
［32］宋芸,乔永刚,吴玉香.6种柴胡属植物核型似近系数聚类分析[J].中国中药杂志,2012(08):1157-1160.
［33］罗小金,等.利用 SSR 标记分析水稻亲本间遗传距离与杂种优势的关系[J].植物遗传资源学报, 2006,7(2).
［34］熊敏,等.华木莲居群遗传结构与保护单元[J].生物多样性,2014.22(4).
［35］PETIT, R., ABDELHAMIDELMOUSADIK and O. AND, Identifying Populations for Consevation on the Basis of Genetic Markers[J]. Conservation Biology, 1998,12(4).
［36］Liu Huan, et al., A Brief Review of 3S Technology Application on Biodiversity Monitoring and Assessment[J]. Science & Technology Information,2014(3).
［37］刘焕,张洪初,唐秋盛.保护遗传学方法在生物多样性监测和评价领域的应用研究[J].科技视界,2014(8).
［38］Mansour, A., Phylip and Phylogenetics. Gene Genomes and Genomics[J].2009.
［39］陈小勇,等.重要物种优先保护种群的确定[J].生物多样性,2002,10(3).
［40］刘焕.应用全球大气输送模型进行数值模拟的研究进展[J].科技视界,2014.
［责任编辑：薛俊歌］