摘要
对青藏高原老芒麦野生种群生态分布特性和形态学变异研究表明:(1)老芒麦在青藏高原地区分布广泛,群落生境可初步划分为高山亚高山草甸型、河谷草地型和森林灌丛型;群落组成以高山红柳+老芒麦+发草、沙棘+老芒麦+蒿类、老芒麦+锦鸡儿+鹅观草、老芒麦+披碱草+多节雀麦4种类型最多。(2)野生老芒麦种群形态学性状具广泛变异,其中与牧草产量和种子产量相关的形态性状变异较大,与分类相关的指标变异程度较小。(3)聚类分析将不同形态的老芒麦聚为三大类群,聚类结果除与海拔有一定关系外,与其地理分布的一致性不明显。(4)主成分分析表明,内外颖长、内外颖芒长、旗叶宽、倒二叶片长、株高、内外稃长、外稃芒长、内外稃宽、穗中部节上每小穗的小花数、穗长、叶色、茎粗、灰度和穗中部节上的小穗数是引起老芒麦形态分化的主要指标。
The ecological characteristics and morphological variation of wild E.sibiricus were studied,the results showed that :(1) E.sibiricus were widely distributed in Qinghai-Tibetan plateau,its habitat types could be classified into three types: alpine and subalpine meadow type,valley grassland type and forest-shrub type;the main community composition included four types: Salix cheilophila var.microstachyoides +E.sibiricus + Deschampsia caespitosa,Hippophae rhamnoides + E.sibiricus + Artemisia;E.sibiricus+ Caragana+ Roegneria,E.sibiricus + E.nutans+ Bromus plurinodes.(2) Natural E.sibiricus produces rich morphological diversity.Among the 30 observed properties,morphological variances most significantly occurred in the traits related with forage and seed yield,as for classification index,there are less variation.(3) Based on the phenotypic characteristics,37 accessions were clustered into three morphological types,and the cluster result don't agree well with geographical distribution but related to altitude.(4) Principal component analysis results indicated that outer and inner glume length,awn length of outer and inner glume,boot leaf width,length of the second leaf from the inflorescence,plant height,lemma length and width,palea length and width,awn length of lemma,floret number in each spikelet,ear length,leaf color,internode diameter,gray grade and spikelet number were the main sources of morphological differentiation of E.sibiricus accessions.
出处
《中国草地学报》
CSCD
北大核心
2010年第4期49-57,共9页
Chinese Journal of Grassland
基金
国家公益性农业行业专项(nyhyzx07-022)
现代农业产业技术体系建设专项资金
国家"十一五"科技支撑计划项目(2008BADB3B07)资助
关键词
老芒麦
生态特性
形态变异
聚类分析
主成分分析
Elymus sibiricus L.
Ecological characteristics
Morphological variation
Cluster analysis
Principal component analysis