以山东省泰安徂徕山、泰安泰山、临沂蒙山、烟台艾山和烟台昆嵛山5个种源63株盐肤木（Rhus chinensis Mill.）的成熟果实为研究对象，对不同种源果实和种子的横径、纵径和侧径以及果皮和种子的百粒鲜质量、百粒干质量和含水量进行了比较；在此基础上，对果实表型性状进行了变异分析、Pearson相关性分析和主成分分析，基于欧氏距离对供试5个种源进行了聚类分析，并对果实表型性状与部分地理-气候因子进行了Pearson相关性分析。结果表明：盐肤木上述12个果实表型性状在种源间存在显著（P＜0.05）差异；烟台昆嵛山种源的果实横径（4.69 mm）、果实纵径（4.53 mm）、种子横径（3.61 mm）、种子纵径（2.93 mm）、果皮百粒鲜质量（0.88 g）、种子百粒鲜质量（1.26 g）、果皮百粒干质量（0.81 g）和种子百粒干质量（1.18 g）最大，果实侧径（2.35 mm）和果皮含水量（8.94%）较大。盐肤木这12个果实表型性状种源间变异系数的均值为24.35%，种源内变异系数的均值为12.99%。Pearson相关性分析结果表明：大部分果实表型性状间的相关性显著或极显著（P＜0.01），且与经度、平均海拔和无霜期的相关性显著。主成分分析结果表明：前4个主成分的累计贡献率为82.62%，第1和第2主成分分别为果实和种子的质量性状和形态性状。聚类分析结果表明：在欧氏距离6.0处，供试5个种源被分成2组，一组仅泰安徂徕山1个种源，另一组包括其余4个种源。综上所述，山东省盐肤木果实表型性状多样性较为丰富，果实表型性状变异以种源间变异为主且受地理位置影响明显；果实和种子的质量性状和形态性状可作为盐肤木优质种源筛选的重要指标；烟台昆嵛山种源为供试盐肤木的最优种源。

Taking ripe fruits of sixty-three individuals of Rhus chinensis Mill. from five provenances of Culai Mountain of Tai'an， Mount Tai of Tai^'an， Mengshan Mountain of Linyi， Aishan Mountain of Yantai， and Kunyu Mountain of Yantai in Shandong Province as research objects， horizontal， vertical and side diameters of fruit and seed， 100-grain fresh and dry masses and moisture content of peel and seed from different provenances were compared； on the basis， variation， Pearson correlation， and principal component analyses on fruit phenotypic traits were performed， cluster analysis on five test provenances were conducted based on Euclidean distance， and Pearson correlation of fruit phenotypic traits with some geographical-climatic factors were analyzed. The results show that there are significant （P<0.05） differences in above twelve fruit phenotypic traits of R. chinensis among provenances； fruit horizontal diameter （4.69 mm）， fruit vertical diameter （4.53 mm）， seed horizontal diameter （3.61 mm）， seed vertical diameter （2.93 mm）， peel 100-grain fresh mass （0.88 g）， seed 100-grain fresh mass （1.26 g）， peel 100-grain dry mass （0.81 g）， and seed 100-grain dry mass （1.18 g） of Kunyu Mountain of Yantai provenance are the largest， and its fruit side diameter （2.35 mm） and peel moisture content （8.94%） are relatively large. The average of coefficient of variation of these twelve phenotypic traits of R. chinensis among provenances is 24.35%， while that within provenance is 12.99%. The Pearson correlation analysis result shows that the correlations among most fruit phenotypic traits are significant or extremely significant （P<0.01）， and their correlations with longitude， mean altitude， and frostless period are significant. The principal component analysis result shows that the cumulative contribution rate of the first four principal components is 82.62%， and the first and the second principal components are quality traits and morphological traits of fruit and seed， respectively. The cluster analysis result shows that at Euclidean distance of 6.0， the five test provenances are clustered into two groups， one group only contains Culai Mountain of Tai'an provenance， while the other group contains the rest four provenances. In conclusion， the diversity of fruit phenotypic traits of R. chinensis in Shandong Province is relatively abundant， and the variations of fruit phenotypic traits are mainly among provenances and are evidently affected by geographical locations； mass traits and morphological traits of fruit and seed can be used as important indexes for selecting good provenance of R. chinensis； Kunyu Mountain of Yantai provenance is the best provenance of test R. chinensis.