基于3种数学方法的粤北油茶果实性状综合评价

Comprehensive evaluation of the fruit traits of Camellia oleifera in northern Guangdong based on three mathematical methods

【目的】探讨确定可行的数学分析方法,为油茶品种科学鉴评提供技术支撑。【方法】本文以广东韶关市主栽的15个油茶品种为研究对象,通过测定其果实性状等指标,采用因子分析、灰色关联分析和熵值法,分别对果实12个性状指标进行综合分析。【结果】1)参试品种果实性状各指标间存在不同程度变异,变异范围为9.10%~54.52%,其中表型性状变异小于品质性状变异;各指标间共有43对达到显著性水平(P <0.05)。2)因子分析结果表明,12个性状指标转化为3个相互独立的综合指标,其贡献率分别为49.865%、15.632%和12.480%,累计贡献率为77.977%,包含了参试品种果实性状的大部分信息。3)灰色关联分析结果表明,不同性状指标权重大小存在差异,其中种仁干质量权重最大,百粒重权重最小。4)熵值法结果表明,单果质量权重最大,果皮厚权重最小。5)在3种评价方法中,采用因子分析进行综合评价时,高州油茶、赣兴48、湘林210、粤韶75-2和华鑫综合表现最好;采用灰色关联分析进行综合评价时,长林53号、华鑫、湘林210、高州油茶和赣兴48综合表现最好;采用熵值法进行综合评价时,高州油茶、华鑫、湘林210、长林53号和赣兴48综合表现最好。6)区分度评判结果表明,因子分析法区分度值最大,灰色关联分析区分度值最小。因此,因子分析法在油茶果实性状评价中可信度较好。【结论】油茶果实性状评价更适合采用因子分析法,高州油茶、赣兴48、湘林210的果实性状综合表现最好。

【Objective】This work aimed to explore and determine feasible mathematical analysis methods to provide technical support for the scientific identification and evaluation of Camellia oleifera varieties.【Method】In this study,15 Camellia oleifera varieties,mainly from Shaoguan City,Guangdong Province,were used as the research objects.By using factor analysis,grey correlation analysis and entropy value method,12 fruit traits were comprehensively analyzed.【Result】1)There were different degrees of variation among the fruit traits of the tested varieties,with a range of 9.10%to 54.52%,among which the variation of phenotypic traits was smaller than that of quality traits,and a total of 43 indicators reached the significant level(P<0.05).2)The results of the factor analysis showed that the 12 trait indicators were transformed into 3 mutually independent indicators,whose contribution rates were 49.865%,15.632%and 12.480%,respectively,The cumulative contribution rate was 77.977%,which contained most of the information about the fruit traits of the tested varieties.3)The results of grey correlation analysis showed that there were differences in the weights of different trait indicators,with the greatest weight of dry seed kernel weight and the least weight of hundred-kernel weight.4)The results of the entropy value method showed that the weight of single fruit weight was the highest and the weight of peel thickness was the lowest.5)Among the 3 evaluation methods,Gaozhou oil tea,Ganxing 48,Xianglin 210,Yueshao 75-2 and Huaxin had the best overall performance when factor analysis was used for the comprehensive evaluation.Changlin 53,Huaxin,Xianglin 210,Gaozhou oil tea and Ganxing 48 had the best overall performance when grey correlation analysis was used for the comprehensive evaluation.Gaozhou oil tea,Huaxin,Xianglin 210,Changlin 53 and Ganxing 48 had the best overall performance when entropy value method was used for the comprehensive evaluation.6)The results of the differentiation evaluation showed that the differentiation value of the factor analysis method was the largest and the differentiation value of the grey correlation analysis was the smallest.Therefore,the factor analysis method was reliable in the evaluation of Camellia oleifera fruit traits.【Conclusion】The factor analysis method is more suitable for evaluating the fruit traits of Camellia oleifera,and the comprehensive performance of Gaozhou oil tea,Ganxing 48 and Xianglin 210 is the best.