基于模糊Borda组合模型的枸杞品质水氮调控效应评价

Evaluation of water and nitrogen regulation effects of wolfberry quality based on fuzzy Borda combination model

为筛选出合适的枸杞(Lycium barbarum L.)水氮供应阈值,利用模糊Borda组合模型评价了不同水氮组合对枸杞品质的调控效应。以三年生‘宁杞5号’枸杞为对象,分析了4个水分调控水平(以土壤含水量占田间持水量θ_f的百分比控制土壤水分上、下限),即充分灌溉W0(75%~85%θ_f)、轻度水分亏缺W1(65%~75%θ_f)、中度水分亏缺W2(55%~65%θ_f)、重度水分亏缺W3(45%~55%θ_f),和4个施氮水平,即不施氮N0(0 kg·hm^(-2))、低施氮量N1(150 kg·hm^(-2))、中施氮量N2(300 kg·hm^(-2))、高施氮量N3(450 kg·hm^(-2)),对枸杞外观品质和营养品质的影响,并构建模型对枸杞品质进行综合评价。结果表明:(1)同一水分调控下中低量施氮可提高枸杞果实果径,较高施氮量提高3.39%~5.19%,随水分亏缺程度加剧枸杞果径呈先增后降变化趋势,其中W0、W1和W2较W3分别提高8.01%、9.44%和6.92%;枸杞果长随施氮量增加呈先增后降变化趋势,N2处理提高最大(4.39%),随水分亏缺程度加剧呈降低趋势,各处理较重度水分亏缺提高2.50%~9.57%;水分调控、施氮水平及二者交互作用均对枸杞果形指数影响不显著。W1N2处理的总糖、多糖、脂肪、维生素C和黄酮含量均最高,分别较其他处理提高了3.32%~16.93%、7.49%~54.72%、6.50%~45.89%、2.29%~47.41%和4.14%~28.34%;W2N3处理的氨基酸含量最高,较其余处理提高了11.12%~86.16%;W0N2处理的蛋白质含量最高,较其他处理提高了7.15%~71.67%;W2N1处理的总酚含量最高,较其他处理提高了8.54%~65.98%。即轻中度水分亏缺、低中施氮量水平时有利于枸杞外观品质和营养品质的形成。(2)各处理枸杞品质的评价值因模型不同存在明显不一致。剔除主成分分析后,隶属函数分析、灰色关联度分析和TOPSIS模型3种单一模型各自的评价结果与除该单一模型外其余2种模型评价结果的Kendall相关系数通过了Kendall-W协和系数事前检验,具有一定的相容性,可进行组合评价。(3)对均值法和模糊Borda评价法构建的组合评价模型进行事后检验发现,模糊Borda组合评价模型与各单一模型评价值的Spearman相关系数高于均值法。(4)对基于模糊Borda组合评价模型的枸杞品质评价值进行回归分析得出,水氮供应阈值分别为348.40~385.08 mm和270.46~298.92 kg·hm^(-2)时,枸杞的综合品质较优。

To determine optimal water and nitrogen supply thresholds for wolfberry(Lycium barbarum L.),a fuzzy Borda combination model was used to evaluate the effects of different water and nitrogen combinations on its quality.Taking the three-year-old ‘Ningqi No.5' wolfberry as the object,four water control levels(the upper and lower limits of soil water controlled by the percentage of soil water content to field water capacity θ_f) were analyzed.Full irrigation of W0(75%~85% θ_f),mild water deficit W1(65%~75% θ_f),moderate water deficit W2(55%~65% θ_f),severe water deficit W3(45%~55% θ_f) and 4 nitrogen application levels of no nitrogen application N0(0 kg·hm^(-2)),low nitrogen application rate N1(150 kg·hm^(-2)),medium nitrogen application rate N2(300 kg·hm^(-2)) and high nitrogen application rate N3(450 kg·hm^(-2)) on the appearance quality and nutritional quality of wolfberry were investigated.The results showed that:(1) Under the same water control,the fruit diameter of wolfberry fruit was increased by medium and low nitrogen application,while the fruit diameter of higher nitrogen application was increased by 3.39%~5.19%.The fruit diameter of wolfberry fruit was increased first and then decreased with the increase of water deficit,in which W0,W1 and W2 were increased by 8.01%,9.44% and 6.92% compared with W3,respectively.The fruit length of wolfberry first increased and then decreased with the increase of nitrogen application,N2 treatment increased the most(4.39%),and decreased with the increase of water deficit,and the severe water deficit of each treatment increased by 2.50%~9.57%.Water regulation,nitrogen application level and their interaction had no significant effect on the fruit form index.The contents of total sugar,polysaccharide,fat,vitamin C and flavonoid in W1N2 treatment were the highest,which were increased by 3.32%~16.93%,7.49%~54.72%,6.50%~45.89%,2.29%~47.41% and 4.14%~28.34% compared with other treatments,respectively.The amino acid content of W2N3 treatment was the highest,which was 11.12%~86.16% higher than that of other treatments.The protein content of W0N2 treatment was the highest,which was increased by 7.15%~71.67% compared with other treatments.The total phenol content of W2N1 treatment was the highest,which was 8.54%~65.98% higher than that of other treatments.The appearance quality and nutritional quality of wolfberry were formed when the water deficit was moderate,and the nitrogen level was low.(2) The evaluation value of wolfberry quality of each treatment was significantly inconsistent due to different models.After removing the principal component analysis,the Kendall correlation coefficient of the evaluation results of the three single models of membership function analysis,grey correlation degree analysis and TOPSIS model and the evaluation results of the other two models except this single model passed the Kendall-W concordant coefficient test,which had certain compatibility and can be combined evaluation.(3) Post-test of the combined evaluation model constructed by the mean value method and the fuzzy Borda evaluation method found that the Spearman correlation coefficient between the fuzzy Borda combination evaluation model and each single model was higher than the mean value method.(4) The regression analysis of the quality evaluation value of wolfberry based on the fuzzy Borda combination evaluation model showed that the comprehensive quality of wolfberry was better when the water and nitrogen supply thresholds were 348.40~385.08 mm and 270.46~298.92 kg·hm^(-2),respectively.