不同施氮水平下番茄叶片氮含量时空分布特征及SPAD无损诊断模型构建OA

To address the current issues of chaotic leaf position selection and poor model applicability in SPAD-based ni-trogen diagnosis of greenhouse tomato,this study took greenhouse tomato as the research object and investigated the spa-tiall variation patterns of SPAD value,leaf nitrogen content,and apparent nitrogen use efficiency(ANUE)at different growth stages and leaf positions under gradient nitrogen concentrations.Through correlation analysis,the optimal diagnos-tic leaf position was identified,and nitrogen content estimation models were established.The results showed that:(1)To-mato leaf nitrogen content increased first and then decreased over the growth stage.The upper and middle leaf positions almost peaked at stage Ⅴ(110 days after transplanting),while the lower leaf position peaked at stage Ⅳ(95 days after transplanting).At the same stage and leaf position,leaf nitrogen content rose initially and then decreased with increasing nitrogen application rates,with the highest value observed under the N3 treatment(450 kg·hm-2)(except lower leaf posi-tion at stage IV).The vertical distribution of leaf nitrogen content generally followed the order(except stage IV):Upper leaf position>middle leaf position>lower leaf position,with more pronounced differences among leaf positions under low nitrogen conditions.(2)SPAD value at all leaf positions also increased first and then decreased over the growth stage,with a spatial distribution order at stage II of upper leaf position>middle leaf position>lower leaf position.(3)Spearman correlation analysis revealed that SPAD value at the upper leaf position were significantly correlated with nitro-gen content during stages Ⅰ-Ⅴ(50-110 days after transplanting),with correlation coefficients ranged from 0.948 to 0.993(P<0.01).At stage Ⅵ(125 days after transplanting),the correlation at the lower leaf position also reached a highly sig-nificant level(correlation coefficient 0.908,P<0.01).Accordingly,the optimal leaf positions for SPAD-based nitrogen di-agnosis were determined as upper leaf position for stages Ⅰ-Ⅴ and lower leaf for stage Ⅵ.(4)The apparent nitrogen use efficiency(ANUE)of tomato leaves was highest at the upper leaf position and lowest at lower leaf position.ANUE in-creased first and then decreased with rising nitrogen rate,and peaked at stage Ⅴ for upper and mid leaf positions.(5)Non-linear nitrogen content estimation models constructed based on SPAD value at the optimal leaf positions achieved coef-ficient of determination(R²)values ranging from 0.879 3 to 0.965 3,with the highest value at stage Ⅱ(65 days after transplant-ing).Model validation results indicated that the stage Ⅰ(50 days after transplanting)model exhibited the best reliability(RMSE=0.90%,RE=6.55%).In conclusion,non-destructive SPAD-based nitrogen diagnosis for tomato should adopt a dy-namic leaf position selection strategy-"upper leaf position in early stages,lower leaf position in later stages"-and the nitro-gen content estimation models constructed for different growth stages can enable accurate diagnosis of leaf nitrogen status.