Application of polyphenol-chlorophyll meter to monitor cotton N nutrition status

Objectives We studied the suitable parameter in diagnosing cotton N nutrition using the polyphenol-chlorophyll meter in the field.

Methods A field experiment was set up in Shihezi, Xinjiang, with three N application rates: N 0, 180, and 240 kg/hm², expressed as N0, N180 and N240. The N fertilizer treatments were topdressed 5 times with drip irrigation. Nitrogen balance index (NBI), Chl, and SPAD values of 20 cotton plants leaves were measured on the third day after topdressing, using both Polyphenol-chlorophyll meter (Dualex-4) and SPAD chlorophyll meter. The total N content of cotton leaves and the nitrate-nitrogen content in 0–20 cm, 0–40 cm and 0–60 cm soil layers were determined simultaneously.

Results With increasing N applicationrate, the total N content of cotton leaves and soil nitrate-nitrogen content increased significantly. The soil NO₃^–-N content in the 0–40 cm soil layer had the closest relationship with the total N content of cotton leaves. The NBI, Chl, and SPAD values of cotton leaves significantly increased with increasing N application rate, and were positively correlated with the total N content of cotton leaves. The correlation coefficients were all above 0.8. The average relative error (RE) of the measured and predicted values of leaf total N content was –4.0% (NBI), –3.1% (Chl), and –5.7% (SAPD), respectively. The NBI exhibited the highest prediction accuracy for leaf N content, with an r-value of 0.9143, average absolute percentage error (MAPE) of 6.91%, and the standard root mean square error (nRMSE) was 8.21%. The model determination coefficient (R²) of the diagnostic index with soil NO₃^–-N content was in the descending order NBI > Chl > SPAD. The NBI was the most sensitive to NO₃^–-N content in 0–40 cm soil layer, with r-value of 0.9116, MAPE of 14.11%, and nRMSE of 17.88%.

Conclusions When using the polyphenol-chlorophyll meter to diagnose N nutrition, we find that the N balance index (NBI) is sensitive and accurate in reflecting cotton leaf N content and the NO₃^–-N content in 0–40 cm soil layer. The average absolute percentage error (MAPE) between the predicted and measured values is as low as 6.91% and 14.11%, respectively, which could meet the requirement for rapid diagnosis of cotton N nutrition.