Effects Of Nitrogen, Phosphorus, And Potassium Fertilizers On Tobacco Internal Quality And Its Fertilization Model

Nitrogen, phosphorus and potassium are the major nutrient in tobacco cultivation. And precision fertilization of the three nutrients to achieve the best tobacco quality has been the key. Though some tobacco advanced fertilization technologies have been studied and promoted in China since the 1980s, there are still many problems. Such as the tobacco external quality and output value are always deemed to a standard to evaluate the fertilization effect; the dynamic imbalance of nutrient demand of tobacco and nutrient supply from soil; fertilizing mainly base on experience of years in tobacco farming, and the lack of reliable scientific basis for quantifying fertilization, and so on. With the development of tobacco agricultural modernization and precision agriculture, traditional agricultural production in the extensive fertilization methods can no longer satisfy the needs of high-quality tobacco leaf production, and it is urgent to establish the estimation standard of fertilization effect base on tobacco quality and to establish the fertilization model which base on the estimation standard.Two pot and two field experiments with the design of orthogonal regression unitized design were performed base on the spatial heterogeneity of soil nutrients with the classical method and geostatistics to study the effects of N, P, K fertilizers on output value and internal quality indexes in flue-cured tobacco leaves, which including single-factor analysis, the first order interactions analysis and the second order interaction analysis of N, P and K. And to establish the empirical model and the fertilization model of the relation of the internal quality indexes to nitrogen, phosphorus, potassium. So the main results were as follows:(1) Establishing an evaluation method of soil nutrient spatial variability. The spatial variability of available N, available P, available K, total N, total P, total K, organic matter content and pH value were evaluated according to the established methods in the two experiment fields. The results showed that the rate of spatial variability of available N, total N, total P, total K, organic matter content and pH value in 2006 were all less than 20%, and their area rate of relative uniformity were 100%. The rate of spatial variability of available P and available K were large than 20%, and their area rate of relative uniformity were individually 82.18% and 94.99%, which large than 80%. The rate of spatial variability of available N, available K, total N, total P, total K content and pH value in 2007 were all less than 20%, and their area rate of relative uniformity were 100%. The rate of spatial variability of available P and organic matter were large than 20%, and their area rate of relative uniformity were individually 91.34% and 95.70%, which large than 80%. Though the spatial distribution of soil nutrients in 2006 and 2007 was not absolutely uniform, the spatial variability was not significant, and the fields'fertility was relatively uniform. The two- year experiment results of orthogonal regression unitized design based on the two fields also indicated that the choice ways of experiment field and the evaluation method of soil nutrient spatial variability based on geostatistics was reasonable.(2) The output value increased first and then reduced under the conditions of 0 ~ 10 g pot-1 purity nitrogen in pot experiment and 0 ~ 105 kg hm-2 purity nitrogen in field experiment, and the total nitrogen and nicotine content of flue-cured tobacco leaves always increased, but the total sugar and reducing sugar content reduced, and the potassium and chloride ion content didn't significantly go up or go down in the same condition. The output value increased first and then slowly reduced under the conditions of 0 ~ 20 g pot-1 P2O5 in pot experiment and under field conditions of 0 ~ 210 kg hm-2 P2O5, but the total nitrogen, nicotine, total sugar, reducing sugar, potassium and chloride ion content didn't significantly go up or go down in the same condition. The output value increased first and then slowly reduced under the conditions of 0 ~ 50 g pot-1 K2O and under field conditions of 0 ~ 525 kg hm-2 K2O, and the potassium, total sugar and reducing sugar content of flue-cured tobacco leaves always increased, but the total nitrogen and nicotine content reduced, while the chloride ion content didn't significantly go up or go down in the same condition. So it was importance not to fertilize over potassium in tobacco cultivation although tobacco was a crop-K.(3) There almost was lutein andβ-carotene in carotenoids component of flue-cured tobacco leaves, which content distribution all ranged from 70 ~ 130μg g-1, and the lutein content slightly higher than the content ofβ-carotene. The amount of carotenoids including neoxanthin, lutein andβ-carotene were less than 250μg g-1. The carotenoids content accumulated significantly at the high N (under the conditions of large than or equal to 6 g pot-1 purity nitrogen in pot experiment and large than or equal to 67.5 kg hm-2 purity nitrogen in field experiment), high K (under the conditions of large than or equal to 18 g pot-1 K2O in pot experiment and large than or equal to 202.5 kg hm-2 K2O in field experiment), and moderate P with the ratio of N∶P2O5∶K2O at 1∶1.3∶3.0.(4)The content of neophytadiene and part of neutral aroma components which degraded from carotenoid increased, the content of neutral aroma components which classified to Maillard reaction products and which classified to cembranoid reduced, and the content of other neutral aroma components which classified to phenylalanine changed in the law was not significant with the increase of nitrogen and phosphors application under the condition of the fertilization in experiment. The content of neophytadiene and the neutral aroma components which degraded from carotenoid increased with the increase of potassium application. The content of neutral aroma components which classified to Maillard reaction products, which classified to cembranoid and which classified to phenylalanine reduced with the increase of potassium application under the condition of the fertilization in experiment.(5) There existed an interaction range between nitrogen and phosphorus, nitrogen and potassium, phosphorus and potassium. The synergistic and antagonistic effects between nitrogen, phosphorus and potassium were shown in the lower and higher range, respectively.(6) The correlation of the total sugar content and the total nitrogen, nicotine content was significant negative, of the total sugar content and the reducing sugar content was significant positive. The correlation of the total nitrogen and the reducing sugar content was very significant negative, of the total nitrogen and the nicotine content was significant positive. The correlation of the potassium content and the total sugar, reducing sugar was significant positive, of the potassium content and the total nitrogen, nicotine was significant negative. The correlation of the phosphorus content and the total nitrogen, nicotine content was significantly positive. The correlation of the lutein content and theβ-carotene content was significantly positive. The correlation of the potassium content and the lutein,β-carotene content was significantly positive, and the correlation of the chlorine content and the lutein,β-carotene content was significantly negative. The correlation of the lutein content and the nitrogen content was significantly positive. The correlation of the potassium content and the neutral aroma components content which classified to phenylalanine, which classified to Maillard reaction products was significantly negative, the correlation of the potassium content and the neophytadiene content, the neutral aroma components content which degraded from carotenoid was significantly positive.(7) The tobacco leaves'inherent quality of treatments in the experiment was scored by the fuzzy comprehensive evaluation, the weights of tobacco quality indexes were identified by analytic hierarchy process (AHP), and a membership function of tobacco leaf quality was established. Regression equations of the relation of tobacco quality index to nitrogen, phosphorus and potassium fertilizer were established base on the fuzzy comprehensive evaluation and then were optimized. And the optimal application in the pot experiment was 4.66~5.84 g pot-1 purity nitrogen, 8.03~10.47 g pot-1 P2O5, 22.79~28.46 g pot-1 K2O. A fertilization model of the nitrogen, phosphorus and potassium fertilizer containing the soil fertility contribution rate, tobacco varieties and other parameters was established based on the optimization of the regression equation of two years pot experiment and was verified by two years field experiment.The innovation points of this research were as follows:(1) Establishing a new choice ways of experiment field in agriculture and a new evaluation method of soil nutrient spatial variability based on geostatistics.(2) Establishing a new analysis method of the three-factor interaction base on frequency analysis. And the interaction of nitrogen, phosphors and potassium fertilizers on the carotenoids of flue-cued tobacco leaves were analyzed by this method.(3) Empirical models of the relation of tobacco quality index to nitrogen, phosphorus and potassium fertilizers were established, and then a fertilization model of the three fertilizers containing the soil fertility contribution rate, tobacco varieties and other parameters was established based on the empirical models.