Hernan R. Tejeda, Ramiro Gonzalez, Alejandro Cox. SQM Santiago, Chile

Transcription

1 Hernan R. Tejeda, Ramiro Gonzalez, Alejandro Cox SQM Santiago, Chile

2 Tobacco production in Chile Objectives of the study Flue Cured Tobacco Nitrogen Trial Effects on Yield and Quality Traits. Burley Tobacco Nitrogen Trial Effect on Yield and Quality Traits Summary and Conclusions

3 Limited tobacco production area, located in two regions Norte Chico ( South; West). Predominant soils: Alfisols and Inceptisols. Central Valley ( South; West). Predominant soil: Alluvial. Mediterranean climate: rainy winter, dry summer Winter rainfall increases north-south from 250 mm to 800 mm. Maximum Summer temp > 30 C. Minimum Summer temp < 12 C Relative humidity below 40%. All crop production under irrigation.

4 Tobacco production is part of intensive crop rotation systems that include cereals, legumes, pastures, sugar beet and other crops. Production of non-tropical fruits, including table and wine grapes, most of them for export markets, are important tobacco competitors for farm soil, irrigation water an other production resources. High fertilizer nitrogen rates have been successful in achieving top commercial yields in other crops, i.e., corn (> 20 mt grain ha -1 ), wheat (> 10 mt grain ha -1 ).

5 Overall objective: to determine maximum Flue Cured and Burley tobacco yield attainable by increasing fertilizer N rates under farming conditions: High solar radiation (PAR) Low relative humidity Low night temperature Controlled water supply. To evaluate the effect of high N rates on: fresh and cured leaf production, quality leaf traits: color, N content, nicotine, reducing sugar, grading. To evaluate the use of field instruments and laboratory determinations to assist in predicting tobacco yield and quality.

6 Materials and Methods (I) FLUE CURED N TRIAL BURLEY N TRIAL LOCATION Chimbarongo, San Fernando Nancagua, San Fernando PLANTING DATE October 22, 2002 November 8, 2004 PLANT DENSITY 1.15m x 0.45m pl ha m x 0.45m19324 pl h -1 EXPER. DESIGN RCBD, 7 treatment, 3 replic. RCBD, 7 treatment, 3 replic. TSP 145 kg P O ha -1 BASE FERTILIZERS TSP 104 kg P 2 O 5 ha -1 SOP 150 kg K 2 O ha SOP 291 kg K 2 O ha -1 CROPMANAGMNT & IRRIGATION Standard practices as by local CCT-BAT company. Standard practices as by local CCT-BAT company. TOPPING Manual, 83 Dat, +application of PRIME TM Manual, 70 Dat, + application of PRIME TM HARVEST February 27, to 22 leaves, 4 priming, a wk apart February 25, Plant cut above soil. CURING Grower s retrofitted barn Grower s barn

7 Materials and Methods (II) Distribution of N Fertilizer Treatments FlueCured Tobacco Trial Burley Tobacco Trial Total Fertilizer N Distribution Total Fertilizer N Distribution Fertiliz Transplt Sidedress [2] Fertiliz Transplt Sidedress [2] N [1] 20 Dat 34 Dat N [1] 21 Dat 35 Dat 49 Dat kg ha -1 kg ha -1 kg ha -1 kg ha -1 kg ha -1 kg ha -1 kg ha -1 kg ha -1 kg ha [1]: Urea [2] Sodium Nitrate.

8 Soil and Plant Analysis Inst. of Agric. Res., La Platina Exp.Sta., Santiago Soil fertility, one composite sample per replication Leaf samples, one per treatment, composite over 3 replications Quality Control Laboratory Compañia Chilena de Tabaco, CCT - BAT Nicotine Reducing Sugars

9 Leaf Color Reading Hydro N Tester (HNT) (similar to Spad-2 instrument) Measures light reflectance, associated to chlorophyll content 7 8 readings / leaf (L6 or L7), 1 leaf / plant, 4 plants/ treatment total 30 readings /treatment and replication. NO3- Concentration in Leaf Petioles Horiba Kardy portable NO3- Reader NO3- read in sap of petioles in same 4 leaves cut from plants after color reading.

10 Green leaf yield (kg ha-1) Flue Cured Tobacco fresh leaf response to N rates and predicted response by an exponential model Exponential Model: Y = (13836)*( N ) 0 < N < 187 R 2 = Small (+) difference in R 2 between the exponential and the quadratic model. - The exponential model gives better fit at the zero and maximum N rates. - A fresh yield increase of 87% over the zero N rate was produced by the maximum N rate (187 kg ha -1 ) N (kg ha-1)

11 Flue Cured tobacco leaf yield and total N% in leaves Leaves Leaf Trait (1) Fertilizer N rate (kg ha -1 ) Sampled Cured Yield [kg ha-1] Fresh Total N [% d.w.] (1): Mean of three replications. -Maximum N rate produced a 55% cured leaf increase over zero N, compared to the 87% increase by fresh leaf. - Total leaf N% increases linearly with respect to fertilizer N rates.

12 Flue Cured tobacco nicotine and reducing sugar in cured leaves Leaves Leaf Trait (1) Fertilizer N rate (kg ha -1 ) Sampled Cured Nicotine [%] (2) Reducing Sugar [%] (2) Red.Sugar : Nicotine (1): Mean of three replications. (2): Cured leaves sampled by treatment and priming. Mean nicotine for each N rate is the weighted average over 4 primings, leaf mass for each prime is the weight factor - Nicotine concentration stayed below 3 %. - Reducing sugar had a slight decrease at higher N rates. - The Reducing Sugar : Nicotine ratio decreases linearly with respect to N rates.

13 NO3 concentration in Flue Cured tobacco leaf petiole Regression equation - NO3 plot data Y = *N *N*Dat 0 < N < < Dat < 112 R 2 = NO3- in petiole (ppm) A steady decline in NO3 concentration over time (Dat) is observed since the first determination at 55 Dat, except for the highest N rates, where NO3 peaks at 62 Dat. N rates have a strong effect on NO3 concentration. Very low levels of NO3 are present at zero N rate Fertlizer N (kg ha -1 ) Days after transplant (Dat)

14 Mean yield of Burley fresh and cured leaves, and percentage of total N in leaves sampled 70 Dat. Fertilizer Total Leaf Yield N Leaf N Fresh Cured kg ha -1 % kg ha -1 kg ha ,121 2, ,374 2, ,622 3, ,530 3, ,816 3, ,897 3, ,169 4,286 Increase (%)

15 Cured Burley Tobacco Leaves (kg ha-1) Effect of fertilizer N on cured Burley leaf yield and total leaf N concentration Y = 7.58 (5.93)*( ) N 0 < N < 420 P< R 2 = N (kg ha-1) N (kg ha-1) Y = 6,365 - (4,271) *( ) N 0 < N< 420 P<0.001 R 2 = N concentration (%) and yield response to N rates follow a similar geometrical pattern represented by the exponential regression model.

16 Burley cured leaf yield (kg ha-1) Observed and predicted Burley cured leaf yield as affected by total N% in leaves Regression equation: Y = 6,882 (6,755)*(0.811)N% 1.68 < N% < 4.43 P<0.001 R 2 = Almost 90% of the observed yield variation of cured leaves can be accounted for by the exponential regression with N% as predictor variable. Under the experimental or similar conditions N% may provide an early indication of cured leaf yield Total N in leaves 70 dat (%)

17 HNT leaf color reading Effect of DAT and N rates on HNT leaf color readings HNT Leaf color reading 800 FLUE CURED TOBACCO 700 BURLEY TOBACCO N (kg ha -1 ) Y = *Dat *Dat *Dat*N 55 < Dat < < N < 187 kg ha -1 R 2 = 0.78 Days after transplant Y = *Dat *(Dat) *(Dat)*(N) 35 < Dat < < N < 420 kg ha -1 R 2 = 0.90 The common statistical model is: Y = A + B*Dat C*(Dat) 2 + D*(Dat)*(N) + e The differences are : (a) estimated coefficients, (b) range of the predictor variables.

18 Commercial Grading of Burley Tobacco Leaves Cured Burley tobacco leaves were graded grouped by N rate and leaf position at harvest. A numerical coefficient between 0.00 (no value) and 1.00 (maximum value) represents the grade. Total yield of graded leaf per N treatment is the sum of the weight of cured leaves at each plant position for the N treatment multiplied by the respective grading coefficient. Across all N rates, the grading resulted in a parallel down shift of the graded yield response with respect to the no graded response. As the N rate increases, the resulting grading decreases while the yield difference tend to remain within certain limits

19 Burley cured leaf (kg ha-1) Effect of Commercial Grading on Cured Burley Leaf Response to Fertilizer N Rates Regression equations, using a factor (grd) to separate graded (grd=1) from no graded (grd=0) leaves: Grd=0 Y = 6637 (4540)*( ) N Grd=1 Y = 6141 (4540)*( ) N 0<N<420 P < The difference between regression coefficients, D = = 496, significant at P<0.001, represents the separation of the 2 response curves. Equivalent yield reduction due to grading tend to be constant. As yield increases the fraction of low grade leaves decreases. Grd=0 Grd= Fertilizer N rate (kg ha-1) 400

20 The Flue Cured and Burley fertilizer trials demonstrated that tobacco yield can be substantially increased using high N rates, as it is the case with other farm crops in the region. In both trials the slope of the yield response at the highest experimental N rates shows that further yield increases may be obtained using even higher N rates. No negative effects of high N rates were observed on quality traits. Leaf color, total N %, nicotine and reducing sugar remained within accepted bounds in Flue Cured tobacco Burley leaf color, total N% and commercial grading were not affected by high N rates.

21 Leaf color, measured with Hydro N Tester (HNT) during the growing season, was affected by days after transplant (Dat) and nitrogen N rates (N) for both tobacco types. The shape of both 3-D relations was similar and could be described by the same regression model. Regression equations differed only in the value of regression coefficients and range of predictors (Dat, N). Grading cured Burley leaves resulted in two parallel regression equations for the response to N rates. The estimated vertical difference of 496 kg ha -1 between the curves is significant at P< The constant differences between response curves could be explained by the fact that at low yields (N rates) a higher fraction of leaves had low grades whereas at higher yields (N rates) a lower fraction of leaves had low grades.

22 Thank you!