DARK FIRE-CURED TOBACCO RESPONSE TO POTASSIUM RATE AND APPLICATION METHOD

Potassium is one of the essential nutrients required by tobacco (1,5), and tobacco takes up more potassium than any other nutrient (6). Tobacco is considered to be a luxury consumer of potassium, as it consumes this nutrient in excess of the physiological needs of the plant (1). Other research has shown that accumulation of fresh and dry weight was greater in tobacco plants that have adequate potassium compared to plants deprived of potassium (5), and burley tobacco market value increased with the application of potassium (2,9). Tobacco plants that are deprived of potassium will show deficiency symptoms including slight mottling and brownish yellow specks, usually near the margin on the tip of the leaf (6).

Previous studies in flue-cured tobacco reported that increased potassium rate had a positive effect on leaf quality (11,14). Other studies have shown no yield or leaf quality response with increased rates of potassium fertilizers in flue-cured tobacco (7,22). Mixed results have also been observed in burley tobacco with increased leaf quality (4) or no yield or leaf-quality response to increasing rates of potassium (8,10). Atkinson and Sims (2) found that potassium rate did not influence yield, but price of cured leaf increased as potassium rate increased in burley tobacco. The observations with no response to added potassium in flue-cured and burley tobacco have inferred that the response depends on level of residual soil potassium (8,22).

Kentucky’s tobacco cropland has had heavy applications of potassium fertilizers (15). Currently, at least one-third of tobacco fields soil tested in Kentucky have adequate potassium levels without adding potassium fertilizer (19). The University of Kentucky potassium (K₂O) rate recommendations for dark tobacco fall between 0 kg ha⁻¹ and 336 kg ha⁻¹, depending upon soil test results (18). Potassium sulfate has been documented to be a better choice than chloride-containing potassium fertilizers that may result in undesirable leaf characteristics (12,15).

Method of fertilizer application is important to ensure that nutrients are placed in a position that will maximize utilization. However, results comparing broadcast and banded potassium applications have been mixed, with band applications resulting in higher yields than broadcast applications (17), broadcast resulting in higher yields than banded in other research (20), and some research showing no significant yield differences between broadcast and band applications (3). Most previous potassium research has been with flue-cured and burley tobacco. The objectives of this study were A) to compare dark fire-cured tobacco yield and quality where potassium was applied at the recommended rate based on soil test K value, half of the recommended rate, or 50% more than the recommended rate; and B) to compare broadcast and banded application methods for potassium on dark fire-cured tobacco.

Field experiments were conducted in 2012, 2013, and 2014 at the West Farm of Murray State University near Murray, KY. Soil type each year was a Grenada silt loam (fine–silty, mixed, active, thermic Oxyaquic Fraglossudalfs) with 1.8% organic matter and pH 6.4 in 2012, 5.7 in 2013, and 6.4 in 2014. Dolomitic limestone (3,360 kg ha⁻¹) was added in early spring of 2013 to raise soil pH to more acceptable levels for tobacco. This soil type is representative of the Purchase area of western Kentucky, a major production region, and these test sites were selected because of soil test results that indicated low soil K values.

Test sites were prepared with the use of standard conventional tillage practices in all years of the study. Nitrogen was applied at 308 kg ha⁻¹ in all years according to soil test recommendations. Soil samples in the test sites were taken in March of each year and submitted to the University of Kentucky soil testing laboratory for analysis. Soil test P levels were 69 in 2012, 47 in 2013, and 70 in 2014, resulting in recommended rates of 56, 134, and 56 kg P₂O₅ ha⁻¹, respectively (18). Soil test K levels were 102 in 2012, 136 in 2013, and 212 in 2014, resulting in recommended rates of 325, 291, and 213 kg K₂O ha⁻¹, respectively (18). Nitrogen and phosphorus were broadcast applied as urea (46–0–0) and diammonium phosphate (18–46–0) to the entire site prior to transplanting each year according to soil test recommendations, and potassium treatments were applied to individual plots with the use of potassium sulfate (0–0–50). All broadcast applications were incorporated using a rotary tiller.

Plants of dark tobacco variety ‘PD7309LC’ were produced in a greenhouse float system with the use of recommended practices in all years. Experiments were transplanted in the field on June 12–13 in 2012, June 20 in 2013, and June 19–20, 2014. In all years, 102-cm row spacing with 81-cm plant spacing was used to achieve 12,100 plants ha⁻¹. Plots were 4 rows wide by 12.2 m long. The entire plot area was treated, but all data were collected from the 2 center rows of each plot.

The experimental design was a randomized complete block design with treatments replicated 4 times. Potassium application treatments consisted of a control that received no potassium, reduced rate (50% of recommended), recommended rate (100%), and an increased rate (150% of recommended) based on the soil test recommendation each year. Potassium rates were either broadcast applied at approximately 1 day prior to transplanting or band applied followed by immediate light cultivation approximately 7 days after transplanting. Potassium fertilizer treatments were applied by hand as either a broadcast or banded application. Broadcast treatments were made on June 8, 2012, June 19, 2013, and June 18, 2014. Band treatments were made on June 18, 2012, June 28, 2013, and June 27, 2014 with a band on each side of each row approximately 15–20 cm from the row. All other production practices were standard according to University of Kentucky Cooperative Extension Service guidelines (16).

Thirty tobacco plants from the center 2 rows of each plot were stalk-harvested in mid-October in 2012 and 2013 and late September in 2014 and housed in a fire-curing barn. Recommended fire-curing practices were used to obtain appropriate leaf characteristics for market (16). After curing, the tobacco leaves were removed from the stalk and placed into 3 stalk positions consisting of lug (lower stalk), second (midstalk), and leaf (upper stalk) and weighed to calculate yield per hectare. Tobacco was graded by a U.S. Department of Agriculture (USDA) grader according to USDA standards for dark fire-cured tobacco (21) and grades were assigned an index value between 1 and 100 (13). Grade index data are a weighted average of grade across stalk positions based on the grade received for each stalk position and the yield proportion of each stalk position contributing to total yield.

All data were subjected to analysis of variance (ANOVA) with the proc GLIMMIX procedure, and means were separated with the use of the Tukey-Kramer multiple comparison procedure at P  =  0.05 with SAS 9.4 (SAS Institute Inc., Cary, NC).

There were no visual differences in growth and vigor observed between potassium treatments during any growing season. Visual symptoms of potassium deficiency were mild and were seen only in control plots that received no potassium (data not shown). Data for 2012 and 2013 were combined and 2014 was analyzed separately because of differences in soil test categories for potassium recommendations.

Yield data for 2012 and 2013 were pooled, as there were no significant year by treatment interactions for potassium rate or application method. Data for 2014 are presented separately.

There was a significant potassium rate by application method interaction on total yield per hectare for 2012 and 2013 (Table 1). Within the banded application method, the 50% and 150% of the recommended rate of potassium (3,452 kg ha⁻¹ and 3,426 kg ha⁻¹, respectively) had significantly higher total yield than the 100% recommended rate (3,001 kg ha⁻¹). Excessive rainfall in 2013 resulted in some small areas of mild water damage within this experiment. Noticeably damaged plots were discarded and not included in data analysis; however, other plots that showed no visual symptoms of water damage but may not have reached full potential could have contributed to this result of the 100% rate, producing less yield than the 50% recommended rate. Although this difference was statistically significant, it is not likely biologically significant. Within the 100% recommended rate, broadcast (3,367 kg ha⁻¹) application of potassium resulted in significantly higher yield than banded application (3,001 kg ha⁻¹). There was no significant rate response within the broadcast application method and there was no significant application method response within 50% or 150% of recommended rates.

There were no significant main effects of potassium rate or application method in 2014, as shown in Table 2. This lack of a significant yield response to potassium even though soil test recommendation was for 213 kg K₂O ha⁻¹ was unexpected. However, one possible explanation may relate to properties of this soil type. The Grenada soil series, as well as some other soil types occurring in the Purchase area of western Kentucky, contain more montmorillonite clay than soils in other areas of Kentucky. Montmorillonite mineral can hold large amounts of exchangeable potassium while only fixing a small percentage (15). This type of clay also may hold large amounts of exchangeable potassium below the 15-cm depth where soil samples were taken, thus resulting in underestimation of the potassium fertility level of this soil (L. M. Murdock, personal communication). It is worth mentioning that the soil test results for 2014 had the highest soil test K of the 3 years in this experiment.

There were no significant effects of potassium rate and application method (P  =  0.2108) on quality grade index in 2012 and 2013, as shown in Table 3. In 2014, there was a significant main effect of potassium rate on quality grade index, as shown in Table 4. The 50% and 100% recommended rate had significantly higher quality (77.70 and 76.80, respectively) than the 150% recommended rate (69.80) when averaged over application method (P  =  0.0004). These differences were also seen in analysis of only the leaf grade index, which has the highest contribution to total grade index (data not shown).

After an analysis based on soil K index was performed, the overall response to potassium (averaged over all potassium rates and application methods) across all 3 years of this experiment became of interest (Table 5). Total yield per hectare was significantly higher in treatments where potassium was applied (3,385 kg ha⁻¹) compared to the untreated controls where no potassium was applied (3,200 kg ha⁻¹). In addition, quality grade index was numerically higher in tobacco that received potassium (65.70) compared to tobacco that did not receive potassium (59.60) when averaged over all years, rates, and application methods, but this quality difference was not statistically significant at the α  =  0.05 level (P  =  0.0993).

According to this research, potassium rate had some influence on dark fire-cured tobacco yield and quality grade index. There were no significant differences in yield of lug, second, or leaf stalk positions in response to potassium rate or application method and all significant yield responses were observed in total yield. Differences in quality grade index were observed in 2014, which was in the medium category for potassium soil test index. These results suggest that current potassium recommendations are at least adequate on low-potassium soils. These data also suggest that there is potential to reduce potassium rates below recommendations without sacrificing yield or quality of dark fire-cured tobacco. Additional studies on low-potassium soils are needed to clarify and refine potassium recommendations for dark tobacco.

The authors would like to thank Bobby Hill and Chris Rodgers for technical assistance in this research. Appreciation is also extended to field personnel at Murray State University for assistance in this research.