Final Report for ONE10-116
The challenge with growing onions on aging muck soils in the Northeastern United States is that the pH can become above the optimum for growing onions, which results in reduced yield and profitability, and environmental contamination. If this pH issue could be addressed via altered management practices, the production of onions on aging muck soils could be sustained and phosphorous (P) loading into water ways reduced significantly. In this study, we investigated the feasibility of applying acid forming fertilizers including mono ammonium phosphate (MAP) and manganese sulfate in a band at planting 2-3 inches below the seed, in order to reduce the pH and improve the availability of P and manganese (Mn). Foliar applications of Mn and P in combination with banding P and Mn were also evaluated. An observational field survey was conducted to better understand the relationship between pH, nutrient availability and yield under highly variable field conditions. Unfortunately, our results showed that banding acidic P and Mn fertilizers below the seed did not have any effect on lowering pH or improving yield. Foliar applications of 2 lb of Mn as manganese sulfate + manganese chloride 33% at the 2-3 leaf and 5-6 leaf stages resulted in an insignificant 30% increase in total yield only when soil Mn was less than 9 lb/A using Morgan extraction, which could increase net return by $1812 per acre. Foliar applications of P had no effect on yield when sufficient levels of P occurred in the soil. Greater than 50% of the sites tested had more than the total recommended nutrients available in the soil in late-July, indicating that there is much opportunity for growers to reduce their fertilizer rates of P and potassium (K) by 50 to 150 lb/A by applying only what is needed according to a soil test. Levels of available nitrate nitrogen (NO3) in the soil in mid-July were also generally very high indicating much opportunity for onion growers to reduce their nitrogen fertilizer inputs. Yield was not correlated to pH, % organic matter (OM), Mn or P as was expected. Instead, yield was most strongly correlated to soil NO3 and tissue levels of boron and zinc (Zn). Levels of K, calcium, magnesium, Mn and Zn were generally very high in the muck soils of Western New York. The nutrient complex in Edwards muck behaved differently than previously reported for muck soils; unlike in other high pH muck soils, Edwards muck had high %OM, and very high levels of available P, Mn and Zn. We will use the results generated from this project to guide further research studies and to make fine-tuned recommendations to onion growers that will help them to become more profitable and better stewards of the land.
The problem with growing onions on aging muck soils is that the pH is above the optimum for growing onions, which results in reduced yield and profitability, and environmental contamination. If the productivity of aging muck soils could be improved by addressing the pH issue via altered management practices, the production of onions on aging muck soils could be sustained and phosphorous loading into water ways reduced significantly.
In the Northeastern United States, New York State is the largest producer of onions ranking sixth in the nation and accounts for 97% of the Northeastern U.S. production, with approximately 10,000 acres and a 3-year average value of 45.5 million dollars. Approximately 80% of the onions are grown on muck soils. Muck or organic soil is defined as one that contains more than 20% and up to 80% or more organic matter. Muck soils are a non-renewable resource that was developed underwater by many generations of plants that were preserved under anaerobic conditions. It takes nature about 500 years to accumulate one foot of muck soil.
Muck soils are highly productive, because these dark soils that are high in organic matter and artificially drained give an early planting advantage and have increased water-holding capacity and availability, and reduced soil compaction, all of which result in higher yields of vegetables, especially onions, compared to mineral soils. Additionally, onions produced on muck soil are of superior quality compared to onions grown on mineral soil, because the high sulfur content of muck soils improves onion pungency and storability.
Muck soils are naturally acidic with the optimum pH for onion production ranging from 5.2 to 5.8. When the pH is above 6.0, manganese (Mn), zinc (Zn), boron (B) and phosphorus (P) are tied up and can become deficient and result in reduced yields. In New York, onions have been grown on muck soils for 85 years or more.
In spring of 2009, the Soil and Water Conservation Districts (SWCD) of Orleans and Genesee counties provided free soil nutrient tests to Elba muck land growers in Western New York. They worked with 8 growers and sampled 21 “fields” or “blocks” which were approximately 10, 25, 50 or 100 acres in size, and often consisted of several fields. Two to 20 sub-samples were taken per “field/block” for a total of 160 sub-samples. Results showed that 65% of the fields/blocks had pH higher than the optimum and 20% had pH between 6.7 and 7.3. The highest pH recorded in a single sub-sample was 7.6.
In this study, there was a correlation between organic matter (OM) and pH; as OM decreased, pH increased. However, it is important to note that of the 116 sub-samples where OM was greater than 40%, 41 or 35% had pH greater than 5.8. Thus, increased pH beyond what is optimum for growing onions on muck is not just a concern where the muck is “shallow” and OM is lower, but can occur on “deep” muck with high OM as well.
It is not surprising that the pH of inherently acidic muck soils has crept up above the optimum pH for onion production as the muck lands have aged. The muck lands in New York were originally drained for agricultural production in the early 1900’s and have now been under intensive vegetable production for 85 to 110 years. In several areas of New York the parent material that lies beneath the muck is calcareous marl, which is alkaline in nature. Muck soils are prone to subsidence, which is the permanent lowering of the surface elevation, a phenomenon resulting from the oxidation of soil OM by aerobic microorganisms, and by wind and water erosion. An estimated rate of soil subsidence on intensively cropped muck soil is one foot every 10 years. As muck soil has subsided over the decades, the growers are now farming deeper layers that are closer to the underlying calcareous marl where the pH is higher. Often, islands of shallower muck and higher pH occur within fields of otherwise optimum pH. When fields are tiled to improve drainage, calcareous marl is brought to the surface and mixed in with the muck, which also results in higher pH. Similarly, marl materials from spoil banks contribute to an increase in pH along field edges. Shallow muck with higher pH occurs in all areas of the state where onions are grown. The oxidation of muck results in the loss of OM, which in turn, results in higher levels of calcium (Ca) and magnesium (Mg), and higher pH.
Typically, soil pH may be lowered by adding sulfur. Unfortunately, soils that contain marl or extensive amounts of free limestone (calcium carbonate), as is the case in some of the muck lands in New York, require large amounts of sulfur, making it cost prohibitive to treat these soils for an effect that would last only temporarily. In a Michigan study, 10 ton of sulfur was added to marly muck and it only reduced the pH by 0.2!
To compensate for the higher than optimum pH, muck onion growers increase the rate of P fertilizer applied, and apply Mn and P to the foliage. Phosphorous fertilizer is applied broadcast and incorporated prior to planting, and when increased rates are used, it results in increased rates of P loading into the water ways. When excess P is applied to muck soil, it is either absorbed by the soil or lost through leaching. It may also find its way to waterways when it is attached to soil particles that are eroded by wind and water. Too much P in the water encourages growth of green algae, which causes eutrophication. In turn, when bacteria eat the algae, they use up dissolved oxygen, suffocating fish and other aquatic life.
A water monitoring project conducted by the SWCD of Orleans County and SUNY Brockport in 2008-2009 identified the Elba muck land as one of eight major sources of nutrient loading into the Oak Orchard water shed with excessive levels of total P (eg. 5 to 12 mg/L) and soluble reactive P (eg. 1 to 2 mg/L).
There is opportunity for growers to improve the productivity of onions grown on aging muck soils with high pH and to reduce P loading into waterways by improving nutrient availability and uptake, especially of P and Mn. This may be achieved by applying these nutrients in an acidic fertilizer band 2-3 inches below the seed at planting and with foliar sprays of Mn and P. A better understanding of the relationship between pH, nutrient availability and yield under highly variable field conditions is needed. The focus of this project was to evaluate, demonstrate and encourage the adoption of sustainable nutrient management practices of aging muck soils with high pH.
It is cost-prohibitive and impractical to reduce pH by applying sulfur to muck soils that contain extensive amounts of marl or free limestone (calcium carbonate), as is the case in many of the muck lands in New York. However, it may be possible to reduce the pH within the plant row by applying acid forming fertilizers such as mono ammonium phosphate (MAP) and manganese sulfate applied in a band at planting 2-3 inches below the seed to improve the availability of P, Mn, Zn and B. Using this strategy, all of the P will be applied to the target area (i.e. within the reach of plant roots) where the pH will be desirable for its availability (i.e. less than 6.0). Ultimately, the amount of P that escapes as a pollutant to waterways will be drastically reduced, because it will not be applied in the row middles where it is unavailable to the crop and free to escape into the environment. Over time, it very well may be realized that with more efficient use of P, rates can be reduced for banding compared to broadcast applications. Fertilizer prices of P have at least doubled within the last 4 years, a trend that is not expected to retreat, thus in addition to the obvious benefits to the environment, using lower rates of P will reduce input costs considerably in the long term.
When pH is greater than 6.0, Mn, P, B and Zn can become tied up and unavailable to the plant. Since B deficiencies are rare and Zn levels tend to be high in New York muck soils, Mn and P deficiencies were more likely to occur on were the focus of this project. Both Mn and P deficiencies result in slow growth, light colored foliage, delayed maturity and bulbing, and a high proportion of thick necks, all of which contribute to reduced yield and bulb quality. It is common practice for muck onion growers to apply Mn, and to some extent, P, as foliar sprays. These types of sprays are heavily marketed by fertilizer salesmen, but without any untreated controls, growers do not really know if they are getting a crop response. A crop response to these sprays would most likely occur on muck where the pH is high and P and Mn are low. Another question is whether foliar sprays of these nutrients alone without an acidified fertilizer band of P and Mn, would be sufficient to correct a nutrient deficiency. If they are, than this would be a simpler technique for growers to adopt compared to banding fertilizer, because banding requires modifications to be made to their planters. In this project, we evaluated foliar sprays of P and Mn with and without banding P and Mn on high pH muck soil to demonstrate if on their own they can induce a favorable crop response.
The soil nutrient results from a survey in the Elba muck land in spring 2009 showed that lower levels of Mn tended to occur where the pH was above 6.0. However, low levels of P were not always associated with high pH. Out of the 39 sub-samples that had high to very high levels of available P, 26 (= 67%) of them had pH of 6.0 or higher. This may be related to individual grower practices. To better understand the relationship between pH, nutrient availability and yield under highly variable field conditions, we monitored pH and available macro-, secondary and micro-nutrients in the soil, and compared them to plant health and yield in several fields over a wide range of soil pH. The data that we generated from this will allow us to fine-tune our recommendations derived from the fertilizer banding and foliar spray trials to different field situations.
To address objectives 1 and 2, we conducted three small plot research trials in commercial muck onion fields with three grower cooperators in Orleans and Genesee Counties in New York, each with slightly different soil type and fertility:
Trial Site No. 1 (CY Farms, Batavia muck): high pH, low OM, low P, low Mn
Trial Site No. 2 (Star Growers, Webster muck): high pH, high OM, high P, medium Mn
Trial Site No. 3 (Mortellaro, Elba muck): low pH, high OM, low P, high Mn
For the standard treatment, the full rates of N, P and K according to a pre-plant Cornell soil test, were applied broadcast and incorporated prior to planting. For the banding treatment, up to 100 lb of P and 4 lbs of Mn was applied 2-3 inches below the seed at planting. For this treatment, full rates of N (minus the amount applied in the band with the P fertilizer) and K were applied broadcast and incorporated. For these studies, N, P, K and Mn were in dry formulations of urea (46-0-0 NPK), mono-ammonium phosphate (MAP: 11-52-0 NPK), potash (0-0-62 NPK) and manganese sulfate 28%. See Table 1 for details. Foliar applications of Mn and P were also included in this trial. Treatments included: 1) broadcast NPK; 2) broadcast NPK + foliar Mn; 3) broadcast NPK + foliar P; 4) banded P+Mn with NK broadcast; 5) banded P+Mn with NK broadcast + foliar Mn; 6) banded P+Mn with NK broadcast + foliar Mn + P. Foliar Mn was applied at 2 lbs per acre as manganese sulfate + manganese chloride 33% liquid in 30 gpa at approximately the 2-3 and 5 leaf stages, using a CO2 backpack sprayer with 30 psi and 8005 nozzles. Similarly, foliar P was applied at 2 lbs per acre as Sol-U-Gro (12-48-8 NPK + micros) dry formulation in 30 gpa three times at the 2-3, 5 and 7 leaf stages.
The trials were set up as a randomized complete block (RCBD) design with 6 treatments and 5 replications. Each treatment replicate consisted of the width of an onion bed (5 or 6 feet wide) with 4 to 6 onion rows per bed by 20 feet long. The grower left the trial area unfertilized. For the trial, broadcast fertilizer applications were spread by hand and raked in, and banded fertilizer was applied with a push cone seeder set to a depth of 3 inches (Figures 1 & 2). Onions were seeded using the grower’s standard planting equipment, in-furrow fertilizer treatment and the onion variety that the rest of the field was planted to. Trials were established on April 21, 22 and 30 for Batavia, Elba and Webster, respectively.
To measure nutrient availability, composite soil samples were taken in mid-June at the 2-4 leaf stage and in mid-August and sent to the Cornell Nutrient Analysis Laboratory (CNAL) for complete nutrient analysis. In addition, samples were collected from within and between the plant rows for one broadcast (trt No. 1) and one banded (trt No. 4) treatment in mid-June, late June, late July and mid-August for analysis of pH. Soil samples were sent to CNAL for analysis in early June and mid-August and were analyzed in-house using the Cornell soil pH kit for the mid- and late-June, and July collections. Results from the mid-June and mid-August samples were compared between CNAL and in-house. In-house results were adjusted to match the CNAL results for the other two sampling dates. Leaf samples were collected for complete nutrient analysis of tissue at the end of July at the 9-10 leaf stage. The inner 4 leaves were collected from 10 randomly selected plants per plot and rinsed with running water before sending to CNAL for analysis.
Stand establishment was estimated at the 2-4 leaf stage in terms of percent of stand compared to the stand in the rest of the field in each row (4 or 6 per bed) per bed row per treatment replicate. To measure plant size, number of leaves and height of the tallest leaf per plant, were quantified from 10 randomly selected plants per plot three times throughout the growing season at the 2-4, 5-6 and 7-8 leaf stages in mid-June, late-June and mid-July, respectively.
To measure yield and bulb size distribution, the trials were harvested on August 22, September 8 and 16 at the Webster, Elba and Batavia sites, respectively. When onions had lodged, 50 plants per plot were randomly selected, pulled and removed from the field where they were windrowed for 4 to 7 weeks before being topped, weighed and graded into small (<1.25-2”), medium (2-3”) and jumbo (3-4”) size classes. Decayed and undersized (<1.25”) bulbs were culled.
To address objective 3, we conducted an observational survey where 5 fields/blocks were selected where the pH varied from 0.6 to 1.4 units based on the 2009 SWCD soil survey results (Table 2). Overall, fields were selected that ranged in pH from 5.0 to 7.3. In each field/block, three mini-plots each of low pH (“A” sites) and high pH (“B” sites) were selected for a total of 30 mini survey plots, attempting to stay within the same variety. Each mini survey plot consisted of 2 beds (10 or 12 feet) wide by 20 feet long. In mid- to late-July at the 7 to 9 leaf stage, composite soil samples were collected from each mini plot and sent to CNAL for complete nutrient analysis. At the same time, the inner leaves from 10 randomly selected plants were collected, rinsed with water and sent to CNAL for complete tissue nutrient analysis. Also, number of leaves and height of the tallest leaf per plant were quantified on 10 randomly selected plants per mini plot. All onion bulbs were pulled from 40 feet of row per mini survey plot on August 18 & 19 (Star), August 24 & 25 (LS 1 & 2) and September 2 (Panek & Mort) and removed from the field and windrowed for 4 to 7 weeks before being topped, weighed and graded into small (<1.25-2”), medium (2-3”) and jumbo (3-4”) size classes. Decayed and undersized (<1.25”) bulbs were culled.
Statistical differences among treatments was determined by General Analysis of Variance (ANOVA) with mean separation by Fisher’s Protected LSD test, ? = 0.05. For the observational field survey, significant relationships among variables including soil and tissue nutrient levels, soil pH, %OM, plant size and yield were determined by Pearson correlations, ? = 0.05.
For the evaluation of banding P and Mn and foliar applications of Mn and P, prices for urea, MAP, potash, manganese sulfate 28% dry, manganese 33% liquid and foliar P, 48% dry were quoted by representatives from Helena Chemical and Crop Protection Services. Keep in mind that fertilizer prices are always subject to change. The costs for materials for each of the six treatments on a per acre basis were calculated based on the rates used for each treatment at each trial location. The cost of application was not included in the analysis, because these would be the same regardless of broadcast or banded applications. Nitrogen and K were applied as a broadcast treatment and incorporated regardless of whether P + Mn was banded, and we assumed that in a commercial operation, the banding applications would be made at the time of planting, thus all treatments required the same number of passes across the field. However, in reality when growers customize their equipment to band fertilizer, this may involve two passes for some operations. Applications of foliar Mn and/or P were assumed to not require an additional pass across the field, because in a commercial operation they would be applied in combination with a pesticide application. When applicable, averages for yield and prices were collected from the New York State Department of Agriculture and Markets Statistics department, Annual Bulletin for Vegetables.
The three field trial sites provided a lot of variability with respect to pH, %OM and available P and Mn. The CY site in the Batavia muck was most representative of a situation where OM is lower (30%), pH is high (7.0), and available Mn (9 lb/A) and P (35 lb/A) are low. We expected that a crop response to banding acidified P+Mn fertilizer and foliar applications of Mn and P was most likely to occur at this site. The Mortellaro site in the Elba muck was most representative of a healthy deep muck where %OM was high (56%), except that the pH was a bit low (4.7, minimum pH should be 5.2), and Mn (98 lb/A) was definitely adequate. We expected the least, if any, crop response to banding P+Mn and foliar Mn and P at this site. Despite the low pH, we did not see a higher potential for toxicity for iron (Fe) and aluminum (Al) at this site. The Star site in the Webster muck was similar to the CY Batavia site, except with higher %OM (52%) and very high levels of P (362 lb/A). We did not add any P in the broadcast treatments, but did add 50 lb/A in the band treatments to achieve the acidifying effect of MAP. We expected to see a crop response to our Mn treatments in Webster. This site was also very high in calcium (Ca), which was 3 times higher than at the other sites. Unique to the Star Webster site was that this muck is Edwards muck, which was formed in herbaceous organic materials and overlies marl (calcium carbonate). Comparatively, the CY Batavia site is Bergen muck, which was formed in well decomposed woody and herbaceous organic material and overlies fine silty deposits, while the Mortellaro Elba site was in Carlisle muck, which was formed in woody and herbaceous organic material and overlies mineral deposits. According to Cornell, levels of K greater than 670 lbs/A is considered to be very high on muck soil and no additional K is recommended. Both the CY and Star sites had very high levels of K (760 lb/A & 940 lb/A) and additional K was only added at the Mortellaro site (50 lb/A). In general, K, Ca, Mg and Zn were very high at all three sites. According to Cornell, soil levels greater than 2000 and 500 lb/A are considered very high for Ca and Mg, respectively, and greater than 1 lb/A of Zn is considered very high potential for toxicity. Soil levels of these nutrients will be discussed further in the observational field survey results.
Seven weeks after planting when the onions were at the 2-3 leaf stage on June 15th, there were no significant differences among treatments for pH, and available Mn, P, K and NO3. The reason that Mn and K were lower than they were in the pre-plant test despite the application of these nutrients may be because the location of the trial ended up being in a slightly different location from where the pre-plant soil sample was taken. The finding that pH was 0.5 to 0.4 units lower compared to the pre-plant results may be due to the acidifying effect of applying fertilizer, or because the trial was in a different location. Notably, P and NO3 were higher than at pre-plant due to the addition of fertilizer. When the data was pooled across foliar treatments, there were no significant differences between broadcast NP and banded applications of P+Mn, although P and K were numerically higher where P+Mn were applied in the band. When nutrient availability was compared between rows and row middles, there were no significant differences, except for P, where there was twice as much available in the row than in the row middle in both the broadcast and banded applications. It makes sense that there would be higher P in the row where it was banded, but it is not clear why this occurred in the broadcast application. Numerically, NO3 was 143 and 158 lb/A higher in the row middles than in the row for broadcast P and banded P+Mn applications, respectively. This could be a reflection of N being taken up by the plants in the root zone in the row first.
On August 18th when the onions were nearing maturity, significant differences occurred between rows and row middles for pH, Mn and P. Where P+Mn was banded, the pH was significantly 0.1 units lower in the row compared to the row middle, but was still very high (6.5) and not significantly lower than when NP was broadcast. Available Mn was 0.4 to 1.8 lb/A higher than it was on June 15th. Where NP was applied broadcast, Mn was significantly 1.2 lb/A higher in the row than in the row middle, which is what we would have expected with the banded P+Mn. P was still significantly higher in the row than in the row middles for both broadcast NP and banded P+Mn applications. P was also significantly higher in the row where P+Mn was banded compared to when it was broadcast. Numerically, K was highest where P+Mn was applied in the band in the row. Surprisingly, available NO3 was still very high late in the season with more available in the row middles than in the rows.
During the growing season, there was no significant change in pH in the row compared to the row middle. On August 18th, where P+Mn was banded, the pH was significantly lower in the row than in the middle, but it was not significantly lower than where the NP was broadcast.
Overall, there was no change in pH when P+Mn was banded, available P was adequate in the row regardless of whether it was banded or broadcast, although it was odd that there was twice as much available P in the rows than row middles in the broadcast NP treatments, and, available Mn remained low and available NO3 was unusually extremely high at this trial site. Such high levels of NO3 are challenging to explain. Soil laboratories can sometimes make a mistake in converting the ppm rate to lb/A for muck soils, because the bulk density of muck is about half that of mineral soil. But, when muck is shallow, as is the case at the CY site, the conversion for mineral soils can be used. Even if the reported level of NO3 was halved, they would still be excessively high (250 to 300 lb/A), especially considering that only 100 lb/A of N was applied at planting. It remains unknown why such high levels of NO3 occurred throughout at this site.
Seven weeks after planting when the onions were at the 2-4 leaf stage on June 15th, there were no significant differences among treatments except for available P. When pooled across foliar applications, banded P+Mn had significantly 52 lb/A higher available P than broadcast N, which reflects the 50 lb/A of P that was added into the band. There were no significant differences between broadcast N and banded P+Mn for pH and available Mn, K and NO3. Similar to the CY site, compared to the pre-plant soil test results, 7 weeks after planting, pH decreased 0.1 to 0.2 units, Mn and K were lower and NO3 was much higher. It is unknown why Mn and P were lower at this site. When nutrient availability was compared between rows and row middles, there were no significant differences, except for available P where rows had significantly higher P than row middles for both broadcast N and banded P+Mn, a phenomenon which also occurred at the CY Batavia site. Perhaps, this is a reflection of P being eroded away from the row middles, while it remained intact in the rows where the plant roots are.
On August 18th when the onions were nearing maturity, there were no significant differences between rows and row middles for pH and available Mn, P, K and NO3. For available P, numerically, there was the same trend as in June with higher P when it was banded compared to broadcast, and in rows compared to row middles. Available NO3 was 34 to 120 lb/A lower than it was on June 15th, but was still very high.
During the growing season, there was no significant change in pH in the row compared to the row middle.
Overall, there was no change in pH when P+Mn was banded, available P was adequate in the row regardless of whether it was banded or broadcast and was generally higher where it was banded (no P was added broadcast), and available Mn remained at relatively low levels.
Seven weeks after planting when the onions were at the 3-4 leaf stage on June 15th, there were no significant differences among treatments except for available NO3. When pooled across foliar applications, broadcast NPK had significantly 103 lb/A more available NO3 and 33 lb/A more available P than banded PK + Mn. Unlike at the other two sites, Mn and K increased from the pre-plant soil sample. When nutrient availability was compared between rows and row middles, significant differences occurred for pH, and available Mn, P and NO3. The pH was actually 0.1 unit higher in the row compared to the row middle when NPK was broadcast and PK + Mn was banded. This was the opposite of what was expected. Available Mn was significantly higher in the row by 8 lb/A and 7 lb/A than in the row middle when PK+Mn was banded and NPK was broadcast, respectively. These results were exactly what we expected, except that the levels of Mn at this site were 14 to 43 times higher than at the other two sites and probably more than adequate. Clearly, Mn is much more readily available at the lower pH. It would have been interesting if we had a site where the pH was optimal. Unfortunately, the pH at this site was lower than we expected it to be based on the soil test results from 2009. Similar to the CY site, we observed significantly higher available NO3 in the row middles compared to within the rows, especially where PK was banded.
It was also fascinating that where the pH was the highest at 6.8 at the Star site in Webster, the soil level of P was 2x and 3x higher than it was at the Mortellaro and CY sites, respectively. This is counterintuitive, because P can become unavailable when pH is high. Such findings have never been reported on muck soils before.
On August 18th when the onions were nearing maturity, significant differences only occurred with available P where the rows had significantly higher P than the row middles when both NPK was broadcast and P+Mn was banded. Available P was higher where it was broadcast. Numerically, NO3 was still higher in the row middles than in the rows. Aside from iron (Fe) doubling from June 8-9 to August 18, we did not see any other major changes in Mg, Ca, Al, Fe and Zn between mid-June and mid-August at any of the other sites.
During the growing season, when pH was pooled across the trial, it did not change. There were also no significant differences between broadcast NPK and banded P+Mn, and rows and row middles.
Overall, we did not see the soil nutrient levels of available P and Mn, and pH respond as we predicted they would to the banding treatments at any of our trial locations.
No significant differences occurred among treatments for Mn and P in the leaf tissue of onion leaves collected at the 8-10 leaf stage. Banding P+Mn and applying foliar Mn and/or P had no effect on the content of these nutrients in the plant. There was a relationship between available Mn and P in the soil and the amount of these nutrients in the leaf tissue. The site that had the lowest available Mn in the soil pre-plant and throughout the trial, CY in Batavia, also had the lowest Mn in the leaf tissue, while the site that had the highest, Mortellaro in Elba, also had the highest levels of Mn in the leaf tissue. According to Millis and Jones (1996), the sufficiency range for onions at half maturity is 50 to 250 ppm. Using these values, Mn was sufficient at the Star and Mortellaro sites, but was deficient at the CY site. Tissue Mn was 2x and 3x higher at the Mortellaro site in Elba than the Star and CY sites, respectively. Soil levels of Mn on June 18 were 14x and 43x higher at the Mortellaro site than the Star and CY sites, respectively. These results suggest that Mn does not accumulate in onion plant tissue proportionately to the levels that occur in the soil.
According to Millis and Jones (1996), the sufficiency level for P in onion leaf tissue at half maturity is 3000 ppm. P can become less available when soil pH is less than 5.0, which was the case at the Mortellaro site (pH 4.6), but there were sufficient levels of P in the soil on June 18 (Table 5) and in the plant tissue (Table 6) at this site despite its low pH. At low pH, Fe can become available and tie up P, but both soil and tissue levels of Fe were low at the Mortellaro site. Tissue levels of P were sufficient and similar between the Star site in Webster and the Mortellaro site in Elba, which demonstrates that addition of P was not necessary at the Star site. Our results suggest that P does not accumulate in onion tissue proportionately to the levels in the soil, because tissue levels of P were similar at the Star and Mortellaro sites, despite the Star site having at least twice the levels of P in the soil (Table 4 & 5).
Interestingly, at the Star Webster site, the nutrient levels of Ca, Al, Zn, B and Na were about double the levels of these nutrients in the plant tissue of the onions from the other two sites. Levels of Mg were about half as much at the Mortellaro Elba site as they were at the other two sites. According to Millis and Jones (1996), the plant tissue sufficiency levels for K, Ca, Mg, Al, Fe, Zn and B are 35,000 ppm, 15,000 ppm, 2500 ppm, 60 ppm, 25 ppm and 25 ppm, respectively, and according to these values, all of these nutrients were slightly deficient.
At the CY Batavia site, no significant differences occurred among treatments in stand establishment, which ranged from 40.1 to 56%. For some unknown reason, 2 of the 3 beds across the width of this trial had very poor stand establishment. We think that the poor stand establishment was a function of the grower’s planting these two beds in a single pass (2 beds per pass with their commercial seeder), because it occurred only in the first two beds of our trial and not in the rest of the field. The growers planted our trial after they planted the rest of their field. The trial was arranged in blocks across the three beds and each treatment fell within each bed across the replications. This may have confounded our results to some extent.
Significant differences in the number of leaves per plant occurred at all assessment dates. On June 12, prior to the application of foliar nutrients, broadcast NP had significantly 0.3 more leaves per plant than banded P+Mn. On June 30, 2 weeks after the first foliar applications of Mn and P, there were no significant differences between broadcast NP and banded P+Mn, although numerically plants had slightly more leaves where NP was broadcast. Plants had significantly 0.6 and 0.4 more leaves after one application of foliar Mn to broadcast NP and banded P+Mn, respectively. One application of foliar P did not significantly increase number of leaves per plant when NPK was broadcast. Foliar Mn+P performed no different than foliar Mn alone when P+Mn was banded. On July 14th after two applications of foliar Mn and P, the number of leaves per plant was only significantly higher when foliar Mn was applied where NP was broadcast. This implies that when soil Mn is low (less than 5 lb/A), application of foliar Mn may increase plant growth when NPK is applied broadcast pre-plant.
Significant differences in plant height only occurred after two applications of Mn and P on July 14th where foliar Mn had significantly taller plants where NP was broadcast. Application of foliar P and banding P+Mn had no effect on plant height in this trial. Overall, onion stands were very thin and plants were very small throughout the season in this trial.
At the Star Webster site, when pooled across foliar treatments, the stand was significantly higher on June 16th where N was broadcast compared to the banded P+Mn treatment. This site had the highest stand establishment of all sites, which was 84 to 94%. There were no significant differences in the number of leaves per plant and plant height at any of the assessment dates, except for plant height on June 17th and July 14th. On July 14th banding P+Mn plus foliar applications of Mn + P had significantly the shortest plants. In general, banding P+Mn and applying foliar applications of Mn and P had no effect on plant size.
At the Mortellaro Elba site, there were no significant differences in stand among treatments, which ranged from 74% to 84%. Significant differences in the number of leaves per plant, only occurred on June 13th when the data was pooled across foliar treatments and broadcasted NPK had significantly 0.3 more leaves per plant than where P+Mn was banded. This was similar to the CY Batavia site. Perhaps applying 100 lb/A of P in the band may be slightly phytotoxic. Similarly, on June 16th when data was pooled across foliar treatments, plants were significantly 1.5 cm taller where NPK was broadcast compared to where P+Mn was banded. On June 25th where NPK was broadcast, plant height was significantly 6.7 cm taller where one application of Mn was applied compared to where foliar Mn was not applied. There were no significant differences when foliar Mn was applied where P+Mn was banded. A single foliar application of P had no effect on plant height. By the end of the season, there were no differences among treatments in plant height among treatments.
At the CY Batavia site, the NP broadcast treatments had numerically higher marketable yield than when P+Mn was banded and had significantly less small bulb weight and higher medium bulb weight. Numerically, foliar applications of Mn increased marketable yield by 30% where NP was broadcast and where P+Mn was banded. In the banded treatments, foliar applications of Mn + P increased marketable yield by 41%. Although not significant, there was a numerical trend that higher yields occurred in response to foliar applications of Mn. Tissue levels of Mn were found to be deficient at this site.
At the Star Webster site, yield was very low with approximately 90% of the bulbs falling into the small and undersized bulb classes with no jumbo sized bulbs. There was very heavy pressure from yellow nutsedge at this site, which undoubtedly contributed to the reduced yield at this trial site. There were no significant differences in yield or bulb size distribution. Numerically, banded P+Mn yielded slightly higher than broadcast N, probably due to the fact that there was no P added to the broadcast treatments. Numerically, the best treatment in this trial was banded P+Mn plus foliar Mn + P due to the highest medium bulb weight in the trial. Although not significant, these results suggest that the onions responded to the added 50 lb/A P despite excessively high amounts reported in the CNAL soil tests. It would have been interesting to see if there would have been a continued crop response to higher rates of added P. Perhaps the CNAL soil test is inaccurately determining the amount of available P at this site. Further research is required to understand the crop needs of P at this site with high pH and high soil P.
At the Mortellaro Elba site, significant differences occurred among treatments for % marketable yield. Banded P+Mn + foliar Mn had significantly the highest marketable yield in the trial, which was significantly higher than banded P+Mn alone and with foliar Mn. Numerically, there were very few differences among treatments and foliar applications of Mn and P, and banding P + Mn had no effect on yield and bulb size distribution at this site where %OM was high, pH was low, Mn was high and adequate amounts of P were available.
Across our 30 sample sites (10 field sites with 3 samples per field), pH ranged from 4.6 to 7.1 with 6 samples falling below the optimum pH (5.2) and 8 samples falling above the optimum pH (5.8) for growing onions (data not shown). Two and three fields had an average pH below and above the optimum pH for growing onions, respectively. During mid-season (July 15 and 28) when the onions were at the 7-10 leaf stage, significant differences occurred among field sites in pH, %OM and all available nutrients (NO3, P, Ca, Mg, Mn, Al, Fe) except K and Zn.
As a general rule, pH increases as %OM decreases, because OM tends to be acidic in nature. However, we did not find a significant correlation between pH and OM in this survey. Although the field that had the lowest %OM (LS 1 B: 19%) also had one of the highest pHs (6.2) in the survey, the site that had the highest pH (Star B: 6.5) also had a fairly high % OM (51%). A likely reason why this field had a high pH despite having high OM is that it is Edwards muck, which overlies marly materials that are alkaline in nature. In fact, sea shells were observed in pieces of marl that had been brought to the surface in shallower parts of the field. Alternatively, the LS and Panek fields were Palms muck, which overlies loamy deposits that are not alkaline in nature. The Mortellaro fields were in Carlisle muck, which are characterized by increasing acidity with depth. Generally, where the %OM was in the 50s, the pH was in the optimal range for growing onions. When the Star fields were removed from analysis, the expected correlation between pH and %OM was significant (R = -0.5503; p=0.0053).
There were significant correlations between pH and all available nutrients except K with the strongest relationships between P (R=0.8118), Mn (R = -0.7169) and Al (R = -0.7278). In addition to having significant relationships with pH, Ca, Mg and Zn also had significant relationships with % OM. Potassium was the only nutrient that had a significant relationship, although weak, with %OM that did not have a significant relationship with pH. As expected, as pH increased, Mn, Fe, Al and Zn decreased. As pH increased, NO3, P, Ca and Mg increased in this study.
Where the pH was below the optimum of 5.2, available Mn, Al and Fe were significantly higher than at all other sites. Numerically, Zn followed the same trend. When pH is below 5.2, these micronutrients increase in solubility and can reach toxic levels. Manganese toxicity can occur when soil levels are 70 to 100 lbs per acre, especially when Al and Fe are also high, which may have been the case in Mort A (Mn: 80 lb/A) and Mort B (Mn: 85 lb/A). Plant tissue Mn was strongly correlated with soil Mn (Table 10: R=0.8394; p=0.0000) and although Mort B had significantly higher tissue Mn than all other treatments, it fell within the sufficiency range (Table 10), and there was no significant correlation between soil Mn, tissue Mn and yield (data snot shown). These results suggest that soil levels of Mn that are 80-85 lb/A using Morgan extraction did not significantly affect onion yield. When Al + Fe exceeds 100 lb/A, which was the case in Mort A (130 lb/A) and Mort B (287 lb/A), it can tie up P making this nutrient less available. Soil levels of P at these sites were slightly lower than where pH was optimal. At the Mort A site where soil Fe was 202 lb/A (Table 10), tissue P was significantly the lowest in the study and deficient, and the yield was significantly lower than most sites where the pH was optimal (Table 11). There were no correlations between Fe in the soil and Fe in plant tissue or with these parameters and yield, soil P or tissue P. However, our results do suggest that when soil levels of Fe are greater than 200 lb/A, especially when Mn is also high, this can result in a P deficiency, which could affect yield. The Mort A site is in Carlisle muck, which overlies mineral deposits, which may explain a “hot spot” of Fe in this field.
When pH is greater than 5.8, Mn can be tied up and become deficient. In our survey, available Mn was numerically about one third lower at the sites where pH was above 5.8 compared to where it was optimal, but the levels of Mn in leaf tissue were not different than where the pH was optimal (Table 10). Although 2 of the 3 lowest yields in the study occurred where soil Mn was low, there was no correlation between soil Mn and yield. Our results suggest that soil levels of Mn that are 30 lb/A or higher are sufficient.
Phosphorus, Ca and Mg can be tied up when pH is below 5.0 or above 6.0. Generally, the availability of these nutrients was significantly lower where pH was below 5.3 compared to where pH was optimal. However, P, Ca and Mg were generally significantly higher where pH was above 5.8 compared to where it was optimal. The field that had the lowest Ca and Mg, LS 1 B, had a pH of 6.2, which is most likely related to this field having shallow muck and consequently the lowest %OM, because Ca and Mg build up as OM decomposes. The reason that the Star fields had the highest levels of Ca and Mg, despite having high pH, was likely a result of the Edwards muck overlying calcareous marl. The reason why the Star sites had such high levels of P (383 & 502 lb/A) despite having such high pH is unknown, and has not been previously reported for muck soil. Normally, high levels of nutrients are due to past nutrient practices. However, Star growers do not have a history of applying excessive rates of fertilizer. Nevertheless, there is still opportunity to reduce rates of P fertilizer applications at this site, which was demonstrated in the small-plot trial that was located at this site. Fortunately, high levels of P are not toxic, but can reduce the uptake of Zn and Mn. Where soil levels of P were high at Star A and B, plant tissue P, Mn and Zn were also all comparatively high (Table 10), which indicates that uptake of Mn and Zn was not negatively affected by the high levels of P in the soil.
Soil levels of Ca and Mg greatly exceeded the sufficiency levels described by Cornell (1996) (Ca: 2000 lb/A; Mg: 200 lb/A) by 9-23x and 7-15x, respectively. It is normal for this to occur on muck soil, because these nutrients naturally build up over time. As organic matter is decomposed, Ca and Mg are released and levels may nearly double over a 10 year period. With the Elba muck land having been farmed for almost 100 years, and the underlying material in some areas containing marl, such high levels of Ca and Mg actually make sense. High levels of Ca and Mg are not toxic, but can reduce the availability of K, but in this study, soil levels of K were also very high and there were no significant differences in tissue K or relationship to yield (Table 10).
Soil levels of NO3 were generally very high for mid-season in all of the fields and ranged from 110 to 502 lb/A. A 450 cwt/A onion crop will mine an average of 75 lb/A of NO3 from the soil and Cornell recommends applying 100 to 120 lb/A of N annually. It is doubtful that the growers applied more than 150 lb/A of N. Either there is an unknown source of N or CNAL made an error converting ppm to lb/A. In mineral soil, this conversion is 2x, based on the fact that on average, there is about 2,000,000 lbs per acre furrow slice (6.67 inches deep) over an acre. For muck soils, on average there is 1,000,000 lbs per acre furrow slice, although this may vary from 500,000 to 1,500,000. Therefore, a soil test in 1 ppm is equal to 1 lb/A and the conversion should be 1x. If this were the case, then the actual levels of NO3 would have been 55 to 226 lb/A, which seems more reasonable. This may explain why the levels of some of the other nutrients, particularly P, K, Mn and Zn were also unrealistically high. The 2x conversion rate may be more appropriate for shallow muck soils (< 18 inches), and the 1.5x rate for medium muck soils (18-48 inches). A weak correlation occurred between soil N and tissue N (Table 10: R=0.4276; p=0.0184) and between soil N and yield (Table 11: R=-0.4585; p=0.0161), which suggests that as soil N increases, yield decreases. The two Star sites had significantly the highest soil N in the study, which also had significantly the lowest yield (Table 11), which suggests that levels of NO3 higher than 380 lb/A may be phytotoxic.
Significant differences in mid-season plant tissue nutrients among fields occurred for all nutrients except for N, K and Al. According to Maynard and Hochmuth (1997), in our survey, N and Cu were within their sufficiency ranges (N: 2.5-3%; Cu: 5-10 ppm), while P (sufficiency 1000-2000 ppm), Mn (sufficiency 10-20 ppm), Zn (sufficiency 15-20 ppm) and B (sufficiency 10-25 ppm) were generally above their sufficiency levels. According to Millis and Jones (1997), P, Mn, Zn and B were generally within their sufficiency ranges (P: 3100-4500 ppm; Mn: 50-250 ppm; Zn: 25-100 ppm; B: 25-75 ppm), while N, K, Ca, Mg, Fe and Cu were generally below their sufficiency ranges (N: 4.5-5.5%; K: 3.5-5.0%; Ca: 1.5-2.0%; mg: 2500-4000 ppm; Fe: 60-300 ppm; Cu: 15-35 ppm).
The Star A field (pH 5.9) had significantly higher P than any other field (6224 ppm), followed by Panek A and B (pH 5.3), which had significantly higher P (A: 4619 ppm; B: 4951 ppm) than the other fields. Even though these levels exceed the sufficiency range, P is not known to cause toxicity. The lowest level of tissue P (2523 ppm), which was also deficient, occurred in one of the sites where the pH was below optimum (Mort A), which also had very high soil levels of Fe (202 lb/A). The lowest tissue level of Ca occurred in LS 1 B which had the lowest %OM and pH 6.2, which was not significantly different than the Mort A (pH 5.0), Mort B (pH 4.8) and Panek B (pH 5.3) sites. The highest tissue level of Ca occurred at the site with the highest pH, Star B, which was statistically the same as Star A (pH 5.9) and LS 2 A (pH 5.6). The highest tissue level of Mg occurred at the LS 2 B site (pH 5.8), which was not significantly different than LS 2A (pH 5.6), Star A (pH 5.9) and Mort A (pH 5.0). The site that had the lowest pH (4.8) had significantly the highest level of tissue Mn (137 ppm) than all other sites. LS 1 B (pH: 6.2; Mn: 28 ppm) and LS 2 B (pH: 5.8; Mn: 43 ppm) were deficient in Mn according to Millis and Jones (1996). LS 2 A (pH 5.6) had the highest level of tissue Fe in the trial, which was not significantly different than Mort A. LS 2 A was the only site where Fe fell within the sufficiency range, all other sites were deficient according to Millis and Jones (1996). Star A (pH 5.9) had the highest tissue level of Zn, which was not significantly different than Star B (pH 6.5) and Mort A (pH 5.0). Interestingly, the Star sites had some of lowest soil levels of Zn. LS 1 B, LS 2 B, LS 2 A and LS 1 A were slightly deficient in Zn according to Millis and Jones (1996), and LS B 1 was the only field that had lower levels of Zn in the soil. Tissue levels of B were deficient according to Millis and Jones (1996) and significantly lower in fields where the pH was below 5.0 (Mort A & B) and where %OM was below 20% (LS 1 B). Mort B (pH 4.8) had the lowest level of tissue Cu, which was not significantly different than Mort A (pH 5.0), Panek B (pH 4.7) and Star B (pH 6.5).
Significant positive correlations occurred between the levels of nutrient in the soil and in plant tissue for N, Ca, Mg, Mn and Al with Mn (R= 0.8394) and Ca (R= 0.8218) having the strongest relationships. Significant positive correlations occurred between soil pH and leaf tissue levels of N, Ca, Mn and Cu. The strongest significant correlation was negative between soil pH and the level of Mn in tissue (R= -0.7020). Although weak, significant positive correlations occurred between %OM and P, Ca, Zn and B.
There were no significant correlations with the number of leaves per plant and nutrient levels in the soil or leaf tissue, except for K (R=0.4666; p =0.0141) and Ca (R= -0.3927; p=0.0427) with nutrient levels in the soil, and Ca (R = -0.4689; p=0.0136) and Mg (R = -0.5036; p=0.0047) with leaf tissue (data not shown). As soil Ca, tissue Ca and tissue Mg increased, the number of leaves per plant decreased. There were no correlations among number of leaves per plant and soil or tissue Mn or P or pH (data not shown). There was a positive correlation with the number of leaves per plant and medium bulb weight (R= 0.6059; p=0.0008) and a negative correlation between leaf number and small bulb weight (R= -0.7266; p=0.0000). The only significant correlation with plant height was with total marketable yield (R= 0.4010; R= 0.0382).
Significant differences occurred among fields for number of leaves per plant and plant height (Table 11). Generally, plants that had the most leaves were where pH was 5.3. In field LS 1, plants had the same number of leaves and no significant difference in plant height where pH was 5.3 and 6.2. In the Star field, plants had significantly more leaves and were significantly taller where pH was 6.5 compared to where it was 5.9, which could be due to a difference in planting date.
Significant differences occurred in yield, and small, medium and jumbo weight, and undersized culls. Yields were average or above the state average yield (360 cwt/A) at the majority of sites. In general, the highest total marketable yields and above state average yields occurred where the pH was optimal. LS 2 B (pH 5.8) and LS 1 A (pH 5.3) yielded the highest and were significantly higher than all fields with below and above optimum pH. No significant differences in yield occurred within fields, except for LS 1, which had significantly lower yield where the pH was 6.2 compared to where it was 5.3.
The most significant relationship occurred among total marketable yield and medium bulb weight (R= 0.8364; p=0.0000) (data not shown). There were no significant relationships among yield and grade with pH, %OM, Mn or P, except for a weak relationship between soil P and small bulb weight (R=0.4450; p=0.0200) and tissue P and undersized cull weight (R=0.4870; p=0.0100). Here, bulb size decreased as P increased, which may be a function of excessive levels of several nutrients and a high pH at the Star fields, and heavy weed pressure. The Cornell soil sufficiency tables state that >220 lb/A of P is considered “very high” and all of the fields with above optimal pH had soil levels of P that exceeded this level (227, 266 & 318 lb/A – Table 9). Normally, P can be tied up when soil pH is greater than 6.0. In the Star fields, levels of P were very high in spite of no additional P being applied in the spring, except for 6 lb/A of starter fertilizer applied in the furrow at planting. The growers did not apply P fertilizer, because the soil test indicated that levels were already high. Tissue levels of P were also high at these sites. It is unknown why such high levels of P occur on this farm.
Significant correlations occurred between medium bulb weight and tissue B (R= 0.4709; p=0.0132), soil NO3 (R= -0.4585; p=0.0161), tissue Zn (R= -0.4752; p=0.0122) and soil Zn (R= 0.6690; p= 0.0001). In this study, there appeared to be increased yield when tissue B was >30 ppm (Table 10 & 11). Where tissue B was lower than 30 ppm, which also corresponded to where the pH was not optimum, yields were less than 400 cwt/A. Onions are generally not thought to be very responsive to B as they are generally tolerant to toxicities and deficiencies are seldom. However, B is subject to leaching loss with high rainfall. Deficiency may also occur under dry conditions, because B is moved to the root zone in water. The CNAL does not offer B analysis as part of their standard soil test, but it would be interesting to compare levels of B in plant tissue to the levels in the soil, and whether plants would respond to fertilizer applications of B.
Interestingly, as soil NO3 increased, medium bulb weight (and yield) decreased. Cornell recommends a maximum of 125 lb/A of nitrogen per year to grow an onion crop. The majority of the fields had higher levels than this with those where pH was above optimum having 2x to 4x higher levels of NO3 available in the soil (Table 9). Although the conversion factor from ppm to lb/A for muck soils may be incorrect, even if such an error was made, some of these fields still appeared to have very high levels of soil N (55 to 251 lb/A), and the feasibility of reducing nitrogen fertilizer rates warrants further investigation.
A similar negative relationship occurred with Zn where the three lowest yields occurred where tissue Zn was 30 ppm or greater (Star A & B and Mort A). Alternatively, yield increased as soil Zn increased. Unfortunately, there was no relationship between levels of Zn in the soil and in plant tissue. Interestingly, we observed the highest tissue Zn in the fields with the highest pH (Star A & B), where we expected lower levels, because Zn becomes less available when pH is greater than 6.0. Zinc can be easily built up and maintained in muck soils over time. In addition to fertilizer applications, Zn is added to muck soils via fungicide use, specifically EBDCs like mancozeb and maneb, which are applied directly to the soil as an in-furrow treatment for control of onion smut, and as a foliar application for control of leaf diseases, especially downy mildew.
Other minor, yet significant correlations that occurred were between weight of bacterial bulb decay and tissue Cu (R= -0.4169; p=0.0305), and between small bulb weight and soil Mg (R= 0.3905; p=0.0440) and tissue Mg (R=0.4862; p=0.0101). Copper bactericides are used to manage bacterial bulb decay and the plants with the highest levels of Cu in their tissue also had the least bacterial bulb rot at harvest. Unfortunately, CNAL does not include Cu in their standard soil test, so we do not know if there was a correlation between the levels of Cu in the soil and leaf tissue. According to the Cornell sufficiency tables, Mg is considered very high when it exceeds 200 lb/A and the levels of Mg in this study were 7x to 16x higher than that, especially in fields with high pH such as in the Star fields. The site with the highest yield, LS 2 B, also had one of the highest levels of soil Mg in the study, which was not significantly different than the level of soil Mg at the Star sites, which had the lowest yields in the trial. Similarly to Ca, Mg builds up over time and these high levels of Mg are reasonable and not harmful.
Based on our results, sufficiency levels for tissue N appeared to be in-line with Maynard and Hochmuth (1997) at 2-3%, while Millis and Jones (1996) sufficiency levels for P of 3100 to 4500 ppm, Mn of 50 to 225 ppm, and B of 25 to 75 ppm, related best to our studies. We found Millis and Jones (1996) sufficiency levels for K, Ca, Mg, Fe and Cu to be too high, as we did not find any yield reductions at our sites where these nutrients would have been deficient, according to them. Since there were no correlations between these nutrient levels in leaf tissue and yield, levels of 29,362 to 32,950 ppm of Ca, levels of 10,974 to 18,154 ppm of K, levels of 1700 to 2437 ppm of Mg, levels of 27 to 54 ppm of Fe, levels of 9 to 24 ppm of Al, and 3.5 to 5.9 ppm of Cu appeared to be sufficient in this study. For tissue Zn, we found Millis and Jones (1996) sufficiency levels of 25 to 100 ppm to be too high, as we had reduced yields when tissue Zn was 30 ppm or greater. Maynard and Hochmuth (1997) tissue Zn sufficiency levels of 15 to 20 ppm were more fitting.
Similarly, soil levels of Mn up to 85 lb/A did not cause toxicity, while levels of 30 to 36 lb/A were definitely sufficient, and soil levels below 20 lb/A may become deficient. In the small plot trials at Mortellaro’s in Elba muck, soil Mn levels as high as 135 lb/A resulted in sufficient tissue Mn and no yield reductions. When soil Zn was about 20 to 25 lb/A, it appeared to be sufficient, but when it was less than 20 lb/A, it appeared to become deficient. These values are much higher than the sufficiency levels of Zn for mineral soils, where greater than 1 lb/A of Zn is considered high. Soil Al and Fe is this study were in line with reported medium toxicity potentials of 50 to 100 lb/A as the levels that we observed were generally lower than 50 lb/A and we did not observe any toxicity. Seemingly excessive levels of Ca (20,000 to 30,000 lb/A) and Mg (2000 to 3000 lb/A) are normal for muck soils that have been in production for about 100 years.
- Table 8. Yield and bulb size distribution for small-plot field trials.
- Table 10. Plant tissue nutrient levels for observational field survey.
- Tables 3,4 & 5. Soil nutrient analysis results for small-plot field trials.
- Table 7. Stand and plant size for small-plot field trials.
- Table 6. Plant tissue nutrient analysis for small-plot field trials.
- Table 9. Soil nutrient levels for observational field survey.
- Table 11. Plant size, yield and bulb size distribution of observational field survey.
This project was the first of its kind. Never before has such a comprehensive study been undertaken to improve the productivity of growing onions on aging muck soils with high pH. Never before in New York has the practice of banding P been seriously investigated. Unfortunately, our results showed that banding acidified fertilizers, specifically, mono-ammonium phosphate (MAP: 11-52-0 NPK) and manganese sulfate + manganese chloride 33% 2 to 3 inches below the seed did not reduce the pH of the soil within the onion row, and had no effect on nutrient availability, crop growth or final yield. Since this practice did not improve yields of onions grown on three distinct muck types in Western New York, growers did not proceed with modifying their planting equipment to place fertilizer below the seed row. One of our grower cooperators saved an estimated $4000 to $6000 in parts and labor expenses to modify a 12-row seeder to band fertilizer. Additionally, this grower will save the annual expense of $8.58 per acre in manganese sulfate, which would total $367 per year on their 43 acres of high pH muck that is cropped to onions.
Foliar applications of Mn resulted in a slight increase in total yield by 30 to 42% only where soil levels of Mn were very low (3 to 4 lb/A) and tissue levels of Mn were deficient (less than 50 ppm). We also observed a significant correlation between soil pH and soil Mn; as soil pH increases, Mn decreases. Thus, it is important for growers producing onions on muck soils with a pH above 5.8 to closely monitor their soil Mn, and to make foliar applications of Mn when soil Mn is less than 9 lb/A according to our study. Making two applications of foliar manganese when needed could increase economic return by $1812 to $2557 per acre when the cost of Mn is considered ($49 per acre), assuming average yields and prices. On CY Farms, this could total $77,916 to $109,951 on their 43 acres of onions grown on high pH muck, assuming that Mn is low on all acreage.
Alternatively, foliar applications of Mn had no effect on yield where soil Mn was greater than 9 lb/A and tissue levels of Mn were sufficient (50 to 250 ppm). In the observational field survey, there was no correlation between soil Mn or tissue Mn and yield, and yields were average or above average when soil Mn was greater than 20 lb/A, and tissue Mn was sufficient. Thus, growers whose onion crops are not likely to respond to foliar applications of Mn (e.g. soil Mn is greater than 20 lb/A) can save $49 per acre by not making these unnecessary applications. At the Star and Mortellaro trial sites, foliar Mn represented 43% and 22% of their fertilizer costs, respectively.
In our small-plot research, we did not observe a crop response to foliar applications of P when the recommended amount of P was applied at planting. Even at the CY site, where tissue P was deficient, we did not see an increase in yield when foliar applications of P were made. Therefore, we do not recommend growers make foliar applications of P, which can save $22.50 per acre. At the CY, Star and Mortellaro sites, this represented 10%, 26% and 12% of their total fertilizer expenses, respectively.
According to Cornell, no additional P and K is required when pre-plant soil test results show > 220 lb/A and > 670 lb/A, respectively, for muck soils. Of our three small-plot trial sites, one site (Webster muck) did not need any additional P, and only one site (Mortellaro) needed a small amount of K (50 lbs). In our observational field survey, 5 of the 10 (= 50%) fields had soil levels of P > 220 lb/A in mid- and late-July, while the rest of the sites had > 100 lb/A of P. All 10 fields (=100%) had > 670 lb/A of soil K in mid- to late-July and three fields had more than double the sufficient level. These results strongly suggest that most muck onion growers can reduce their spring applications of P and K fertilizer. Reducing P and K fertilizer rates by 50 to 100 lb/A could result in savings of $17 to $34 per acre per nutrient. Applying only the necessary amount of P fertilizer according to a soil test will also greatly reduce P loading into the waterways, thus reducing water pollution and eutrophication.
The nutrient complex in the high pH Edwards muck in Webster behaved differently than previously described for muck soils. First, the typical relationship between pH and %OM, pH increases as %OM decreases due to the acidic nature of OM, did not exist here. Instead, the pH ranged from 5.9 to 7.0, despite having greater than 50% OM. This was likely related to the Edwards muck overlying calcareous marl, which is alkaline in nature.
Second, typically, soil P becomes less available when pH is greater than 6.0 in muck, but in Webster, soil P in the spring was 7x to 10x higher than it was in the Elba and Batavia mucks, respectively. In fact, with 325 lb/A of P available prior to planting, no additional P was needed. The highest levels of soil P in both our small-plot trials and the observational field survey occurred in the Webster muck, which corresponded to the highest levels of P in leaf tissue. Typically, high levels of soil P occurs due to several years of excessive fertilizer applications. Since the grower in Webster does not apply P excessively, it remains unknown why such high levels of P occur in this muck pocket. In our small-plot trial, a yield increase of 57% in medium bulbs and 8.3% in total yield did occur where 50 lb/A of P was banded in the row at planting compared to the broadcast treatment, where no P was applied. This result suggests a benefit of applying a low rate of P in the spring even in soils that have adequate P, because P is less available in early spring due to cold soils and the onion plant roots being too small to mine nutrients from the row middles. In this case, the cost of 50 lb of P, $32.47 per acre, resulted in up to $515 per acre in increased yield. Further research is warranted to optimize the rate and application method (band vs. broadcast vs. furrow) of P at planting in soils with high P.
Third, in addition to P, soil levels of K, Ca, Mg, NO3 and Zn were also very high and the highest of the small-plot trials and in the observational field survey in the Webster muck. There were also correlations between soil and tissue levels of all of these nutrients, except for K and Zn. Levels of Zn were also unusually high, because Zn also becomes less available when pH is greater than 6.0. It is unknown why such high levels of nutrients occurred at this site.
In the observational field survey, yield was not related to pH, %OM, Mn or P as we expected. Instead, yield was most strongly correlated to soil NO3 and tissue levels of B and Zn. Interestingly, in the observational study, yield decreased as soil NO3 increased. Although there may have been some issues with the accuracy of the CNAL test results for NO3, the majority of the sites from both the small-plot trials and the observational study had 2x to 5x more than the maximum of 125 lb/A of N that Cornell recommends per year for onions. The feasibility of reducing nitrogen fertilizer rates warrants further investigation. Yield increased as the tissue level of B increased; in these studies, it appeared that yield was reduced when leaf tissue B was < 30 ppm. Alternatively, when tissue levels of Zn were > 30 ppm, yield was reduced.
The results of these extensive studies enabled us to fine-tune our sufficiency values for soil and tissue levels for some of the nutrients that will allow growers, Extension professionals and crop consultants to improve fertility recommendations for onions grown on muck soils in Western New York (Table 12). The sufficiency levels of nutrients in plant tissue described by Maynard and Hochmuth (1997) were reasonable for N and Zn using the CNAL test, while the sufficiency levels described by Millis and Jones (1996) were reasonable for P, Mn and Cu. We found Millis and Jones (1996) to be too high for K, Ca, Mg, Fe and Cu, because according to their levels, these nutrients would have been deficient in most of our samples, and we did not find any correlation with the levels of these nutrients in plant tissue and yield. We also identified that higher levels of Mn and Zn can occur in muck soils without any risk of toxicity.
Since the CNAL reported such high soil levels of several nutrients, we sent three soil samples from the observational study representing below optimum, optimum and above optimum pH for growing onions on muck soils to 5 different soil laboratories, including CNAL, Agro-One (Ithaca, NY: conducts commercial analysis for Cornell), Michigan State University (East Lansing, MI), Spectrum Analytic (Washington Courthouse, OH) and A&L Analytics Lab (Memphis, TN). Results with respect to pH, %OM, P and K were very similar between CNAL and Agro-One, but varied widely among the other laboratories. For the same soil sample, P and K ranged from below optimum to very high and recommendations ranged from 0 to 125 lb/A for P, and 0 to 265 lb/A for K (data not shown). Similarly, lime recommendations ranged from 0 to 3.8 ton/A. Since the sufficiency ranges that Cornell recommends were based on the soil tests conducted at CNAL, it was recommended that onion growers in New York send their soil samples to Agro-One, because they have calibrated their analytical methods with CNAL in order to use Cornell recommendations.
Education & Outreach Activities and Participation Summary
The concept of this study was presented at the annual Elba muck onion twilight meeting on August 2, 2010 to 34 onion growers, industry representatives, and Cornell professionals.
A newspaper article, “Researchers descend on muck for projects” published in the Batavia Daily News on August 31, 2010, featured this project among others that demonstrate the overall goal of sustainable onion production on muck soil. (http://thedailynewsonline.com/news/article_bed5d074-b468-11df-891a-001cc4c03286.html).
Conclusive results from this study will be shared with onion growers and allied industry representatives in 2012 via a newsletter article in the Veg Edge Weekly newsletter of the Cornell Cooperative Extension Vegetable Program, which has a readership over 900 across 23 counties in New York and will reach the majority of large-scale onion growers in the Western and Upstate regions of New York. Results will also be shared at the Annual Elba Muck Twilight Meeting on August 2, 2012 in Western New York, where typically 30 to 35 onion growers and allied industry professionals are in attendance. Finally, Hoepting will discuss results individually with grower cooperators and other interested growers and crop consultants.
Hoepting will incorporate the results of this study into her analysis of CNAL soil tests and fertility recommendations for onions grown on muck soils.
The cost of urea, MAP and potash that were quoted from Helena Chemical were $600, $675 and $685 per ton. In the standard treatment where NPK was broadcast, the total cost of the required nutrients according to the Cornell soil test were $142 (Table 13), $65 (Table 14) and $169 (Table 15) per acre at the CY, Star and Mortellaro sites, respectively. These differences reflect the wide differences among the baseline levels of available nutrients from field to field and represent the value of conducting soil tests on a regular basis. If a grower were to apply the same rate of P and K, because the crop needs for P and K are relatively equal for onions, CY Farms would have applied an additional 150 lb of potash, which would have cost $83 per acre, and Mortellaro would have applied an additional 100 lb of potash, which would have cost $52 per acre. Star Growers had adequate levels of P and K in their soils and did not have to add any additional P or K, a savings of $148 per acre had they applied 100 lbs each of P and K. On their 80 acre farm in the Webster muck, Star Growers saved $11,840 in 2010 by adjusting their fertilizer applications according to soil test results.
Two foliar applications of Mn cost $49.32 per acre (Table 12, 13 & 14). Our results revealed no significant yield increases from making foliar applications of Mn at any of the sites, but especially at the Star site in the Webster muck and at the Mortellaro farm in the Elba muck, where pre-plant soil tests showed higher levels of Mn (Star: 27 lb/A; Mortellaro: 98 lb/A) compared to the CY site where Mn was 9 lb per acre (Table 8). Star and Mortellaro could save $49 per acre by not applying Mn when it is not needed, which represents 43% and 22% of their fertilizer expenses, respectively (Table 14 & 15). At the CY site where soil Mn was lower, the yield where foliar applications of Mn were made increased yield by 30% (Table 8). For a crop that yields 365 cwt per acre (NYS average 2006-2010), a 30% increase in yield when onions are sold for $17 per cwt (NYS average 2006-2010) increases economic return by $1861 per acre. Every dollar invested in Mn could return $38 in increased profit due to increased yield. Our results suggest that when soil levels of Mn are 20 lb/A or higher that foliar application of Mn is not worth it. When soil Mn is 20 lb/A or lower, there is a greater chance that an onion crop will respond to foliar applications of Mn with increased yields and is well worth its cost.
Three foliar applications of P cost $22.50 per acre (Table 13, 14 & 15). Our results revealed no significant differences in yield between broadcast NPK alone and with foliar applications of P (Table 8). Pre-plant broadcast and incorporated applications of P according to the onion crop needs as determined by a soil test were sufficient deeming additional foliar applications of P unnecessary. By not making foliar applications of P, growers can save $22.50 per acre, which represents 10%, 26% and 12% of the total fertilizer expenses for CY, Star and Mortellaro, respectively.
Our studies showed that banding an acidified formulation of P fertilizer and manganese sulfate 2-3 inches below the seed had no effect on changing the pH within the row and generally did not result in increased yields. Banding fertilizer was proven to not be beneficial in this study, and thus, it will not be recommended for onion production on muck soils in Western NY. This will save growers the one-time cost of modifying a seeder to band fertilizer, which could cost approximately $4000 to $6000 in parts and labor for a 12-row seeder, and $8.58 per acre in manganese sulfate (Table 13, 14 & 15). There is no benefit to applying Mn as a broadcast treatment, because it is immobile in muck soil.
Because P was so high at the Star site in Webster, a low rate of P (50 lb/A) was only added in the band treatments and not in the broadcast treatments. Although not significant, the banded P treatments had 57% higher medium bulb yield and 8.3% higher total yield. Since levels of soil P were also significantly 51 lb/A higher in the band treatments (Table 4), we suspect that this difference in yield is related to the addition of P. In this case, the investment of $32.47 per acre (Table 14) in applying 50 lb of P could translate into an additional $515 per acre using the NYS average yields and prices; for every $1 spent on P, and additional $16 were made in higher yields. It is unknown whether the application of 50 lb of P broadcast would have the same crop response, because this treatment was not trialed.
- Table 15. Economic analysis for Mortellaro in Elba.
- Table 14. Economic analysis for Star Growers in Webster.
- Table 13. Economic analysis for CY Farms in Batavia.
One of our grower cooperators, CY Farms, was very interested in modifying their onion seeder to band acidifying fertilizer below the seed. Since our research results showed that fertilizer banding was not beneficial, this farm saved an estimated $4000 to $6000 in parts and labor expenses to modify a 12-row seeder to band fertilizer. Additionally, they will save the annual expense of $8.58 per acre in manganese sulfate that would have been banded, which would total $367 per year on their 43 acres of high pH muck where onions are grown.
This study demonstrated that there is tremendous opportunity to save on P and K fertilizer inputs. In our observational field survey, 50% and 100% of the fields had soil levels that exceeded the sufficiency levels for P and K, respectively, in mid- to late-July. According to our pre-plant soil tests, no additional P and K were needed in the Webster muck, so Star Growers did not apply any, and saved at least $148 per acre had they applied 100 lbs each of P and K. On their 80 acre farm in the Webster muck, Star Growers saved $11,840 in 2010 by adjusting their fertilizer applications according to our soil test results. Since 2010, our grower cooperator, Matt Mortellaro, started sending his soil samples to Cornell. In 2011, he needed to apply only 50 lb of K on just over 100 of his 250 acres of onions, representing a savings of $17 per acre and $1712 on his farm, had he applied 100 lb of K. Reducing P and K fertilizer rates by 50 to 100 lb/A can result in savings of $17 to $34 per acre per nutrient.
This study indicated that nitrogen levels in muck soils were generally high. Recently, other Cornell studies, including NESARE funded project ONE-07-072, have shown that New York muck onion growers can reduce the rates of nitrogen fertilization without reducing yield, while reducing bacterial bulb decay and onion thrips pressure, and thus insecticide applications. In 2011, Matt Mortellaro reduced his nitrogen rate by 30% from 125 lb/A to 75 lb/A on 30 acres of onions, which saved him $15 per acre and a total of $450. In addition, he grew early yellow transplants in this field without a single insecticide spray to manage onion thrips. He claims that in this field, “I spent the least amount of money to grow my best onions!” We are aware of two other onion growers who experimented with reducing their rates of nitrogen in 2011, and both were pleased with the results. We expect that more onion growers will experiment with reducing nitrogen fertilizer each year, and gradually onions in New York will be grown with less nitrogen and less nitrogen will leach into the waterways.
In 2011, several fields in the Elba muck were tiled. During this process, the underlying soil containing calcareous marl which has a high pH was brought to the surface and blended with the muck soils. In one field, the resulting pH of the blended soil was 7.6, which is very high. In 2012, Hoepting advised growers that are at risk for Mn deficiency due to high soil pH to make foliar applications of Mn at the 2-3 leaf and 5-6 leaf stages. This is not expected to increase yield to its full potential, but should elicit somewhat of a crop response, according to the results of this study.
Ideally, in 2012 and beyond, based on the results of this study and consequent recommendations, growers who have high levels of Mn in their soils will not make foliar applications of Mn, and realize savings of $49 per acre in unnecessary input costs. Similarly, it is hoped that growers will not bother with foliar applications of P, as this study found them to be completely unnecessary.
Areas needing additional study
In these studies, CNAL reported soil levels of several nutrients to be very high, especially NO3. This finding requires further investigation of the conversion factors from ppm to lb/A that CNAL uses for muck soils. In mineral soil, this conversion is 2x, based on the fact that on average, there is about 2,000,000 lbs per acre furrow slice (6.67 inches deep) over an acre. For muck soils, on average there is 1,000,000 lbs per acre furrow slice, although this may vary from 500,000 to 1,500,000. Therefore, a soil test in 1 ppm is equal to 1 lb/A and the conversion should be 1x. The 2x conversion factor may be more appropriate for shallow muck soils (< 18 inches), and 1.5x for medium muck soils (18-48 inches). First, it should be confirmed that these conversion factors make sense for the bulk densities of muck soils in New York, and then the appropriate conversions should be applied to ensure that muck soil samples in New York are interpreted correctly by Agro-One.
Although there may have been some issues with the accuracy of the CNAL test results for NO3, the majority of the sites from both the small-plot trials and the observational field survey had 2x to 5x more than the maximum of 125 lb/A of N that Cornell recommends per year for onions. The feasibility of reducing nitrogen fertilizer rates in onion production warrants further investigation. Cornell entomologists, Hsu et al. (2012) investigated the effects of reduced nitrogen on onion thrips pressure, and found that when applied N was reduced from 125 lb/A to 94 lb/A (= 25% reduction), the time it took for the onion thrips to reach the spray threshold of 1 thrips per leaf was delayed by 1 week, resulting in 2-3 fewer sprays per season without any difference in yield. Hoepting is currently investigating the relationship between nitrogen fertilization and bacterial bulb decay of onions with preliminary results suggesting that reduced N results in less bacterial bulb decay. Through these studies and others, the feasibility of reducing nitrogen applications will be investigated and promoted. In the future, we hope that nitrate loading into the waterways from the muck lands where onions are grown will be reduced.
At the Webster muck site, our results showed a slight yield response of 57% higher medium bulb yield and 8.3% higher total yield when 50 lb/A of P was banded below the seed, despite there being extremely high levels of P in the soil prior to planting. Further research is warranted to optimize the rate and application method (band vs. broadcast vs. furrow) of P in soils with high P.
Since the Webster site on Edward’s muck responded differently than previously reported for muck soils, had unexplainably high levels of soil nutrients, and the most variability in test results among the soil laboratories, further research is necessary to optimize the nutrient management for this type of muck soil.
Although in our studies, we did not find a direct correlation with soil Mn and yield, some fields had seemingly very high levels of Mn. We did find a correlation with yield and levels of tissue Zn and B, and we found seemingly high levels of Zn in the soil. Unfortunately, CNAL does not include B in its standard soil analysis package, so we were not able to relate tissue B with soil B and yield. Therefore, it would be worthwhile to conduct focused studies to determine the effect of soil and tissue Mn, Zn and B on yield, and also to develop accurate sufficiency levels for soils of these important micronutrients for onion production in the different muck soils of New York.
We found a minor, yet significant correlation between weight of bacterial bulb decay and tissue Cu (R= -0.4169; p=0.0305). Interestingly, copper bactericides are used to manage bacterial bulb decay and the plants with the highest levels of Cu in their tissue also had the least bacterial bulb rot at harvest. Unfortunately, CNAL does not include Cu in their standard soil test, so we do not know if there was a correlation between the levels of Cu in the soil and leaf tissue. Since bacterial bulb decay is such an important disease of onions with few effective management strategies, the relationship between soil Cu, tissue Cu (leaf and bulb) and applied copper bactericides should be investigated for management of bacterial diseases in onions.