Evaluation of Sulfur Mineralization and Availability in Soil and Manure and Sulfur Content in Plants

INTRODUCTION

The response of corn grain yield to sulfur fertilization has been one of the major factors for increased productivity and profitability in some cropping rotations. Current projects on sulfur timing, rate, and placement have clearly demonstrated the need for sulfur in areas with low organic matter as do cropping rotations with high amounts of residue remaining on the soil surface. While some of this work is still ongoing there have been some questions raised that have not been fully covered by current research. One question is, are there any differences in the availability of sulfur from different fertilizer sources. For example, fertilizer retailers are limited on what the sell to farmers and in some cases their only option is elemental sulfur. This product has been used for years to supply sulfur to crops. However, the elemental sulfur needs to be mineralized to sulfate before it is available to crops. This mineralization is mediated by bacteria, Thiobacillus thiobacteria. From previous work we know that the activity of Thiobacillus tends to be low when soils remain cool. In fact, the optimum temperature is above 80oF and even at these temperatures the oxidation of elemental sulfur can take 30 days. This can cause problems in Northern parts of Minnesota when soils warm slower and have caused some researchers to question whether Thiobacillus has much activity. It is also thought that there is more activity from fungi converting elemental sulfur in these soils. Fungal conversion occurs at a slower rate than bacterial mineralization. Research should be conducted to understand the mineralization and activity of Thiobacillus in different soils across the state.

A second question is whether sulfur fertilizer is necessary when livestock manures are applied to fields. Some consultants have reported responses to sulfur application even in manured fields. This suggests some limited availability of sulfur in manures. There is a lack of data in Minnesota to sufficiently answer these questions. Most research on availability values from manure focus on nitrogen, phosphorus, and potassium. A simple incubation study would be sufficient in order to determine if there is limited availability of S in manure and if further fertilizer should be applied.

Third, there is increased interest in using plant analysis to determine “hidden” nutrient deficiencies. However, the benchmark values for the plant analysis are 20 to 30 years old. Seven strip trials were conducted with funding from the Minnesota Soybean Research and Promotion Council starting in 2008 and ending in 2009. These studies examined interactions between nitrogen, phosphorus, and sulfur applied as starter fertilizer. In a number of the trials there was a significant corn yield response to sulfur. In the soybean trials only one responded to sulfur. Early plant samples were collected from this study but most benchmark values are available for later sampling at the R2 growth stage for both corn and soybeans. These samples were collected from all of the studies with the intention of analyzing the data to compare sulfur data with benchmark values to establish if they are current with current hybrids.

Fourth, the recycling of sulfur in residues in not fully understood. Typically, the return of crop residues is considered beneficial as it supplies carbon and recycles nutrients that may be available for following crop. Corn residues generally have high amounts of carbon and when compared to nitrogen the amount of N is typically not enough to decompose the residue. In this case nitrogen is then used from the soil, rendering it unavailable for crop response. From previous work we know that soil organic matter levels are important in determining where a sulfur response may be more likely. However, we do not know how great of a role the residue may play in this effect.

Finally, previous work has shown that the recommended soil test for sulfur is not a good indicator of where crop responses may occur. Other tests are available and used in labs. For instance labs in Minnesota use KCl or CaCl2 to extract sulfur and report the number to growers, but we do not know how adequately these values will correlate to crop response.

Main Objectives

1. Study the effect of sulfur cycling from corn residue on the following years’ crop
2. Establish whether different sulfur soil tests can be correlated to grain yield
3. Evaluate the sulfur supply potential from soil organic matter in various Minnesota soils
4. Determine the mineralization potential of elemental sulfur in Minnesota soils
5. Determine the availability of sulfur in livestock manures
6. Evaluate the effectiveness of current benchmark S values under current hybrids and growing conditions.

SULFUR RATE AND TIMING STUDIES

Experimental Methods
Trials were established at five locations (Table 1) that varied in fertilization history and previous crop. Soybeans were the previous crop at the New Ulm, Theilman, and Montgomery location while it was corn at Renville and Otisco. The New Ulm and Renville locations had a previous history of manure application although it had not been applied for a few years at the Renville site. Tillage was a chisel or disk ripper at New Ulm, Renville, Montgomery, and Otisco followed by spring field cultivation and Theilman was managed with no-tillage. In comparison the highest amount of surface residue was left at the Renville and Otisco sites followed by Theilman, New Ulm, and Montgomery contained the least. Corn hybrids varied by location. Plot size measured 20 feet wide by 40 feet long at all locations except for Theilman which measured 15 feet wide and 80 feet long. The large plot size was divided into a half for a second year study on the same plot areas.

First year treatments consisted of a control (no sulfur fertilizer) and four sulfur rates of 10, 20 ,30 and 40 lbs of S per acre. Sulfur rates were broadcast applied on the soil surface at two timings of within 3 days after planting and at approximately the V3 to V4 growth stage. For the later application the fertilizer was broadcast between the corn rows to lessen the risk of fertilizer entering the whorl and causing leaf burning. Ammonium sulfate (21-0-0-24S) was the broadcast sulfur source used because it is a highly available source of the nutrient. In order to separate out potential response to nitrogen, additional rates of fertilizer nitrogen were applied as ammonium nitrate to equalize the nitrogen applied with the highest sulfur rate across all treatments for both application timings.

For the second year at a particular location, the previous rates x timing plots were divided in half. On one half a single rate of 25 lbs. of S per acre was applied as ammonium sulfate while the 4 other half received nitrogen only to balance the rate applied with ammonium sulfate. All fertilizer was applied in the spring following soil sampling and before planting. Each side by side with and without sulfur plot was used to compare yield response to the previous sulfur application. Relative yield was calculated by dividing the yield without sulfur with the yield with sulfur.

Composite soil samples were taken prior to the first sulfur application during year 1. Soil samples were collected to a depth of 2 feet and separated out into 0-6, 6-12, and 12-24 inch depth increments. The 0-6 inch soil samples were analyzed for phosphorus, potassium, soil pH, and organic matter content. All depths were analyzed for soil sulfur content. Leaf chlorophyll readings were taken at silking (R2 growth stage) from the ear leaf. For the first year studies, at physiological maturity six plants from each plot were sampled, the ear removed, and the stover weighed and analyzed for total S to determine plant stover yield and S uptake at the end of the season. Stover yield per acre was calculated by multiplying the plant population from each plot (as taken from the harvest rows in the spring) multiplied by the average individual plant stover weight. Gain samples were collected after harvest to determine sulfur content in the grain. Soil samples were taken in the spring of year two and were collected from the top two feet at Renville and Theilman to measure soil sulfur carryover from the previous year.

Statistical analysis was conducted using the Proc MIXED procedure in SAS for fixed treatment effects of sulfur application rate and timing of application. Trial data was analyzed as a complete factorial with the no sulfur plots included with data from the pre-plant sulfur rate treatments. When the statistical analysis indicated a significant effect on main treatments (P<0.10) treatment means were compared using least significant differences. When the analysis indicated a significant main treatment effect of sulfur rate regression analysis was used to fit a model to the data to determine optimum sulfur application rates.

RESULTS AND DISCUSSION

Results from the initial soil test data are listed in Table 1. Soil P and K levels were generally High to Very High at all locations except for soil P at Theilman which was medium. Soil pH levels did not vary greatly between locations but did not appear to affect results. Soil organic matter content ranged from 2.3 to 7.2% representing a low to high range in the mineralization potential for sulfur within a soil. Soil sulfate levels varied widely but were the lowest for the Renville location. Since the soil sulfate test was not the same as used in the Minnesota recommendations a comparison to current recommendations for sand could not be done with any certainty. However, past work has noted that the recommended sulfur soil test performs poorly on fine textured soils.

Leaf Greeness and Above Ground Biomass
Sulfur deficiencies can be visually seen in fields as a yellowing of the leaf and plant tissue. Photograph 1 shows the typical S deficiency symptoms in the early season consisting of yellow interveinal chlorosis of the upper leaves of the plant while Photograph 2 shows late season deficiencies in rows with and without sulfur. Chlorophyll meters can be used to assess the greenness in the leaves during the growing season and have been shown to relate to yield. In general the relative differences in numbers need to be around 90% for us to expect a yield difference. Sulfur rate significantly (P<0.10) impacted leaf greenness at all locations except for at Montgomery (Table 2) and there were no  differences in application timing. In addition, there were no significant interactions between rate and timing indicating that any rate responses were the same regardless of when the sulfur was applied.

When the effect of rate was significant the effect was mainly due to the initial increase from the 0 to 10 lb. sulfur rate. The only exception was the 40 lb. rate at Renville which was no different than the 0. However, this rate is much higher than what would typically be recommended even in coarse textured soils. Therefore, the effects of the 10 and 20 lb. rates are more relevant and important for corn production systems. The 30 and 40 lb. rates were included in order to analyze the data with a curve fitting procedure but since the greenness was maximized at all locations with the 10 lb. rate it is hard to fit a response curve to the data. At this time it is clear that 10 lbs of S maximized greenness when applied either at planting or as an early side-dress application. On a relative basis, the control plots were 94, 88, 94, and 90% of the greenness of when sulfur was applied at the New Ulm, Renville, Theilman, and Otisco locations, respectively.

Plant biomass data was collected to assess luxury uptake in corn stover. The effect on stover uptake is important since sulfur is an important part of organic systems and thus is also needed for microbes to decompose corn residues similar. On average a carbon to nitrogen ratio between 200 and 400:1 would indicate that there is enough sulfur in the residue to break it down and will not add to or draw down sulfur in the soil. Thus for crops that produce large amounts of residue with high carbon contents, such as corn, there can be a high demand for not only nitrogen but also for soil sulfur that can potentially lead to tie up of sulfur in the soil. The measured data from 2009 found that the ratios were 258:1 at New Ulm, 323:1 at Renville, and 236:1 at Theilman. In addition, there was no evidence that sulfur application increased the amount of sulfur in the biomass (not shown). It appears likely that at these locations the return of corn residue to the soil likely would not result in an increase in sulfate within the soil. In contrast, a study conducted at two soybean locations in 2008 found the ratio of S in soybean stover to be about 125:1 (Kaiser, unpublished data) indicating the potential for sulfur to be returned following soybean. The lack of sulfur mineralization may lead to reasons why sulfur may eventually become more deficient with successive years of corn grown in the same field. If mineralization remains constant any residual sulfate could be depleted in the soil through S removal in grain. Thus the change in crop rotation from corn-soybean to corn-corn may necessitate further changes in fertilization strategies similar to changes in nitrogen rates.

Plant stover yield was not affected by sulfur rate at any location and the only significant difference was between application timing at the Theilman and Otisco locations (Table 3). In this case plant stover production was 0.4 tons per acre less when sulfur was applied as an early side-dress than the application near planting and a 0.3 ton per acre increase at Otisco. The decrease is odd since there was no effect of rate on stover yield, but may indicate there is some impact on sulfur in the early season that may affect plant growth later. However, there was no evidence of a similar effect on greenness and may not be a similar effect on yield.

Stover production was the lowest at the Renville, Theilman, and Montgomery locations averaging around 3.0 to 3.4 tons of dry matter per acre while nearly a ton more was produced (~4.0 tons dry matter per acre) at the New Ulm and Otisco sites. This is likely due to the higher fertility levels and more recent manure application at the New Ulm site or could be due to differences between the hybrids planted (not shown). Since uptake is generally largely influenced by plant weight it appears unlikely that there would be luxury uptake of sulfur in the plant stover. There were significant differences in sulfur uptake at two of the 2009 locations, New Ulm and Theilman (data not shown). Overall this difference was only 1-2 lbs. of additional S taken up in the plant. At Theilman this increase was mainly due to increased plant mass, but at New Ulm the effect seemed to be more related to higher S concentrations in the plant. Overall this effect was minor and would not be expected to have major implication in S carryover to the next year. It still would be expected that little sulfur would become available from the residue for the following year.

The ratio of carbon to sulfur was studied to determine whether the application of sulfur would affect the amount cycled in the corn stover. As a general rule of thumb, a ratio less than 300:1 would indicate that sulfur would be released upon breakdown of the residue, greater than 400:1 sulfur in the soil would be tied up (immobilized), and there would be no release or immobilization if the ratio is between 300 and 400:1. The ratio of sulfur to carbon was generally around 300:1 at most sites except for the two field locations in 2011 which sulfur concentrations in the stover were less and the analysis would indicate that sulfur may be immobilized (Table 4). Analysis of the data indicated that the ratio did change with sulfur fertilizer rate for the New Ulm, Montgomery, Otisco, and Alden locations. As a general trend, the C:S ratio decreased with increasing rate of sulfur applied. This increase was seen up to the 20 lb S rate where there was no difference or a slight increase in the ratio. This effect has some significant implication for the cycling of sulfur to the following crop. The changes were not large in magnitude but significant enough to supply some sulfur to the following crop. This effect would not have had a significant impact on the Alden site where the concentrations were well above the 400:1 threshold. For New Ulm and Otisco the levels went from the range were we would not expect any release or immobilization to where we the release of S would be likely. The reason for the difference between the 2011 data and the other years is not clear. The samples from 2011 were run at a different lab but were run on the same machine therefore we would not expect the degree of variation found. The differences could be related to the year itself.

Corn Grain Yield and Harvest Moisture
First year corn yield was significantly (P<0.10) increased by sulfur rate at Renville, Theilman, and Otisco (Table 5) and was never impacted by timing. A yield response at Theilman was expected since deficiencies have been noted more commonly in southeastern Minnesota and research in Iowa has shown a yield increase from sulfur applied to similar silt loam soils. Other research has shown that the potential for a yield increase from sulfur tends to increase when soil organic matter levels in the top six inches is 3.0% or lower which would be consistent with the Otisco location. A yield response was not expected at the Renville location since the organic matter levels were near 5.0%. In this case the soil will generally mineralize enough sulfur but examination of the sulfur soil test values indicates that the site had the lowest residual soil test levels of all locations. Previous history on this location is multiple years in corn and a past, but not recent, manure history. At both the Renville and Theilman locations the 10 lb rate maximized corn yields, but there was a significant yield increase up to 20 lbs. at Otisco. The difference in the rates could be attributed both to low soil organic matter levels and a previous corn crop. More locations with similar cropping system and organic matter levels would be beneficial in order to determine if rates need to be adjusted for both previous crop and soil organic matter levels.

We can theorize from the rest of the data that the 10 lb. application rate would have been enough at this location as it was at other sites confirming previous data. It appears that 10 lbs. applied as broadcast may be an optimal rate for corn following soybeans on soils with less than 4.0% soil organic matter or continuous corn fields regardless of organic matter levels. However, the data from Otisco has raised the following questions 1) do rates need to be increased for continuous corn when soils have less than 4.0% organic matter; and, 2) if rates need to be increased for continuous corn in corn-soybean rotations should rates be higher when organic matter is less than 2.0%? It was clear that when there was a rate response it did not matter whether the sulfur was applied at planting or as an early side-dress at V3 to V4.

Data from Figure 1 shows the effect of greenness determined with the SPAD chlorophyll meters on yield by location. At New Ulm and Theilman there was no relationship between ear leaf greenness and yield. At Renville, Montgomery, and Otisco yield increased as leaf greenness increased. This data followed a steady linear relationship. It should be noted that the relationship at the Montgomery location was mainly due to the low values of one control plot within the location. The lowest SPAD readings were likely attributed to no- or low sulfur application rates at the other sites. Visual differences were also noticed in an aerial photograph (Photograph 3) from the Renville location taken on June 3 in which the control plots are clearly visible in the field, but no other differences can be seen between rates. Seeing that this also was affected by yield the SPAD meters could be used to assess fields for possible sulfur deficiencies. However, nitrogen deficiencies will also affect the greenness of the plant thus making it somewhat difficult to diagnose at certain times of the year. In addition, diagnoses may be made after the fact when yield losses have already occurred. The other important point with this data is that the SPAD meters at used in nitrogen studies to determine the effect of rates on greenness. Since we applied excess N in these trials any greenness differences were likely due to a sulfur effect. This illustrates that sensing methods may my not be able to pinpoint specific nutrient deficiencies and that additional data should be collected from particular sites in order to make sure the nutrient is actually a problem. Late season sulfur deficiency can easily be mistaken for nitrogen and the only way to differentiate may be through plant analysis and additional nitrogen application will likely not help to increase yields.

Corn grain moisture was impacted by sulfur application at Renville, Theilman, and Alden (Table 6). At Renville and Alden, harvest moisture was lowered by the 10 lb and 20 lb S rate. There was no significant decrease in harvest moisture beyond 20 lbs. of S at Renville, but grain moisture was slightly less at Alden with the higher rate. The effect of timing was also significant at Renville but the difference was due to the control plot only included with pre-plant sulfur data. Since growth was not influenced at this location the effects on grain yield could not be tied to faster growth or larger plants. At Theilman grain moisture was less when sulfur was applied. However, the effect was not as large as was seen at Renville. The largest differences were seen with the highest sulfur application rate but there was negligible difference between that rate and the other sulfur rates. At Montgomery and Medford sulfur timing influenced grain harvest moisture. However, this effect could be attributed to higher moisture in the no sulfur control of the in-season application which suffered significant yield losses. The effects on grain moisture coupled with yield increases can be extremely important on not only increasing profit from yield but also on saving on drying costs in the fall. Optimal rates for yield should provide enough sulfur to decrease harvest moisture.

Sulfur Carryover Effects on Corn and Soybean Yield
Crop yield for a second crop following the initial sulfur application were measured to study the carryover effects of sulfur on corn and soybeans. Three corn and two soybean locations were studied. In these plots sulfur was applied to half of the old rate plot to compare sulfur response to a fresh application of fertilizer. Table 7 summarizes yield data from the studies. The effect of timing of the first year’s application was not significant so it’s affects along with the interaction between the first and second year sulfur application are not considered.

Effects from the first year’s application of sulfur could only be clearly seen at the Renville locations. At this site, both the rate applied the previous and current year significantly affected corn yield. The interaction between the two was significant and indicates that the majority of the rate effect from the previous year’s application occurred only when sulfur was not applied for the current crop. This response is reasonable since the 25 lbs applied for the second growing year should have been enough for the crop. While either of the rates applied at the Otisco site were not significant, there was a similar interaction found between the rate applied year one and year two. This interaction is again due to a rate response from the previous sulfur application but only when sulfur was not applied the current year. At Renville, corn yield responded to a maximum of 20 lbs of S per acre applied for the previous year. The amount was less at 10 lb S per acre at Otisco. The response at Otisco was interesting as it took 20 lbs of S to maximize yields in year 1. There was a significant amount of plot variability at the Medford corn site and the two soybean sites, but neither significantly responded to the application of sulfur.

Soil samples were collected in the spring of the following growing season from the locations were a second crop was grown. Three soil tests were studied for their ability to predict final corn yield. These tests were the mono-calcium phosphate test (MCP), sulfate sulfur extracted with potassium chloride (KCl), and extraction with calcium chloride (CaCl2). Samples were taken from a 2- depth to look at sulfur in the profile. This sampling is  recommended by neighboring states. In comparing the tests, it was found that there was no direct relationship between any of the three when compared to each other. Because of that data was summarized only for the MCP test. The soil tests levels from the MCP tests were not affected by any of the treatments (Table 8). In fact, the levels generally remained constant. When the test was compared to relative corn and soybean yields, there was no relationship between the two (Figure 2, soybean data not shown). None of the tests appeared to be a better index of crop response compared to the  others. The KCl test did exhibit a slightly increasing trend with increasing yield for one of the locations (Figure 3), but was not better than the other two when compared across locations. This data supports other data in that the sulfur soil test does not work in medium or fine textured soils. Other research in Minnesota has found a better relationship between yield and organic matter levels. This relationship is better as organic matter level gives a better picture of the potential for sulfur mineralization in a soil. The soil test is just a snapshot in time and appears to not be greatly affected by previous fertilizer applications. Based on this data a soil test for sulfur would still not be recommended for Minnesota.

CONCLUSIONS

The data from this study has shown clear benefits for sulfur applied to corn. Rates needed to maximize yields have been 10 lbs. of S per acre broadcast, but further data has shown that more may be needed under some circumstances. It has also been shown that delaying application of sulfur up to the V3 growth stage did not have any negative impacts on yield. Grain harvest moisture was also decreased when sulfur was applied, but the effect may not be as great if application is delayed to the V5 growth stage. The limited second year data has shown that sulfur may carry over from year to year and this carryover can affect yields. Sulfur soil test was shown to not be affected from S fertilizer. However, the mineralization potential of S from corn stover was increased when sulfur was applied which may explain a possible source for mineralized sulfur for the following year’s crop.

Benchmark Sulfur Plant Tissue Values

Interest in plant tissue analysis has created a need to determine if optimum tissue S concentrations are similar to benchmark values being currently used. A study funded by the Minnesota Soybean Research and Promotion council was established in 2008 and 2009 to investigate the effects of N, P, and S fertilizers applied in combination on corn and soybean yields using replicated strip trials within farmers’ fields. Field sites  consisted of different soils ranging from fine sandy loam to silt and clay loam soils. This study found corn yield responded to sulfur especially in field areas low in organic matter. In addition, there was a significant yield increase to sulfur at 1 of the 4 locations. Using the yield data from this study, we propose to look further at in-season tissue tests to evaluate their effectiveness in predicting critical levels for crop response. Corn ear leaf and soybean trifoliate samples were taken at the R2 growth stage. Because of limited funds, we were not able to analyze these samples at the time of collection; they were stored for later analysis. One corn and one soybean location have been analyzed, but the other 5 have not. Small plant samples were also collected at the V5 growth stage and analyzed. These numbers could be compared with the mid-season sampling and both would be valuable to determine if current values used for sufficiency levels for corn ear leaf tissues are relevant.

Figure 4 summarizes plant tissue data from the sulfur mineralization study. Early and midseason plant tissue concentration was significantly correlated to yield response to S for both corn and soybean. For corn, the correlation was better with the ear leaf tissue than for the early plant tissue. Similar effects were seen with soybean but the overall correlation was poorer than for corn. The critical levels determined for corn fell within the currently used ranges of 0.15 to 0.50% S at for both sample timings. However, the ear leaf tissue was on the low end of that range. It is likely that our range would be less than the currently utilized range for the midseason sampling. For soybean, the trifoliate concentration found in this study fell within the range of the established sufficiency range between 0.21-0.40% S. There is no current established range for soybean V5 whole plant sulfur concentration. Even though the predictability may not be the best, the values currently used should still be adequate to judge the sufficiency of sulfur for corn and soybean tissue.

Sulfur Availability from Animal Manure

Manure and fertilizer was applied to two soils and incubated in growth chambers. The two soils consisted of a Seaton silt loam and Lamont sandy loam. Soils were collected from the upper six inches of the soil surface, sieved through 2 mm, and mixed in a cement mixer prior to treatment application. Treatments applied were fertilizer N (finely ground urea) and three manure sources. Two of the manure sources were from a dairy farm one of which came from an outside pen and the other from a sand separated bedding system collected after the final separate before the manure was sent to an outside lagoon for storage. The third manure source was collected from a secondary storage pit within a swine finishing farm. The manure was mixed thoroughly then a representative sample was taken from each source to determine nutrient concentration. Results from that analysis are listed in Table 9.

Manure and fertilizer was applied at and equivalent of 200 and 400 lbs of total N per acre assuming incorporation to a depth of 6 inches. An additional rate of 600 lbs of total N as fertilizer was applied. Soil bulk densities were estimated at 1.3g/cm3 for the silt loam soils and 1.5g/cm3 for the sandy loam. Manure and fertilizer treatments were applied to bags containing 1.4 kg of air dried soil. Soil and fertilizer or manure was mixed using a small fertilizer mixer. After mixing the soil was divided into 200g amounts and put into 5 separate cups representing individual sampling dates of 7, 14, 28, 56, and 84 days. Incubations were conducted in clear plastic jars measuring 3.25” in diameter. Soil was packed into the jars to the desired bulk densities (listed above) allowing for at least 0.5 inches of head space for air exchange. Lids were placed on the jars to prevent movement of soils into and out of each vessel and to prevent rapid water loss. To facilitate for air exchange, a small hole was drilled in the middle of each container lid. All treatments were replicated 3 times.

Soil incubations were done in a growth chamber in the dark and maintained at 75oF. Soil moistures were maintained at approximately 80% field capacity for the duration of the experiment by monitoring cup weights and adding water according to mass loss over time. Volumetric water content for the soil at field capacity was determined to be 31.3% in the silt loam and 16.0% in the sandy loam. At the specified times individual cups were removed from the incubation chamber and half of the soil was separated to be oven dried and the other half was frozen. The frozen half of the soil was saved and later thawed and analyzed for ammonium and nitrate content on a moist basis. The dried portion of the soil was weighed before and after drying to determine soil moisture content at the sampling date. Statistical analysis for significant treatment effects within each sampling date and soil type were assessed using the PROC GLM procedure in SAS. Patterns of nitrate accumulation and ammonia depletion were analyzed using the PROC NLIN procedure in SAS.

Results and Discussion

Figure 5 shows soil ammonium (NH4 +) decline over time represented as the percent of total N applied. Calculations were made by subtracting the ammonium or nitrate value of the control (no treatment) from the values from the treated soils. The difference was then converted to lbs of N per acre. Ammonium in the soils steadily declined over time. In no case did the ammonium concentration represent more than 20% of the total N applied. Decline in soil ammonium was faster in the silt loam soil than the sandy loam. After two weeks there was little to no ammonium remaining in the 200 lb N rate treatment for any of the N sources. For the higher rate it took about 2 months for the ammonium concentration to drop to baseline levels in the silt loam soil. The difference in timing may be due to a limited capacity of N fixing bacteria to convert ammonium to nitrate. In all instances fertilizer treatments tended to have higher amounts of ammonium remaining in the soil compared to any manure source.

Conversion appeared to be slower for the sandy loam soil. For the 200 lb rate it took 2 months for the ammonium levels to reach baseline levels while the 400 lb treatments still had elevated levels after 3 months. For manure sources there did not appear to be any difference in the level of ammonium in the soil over time. The only exception was in the silt loam soil where liquid swine manure tended to have higher amounts of ammonium than the other two sources. This result is not surprising since the amount of ammonium-N was greatest in fertilizer, followed by liquid swine manure, and both dairy sources were approximately similar (Table 1). The difference in conversion rates could be attributed to soil water holding capacity. Since soil microbes need soil water in order to convert ammonium to nitrate the greater water holding capacity of the silt loam could explain differences between the soils. It is interesting that the conversion rate may be slower even though leaching potential would be higher in the sandy loam soil. The difference in conversion also could potentially lead to decreased availability of the nitrogen in the manure in the sandy soil relative to the silt loam. This is mainly true for fertilizer but for manure most of the N was converted by the end of the first month. We did not analyze any of the samples taken at treatment initiation to determine the amount of ammonium at time 0. However it does appear that the conversion rate is rapid if optimal conditions exist and that large amounts of N may be converted within the first week after application.

Soil Nitrate accumulation is given in Figure 6. Nitrate accumulation is presented as a percentage of the total N applied over time. For the sandy loam soil nitrate accumulation was more gradual over time compared to the silt loam soil where nitrate levels increased rapidly and evened off after 1 to 2 weeks. In general more nitrate was accumulated with the low N rates and in the silt loam soil. As N rate increased it appeared that more N may have been lost likely by ammonia volatilization. The difference between either soil type or rate tended to be about 20% difference in apparent N recovery in the soil. For the sandy loam soil there still was some ammonium in the samples at the end of the incubation thus more nitrate likely would have accumulated with longer incubation times. A 600 lb N rate comparison was included and the data is included in the 400 lb graphs. Similar to differences between the 200 and 400 lb rates, N recovery was less at 40% with the 600 lb rate indicating that as N rate increases the efficiency of recovery decreases. Nitrate accumulation was always the highest for the fertilizer treatments for comparable N rates. In most cases the manure collected out of the sand separated dairy system had the highest accumulation of nitrate over time. At times the liquid swine manure was similar to the sand separated dairy while the lowest nitrate accumulation was seen in the pen pack dairy manure. These results somewhat follow the amount of ammonium-N in the manure from Table 1 which would represent the fraction of N that would be readily used and most comparable to fertilizer.

Figure 7 shows that soil Bray-P levels were constant over time for a silt loam and sandy loam soil  represented as the change of soil Bray-P concentration (mg/kg). Calculations were directly made from the treated soils minus the control soil (no treatment). Bray-P concentration was lower for the 80 lbs total P rate than 160 lbs total P rate. For manure sources liquid swine manure tended to have higher amounts of Bray-P than the other two sources. This is because the amount of Bray-P was greatest in liquid swine manure, followed by both dairy sources (Table 9.). Overall, Bray-P in the soils constantly stabilized over time.

Figure 8 shows that soil K levels were constant over time for a silt loam and sandy loam soil represented as the change of soil K concentration (mg/kg). Calculations were directly made from the treated soils minus the control soil (no treatment). Soil K concentration was lower for the 120 lbs total K rate than 240 lbs total K rate. For manure sources liquid swine manure tended to have low amounts of soil K than the other two sources although the amount of K was greatest in liquid swine manure, followed by both dairy sources (Table 9.), and initial K concentration for both dairy sources were similar (Table 9.). Overall, soil K concentration for the each treated soil was similar over time.

Figure 9 shows that the change of soil sulfate-S levels over time for a silt loam and sandy loam soils. Calculations were directly made from the treated soils minus the control soil (no treatment). Soil sulfate-S concentration was higher for the 30 lbs total S rate than 15 lbs total S rate. For manure sources liquid swine manure tended to have low amounts of soil sulfate-S than the other two sources. The amount of S was greatest in liquid swine manure, followed by both dairy sources (Table 9.), and initial S concentration for both dairy sources were similar (Table 9.). Overall, soil sulfate-S concentration in the soils was similar over time.

The comparison between the liquid swine manure and sand separated dairy indicates that in this instance the availability values for normal dairy manure likely are not sufficient for the sand separated manure. This is logical since this manure source contained a low amount of solids similar to swine manure (Table 9). Table 10, 11, 12, and 13 summarizes the fertilizer equivalency, the ratio of the amount of nitrate, Bray-P, K, and S mineralized from manure compared to fertilizer, for the different manure sources, while Table 14 summarizes the apparent N recovery in nitrate for each source at the end of the incubation time. The data in Table 6 is interesting in that it shows that for fertilizer when N rate is increased less N is recovered. There were differences between the soils and it appeared that the sand lost more N with the 400 lb N rate but the other two rates were no different between each soil. It is unclear if the difference between the soils at 200 lbs N is real or if some outside factor influenced the result. For each manure source, there was not an appreciable difference between the application rates, but again the soils tended to differ with lower recovery in the sandy loam. Some caution should be taken in using this data for calculating first year availability. Since the incubations were conducted in a small container with relatively large surface area it is likely that more N may be lost through volatilization than if it was point injected below the soil surface in the field. The data from Table 10 and 13 likely represents a better instance for N and K availability and the data from Table 13 likely represents a better instance for P availability. In all cases the sand separated manure had the highest fertilizer equivalency. Swine manure was second followed by dairy  manure. The expected values for swine manure are higher however the samples were collected from a holding area in the facility before the manure went to the lagoon which likely affected the results. In general the solids content (Table 9) was higher than expected likely due to where the sample was collected. We would assume that the values would be higher and similar to those for the sand separated manure source. The numbers were variable between the different application rates for each manure source. However, we have yet to do any model fitting to the data which likely will change the ending values and fertilizer equivalency. Minimum (Min.), maximum (Max.), and average (Avg.) soil test NO3-, NH4+, P, K, and S from sampling Date1 through Date 5 during the incubation study were summarized in Table 15.

Conclusions

1. The availability of sand separated manure is likely higher than that of normal scrape and haul or pen pack manures. For the source tested using first and second year N credits from book values likely will underestimate first year and overestimate second year availability. The difference between sand separation manure and scrape and haul is likely due to the amount of particulate matter separated out with the sand which would greatly affect potential carryover or second year availability by possibly lowering the organic N
   content in the manure.
2. Application rate can significantly affect the amount of N recovered. High application of fertilizer N and in some instances manure lead to more N being volatilized. This effect would be less likely for a point injection below the soil surface.
3. There may be some difference in the amount of N that can be recovered by differing soil types. While this data does support this statement it would be difficult to know if the same effect can be seen in the field and likely would have to be tested to determine if N availability would change based on soil type. At this time it is not recommended to change the availability based on soil maps. For manure it appeared that all the ammonium N was converted well within the time frame of the study so the previous statement may only pertain to very high rates of N applied as fertilizer.