Case Studies: Field D

Field History

Field D has variable topography and texture and is located in southwestern Québec. The crop rotation is three years of alfalfa hay followed by corn B soybean B seed barley B corn. The field received between 30 to 40 t/ha of solid dairy manure in the fall of 1996, before the first corn crop. Table 1 gives other management practices. The 6 ha field was sampled on a 40 x 40 m grid pattern resulting in 80 geo-referenced soil samples. This gives a density of 13 samples per ha compared to 1 per ha for commercial sampling grids. Monthly rainfall distribution for 1997-1999 is given in Table 2. In the years 1998 and 1999 the spring was drier than normal, which may have affected seed germination and the final yield - the yield maps of Soya and barley showed patterns of high and low yield areas. In 1997 the months of May and June had somewhat lower precipitation than the norm but had higher levels in July. The corn yield maps did not show any distinct patterns, however the producer remembers that there was an early frost that year, on September 24, and the corn hybrid did not reach its full yield potential. This may partly explain why there were no distinct high and low areas as in the following years. The field is outlined in green in the aerial photo . The pale area in the southeastern section is a small ridge that cuts through the field.

Table 1. Crop Management Practices


Fertilization ha-1

solid dairy manure

plant population ha-1

Corn 1997 28 kg P and 40 kg N applied in the band at seedingside-dress 90 kg N 30-40 t ha-1 applied the previous fall 74,000
Soybeans 1998 none none 45 kg seed ha-1
Barley 1999 30 kg N, 17 kg P and 58 kg K broadcast before seeding, 50 kg N broadcast at 4 leaf stage,growth regulator applied in July none 170 kg seed ha-1

Phosphorus & Potassium

Both potassium (K) and phosphorus (P) were mapped according to the ranges for field crops given by the CPVQ (2nd edition, 1996). Figure 1  shows the P distribution across the field when all 80-sample points are used. Most of the field is in the "good to excessively rich" range for P. Figure 2  which shows no variation in P values was determined from only 6 sample points (1 per ha). In this case there was no obvious benefit in the grid sampling for P, the application rate will be uniform. For K, only about 5% of the field had a medium level of K while the rest is in the "rich" range. Both Figure 3  (80 points) and Figure 4  (6 points) show similar K distribution patterns. Crop yield does not appear related to either P or K patterns on the field.

Table 2: Monthly Rainfall and Temperature distribution





average 1961-19990


















































































** Growing season average temperature for each year

Crop Yield Patterns B Unaltered

Figures 5, 6  and 7  give the different crop yields for three years after they have been filtered to remove yield points that are unnaturally high or low due to combine error. From the maps certain high yielding and problem zones can be distinguished. This is the first step in recognizing the causes of yield variability. Using Table 2 in Yield Maps  as a guideline, distinct patterns on the fields are identified as either inherent soil or topography related or management induced. The corn yield (Figure 5 ) in 1997 shows little variation but does have "lines" on the field, which are most likely due to combining errors. In Figure 6 , the GPS monitor had an error in its signal, which is obvious as the field changes shape in one section. As well, there appears to be low soya yields as a result of compaction or drainage along the field edge. This yield reduction is not due to shading from trees as there are no trees along any of the field edges. Figure 7 , which is the yield map of barley, shows definitive low yielding zones as well as some problems in the headlands of the field due either to compaction or combine error. Frequently, problem areas may have more than one reason for low yield. A cursory observation shows that soya and barley low yielding areas correspond though the patterns are not as distinct in soybeans (Figure 6 ). Table 3 gives the average yield from 1997 to 1999.

Table 3: Crop and yield

crop and year

yield, t/ha







Crop Yield Patterns B Normalized

The yield patterns in Figures 8, 9  and 10  are based on the average crop yield for each year. Each of the yield points (from the combine) was divided by the average yield and multiplied by 100 to give a percent. The yield maps were then divided up into three unequal areas: lowest yield to 90% of the average yield, 90 to 110% of the average yield and greater than 110% of the average yield. The corn yield in Figure 8 , shows that most of the yield was average to above average. Remember that the producer said there had been an early frost so the maximum yield was not obtained. Low yielding areas occurred mainly in the headlands. Figures and 10  show the normalized yield patterns for soya and barley and the contour lines for the K: Magnesium ratio. The very high K: Mg ratio is a suspected cause of low yield to be discussed later.

Spatial, Temporal and Classified Management Maps

Figures 11, 12  and 13  are used to determine management zones on the field. For a field to be separated into zones that will be managed differently the field and yield must exhibit "stability". That is the low yielding areas are low yielding for all crops as well high yielding areas are high yielding every year, and for every crop. If there is no consistency in yield either in time or space, then it is difficult to apply precision farming techniques B the producer has to be able to plan ahead. Figure 11  shows where in general there are low, medium and high yielding areas over the total three years. The three maps of normalized yield for corn, soya and barley were summed together and then divided into three classes of low, medium and high yielding areas. An area could be a sum of high plus high plus medium, giving a high yielding zone. Figure 12  looks specifically at the stability of the yield in time, whether it is consistently low, medium or high regardless of the crop. From the map, about 1/3 of the field shows stable yielding areas B the rest of the field could be high one year, low the next and high again B this would result in "unstable yields".

The Classified Management Map, Figure 13 , is the sum of the spatial yield map and the temporal yield map B which shows areas that are consistently low, medium or high yielding for all three crops and in all three years. Two areas are high yielding and stable B these areas could perhaps be managed more intensively so as to push the maximum yield potential. Only a few areas are constantly low yielding, at the southeastern end of the field, which may be due to compaction (see Figure 19 ). About 2 of the field is "unstable" either in time or space or both. Perhaps with more years of yield data and if the years were separated into wet and dry, the yield patterns could be identified as stable (under certain conditions) and thus "manageable".

Soil Physical and Nutrient Maps

The K: Mg ratio should be around 0.5 for optimum crop growth. If the levels of K are too high, K may interfere with Mg uptake by the crop especially when the crop is stressed due to drought or other problems. Both soybeans and barley show distinct low yielding areas in the same location where the K: Mg ratio is 3.0 or higher. Additionally in later discussions with the producer, he observed that the same region also consistently gave lower yields in alfalfa. The high yield areas for both crops occur where the ratio is less than 1.5. Observing Figures 3  and , it can be seen that the K levels are generally uniformly rich and it is the availability of Mg that varies in the field. There is more Mg in clay soils than sandy textured soils. In Figure 9 , soya appears to have reduced yields along the northern field edge due to compaction or drainage. The K: Mg ratio (Figure 14 ) appears to have a strong correlation with soya and barley yield when the crops are under stress (drought). But as always many factors will affect the final yield. The southeastern corner has a low K: Mg ratio (which should indicate a high yielding area) but the yields in that area are either medium or low yielding. This may be due to the effects of drainage or compaction (Figure 19 ).

The higher clay areas (Figure 15 ) correlate with the higher yield areas which is probably related to nutrient retention and the fact that clays contain more Mg than sands (lowers the K: Mg ratio). The clays may also have an effect on the moisture holding capacity of the soil, which could impact seed germination and crop growth. Figure 16  gives the % sand distribution over the field. In general sands will retain less nutrients and less water. The soil textural class ranges from a silty clay loam to a loam to a sandy loam. The silt % was highest in the northern section of the field (Figure 17 ).

The field was also mapped (Figure 23 ) for electrical conductivity (EC) as it has been found that EC can match yield patterns (see Appendix II  for more details). EC reflects a combination of soil properties including water holding capacity, cation exchange capacity (clays) and organic matter. Initially both Figures 11  (spatial trend map) and Figure 23  (EC map) show similar patterns (reflective of yield) but when examined more closely the EC level of 7.9 to 19.1 indicates not only high but medium yields as well. At the southwestern portion of the field, the same EC level also reflects low yielding areas. It is critical to examine closely all the possible factors that affect yield. Technology such as EC mapping can be useful but only when there is a good consistent relationship between yield and EC and when something can be done to manage the problem. The % sand cannot be changed in a field.

Figure 18  shows the soil moisture content in early August 2000. However it does not correlate well with the EC map except in the northern section of the field. In this case the EC map does not reflect the water holding capacity of the field. The bulk density is shown in Figure 19 . The headlands, where machinery is turned, have higher bulk density levels. The extremely high values in the southern part of the field, along the edge did have an effect on the crop yield. The organic matter levels (Figure 20 ) in the field are good but show no clear relationship with yield. The southeastern edge of the field with high organic matter levels also correlates with a low bulk density. This is expected as increasing organic matter increases the porosity of the soil thus decreasing bulk density. But this area could also be wetter posing a problem at seeding.

The soil pH can be seen in Figure 21 . On the whole the pH level is good for all of the crops. The pH pattern does not correspond with any large-scale yield patterns. The high pH level at both ends of the field, may be a result of where the lime trucks turn and thus extra lime material is dropped there. The producer limes regularly generally every five years. The topography on a fine scale is given in Figure 22 . Although the elevation map does not match the yield patterns, the lower southeast corner has a higher organic matter content and observed to be wetter when soil sampling.

The aluminum distribution (Figure 24 ) reflects the parent material that formed the soil. On the north side of the small ridge, the low zone of aluminum was located in the same area with the K: Mg ratio problem. The P saturation % (Figure 25 ) was between 5 to 10% over 90% of the field. This "high" level of saturation may be a problem in coming years (restrictions on P additions to the field). The calcium and magnesium (Figures 26  and 27 ) are levels are good for crop growth. Ammonium and nitrate levels were determined in August/00 (Figures 28  and 29 ). The levels of ammonium were low which was normal for that time of year. However the nitrate levels were higher in a central strip that approximates a low barley yield zone - it may be due to drainage problems.

Conclusions & Recommendations

In summary, the soybeans and barley yield maps did show signs of yield stability, however the corn yield tended to mask the effect. From this case study it can be seen that the factors that affect yield are complex and many. The major problem is that they are not constant from year to year or for all crops. However, there are two zones that could potentially produce higher yields if more fertilizer was applied. The high levels of K should be reduced (K fertilizer levels should be kept to a minimum) so as to balance the Mg present in the field. Additional magnesium should be applied to those specifically low yielding zones that are seen on the normalized yield maps for soya and barley and Figures 11  and 14. Magnesium is an essential crop nutrient and the addition could potentially improve crop yields especially in dry years. Phosphorus levels are high (in the rich zone) and should be monitored. Very high P levels could impact zinc uptake by corn and cause a reduction in yields. This possibility would have to be explored further by tissue or soil analysis for zinc levels. The high bulk density zones that appear to be reducing crop yields in the headlands could be corrected by sub-soiling and by careful tillage practices.