Pulse Beat, Soybean Fertility

Preliminary Results of Potassium Management of Soybean Production in Manitoba

Megan Bourns, MSc Candidate, Department of Soil Science, University of Manitoba

HISTORICALLY, POTASSIUM FERTILITY may not have been at the forefront of a typical Manitoba farmer’s nutrient management concerns given the extent of K-rich clays in the province. However, potassium fertility management is, and will continue to become, demanding of attention – and soybean production has a lot to do with this.

In Manitoba in recent years, there has been a very large and rapid expansion in soybean acres. In fact, soybean now occupies more than a quarter of the province’s annual crop land. This expansion, coupled with soybean’s high K removal rates in the grain at harvest (1.1–1.4 lb K2O/bu), has changed the total amount of K being removed from Manitoba soils over time. There has been a large increase in the last several years, and most of that increase is accounted for by the change in soybean acres (Figure 1). The expansion in acres, high K removal rates and increasing genetic yield potentials likely explain the increase in incidence of K deficiency symptoms in soybeans in recent years, especially as production expands into lighter textured soils that are inherently low in potassium.

According to the Manitoba Soil Fertility Guide, current soybean K fertility recommendations are identical to those for crops such as spring wheat and canola – both of which do not remove K to the same extent as soybean crops (Table 1). In addition, these thresholds and rates are lower than what is currently recommended in neighbouring soybean producing areas such as North Dakota, Minnesota and Ontario. With the growing prevalence of soybeans in Manitoba’s crop rotations, the increase in incidence of K deficiency symptoms in soybeans and the lack of comprehensive historical research for soybean K fertility in the province, it was time to reassess and update these recommendations.

The objectives of this two-year K fertility project were to determine the frequency of yield response to added potassium fertilizer across a range of soil test K (STK) levels, and to assess the effectiveness of different potassium fertilizer rate and placement combinations as a means to improve soybean seed yield. In order to achieve these objectives, two sets of trials were established: one set in field-scale on-farm trials and the other in intensively-managed small-plot trials.


The on-farm trials were established in conjunction with MPSG’s On-Farm Network to characterize yield response across a range of STK levels. In total, 20 site-years were established with background ammonium acetate exchange­able STK levels ranging from 52–235 ppm. Each site was a replicated strip trial with one treatment of either 60 lb K2O/ac banded away from the seed, or 120 lb K2O/ac broadcast and incorporated. Treatment strips were randomized and replicated along with untreated control strips.

Three of 20 sites responded statistically significantly, two being yield increases and one a yield decrease (Figure 2). Only two significant positive responses were found. While one of these responses was at a site that had a background STK level less than 100 ppm, the current threshold for recommending an application of K fertilizer, the other was at a site that had well over this level of K in the soil. A higher frequency of response was expected, with more positive responses being anticipated at the sites that were at, or below, 100 ppm STK.

Preliminary Conclusions

Two sites showed a statistically significant yield increase in response to potassium fertilizer, and one site had a significant decrease in yield. The frequency of response and behaviour of responsiveness at individual sites was not as predicted based on background ammonium acetate soil test K levels. No agronomically, or statistically significant relationship was found between back­ground ammonium acetate K levels and relative yield.


Small-plot trials were established in commercial fields, with seven site-years in total. At these sites, less than 100 ppm background STK level was desired to increase the likelihood of K responsiveness. The small-plot trials compared six combinations of K rate and placement: 30 or 60 lb K2O/ac sidebanded and 30, 60 or 120 lb K2O/ac broadcast and incorporated, as well as a control with no added K.

Deficiency symptoms were observed throughout the season at multiple locations in both 2017 and 2018. Early season deficiency symptoms showed at some locations at V2–V3 stage (Figure 3a). Symptoms were also identified at seed fill, and in some cases these symptoms persisted to leaf drop (Figure 3b). This indicated that the sites were, in fact, low in K. However, at harvest no significant yield responses were found to any treatment at any site. The lack of yield response was surprising, especially given the low background STK levels at these sites, and the presence of deficiency symptoms.

There was no agronomically or statistically meaningful relationship found between background ammonium acetate STK and relative yield regardless of placement and rate. No meaningful relationship between background levels and yield were identified regardless of method of ammonium acetate extraction, on a moist or dry basis.

Preliminary Conclusions

Optimum rate and placement of potassium fertilizer could not be determined as there was no yield response to added K. No statistically significant, or agronomically meaningful, relationship was found between background ammonium acetate STK levels and relative yield. Tissue data, seed oil and protein, and seed K concentration will be further analyzed in the coming months. In addition, a chemical study will be carried out to better understand K retention and release from these sites.


There were three major challenges encountered that could have influenced the ability to detect differences in yield as a result of K treatments. Both 2017 and 2018 field seasons were very dry, the effects of which were definitely noticeable in the light-textured sandy soils where the small-plot sites were established. If moisture was the greatest yield-limiting factor, and full yield potential was not being achieved, then the demand for K may not have been as high.

Another challenge was the variability across the sites, which operated at a very small spatial scale (Figure 4). The differences in soil test potassium from one location to another across a site area were much greater than anticipated. The lack of consistency that results from this variability also influenced the ability to detect yield effects of the K fertilizer treatments.

Lastly, ammonium acetate soil test K was not a reliable indicator of potassium response. The ammonium acetate test for exchangeable potassium was relied on to select sites that were likely to be responsive to potassium addition, based on current threshold levels from the Manitoba Soil Fertility Guide. However, potassium responsiveness was not accurately predicted based on the spring background ammonium acetate potassium levels at these small-plot sites.