Determine and revise estimates of the nutrient uptake and removal of crops commonly grown in western Canada. Develop a user friendly on-line and mobile app for determining nutrient uptake and removal estimates.
The project objective was to develop new estimates for crop nutrient uptake and removal, based on analyses of grain and biomass samples collected from commercial fields and values from existing literature. Through the collaborative efforts of Nutrien Ag Solutions and Manitoba Agriculture, over 2200 grain and biomass samples were collected from across the three prairie provinces from 2020 to 2022, and analyzed for macronutrient (N, P, K, S) and micronutrient (Cu, B, Zn) uptake. Results confirmed that some of the nutrient uptake and removal values are aligned with existing estimates (e.g., CFI Nutrient Uptake and Removal Guidelines for Western Canada, 2001) whereas others needed revision. Importantly, existing ranges for nutrient removal do not capture the full extent of the observed variability in nutrient uptake and removal, underscoring the importance of using any uptake and removal estimates together with regular soil testing for informing fertilizer management decisions. The survey provided new uptake and removal estimates for micronutrients (boron, copper, and zinc).
Additionally, we conducted a review of existing literature and research relevant to Western Canada. In addition to peer-reviewed publications, online university repositories were searched for relevant theses and dissertations. Other databases included the ministries of agriculture and commodity groups from across the prairies. Research groups also were contacted for raw field data. We developed a list of criteria to include (or exclude) sources of information with the goal of only using data deemed relevant and robust. Together, these sources proved to be a rich source of information for corn, oat, wheat, winter wheat, canola, soybean, faba bean, lentil, and pea; however, search criteria were not met for barley, durum, flax, mustard, and chickpea, and these were not included. The data accessed for some crops was abundant. For example, for spring wheat, we acquired 2596 N, 858 P2O5, 84 K2O, 372 sulfur, 84 boron, 116 copper, and 196 zinc data points. In contrast, we found only 24 data points for oat for each of the nutrients (i.e.,168 total data points).
When comparing the literature values, which represent data largely from small-plot experiments, to the survey data, the literature-based yield data were frequently within 10% of the survey data for wheat, oat, canola and lentil, but were lower for corn (124 versus 150 bu·acre-1) and higher than the survey yields for the remaining crops. It is probable that yield losses in commercial harvesting operations versus hand-harvested research plots account for the higher yields reported in the literature.
In general, the literature revealed similar or lower estimates for nutrient removal (i.e., nutrient in the grain) than from the survey with a few notable exceptions, one of which was S removal in canola. Whereas the survey estimate (mean) was 0.19 lb S·bu -1 canola, the literature-based estimate was 0.42 lb S·bu -1. Given the importance of S in canola production, and the variance in the estimated removal, we conclude that it is prudent to opt for a higher estimate of S removal to avoid potential deficiencies. Similar results were observed for S removal in flaxseed. Additionally, survey values for potassium removal in soybean were lower than the literature-based estimates (mean 0.89 lb K2O·bu-1 versus 1.29 lb K2O·bu-1), which may warrant further investigation.
A survey of nutrient uptake of forages (alfalfa, clover, forage grass, barley silage, corn silage) in the published and grey literature was conducted. Although a number of relevant reports were accessed, parameters reported in the various studies were inconsistent, including growth stage at which biomass or nutrient uptake was determined. Moreover, although some studies included nutrient uptake, the vast majority focused on yield, and far fewer studies included both yield and nutrient uptake. Ultimately, we were unable to compile consistent and coherent data to justify any modifications to the existing nutrient uptake and removal guidelines for forages published in the 2001 CFI guidelines.
- Regular soil testing is a critical tool for assessing current soil nutrient status and determining appropriate fertilizer application rates to achieve crop yield goals.
- Nutrient removal estimates are an additional tool for assessing nutrient addition required to maintain or build soil fertility levels. Crops are not able to extract all the total nutrient available in the soil (i.e., soil plus added fertilizer) and thus estimates of available nutrient(s) should be greater than estimates of nutrient removal to maintain soil nutrient levels.
- Revised nutrient removal values developed specifically for crops grown in Western Canada provide estimates of nutrient removal, although it is essential to recognize that due to known variability in the data (including weather-associated variability), these estimates provide guidance but should not be viewed as prescriptive.
- For some crops and nutrients, lower grain concentrations observed in the current survey suggests that management practices and modern varieties have resulted in improved nutrient use efficiency on a per bushel basis, although higher yields remove more nutrients on a per acre basis.
- Although the survey data indicates lower sulphur (S) removal by canola (and flax seed) than previous estimates, given the importance of S in canola production, and the variance in the estimated removal, it is prudent to opt for a higher estimate of S removal than suggested by the survey data to avoid potential S deficiencies.
- An on-line calculator and an Excel-based calculator have been developed. The calculators use the 75th percentile of the survey data as the nutrient coefficient, with the goal of limiting the risk of underestimating nutrient removal. The 75th percentile represents that point at which 75% of the survey values were below the coefficient value and 25% of the values were above the coefficient value.