An on-farm approach to monitor and evaluate the interaction of management and environment on canola stand establishment
Date: May 2021
Term: 3 years
Status: Completed
Researcher: Christiane Catellier, Indian Head Agricultural Research Foundation
SaskCanola Investment: $87,054
Total Project Cost: $87,054
Funding Partners: n/a
Project Summary
Key Messages:
Canola cultivar was the most influential management variable on percent emergence and early growth rate, and also significantly impacted the spatial uniformity of emergence in this observational study.
Seeding dates had a consistent and significant influence on emergence; however, the effect of seeding dates was mainly a function of environmental conditions.
All environmental variables that were measured influenced the emergence response and had additive and sometimes interactive effects with management variables.
Temperature and heat units were consistently more influential on emergence than precipitation and moisture.
Improved stand establishment is one of the key factors in achieving the canola production goal of 52 bu/ac average yield, specified in the Canola Council of Canada's strategic plan. Rapid and uniform canola emergence ensures the crop yield potential by reducing the time window of seeds and seedlings’ susceptibility to insect, weed, and soil-borne disease pests, and by establishing a uniformly developing crop that can be managed efficiently.
Many canola research projects have been conducted to improve our understanding of agronomic and management practices in canola production systems. Management factors that have been shown to be important in canola emergence and stand establishment include seeding speed and depth, seeding date, seed size, row spacing, crop rotation, and stubble management. However, the results of these studies have been inconsistent across western Canadian climatic conditions where environments are highly variable and can be significantly yield-limiting. In fact, agronomic field experiments are usually designed to isolate the responses to specific treatment effects while factoring out the variability from environmental conditions. Yet, under many circumstances, the influence of management depends on other management factors or on environmental conditions such as seedbed moisture and temperature. Thus, an observational on-farm study was conducted to improve our understanding of the interactive effects of management and environment on canola stand establishment.
This observational study was conducted on commercial canola fields in collaboration with six local producers around Indian Head, Saskatchewan, from 2018 - 2020. Environmental and agronomic data were collected from several fields throughout the growing season and the producers provided their individual crop management data. The additive and interactive effects of management and environment on the speed, temporal uniformity, and spatial uniformity of canola emergence were examined using forward-selection multiple regression and competing models approaches.
Table 1. Replication at the sample site, field, and operation level in each growing season throughout the study.
1 Information regarding fertilizer form, placement, and timing was not provided in enough detail in most cases to include in the analysis
2 Producer-reported row width was used to calculate plant density but was not included in the analysis
Environmental variables were averaged or totaled for the pre-seeding date period and over post-seeding date intervals of 7, 14, and 21 days for each sample site. Canola emergence rate, early growth rate, and spatial uniformity of emergence were modeled using a number of management and environmental variables including cultivar, soil fertility, soil moisture and temperature, growing degree days, crop rotation, seeding date, seeding depth, and residue cover along with many others (Table 2). Mixed effects models were used to account for unbalanced and nested effects resulting from the experimental design. Emergence rate was modeled as percent emergence with days after seeding as an indicator of both how quickly germination is occurring, and which proportion of seeds become established seedlings. The early growth rate was modeled as the average growth stage with days after seeding. This indicates how quickly seedlings are developing as they come out of the ground. The standard deviation of plant density, regressed against plant density to account for non-independence, was used as an estimation of the spatial uniformity of emergence.
Emergence Rate
The base model consisted of the curved relationship between percent emergence and days after seeding, with no additional management or environmental variables. Forward-selection modeling was utilized to determine whether the addition of explanatory variables (management or environmental) had a significant effect on the base model response. When the effect of each explanatory variable on canola emergence was examined individually, greater cereal frequency and reduced canola and pulse frequency in rotations were associated with higher percent emergence, which could be explained by the prevalence of soil-borne diseases in fields with tight rotations of broad-leaf crops. Higher percent emergence was also associated with additional insecticidal active ingredient in the seed treatment, greater seeding depth, lower seeding rates, higher applied phosphorus (P) rate and residual topsoil nitrogen (N), and with higher soil organic matter content and cation exchange capacity. Higher temperature and heat units along with greater precipitation and soil moisture before and after seeding were also associated with higher percent emergence.
As the explanatory variables were highly intercorrelated, a competing models approach was utilized to assess the relative importance of each explanatory variable and whether different combinations of management and environmental variables had additive or interactive effects. Models were ranked from lowest to highest AIC and the model weight was calculated using the difference in AIC relative to the top model. AIC is a measure of prediction error, and so the model weight indicates the probability that a model is the best representation of the emergence response. The top-ranking model was weighted significantly higher than other models, revealing that there was 80% probability that “cultivar” and “average soil temperature 14 days after seeding” were the most influential management and environmental variables on canola emergence rate, respectively. Cultivar was included in the top four models in combination with four of the different temperature variables. All other management variables were weighted much lower. The top-weighted model also had a significant interaction between cultivar and soil temperature, indicating that cultivars’ emergence rates were differentially affected by soil temperature.
Early growth Rate
The base model consisted of the linear relationship between growth stage and days after seeding. When the effect of each explanatory variable on early growth stage was examined individually, a greater frequency of canola and lower frequency of cereals in a crop rotation were associated with fewer days to emergence and a slightly higher growth rate of canola. Cultivar had a surprisingly significant impact on growth rate. Higher seeding rates were associated with longer days to emergence but higher growth rate. This could be explained by seedling competition. Later seeding dates were associated with fewer days to emergence and much higher growth rates. Higher applied N and residual topsoil N were related to a slightly earlier emergence and higher early growth rate, while higher rate P had little impact on days to emergence and the opposite effect on early growth rate. Higher air temperature, soil temperature, and heat units before and after seeding were associated with earlier emergence and higher early growth rates. Greater precipitation before seeding was associated with a faster growth rate, while higher soil moisture was associated with a delayed start and slower growth rate. According to top ranking model, there was 99% probability that cultivar and average soil temperature 21 days after the seeding were the most influential management and environment variables on canola growth rate. The top-ranked model again included a significant interaction between cultivar and soil temperature.
Spatial Uniformity of Emergence
When studying the effect of individual management and environmental variables on spatial uniformity of emergence, the cultivar was again a primary variable. An additional insecticidal active ingredient in the seed treatment, longer hypocotyl length (deeper seeding depth), higher N fertilizer rate, greater air and soil temperature and heat units had positive effects on spatial uniformity. Model weights were lower overall when competing models were compared, indicating that the prediction of spatial uniformity was more uncertain than emergence rate or growth rate, and that the variables measured were more equal in their influence on spatial uniformity. Environmental variables were more influential than management variables overall.
In summary, this observational study identified canola cultivar and post-seeding date temperature as the two most influential management variables that consistently impacted emergence response variables. Seeding date was also consistently and significantly influential on emergence, however the effect was not additive when combined with environmental variables, indicating that the effect of seeding date was mainly a function of environmental conditions. Nearly all the environmental variables measured consistently influenced the emergence response and had additive and sometimes interactive effects with the management variables. Temperature and heat units were consistently more influential on emergence than precipitation and moisture. Overall, the study was valuable in demonstrating the potential of on-farm observational studies in agronomic research. Future expansion of this observational study to different agricultural production regions would benefit producers.
Acknowledgment:
The cooperative involvement of local producers was invaluable in the completion of this project.