Development of Management Zones for Site-Specific Fertilization in Mustard Fields ^†

Abstract

Smart-farming practices can help to overcome the impacts of soil and crop variability on crop yield. This article studies the impact of fertilization treatment on the production of a major oilseed crop, mustard (Brassica campestris L.), in an arid region of Pakistan. Soil-nutrient sampling was performed, using 20 × 20 m grids to characterize and quantify the variation of mustard yield and soil nutrients. Based on the results of a significant correlation between yield and soil nutrients, management zones (MZs) were developed that use Variable Rate Fertilization (VRF) to increase crop production and profitability while reducing environmental risk.

1. Introduction

Sustainable agricultural-production systems are necessary to feed the global population, as it is projected to increase by 35% in 2050, which requires a 70–100% increase in food production, and there is an insistent demand to develop this production more efficiently [1].

Mustard (Brassica) is a major oilseed crop grown in Pakistan, but production capacity is limited in Pakistan, despite having huge potential. Therefore, edible-oil import is necessary, which hast cost millions of dollars for many years [2].

Agriculture management in semi-arid and arid regions faces a lot of challenges, because soils are highly variable due to unpredictable rainfall, in terms of both amount and intensity [3]. The precise application of fertilizers can be improved by site-specific nutrient management (SSNM), especially for crops such as rice, maize, wheat, etc. SSNM is a component of precision agriculture (PA), and it helps in improving Nutrient Use Efficiency (NUE) in soil and crops [4].

Therefore, it is necessary to implement efficient techniques to accurately measure variations in soil properties within the field and to draw precise Management Zones (MZs) [5]. To achieve this goal, it is important to explore the crop-production potential, quantify yield gaps, identify limiting variables for enhancing mustard yield, and develop MZs for the precision application of fertilizer.

2. Methodology

A mustard field (1 ha, 33°6′50″ N 73°0′57″ E) was selected during the mustard growing season 2021–2022 at the University Research Farm “Koont”, PMAS-Arid Agriculture University Rawalpindi, to investigate the spatial variation effect on mustard yield. The selected field had been under uniform management for the past few decades. Soil variation was studied in the selected field, using grids and soil-nutrient samples, e.g., nitrogen (N), phosphorus (P), and potassium (K), which were recorded with the help of a proximal sensor from each georeferenced grid-point. Yield was measured manually with a hand sickle. The collected data were organized for statistical and geostatistical analysis. The proposed methodology is shown in Figure 1.

Descriptive statistics and multivariate geostatistical methods were used to determine the existence and pattern of spatial variability in the selected fields. Descriptive statistics were utilized to calculate the coefficient of variance (CV), which defines the existence of spatial variability in the mustard field, by using Minitab 21 statistical software (Minitab, LLC, State College, PA, USA). Geostatistical analysis was performed using ArcGIS 10.8 software, and semivariogram parameters were estimated, which were used to define the pattern of variability by the kriging interpolation technique. Different models, such as the exponential, circular, spherical, and Gaussian semivariogram models were tested, to find the parameters of the semivariogram nugget, sill, and range [6].

3. Results

3.1. Interpolation of Soil Nutrients and Mustard Yield

There was significant spatial variation in the mustard field, as evidenced by the kriged interpolated maps of sol nutrients and mustard yield, requiring the establishment of MZs for site-specific nutrient management. Higher values of nutrients were mostly observed on the south sid, and lower values on the central and north sides of the field, as indicated in Figure 2.

The visualization of the maps revealed that variation in soil properties caused substantial variation in yield. The high-yielding values were observed on the north-south side of the field, and low values were on the northwest side of the field. The results of interpolated maps suggested moderate-to-high variation in nutrients and mustard yield.

Delineation of MZs for Mustard Crop

The clustered soil nutrients and mustard-yield data were imported into ArcGIS 10.8 software to analyze and develop the MZs. The developed MZs, on the basis of clusters, were classified into four productivity zones: very poor, poor, medium, and good for mustard, as shown in Figure 3.

The delineated MZs showed different productivity levels within a field, which required variable-rate fertilizer-application to improve the mustard yield. A higher productivity level was observed on the southeast side of the field. The southwest and the middle of the field had average productivity, whereas the northern side had extremely low productivity. Developed MZs can be used for site-specific soil- and crop-sampling, significantly reducing farm-production costs while protecting the environment from the harmful effects of chemical fertilizer. The authors in [7] summarized how the developed MZs can reduce the over- and under-application of fertilization for wild blueberries, which reduces the cost of production of the farm and increases the yield. Up to 40% of agrochemicals was saved by the accurate appliication of fertilizer according to the developed MZs [8].

4. Conclusions

Soil-sampling planning and management of mustard crops is significantly assisted by an understanding of soil parameters. Cluster analysis could provide a method for planning and organizing the spatial variability of soil parameters and crop production within individual fields. Input-use efficiency, production costs, and environmental advantages could all improve with the implementation of MZs-based variable-rate fertilizers. To better understand the effect of climate change on crop-production variability and to determine whether or not MZs have the opportunity to conduct site-specific nutrient management, it would be useful for future research to correlate soil attributes with climate change.