GPS-Based Soil Fertility Mapping of a Development Block: Macro- and Micronutrient Status with Ordinary Kriging

This M.Sc (Agriculture) thesis in Soil Science and Agricultural Chemistry builds a GPS-referenced soil fertility map of one development block, so that fertiliser advice can move from district-wide blanket recommendations to village- and field-scale guidance.

About 150 surface soil samples (0–15 cm) are collected on a roughly 1 km grid with GPS coordinates recorded at every point, and analysed in the departmental laboratory by standard methods: pH and EC (1:2.5 soil–water), organic carbon by Walkley–Black wet oxidation, available N by the alkaline permanganate method (Subbiah & Asija, 1956), available P by Olsen's method (Olsen et al., 1954) — or Bray's method for acid soils — available K by neutral normal ammonium acetate with flame photometry, available S by CaCl₂ extraction with turbidimetry, and DTPA-extractable Zn, Fe, Cu and Mn (Lindsay & Norvell, 1978) on an AAS.

Results are classified as low, medium or high using the ratings adopted for Soil Health Cards, summarised by Parker's nutrient index, related to each other by correlation, and then interpolated with ordinary kriging in R (gstat) after semivariogram modelling, with maps finished in QGIS. The thesis ends with nutrient-deficiency maps and zone-wise fertiliser recommendations.

The bundle provides the sampling protocol, lab method list with citations, rating tables, geostatistical workflow, thesis chapter plan, a 12-slide seminar outline and a viva bank. No analytical values are included — your lab data make the thesis.

Abstract

Soil nutrient status varies considerably within short distances because of parent material, topography, cropping history and fertiliser practices, yet fertiliser recommendations are often made for large areas. This study assesses the fertility status of surface soils of a development block using GPS-referenced grid sampling, standard laboratory methods for pH, EC, organic carbon, available N, P, K and S and DTPA-extractable micronutrients, and maps their spatial variability using ordinary kriging. Nutrient index values summarise block-level fertility, and deficiency maps support site-specific nutrient recommendations.

Introduction

The Soil Health Card scheme made soil testing a national programme, and GPS-based sampling allowed results to be tied to locations. Geostatistics adds a further step: from point values it estimates nutrient levels at unsampled locations along with the uncertainty of those estimates, producing continuous maps that planners and extension workers can use.

Review gap

Many fertility-status studies report only means and percentage deficiency, without describing spatial structure. Where kriged maps exist, they are often at district scale, too coarse for block-level extension planning, or cover only macronutrients. A block-level study covering macro- and micronutrients with semivariogram analysis fills this gap for the selected area.

Hypotheses

• Available nutrients and organic carbon show moderate-to-strong spatial dependence at the block scale.
• A measurable proportion of the block is deficient in at least one micronutrient, particularly zinc.
• Soil pH and organic carbon are significantly correlated with the availability of several nutrients.

Feasibility

• Laboratory: all methods use standard equipment in SAU soil-testing labs.
• Software: R and QGIS are free.
• Field: a block of this size can be sampled in three to four weeks after kharif harvest with a field assistant.
• Time: 26 weeks, including ORW approval, sampling, analysis, mapping and thesis writing.
• Economic: chemicals, AAS time and fuel for field travel are covered by the departmental PG contingency grant; no special equipment has to be purchased.

Farmers in the study block apply fertilisers largely by habit or by blanket recommendations, often over-applying nitrogen while ignoring secondary and micronutrients. Soil test data exist as individual cards, but they are rarely combined into maps that show where deficiencies cluster and how nutrient levels change across the block. Without spatial information, extension workers cannot prioritise villages for zinc or sulphur application or target organic-matter improvement where carbon is lowest.

This thesis addresses that gap by collecting GPS-referenced surface soil samples across the block, analysing them for major, secondary and micronutrients with standard methods, quantifying spatial variability with semivariograms, and producing kriged fertility and deficiency maps with zone-wise fertiliser recommendations that the KVK and the agriculture department can use.

In scope

• One development block; about 150 surface samples on an approximately 1 km grid, adjusted to cultivated fields.
• Chemical fertility parameters listed in the objectives; soil texture by feel/hydrometer for a subset as supporting data.
• Descriptive statistics, correlation, nutrient index, geostatistics (ordinary kriging) and GIS mapping.

Out of scope

• Subsurface horizons and full soil profile classification.
• Plant tissue analysis and crop-response trials to recommended doses.
• Remote-sensing prediction of nutrients from satellite imagery (noted as future scope).
• Heavy-metal contamination assessment.

Research design

A descriptive, spatial survey design: systematic grid sampling, laboratory analysis, classical statistics and geostatistical interpolation.

Study area and sampling

• Block selected in consultation with the advisory committee and the district agriculture office; boundary from official administrative data digitised in QGIS.
• Grid: a 1 km × 1 km fishnet in QGIS; one composite sample per cell from a cultivated field nearest the cell centre.
• Composite: 8–10 cores in a zig-zag pattern from 0–15 cm with a stainless-steel auger (avoiding bunds, manure pits and recently fertilised spots), mixed, quartered to about 500 g.
• Records: GPS latitude/longitude, village, crop grown, irrigation source, date. Samples coded S001–S150.
• Timing: after kharif harvest, before rabi fertiliser application.

Laboratory methods

Quality control: reagent blanks, duplicates for every tenth sample and a laboratory reference soil in each batch.

Analysis

1. Descriptive statistics, CV, skewness; log-transform skewed variables.
2. Rating into low/medium/high; nutrient index = (1×Nl + 2×Nm + 3×Nh) ÷ Nt (Parker et al., 1951); < 1.67 low, 1.67–2.33 medium, > 2.33 high.
3. Pearson correlation among properties.
4. Semivariogram modelling (spherical, exponential, Gaussian), nugget:sill ratio for spatial dependence (Cambardella et al., 1994: < 25% strong, 25–75% moderate, > 75% weak).
5. Ordinary kriging; leave-one-out cross-validation (ME, RMSE).
6. Classified maps and deficiency-area statistics in QGIS.

Timeline (26 weeks)