Ask An APNI Expert: Precision Ag Models

Question: How can I develop a model for precision agriculture? 

Submitted by Tolera Goshu, Addis Ababa, Ethiopia

To answer this question, we need to first define “precision agriculture”. Precision agriculture (PA) according to the International Society of Precision Agriculture is “A management strategy that takes account of temporal and spatial variability to improve sustainability of agricultural production.” Notice that the word “technology” does not appear in the definition. Many people assume that PA is going to involve technology – sensors, satellites, computers, etc., but according to the definition, PA is about variability – how to measure it and how to manage it. Technology gives us tools to do this but developing a model for PA doesn’t depend on it. 

Crop performance can vary greatly across a landscape due to natural changes in soil properties plus the influence of field management history.

Now, let’s talk about variability. Temporal variability refers to weather driven variation in crop performance. We all agree that production, particularly in rain-fed agriculture, can vary considerably from one year to the next. So, a PA approach would never assume that yield potential and the inputs required to achieve that potential are the same each year. Spatial variability refers to changes in the field that affect crop performance. These can be naturally occurring changes such as soil type, texture, or slope, or they can be due to management such as lack of uniformity in input applications. 

So, where do we start to develop a model for PA? Yield history is information that every farmer regardless of size will know. Yield monitors are used for this in mechanized agriculture, but a smallholder farming 1-ha knows which part of the field consistently yields better than others and which part always seems to have problems. A simple sketch of the field is a great place to start. The farmer outlines “zones” of the field where consistent yield differences are observed.

The next step is to determine the factors causing these differences. Some factors can be observed by the farmer. Are there differences in soil texture that affect nutrient and water availability? Are there physical soil differences (rocky, hardpan, lack of topsoil) that affect crop growth? These areas can be managed for lower yield potential compared with other parts of the field. A common source of variability is heterogeneity in soil fertility, which cannot be seen and need to be assessed using soil sampling. The PA approach can begin by collecting separate samples from within the farmer-defined zones from the yield map. This will allow the farmer to have variable fertilizer recommendations to optimize yield within each of the zones – leading to more efficient production overall. 

Answer contributed by: Dr Steve Phillips, Principal Scientist, APNI, Benguérir, Morocco 

Submit your plant nutrition related questions via our website https://growingafrica.pub/ask-an-APNI-expert Selected questions, along with answers from our Institute’s experts, will be published in future issues. 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.