Yield data: more than a pretty picture

It is mid July, which, for the Australian cotton industry, means that picking is still going on for some growers. A different group of growers however is getting ready to start planting cotton again in three weeks time; the introduction of Bollgard III has brought these two events awfully close together! Most growers only have two priorities in-between two cotton seasons: finding time for a well-deserved holiday and working out a plan for the next crop. So what historical information do you base your crop management decisions on?

Quite often, crop management decisions end up getting dictated at least partially by old habits and environmental conditions but, more and more, data gathered from previous seasons gets analysed and used as well. As more and more data becomes available from different sources, it is important however to distinguish between the types of data that can be used to support a certain decision-making process. In this article I will focus on yield data and the role it can play in planning your next crop.

Up until recently, yield data was quite often only used to either confirm an existing belief or show what a grower already knew himself. This type of yield data service, also known as a 'pretty picture service', is getting less and less accepted in the marketplace as growers are starting to look at the ROI of their Precision Ag investments. It is common knowledge that farmers around the world want to reasonably be able to expect at least a $5 return for every $1 they invest; why should Precision Ag be any different?


In the case of yield data analysis there are a certain number of basic services that get performed on raw data: data gets calibrated, clipped to field boundaries, cleaned up and polygons get smoothed out in order to get a yield map which can be used for further analysis. From there on, it depends on your provider as to how much you actually do with your yield data and what your ROI for this process will be. Similar to NDVI imagery, you can look at yield data in two different ways: you either analyse the in-field variability, which are the differences in yield within a field in the same season, or analyse the trends that develop over time, which are the yield differences in the same field over a number of seasons.


Once the initial analysis has been performed, variability can be analysed in combination with any number of other data sources: NDVI Imagery, EM survey data, soil nutrient analyses, etc. This has to be a process however where the grower, consultant & technology service provider work together closely in order to get to a good end result. By analysing the data, looking at the crop management   strategy and combining that with some in-field ground-truthing, the reasons behind variability can be found.

Once the reasons behind variability in yield have been uncovered, a plan to increase the total field profitability can be formulated. When trying to lift the average yield for example, it is commonplace to aim for improving some of the lower performing zones within a paddock in order to lift the average yield. In some cases you will find however that lower performing zones are very hard to improve as there are certain constraints such as soil type, soil depth etc. In this case the plan might be to start applying nutrients at a variable rate, where the zones with soil constraints are getting a lower amount, only getting enough nutrients for the yield potential they have, and to put more nutrients in those zones where the performance is already relatively high. The reasoning in this scenario is that the good yielding parts of the field can still perform better whilst the lower yielding zones are already performing as well as they can. Although counter-intuitive for most growers, it has proven to be an effective strategy in some cases; it definitely proves that you need data to make these kinds of decisions... if you don't measure, you don't know!

Whilst growing the next crop, there will of course be environmental and other factors during the season that will play a role that cannot be predicted ahead of time; but with good data and an even better analysis, you can at least be prepared to face challenges with as much knowledge as possible.  All sorts of crop management decisions can be influenced by data from previous seasons, such as variable rate nutrition, irrigation scheduling and growth management to just name a few. With the right people, knowledge and tools, you can make agricultural data help you in your decision-making process.


It is important to record and store as much farm data as possible, but not every grower has the capability to record every kind of data. In the Australian cotton industry for example, there are still a significant number of basket pickers in operation that do not have the capability to record yield data. Whatever your recording capability might be, every bit of data is worth saving, and just because you can't record everything on your farm yet doesn't mean you shouldn't start building a (limited) database today.


If you want to know more about adopting precision technology in your farming operation, please reach out to me on the details below.

Written by: Reinder Prins. Reinder Prins is the Market Development Manager for Cotton Growers Services and as such is responsible for the Precision Agriculture initiatives within the organisation. Reinder can be contacted on +61 427 - 808 489 or via e-mail on reinderprins[at]cgs[dot]com[dot]au. This article was first published on www.precisionag.com