For in crop variable rate application of fertiliser and manure you need reliable data to base your applications on. Is it better to rely on variation in biomass or on variation in soil conditions? Experts from Wageningen Universiteit & Research are trying to find that out.
Several farmer participants of the Dutch experimental ground for precision farming (NPPL) and experts from Wageningen University & Research (WUR) are trying to develop better strategies for in crop variable rate application of fertiliser and manure.
They currently focus on grassland as effective grassland management is quite different from the management of other crops. It is harvested several times per year, either mechanically or by grazing cattle, and fertiliser and manure are also applied several times per year. This makes it stand out from arable crops but on the other hand, the challenges are the same: do I rely on variation in biomass or on variation in soil conditions for my variable rate applications of fertiliser and manure?
The WUR experts say it will take another three to five years before the scientific answers and proof for these questions are available.
Text continues underneath image
Because grass, but also grass-clover mixtures and crops like alfalfa, are harvested several times per year, their management differs quite a lot from that of ‘regular’ arable crops.
“You’ll be amazed by interaction between several yields and by the recovery of grassland after it is harvested”, says WUR researcher Bert Philipsen. “The yield potential not only depends on the soil and crop conditions, but also on the quantity and quality of previous yields. What we for instance see is that heavy yields result in less heavy next yields and vice versa. These interactions between yields make it very challenging to manage it right, let alone perform site-specific or variable rate management. I think it will take another three to five years before we can manage this effectively. Until then, I advise farmers and contractors to focus on more precise management instead of on variable rate management. For instance, by combining data from amongst others (scientific) growth models, protein predictions, drone and satellite images and the soil.”
Fellow WUR researcher Idse Hoving feels that farmers and contractors can best rely on existing scientific models, algorithms and advice for applying fertiliser and manure. “Such advice exists in The Netherlands and definitely also in other countries. If you apply fertiliser and/or manure per field according to such an advice, then I’m sure you utilise the major part of the potential of your soils and crops. Moreover, since such an advice takes into account the organic matter balance and the CEC (Cation Exchange Capacity) of the soil.”
CEC is the total capacity of a soil to hold exchangeable cations and it influences the soil’s ability to hold onto essential nutrients and provides a buffer against soil acidification. Soils with a higher clay fraction tend to have a higher CEC.
Text continues underneath image
Hoving has another advice: “It can be common practice to mix soil samples but try not to. Take soil samples on those spots that, according to the field history and to your own experiences, are likely to differ/vary. A mobile soil scanner also is perfectly suitable for this. Keep track of the (site-specific) applications of fertiliser and manure, and of your yields. By collecting these data consistently, your insights and knowledge will gradually improve and become more reliable and viable.”
Bert Philipsen on his behalf warns not to rely too much on the relation between soil sampling and scanning and fertilisation. “There’s guidelines for the soil nitrogen supply capacity in relation to the percentage of organic matter. For grass on sandy Dutch soils, the soil nitrogen supply capacity typically is 25-30 kilos per percent organic matter and for clay it is 10-15 kilos of N per percent organic matter. In reality however, the true supply capacity highly depends on (soil) temperature and soil moisture and that makes the relationship between organic matter and nitrogen supply capacity quite unmanageable.”
Text continues underneath image
Jouke Oenema, also a WUR researcher but focussing on soil fertility and manure application, says that establishing links between soil, fertilisation, biomass and yields is difficult because of the lack of stability. “What are the right indicators? How do you determine the relationship between, for example, biomass, soil type and condition and fertilisation? And how reliable are those relationships and measurements?”
Do you fertilise spots within fields with less biomass more or rather less? Higher rates should lead to a more uniform crop, but perhaps the yield potential of the soil is the culprit and more fertiliser or manure is a waste of money. Maybe it’s better and more valuable to try and divide the available (and permitted) amount of fertiliser and/or manure better over the total acreage than to vary it within fields.
At Dutch farmer participant of the Dutch experimental ground for precision farming (NPPL) Ad van Velde, it was noticeable that grass yields (dry matter and protein content) on long-term grassland were consistently 20% higher than yields on former arable crop fields. Field history therefore also is a very important factor to consider. Watch the video.
Several NPPL-participants and their experts have found that a suspected correlation between soil electrical conductivity (EC) in the 0 – 30 cm top layer and yield quantity and quality isn’t always present. Several fields have been scanned with a soil scanner, a Veris MSP3 in this case, and although the scanner indicated differences in EC, this difference wasn’t always noticeable in grass yields evaluated with a NIR-sensor on forage harvesters.
WUR specialist researcher Herman van Schooten says: “On particular fields, a higher EC resulted in a lower yield in kilos of dry matter per hectare. That is unexpected and illogical. We also found a large spread and no correlation between the percentage of organic matter and the site-specific yield on fields. We suspect that the dry conditions during the soil scanning in January 2019 and also the small height differences of the fields of no more than one meter, may be causes of the lack of expected correlation.”