Whole farm test additional

The aim of whole-farm soil testing is to give a better definition of the spatial variability (paddock by paddock) in soil fertility, Joanna Cuttance writes.

Farmers need to ensure they are applying the right amount of nutrients in the right place at the right time.

To achieve this many farmers generate fertiliser plans by sampling a limited number of paddocks which represent the soil type and farm use of the whole farm, but some have extended this approach by doing whole-farm testing.

Ravensdown chief scientific officer Dr Ants Roberts said the purpose of whole-farm testing was not to get a better average fertility status per farm management unit (FMU) by taking more samples to calculate an average, but to give a better definition to the spatial variability (paddock by paddock) in soil fertility.

It was not a replacement for the normal soil fertility monitoring programme which used multiple sampling transects in FMUs, for example, effluent versus non-effluent blocks, contrasting soil types, and different land slope categories.

Whole-farm soil testing was appropriate for hill country farms, along with intensive dairy farms, though the intensity of soil sampling might be different between farm types, Roberts said.

Whole-farm soil testing sought to identify those paddocks which were below optimal soil fertility for pH, or one or more nutrients. Those paddocks could then be given capital lime or P to lift those levels, and for K a higher than maintenance K level might be appropriate.

Generally but not always, S was applied regardless of soil fertility level unless very high, which could happen on high anion storage capacity soils.

As well as identifying paddocks deficient in nutrients, Roberts said whole-farm testing also identified paddocks which were above optimal, therefore the nutrients in excess of long-term requirement could either be withheld or less than maintenance applied. Maintenance nutrient levels are defined as ‘the rates of nutrients required to maintain the status quo soil fertility level’.

With whole-farm testing, the main soil fertility criteria looked at was the Olsen P status and the soil pH. The other nutrients like K, S, Ca and Mg could also be assessed. K and S were mobile nutrients in the soil and did not lend themselves to capital application when the levels were low enough that they could limit pasture production.

“Fertiliser mixes get quite complicated if you try too many different rates of P, K and S,” Roberts said.

The Ravensdown team recommended whole-farm soil testing was conducted once every five years or so, with the intervening years having the normal monitoring transects by farm management units to keep track of soil fertility trends. Although, he said, there were some farmers who insisted on carrying out whole-farm soil testing every year or every couple of years.

Averages tell the story

The basic problem with soil sampling is that there is a lot of error associated with the results, soil scientist Robert McBride says.

If soil tests were taken properly, depending on the nutrient, the errors were typically +/-15% to 30%. This meant if the Olsen P result was 25, the actual soil level was probably somewhere between 20 and 30.

“If you were to keep sampling the same transect over and over, you’d get different results each time,” he said.

However, McBride said if you averaged those results, it would be about right.

This was why the standard protocol had been to sample the same transects in the same paddocks every year.

“By averaging multiple transects in a given block, it gives a better indication of the actual levels, and by plotting those averages over time, you can see trends,” McBride said.

Even though none of the soil tests were very accurate by themselves, used together, and particularly used together over time, farmers could get a good idea of where things were. The standard procedure to use established transects and blocks was an effective way to deal with variability.

McBride said the problem with all-paddock or all-farm testing was that each sample was treated as if it accurately represented the nutrient levels in that paddock. He said this was not the case.

“Assuming the sample was taken properly, which is a dangerous assumption, it is +/- 15% to 30%.”

If every paddock had an actual Olsen P level of 25 for example, the test results would come back ranging between 20 and 30. The ‘30’ paddocks would get fertilised less, and the ‘20’ paddocks fertilised more, he said, even though in reality they were exactly the same.

Quality not quantity

Farmers do not have to sample soil every year, but it depends on what they were wanting to do, soil scientist Robert McBride said.

If everything was fine and farmers were just trying to maintain nutrient levels, they could get away with not sampling every year.

If farmers were trying to build fertility and spend lots of money – the more information the better.

It must be kept in mind that the quality of the samples was more important than the quantity or frequency.

“Poor soil sampling leads to inflated results, and you can’t make good decisions with bad information.”