Opinion: Why farmers need field trials

Farming benefits from field trials but they are more expensive than modelling, Dr Jacqueline Rowarth writes.
THREE KEY FACTS
Dr Jacqueline Rowarth, adjunct professor Lincoln University, has a PhD in soil science and is on the board of directors of DairyNZ, Deer Industry NZ and Ravensdown.
OPINION
Does it matter if a conference is cancelled because of a lack of support? Yes, if it is a signal that bad forces are at play.
In a recent case, the longer-term implications for New Zealand’s goals are significant and negative.
Since 1987, soil scientists, rural professionals, and keen farmers have started the year at Massey University’s Fertiliser and Lime Research Centre Workshop.
Initially, the workshop was an invitation-only event, but its importance as a forum for discussion, challenge and exchange of ideas grew, and so did the attendance.
Covid didn’t stop it – the forum went online.
But the stress in the science and agri-business system this year has meant that the 2025 workshop (now under the name of Farmed Landscape Research Centre), has been rescheduled for 2026.
The cancellation of the 2025 workshop matters for its own sake, because the traditional robust debate and consequent improvements in future research and understanding, as well as the technology transfer, will be absent.
It is also a red flag for the state of the industry.
Uncertainty is high and morale is low — neither are good for creativity and progress, whether in the laboratory or on the farm.
And it is an alert for the negative impact on the Government’s goals of doubling the value of the export industry while reducing environmental footprint.
The workshop was postponed with the explanation that “many of our long-term supporters are undergoing internal reviews amidst the current challenges in our industry. Thus we’re giving our sponsors, presenters and delegates alike, some space to reset before returning to our planning for 2026.”
Interpretation — in my view — no money for sponsorship, no funds for conference attendance, and no time for writing papers.
And what of actual research funding?
The figures for redundancies have been in the news; more are in the offing.
Expenses are being cut across the board and the squeeze has already been felt.
The New Zealand Grassland Association Conference held in Oamaru at the beginning of November attracted 214 delegates.
Pre-Covid and budget cuts, nearer 300 would have been the expectation.
Further, papers based on models and surveys, rather than field trials, left some farmers wondering whether any of the research would actually be able to be implemented on their farm with the benefits visible in the bottom line.
The problem is that field trials are expensive, and it is difficult to make them applicable to more than one location.
The temptation is to build a model where parameters can be changed to fit more than one scenario.
But models are based on assumptions, and all have constraints and limitations.
A new book, Escape from Model Land, makes the point.
Dr Erica Thompson, University College London, but currently a visiting senior fellow at the LSE Data Science Institute, says “Whole careers can be spent in Model Land doing difficult and exciting things. Except these things are not real. Or rather, they do not apply to the real world — a delusion that has led governments and businesses into trouble”.
Thompson suggests that Model Land is a great place for theorists such as economists, climatologists, financiers, and political scientists because models are entirely controllable.
Researchers set the parameters, run their tests, and write with confidence about their results.
But Model Land is not real life and taking shortcuts because of funding will ultimately lead to problems.
Models in agriculture require testing on-farm and the risk and economic implications factored in.
The year the Fertiliser and Lime Research Centre Workshop started, “a meteorologist working for the British Broadcasting Corporation reassured a concerned viewer that rumours of an approaching hurricane were unfounded. Hours later, 22 people had been killed and billions of pounds of damage done by highly unusual hurricane-force winds”.
Thompson describes the 1987 weather forecast as both the most famous and the worst in British history.
Part of the problem was incomplete data in parts of the North Atlantic, but part of it was misplaced confidence in the model.
In agriculture it is difficult to imagine how a model ever could be complete because the climate is changing, the weather is variable and people are erratic — including in purchasing behaviour.
Economists and political scientists, and the models they run, have been influential in creating the current state of the economy, which is what has led to the searing cuts in the Government Budget, and hence to the science budget.
But the Government still has the goal of doubling the value of the export economy.
At the same time, the Climate Change Commission, having used models (with inherent assumptions, constraints and limitations), is suggesting that to meet nationally determined contributions of greenhouse gases in the next five years, sheep and beef farms should reduce by 72,000ha per year and dairy farms should by 3000ha a year.
More forestry?
And although high-value horticulture might replace some dairy, if it were feasible, the farmer would already have made the change.
High value carries risk.
That is why scientific research, robust debate and technology transfer are so important — and should include economists and political scientists as well.
The advances we need are not going to happen without it.