Operation Overseer: farmers learn to live with new nutrient limits

Precision farming systems like this N-Smart sprayer can dose paddocks exactly to help in hitting new nutrient limits.

Precision farming systems like this N-Smart sprayer can dose paddocks exactly to help in hitting new nutrient limits.

Canterbury's farmers are being wrapped up tight in new environmental regulations. But how is New Zealand's latest stab at "world best" performance-based rule setting really working out? JOHN McCRONE investigates.

Canterbury farmers know all about Overseer version 6.2. The shiny new world of farming within environmental limits rests on this one bit of software for modelling nitrate leaching. And suddenly in April comes an update which produces haywire results.

Ashburton crop and dairy farmer Ian Mackenzie – until recently, Federated Farmers environment spokesperson – says overnight, just because of tweaks to the program, farmers found the calculations required by the new Canterbury Water Management Strategy (CWMS) gave completely different outcomes.

"Many people's nitrogen loss numbers doubled. Some people's numbers went down, but a lot more went up, and it just blew everybody's faith in the model to pieces."

Mackenzie says Overseer was originally designed for the high rainfall and heavy soils of the North Island. The software has needed steady upgrading to make sense in the low rainfall, often stony setting of Canterbury. But there were obvious nonsenses in the latest results.

"We have a neighbour who grows blackcurrants. Their currants went from 6 [kilograms of leached nitrogen per hectare per year] to 45. When you delved into why that happened, Overseer was assuming blackcurrants were the same as grapes."

Ex-Federated Farmers environment spokesman Ian Mackenzie in a panel discussion on irrigation.
Gerald Piddock

Ex-Federated Farmers environment spokesman Ian Mackenzie in a panel discussion on irrigation.

Grapes of course take up far less fertiliser, scoffs Mackenzie. So it would indeed look like the amount planned to be used on a blackcurrant crop would drain through the root zone, sink into the water table, spread out to cause algal blooms and water problems in the surrounding streams, rivers and lakes.

So garbage in, garbage out, with a computer program, as they say. And yet this is the tool which is meant to be underpinning New Zealand's latest grand adventure in performance-based regulation.

It is how we do things. We now have a whole history of such legislative reforms, from our fishing quota system and liberalised Building Act to our famous Resource Management Act (RMA).

Elsewhere in the world – Europe in particular – agriculture's impact on the environment is being dealt with by strict input-side controls. Legislators make rules about how much fertiliser farmers can toss on the landscape, how many head of stock per hectare they can run. Things are done by the book.

New Zealand, as part of the great leap forward of the 1980s Rogernomics era, believes instead in output or effects-based regulation. When it comes to an issue like water quality, an industry is set its broad goals, the general targets it needs to achieve, then left to work out for itself how best to meet those expectations.

And of course it is a better way to do things, says Mackenzie. He, like most farmers, shudders at the thought of a prescriptive European Union style approach to agriculture. Being judged on outcomes creates the space for innovation and flexibility.

Soil consultant Doug Edmeades says Overseer was employed out of political expediency.
Mark Taylor

Soil consultant Doug Edmeades says Overseer was employed out of political expediency.

Think of how often Canterbury has had to shift its production focus over the past 60 years in going from wool to lambs to crops to dairy. Horticulture is probably coming next. Regulation in the modern world has to leave business as free as possible to pursue the ever-evolving opportunities.

So Mackenzie says all Canterbury's farmers want is a general social agreement on what the plains can take when it comes to nutrient discharges, then be allowed to organise their activities within these clearly measured goals.

The CWMS being produced by Environment Canterbury (ECan) was meant to achieve that. Yet Overseer, the soil leaching model developed jointly by AgResearch, the Ministry for Primary Industries and fertiliser company Ravensdown, seems just too flaky a basis for the new regulatory system.

Already its whims have produced unfairnesses in ECan's decision making.

Mackenzie says allocations for nutrient discharges are now being determined catchment by catchment. His local Hinds Plains zone was modelled by Overseer as having a total current nitrogen loss of 4600 tonnes a year.

Because the area is red zoned – already past acceptable limits – a goal was set to cut that by around a third. That was agreed to by farmers through the zone committee process and a new limit of 3400 tonnes was written into the plan.

But then along came this latest version of Overseer and suddenly overnight every farmer was reckoned to be producing about double their original estimate. Yet ECan stuck to its own earlier version of the Overseer figure for the catchment target, effectively halving what farmers could afford to produce.

Alright, this might sound like a win for the environment, just hard luck for the farmers, says Mackenzie. The greenies will be applauding. And ECan's argument is that once it had plugged a catchment figure into the regulations, it had to treat it as law.

But do you follow what is going on, asks Mackenzie? It is models on top of models. A virtual exercise.

Explaining the use of lysimeters: demonstration at a reference dairy farm at Eyrewell in North Canterbury.
Howard Keene

Explaining the use of lysimeters: demonstration at a reference dairy farm at Eyrewell in North Canterbury.

Even the 4600 tonne figure for what agriculture is supposedly producing around Hinds is a modelling prediction being made by Overseer, not a physical measurement. It is what Overseer says ought to be happening as a result of the farming practices that have been noted on some catchment-wide spreadsheet.

So it is bad enough that farmers are being ruled by what is an educated guess on their environmental impacts. However to allow a mismatch of software versions to force a further cap on farming is really not on.

Ad Feedback

"Most of the farmers down here are thinking, well look, we'll do what we can. But this process is a bit of a bloody nonsense when it comes to its focus on the numbers coming out of Overseer. It's such a moving beast that it's very hard for the rational farmer to work out what's being required of him."


It is not hard to find critics of Overseer. Doug Edmeades of Waikato consultant agKnowledge was the government's chief soil and fertiliser scientist when Overseer was first being developed in the 1980s.

Edmeades sees it as a straight case of political expediency. ECan was under pressure to get water quality laws through so Canterbury could increase its farm production. The Government had installed commissioners and suspended environmental laws to hurry things along.

ECan officials had to grab whatever regulatory instrument was available, whatever might look like it was talking hard numbers, to push through the desired performance-based framework.

"Overseer was over-sold. People said we've got a tool so let's use it. I think the mistake was as fundamentally simple as that," Edmeades says.

Edmeades complaint is that Overseer is a world-class program, but it was designed for an entirely different purpose. It was meant for making predictions, for letting farm consultants model alternative farming scenarios.

"It was an expert system to do what-if analyses," he says. You could plug in changes to fertiliser application rates and irrigation schedules, or even add in fixtures like wintering barns and feeding pads, to see if it made a farm's nutrient performance better or worse. And as such, the results were qualitative rather than quantitative.

Edmeades says even with perfect knowledge of a farm's inputs, Overseer could only estimate nitrate leaching numbers to within plus or minus 30 per cent. And given the normal uncertainty about input measurements, Overseer's errors often shot out to 100 to 200 per cent.

This was fine for exploring farm practices, says Edmeades. "It can tell you whether you're going in the right direction if you do this or that. It's qualitative, giving you the direction of any change."

However ECan is trying to employ Overseer as a quantitative tool of regulation. The model is being used to make claims about what must be happening out in the real world so that hard limits can be placed on the region's farming.

Edmeades agrees that regulating agriculture on its proven environmental impact would be brilliant, world-beating, if New Zealand can do it. But he sees too much softness in the science being employed.

So the case against Overseer looks damning. But then while farmers like Mackenzie are certainly grumbling, the complaints over the release of V6.2 also seem oddly mild.

Normally you might expect tractor protests in the street, muck-spreaders being backed up to the steps of ECan's offices, when something like this happens to the farming community.

And ECan officials say the muted response is evidence for how well the CWMS is working at a deeper level. The new regulations on nutrient loss are one thing. But what is really significant is how the very culture of farming in Canterbury is being changed. There has been a mindset shift that is going to allow this performance-based approach to work, let it shrug off any teething problems.


Ken Taylor, ECan's science director, says the major criticism of those like Edmeades is that if you want to know the hard numbers for a farm's impact on the environment, then you have no choice but to get out there and make some real world measurements. However desktop modelling the effects is in fact likely to be more accurate.

Look at it like this, Taylor says. The real world is messy. Infinitely varied. The same paddock can have five different kinds of ground. "You'd need to take so many different measurements if you wanted to assess what was being leached."

Zone committees have been bringing rural communities together for water and nutrient limit  decisions.

Zone committees have been bringing rural communities together for water and nutrient limit decisions.

It would be fine if you could put a tray under a whole farm, capture all the run-off. But dotting the Canterbury countryside with enough probes to keep a proper running tab on its leaching rates – lysimeters, fluxmeters and other instruments planted every few metres – would be an impossible cost.

In practice, any direct measurement system would be so thinly spread that its accuracy would be spurious, says Taylor. So spending your money on predictive modelling is going to buy you better quality information in the end.

Overseer will improve, says Taylor. Millions are being invested to build more elements into the software – like the differences between blackcurrants and grapes.

And there is validation of Overseer's assumptions taking place. Taylor says Canterbury has some half dozen reference farms where farm practices and leaching rates are being meticulously monitored. There is the maze of instrumentation buried in paddocks to check the information going into the models.

So ECan realises Overseer is a work in progress. "It's less than perfect." But Taylor says the program already does do a credible job of being able to put a figure on nutrient losses from a detailed description of a farm's operations.

And more importantly – what many of the critics may have missed, he says – is how Overseer is the key to tying everything together. The fact that it is also a predictive tool, that it can both tell you what you are doing wrong, and then how to fix it, is why ECan can claim to be doing something rather clever and novel with its new regulatory framework.


ECan regional planning manager Brett Aldridge, who used to be ECan's compliance boss, wants to talk about the deeper mindset shift that the CWMS involves.

Aldridge says the old ECan was expected to be the cops. Its job was to set and enforce the environmental rules. Farmers were then the private property owners who naturally resented being told what to do. That made regulation hard work.

"In the old days, we'd drop a regulation. But unless ECan was knocking on your door, saying, 'Well jeez, you need to comply with that,' nobody did a thing."

Aldridge says this adversarial approach could work when the RMA was being used to deal with "point-source" problems like water theft, effluent discharge or unauthorised wetland clearance. If farmers broke the law, it was easy to spot and take them to court.

However nutrient run-off is something that is happening everywhere in a creeping, diffuse fashion. So the CWMS needed to start with a community buy-in, says Aldridge. ECan had to foster a cultural change where Canterbury farming had the incentive to be self-policing.

Methven farmer Craige Mackenzie says Overseer can capture good practice.
Mike Scott

Methven farmer Craige Mackenzie says Overseer can capture good practice.

This is why the most important piece of the CWMS puzzle is not Overseer, but in fact something else – a set of industry-defined Good Management Practices (GMPs).

The GMPs seem to have gone under the radar, not attracting much public attention, says Aldridge. What they are is a collection of standards for the different kinds of farming – sheep, dairy, pigs, crops, even life-style blocks – that spell out how it is reasonable to operate a farm in a way to minimise nutrient run-off.

These GMPs were largely drawn up by the industry sectors themselves – Dairy NZ, NZ Pig Farmers, the Foundation for Arable Research, Beef and Lamb NZ.

So written into Canterbury's regulatory system now, says Aldridge, are each farming sector's own expectations of how high its individual farmers need to aim. And the beauty of it is that the industry leaders have no reason to go easy.

Aldridge says everyone knows agriculture can only continue if it has a social licence. The era of water quality controls is here. So as a matter of self-interest, to maximise its own production, each sector needs all its members to be wasting as little of the collective nutrient allocation as possible. "The industry has to drive its own headroom effectively."

The GMPs give lists of to do's that ought to be standard on a modern farm, like fenced off waterways, monitored sprinklers, riparian planting. Aldridge says the GMPs are meaningful as few farms can yet meet them. All will have to lift their game.

Then these GMPs are tied to FEPs, or Farm Environment Plans. Every year farmers will have to prepare an audited plan that shows how they are faring in terms of their industry GMPs.

The marks will count. A D or a couple of Cs in a row will flag a farm for attention, says Aldridge. "You'll risk getting kicked to compliance."

Thus what the CWMS is about is the setting up of a hierarchical regime based on agreements about actions. The GMPs define global ways of farming which target the region's water quality goals, and individual farmers, through the FEPs, have to benchmark their farm operations against those standards.

Overseer then fits in as the bit that makes it all hang together. It gives the numbers that can relate both the GMPs and FEPs back to the third part which is the CWMS's various catchment-level targets.

Aldridge describes how it might work. You are a farmer talking to your auditor for the annual check. Your catchment number might be 24kg of allowable N loss, so you run your FEP model with its checklist of the good management practices you are applying and find that either Overseer says you should be about right and can relax, or no, Overseer predicts you stand at 56kg and need to start doing more.

And here is where Overseer's value as a modelling tool comes into play. All the necessary information is already gathered. You can begin making tweaks to your farm plan to discover what combination Overseer says will give you the right outcomes.

Aldridge says in this light it doesn't matter if the Overseer numbers are a little soft because it is about comparing individual farmers to a general shared standard and then prompting the changes that are likely to result in the right average environmental trends for a water zone.

If over time the system is not producing – the real world numbers from monitoring nitrate and phosphorous contamination in the waterways are not improving – then ECan has the grounds to go back and tighten up the rules still further.

But Aldridge says the point is that the CWMS is being set up as a collaborative framework where the regulators and the farming industry are all in it together. The nutrient loss goals are being socialised so that regulation feels like a team effort.

As part of the new system, ECan is encouraging farmers to form local groups to share their nutrient allocations. It makes sense. A co-operative holding a larger catchment-scale consent can make trade-offs between its members. Discharge rights could be swapped about as needed.

Yet Aldridge says it will be another way that farmers come to view themselves as part of a collective industry where it is then fair to exert peer pressure on those not coming up to the mark. In the traditional farming world, you don't go sticking your beak in someone else's business.

"But where we want to get to is where farmers know what they need to do, and so when [a neighbour] down the road is not doing what they should, then there is the community pressure. People will say, hey, that's not what we do here. You're letting us down."

It is the collaborative spirit which is why Overseer V6.2 is causing less of a ruckus than might be expected, believes Aldridge. Canterbury's farmers are engaged and committed to working through the process.

ECan has had to change too. "Once upon a time, [for ECan] just getting out a plan was a success. Nowadays, if we don't get a plan out with community buy-in and ownership, we don't consider that a success."

So hiccups like the Hinds' catchment target issue are going to get "revisited". Some solution will be found as the CWMS continues to roll – probably just writing in a new clause to make any necessary catch-up adjustments automatic.


Perhaps it is all too rose-tinted. New Zealand has had experience of bringing in these kinds of bold performance-based regulatory systems – the intellectually better way of doing things – and then spending the next 20 years complaining they are not really working.

Or perhaps ECan is right that the same experience means New Zealand is getting better at creating such frameworks. The CWMS represents another step forward because of the way it is managing to socialise its goals.

And there are farmers who are amazingly optimistic about the potential. Craige Mackenzie, a crop and dairy farmer near Methven, who is also chair of the Precision Agriculture Association, runs one of Canterbury's reference farms.

Mackenzie has the arrays of lysimeters and fluxmeters to create a measuring "tray" under his property. He has done the geomagnetic mapping of his fields to find the stony parts that leach fast, the heavier spots of clay that can bind nitrates. He has installed the latest irrigation rigs that can dose the land with exactly what it takes to reach the root zone and no more.

Mackenzie says his operation is so controlled he gets no leaching. Zero loss for the last few years. "We put on nitrogen little and often, and water little and often, so it all gets used within the crop. The measurement profile goes down 60cm and every 10cm we can see exactly where the water's tracking. By doing that, we can flat-line those bottom depths."

Mackenzie says this is the sort of good management practice, the information about precision techniques, that can then be built into the Overseer model. And at the click of a button, Overseer will also be able to quantify the dollar savings on a farm's irrigation and fertiliser bills. Mackenzie says he himself uses 30 per cent less water now.

This is the benefit of creating a regulatory framework based on a tool which also spells out the corrective actions. It turns it around from farmers having limits imposed upon them to feeling in charge of their efficiency and thus well equipped to meet good environmental outcomes.

So that is why Overseer V6.2 is being treated more as an irritation than a crisis. The muck spreaders can stay parked for the moment. There is no cause for general revolt.

Another version of Overseer is due out shortly. Those numbers could shift yet again. In Canterbury, it may still be a case of keep watching this space.

What is nitrate and why is nitrate leaching important?

Nitrate is NO3, nitrogen bound to oxygen in a soluble form that plants can take up with their roots. On farms, it comes from both added fertiliser and stock effluent.

If nitrate is not completely absorbed by a growing crop or a paddock of grass, the excess can be washed out of the root zone and into the water table where eventually it ends up in streams and lakes, causing soupy algal growth, or at a high enough level, toxicity in drinking water.

Phosphorous is the other troublesome nutrient for farmers. But this binds to top soil and so escapes farms by erosion and the sediment reaching water bodies. So it is more an issue for dryland sheep farmers in Canterbury.

Nitrate leaching is usually worst in winter when crops are not growing, but heavy summer storms can also cause sudden losses.

More water stories:

NZ irrigation and its guilty secrets

Lake Ellesmere clean-up deal explained

Water management targets proving too ambitious

Community involvement is crucial for water plan

 - The Press

Ad Feedback
special offers
Ad Feedback