Looking at different scenarios to try and reduce nitrogen (N) leaching and greenhouse gas emissions (GHG) using the Enviro-Economic Model (E2M) got Golden Bay farmer Robert Rosser thinking about what he could change on the farm, but he thinks the technology is still a work in progress.
The family farm at Upper Takaka was used as a case study for the model, using its 2016-17 season DairyBase figures, with the goal of producing different options that could reduce nutrient losses, irrigation use and greenhouse gas emissions without impacting on farm profitability.
Robert and his wife Cindy own 25% of the farm which milks about 550 crossbred cows on a 165-hectare milking platform, with an 82ha runoff across the road and an extra 97ha planted in forestry on steep surrounding hills. It bounds the Takaka River and lies in the sensitive Mt Arthur marble recharge zone feeding the Te Waikoropupu Springs which have a pending conservation order and protocols yet to be decided for water use.
The farm is in the top 5% for farm profit in the Top of the South and Robert has long had the goal of reaching carbon zero with the help of the forestry. They have registered forestry for carbon credits, though are still waiting for the results.
‘Some farms handle droughts better than other farms and some handle wet better than others, so it’s hard to fit all farms or to take all that into account. But it is good for getting farmers thinking about how we can be better farmers and better for the environment.’
Robert had several factors he wanted to retain in each of the model’s scenarios – the non-negotiable limits – including feeding pellets in the dairy to provide minerals to the herd and also irrigation which uses water for about 40 days of the season on 114ha. It includes three pivots and though one of the goals of the study was to reduce irrigation, Robert says the farm simply won’t grow grass on those paddocks without irrigation during dry summers, despite the 2.5-metre rainfall.
The rainfall often comes in deluges which along with the soil type such as poor-draining pakihi on some areas, supplementary feed, N inputs, stocking rate and irrigation, leads to increased N leaching. The farm’s N leaching levels are below average for the Takaka catchment, with the higher leaching blocks on irrigated and effluent land. One of the goals in the model then was to reduce the N leaching.
The E2M uses marginal cost/benefit relationships to drive its economic optimisation. It assesses the change in profit from each additional input such as kilogram of drymatter (DM), kg N or bought-in feed by comparing the marginal cost with the marginal return. The result is the least-cost combination of varying inputs that gives the best economic outcome. For this analysis, the model used $6/kg milksolids (MS).
Four alternative farm systems were considered by the model including reduced N use, different pasture mixes to reduce reliance on irrigation and the use of the runoff block versus a self-contained system on the milking platform. The amounts of inputs, such as bought-in feeds, supplements harvested, areas of summer crops, stocking rate and per cow production were varied by the model, based on the energy feed supply and economic optimisation.
While Robert and Cindy had their non-negotiable limits in the modelling, cow numbers were relaxed which led to successive reductions in the number of cows milked, but also allowed heifers to switch to twice-a-day milking when feed supplies allowed.
When the model reduced cows, it did so because the figures showed it was more profitable to do so. By dropping cow numbers, it lifted per cow production, reduced the area of irrigation and the need for hay and silage from the 82ha runoff area. That feed was instead fed directly to replacements and dry cows.
Different pasture combinations –– in this case a prairie grass and caucasian clover mix – were then introduced to minimise the potential impact of dry spells and to allow the accumulation of higher covers without losing quality. In three of the four scenarios, the runoff was considered a separate entity where stock was grazed at commercial rates and surplus feed sold to the dairy enterprise.
The outputs from each scenario were then put into Overseer to calculate changes to nutrient losses and greenhouse gas emissions from the milking platform and 22ha of trees attached to it.
To produce a set of scenarios for the farm, a base farm was created that reproduced the figures on the Rosser farm. In that, the farm milked 544 cows for 412kg MS/cow, spread 200kg N/ha, grew 11ha of summer crop and bought in 490 tonnes of feed each year. It added in the grazing costs on the runoff, feed produced on the runoff, silage on the milking platform and other costs. The base model for all the scenarios required either 80ha of exotic hardwood, 154ha of pines or 359ha of native plantings to offset its agricultural GHG.
In the first alternative farm system or scenario that looked at optimal N use, production was fixed at 412kg MS/cow and the farm retained the full 114ha of irrigation. The model was allowed to vary cow numbers between 484 and 544, vary N, keep bought in feed no higher than the base farm, have a summer crop up to 11ha and could choose whether or not to graze stock off the milking platform.
Using these fixed figures and information, the model chose to reduce cow numbers by 60 to 485, bought in less pellets, reduced summer crop and reduced N use by 59%. According to the model, this would increase profit by 3%, reduce N leaching by 25% and reduce GHG by 11%.
In the second scenario, the model looked at optimal herd size and irrigation. Production was again fixed at 412kg MS/cow, but cow numbers could be varied to optimise profit and the other information was the same as in the previous scenario, except irrigation which could be varied up to a maximum of 114ha. Using this information, the model reduced cow numbers to 450, made 20t more silage on the milking platform in early summer, reduced N by 74%, bought in 180t less pellets, purchased 120t less balage, reduced summer crop by 5ha, grazed yearlings at home, while all R2s were grazed off along with 210 cows in winter. The irrigation area was reduced by 47ha to 67ha. The result from the model indicated a 5% increase in profit, 35% reduction in N leaching and 18% reduction in GHG.
For the third scenario, it assessed an alternate pasture sward, introducing more drought-tolerant species on up to 47ha such as the prairie grass and caucasian clover. Cow numbers were fixed at 450 and irrigation at 67ha. Production per cow could vary along with timing of N, while bought-in feed could not exceed the base farm and summer crop could be up to 11ha. It could choose whether or not to graze animals off the milking platform. In the model’s response, it kept heifers on twice-a-day milking and therefore improved their production to 275kg MS/heifer while cows produced 448kg MS/cow. It made 95t of balage on the milking platform, reduced N by 74%, reduced pellets by 140t, bought in no balage and hay from runoff, grew no summer crop and grazed all yearlings and R2s on the milking platform, with 140 cows grazed off over winter. According to the model, this would increase profit by 17%, reduce N leaching by 32% and reduce GHG by 26%.
The final scenario looked at an alternate pasture sward again as well as being self-contained. The fixed information included 448kg MS/cow, 72ha of prairie grass and caucasian clover, 140t of pellets and 67ha of irrigation. The runoff was added into the system and grew all the feed requirements and was self-sufficient in grazing, though runoff costs were assumed to be the same as the assumed grazing bill.
The new farm area under this scenario was 269ha. The model could then choose cow numbers, bought-in feed up to base farm figures, up to 11ha of summer crop, whether or not to graze stock off and the amount of irrigated land.
From that information, the model’s response reduced cow numbers to 405, made 160t of balage on the milking platform, reduced N by 80%, made no balage on the runoff and no summer crop, grazed all stock at home and gave greater flexibility to pasture cover with a greater area of the prairie grass mix. The result through the model was a 27% increase in profit, 10% reduction in N and 9% reduction in GHG.
Robert says they initially became involved in the project to look at ways to reduce nitrate leaching as well as their farming footprint and to increase their environmental awareness.
It was a long, time-consuming process and though it raised good discussion points, he questions the practicality of the model outcomes and by the time the study was completed, the figures were out of date. That was due to the project itself which needed time for aspects such as community engagement and funding, rather than the model itself.
In regard to the model’s outcomes, Robert says that for starters, he would not want to use prairie grass which he considers outdated and says it makes sense to use the runoff for grazing and supplements as part of the business since they have it there and it is attached to forestry on the hills. To bring it into the milking platform like in the final scenario would require a $200,000 bridge.
“I reckon if you include the runoff, the N loss and everything would be a lot less because all the figures would come right down,” he says. “I’m not convinced it should be self-contained on the milking platform.”
He says every farm is more complex than the information used in the model and for all models, there’s different ways of putting the figures in to get different answers which makes it difficult for farmers to judge. Not only is every farm different, but even the cows and he says the size of cows alters the number of cows per hectare and feed.
It also doesn’t yet address variables sufficiently such as drought or wet springs which alter just about everything from pasture growth and supplements fed to the cows, to milk production.
“Some farms handle droughts better than other farms and some handle wet better than others, so it’s hard to fit all farms or to take all that into account. But it is good for getting farmers thinking about how we can be better farmers and better for the environment. And there’s different ways for us all to do that.”
One of the model’s scenarios for the farm suggested spreading urea in June and July, but he says that would be pointless on their farm because it just wasn’t warm enough in those months to make use of it.
He also thinks there is still more research to be done on how to farm sustainably before models can be really useful and farmers are already changing their farm systems to tackle environmental issues, which takes time.
“It’s not going to happen over night and I think science will get better and models will be tweaked,” he says. “Fertiliser is getting better all the time and we’re testing every paddock so people will use less fertiliser because they are only using it where they need it. It all takes time.”
Improving water quality on their dairy farm has been an example of dairy farmers needing time to change because their changes have happened over a few years as income allows. They have an environmental plan to work on and have built bridges over dry streams, carried out a three-year upgrade on their effluent system to include a vibration screen and large storage tank, and fenced streams.
“You can’t do it all in one big step because of costs.”
Robert says they have just about enough trees to be carbon zero, depending on how an equation is finally decided. But he says not every farmer has 50ha of hill country that can be planted in trees and planting trees on productive land doesn’t make sense financially, such as the flat runoff attached to the forestry.
He will be taking some ideas from the model, such as reducing cow numbers and fertiliser, especially to get N leaching down. For the past two years they have been using N-Protect which is a coated urea that keeps more nitrogen in the soil for plant uptake, so N leaching will be less than the 2016-17 figures used for the model. They usually put on 200kg N/ha over the farm and now they will look at taking that down to 160kg/ha with future reductions planned to 90kg/ha.
And that’s where the model has been useful in showing what the effect might be if you change certain aspects in the system, he says. Because of the variables and scientific unknowns, they will look at trying any changes carefully and slowly.
“In the future we may get pushed to do some of these changes, but we’re not there yet and there’s a lot of work to do, a lot of unknowns.”
He acknowledges there is also work to do on dairy farmer mentality toward reducing cow numbers and changing farm systems and like the research and modelling, it is a work in progress.
The full reports can be viewed at https://www.landcare.org.nz/current-project-item/farm-systems-project
Finding environmental solutions on a dairy farm is often associated with less production and less profit, but a new predictive model provides options to alter the farming system for a good environmental outcome and often more profit.
The Enviro-Economic Model (E2M) analyses a farm’s system and by adding many options, it can recalculate the figures to provide the optimal economic solution and the outcome such as profit, nitrogen (N) leaching and greenhouse gas (GHG) emissions.
E2M has been developed by agricultural analyst Barrie Riddler and is used on the Lincoln University Dairy Farm. Now there have been case studies conducted on farms in conjunction with NZ Landcare Trust, Fonterra and DairyNZ.
NZ Landcare Trust coordinator for the Top of the South, Annette Litherland, says it is all about optimising a farm’s system. Farmers can decide they want to retain aspects of their farm system, such as cow numbers or adding a certain feed into the cows’ diet, then the model provides scenarios for the best outcome. It does that by assessing the change in profit from each additional individual input such as kilogram of drymatter, kilogram of N or bought-in feeds.
“This is a tool a farmer could use to interact with his or her farm system and come up with the optimal farm system for the changes that are facing us with the environment. E2M can do this much faster than other modelling tools currently available.”
For the Rosser farm in Golden Bay used as a case study, the data supplied for the model came from a Farmax simulation for the farm using DairyBase information as well as farming details from the family. The E2M model then came up with four alternative farm systems that reduced N use, used different pasture mixed to reduce reliance on irrigation and the use of the runoff block versus a self-contained system.
The amount of inputs such as bought-in feed, supplements harvested, areas of summer crops, cow stocking rate and per cow production level were varied by the model based on the energy feed supply and economic optimisation.
Annette, who has a background in agricultural science, says feedlots overseas use models all the time, but New Zealand farmers have been reluctant in the past because they can’t see how a model could predict what happens on a farm. Yet she says modelling is powerful and this particular model is a way for farmers to bounce their ideas around to find economic solutions to meet nitrate leaching and GHG reduction targets.
Whenever a scenario changes in farming, such as milk price, farmers usually have goals and change their system to cater for it, such as dropping inputs for a lower milk price, she says. Now they need to look at changing farm systems for better environmental outcomes and she says that requires some left field solutions such as modelling.
“Can you achieve what you need to do and still make the same profit or more? This gives you a system to examine many situations.”
As Annette explains, the model assumes cows are going to eat grass, produce milk and get in calf, so if an extra kilogram of feed was added into the system one way or another, the marginal economic result of that extra kilogram can be calculated, or any other factors in the farming system.
In the Golden Bay case study, Annette says the Rossers wanted to carry on feeding supplements in the shed, so the model could reduce it but not drop it completely. They also retained the production per cow and irrigation area, but allowed the model to change cow numbers.
“We allowed it to reduce cows by up to 60 and it said it would reduce cow numbers by 60 because it was more profitable. And it didn’t have to reduce N so much because it had reduced cow numbers.
“So then we looked at reducing cow numbers by up to 90 and it bought in less feed and had extra grass which meant more silage and so more feed and we could keep more stock at home and less grazing.”
At about $3000 to use the model, she says it is not cost-prohibitive for farmers to get modelling done on their own farm system. All they need is figures from Dairy Base and/or their farm accounts, plus other factors such as pasture growth, to look at various options for their farm. It’s particularly easy for those farmers set up in Farmax. Annette says it is really important that the farmer works with the modeler to try practical options that they think might work for them and then let the model optimise those options.
“This is a tool a farmer could use to interact with his or her farm system and come up with the optimal farm system for the changes that are facing us with the environment.
“You don’t need to change your whole system overnight. You’d make some changes the first year and step it through and see how close it’s coming to the model.”
In time, the model is expected to be expanded to include aspects such as soil and forestry get it to work out N leaching as well which is now being worked out through Overseer. But first it needs more funding and to achieve that it needs more farmers on board, she says.
The full reports can be viewed on https://www.landcare.org.nz/current-project-item/farm-systems-project.
This article is free to view because it is a topic of high importance. This article was published in New Zealand Dairy Exporter magazine. For less than $10/month, you can receive this detailed information to help improve performance within your business. nzfarmlife.co.nz/country-wide/
Supporting New Zealand journalism.