This year is set to be the best yet in terms of performance for the systems research herd at UCD Lyons Estate. The herd is on track to produce 610kg MS/cow on average and the overall empty rate, after just 10 weeks of breeding, was a remarkable 3.5% when the three cows that were not submitted for breeding are excluded.

The project was set up in 2015 to demonstrate high-input high-output grass-based systems. Finbar Mulligan is the current lead researcher with several research and farm staff involved.

Finbar says the fundamentals of the system are grass and fertility. Essentially, these are no different to the standard Moorepark blueprint. The difference in Lyons is that cows are bred for more milk, fed more concentrates at 1.5t meal per cow on average and are stocked higher on the milking platform.

The overall empty rate, after just 10 weeks of breeding, was a remarkable 3.5%

There were 57 cows in the systems herd this year, giving an overall stocking rate of 2.33 cows/ha and a milking platform stocking rate of 3.27 cows/ha. The 7ha of silage ground is not grazed by cows. This is in a red clover and hybrid/tetraploidsward, which is cut for silage three times during the year and grazed by sheep in the autumn. This land is growing approximately 15t DM/ha of silage per year and is reseeded every three years.

The grazing ground has grown 13.6t DM/ha to-date and the most recent farm walk showed an average farm cover of 711kg/ha. Cows were housed on 11 November and are on a diet of grass silage and 1.5kg of meal. When the Irish Farmers Journal visited the farm last week, 35% of the herd was dried off. This year, 75% of cows are getting teat sealer only at dry off.

As can be seen in Table 1, the production per cow for 2020 is likely to be 610kg MS/cow when all cows finish lactation. Ten cows will be milked through the winter, with the last batch dried off on 26 January; the empty and cull cows will be sold at that stage. At 610kg MS/cow, the 2020 performance will have been the best yet at Lyons and closest yet to the target of 625kg MS/cow. As can also be seen in Table 1, 2018 was a bad year for the farm.

Finbar says: “We had a severe snowstorm in March and that set the herd back considerably, as we were in one of the worst-affected regions. This was followed by a severe drought and to top it off, we had a bad outbreak of lungworm in the cows also.”

Extra meal was fed in June this year due to a mini drought, but less meal has been fed than normal this backend because cows were in good body condition score, so the total meal will be 1.5t/cow.

There are set guidelines for how much meal is to be fed depending on the stage in lactation, but Finbar says they will deviate from this depending on conditions, such as drought or body condition score. For example, cows are currently being fed 1.5kg of meal per day, while the template says they should be on 3kg/day.


Fertility has been excellent for the last couple of years, with 2020 performance exceptional. It is fair to say the first few years were disappointing, with low conception rates to first service, six-week in-calf rates and subsequent six-week calving rates.

This year’s 21-day submission rate was 91% but was 98% at 22 days. The conception rate to first service was an impressive 74% and the six-week in-calf rate (excluding heifers) was 87%, well past the 75% target. The overall empty rate reported by the UCD team was 9%, but this includes three cows that were deliberately not bred, so it drops to 3.5% when these are excluded. The length of the breeding season was 10 weeks.

Seven cows were bred to beef and the remainder bred to high-EBI genomic Holstein Friesian bulls. There were 11 dairy AI bulls used in total. All AI was practised, with no stock bulls used. According to Alan Fahey, genetics lecturer at UCD, the reason for the good fertility is the herd has been bred for high fertility.

Winter housing at Lyons. Milkers are currently getting silage ad-lib and 1.5kg of meal.

“The overall EBI of the herd is €206, which puts it in the top 1% nationally. The milk sub-index at €69 is also in the top 1%, while the fertility sub-index at €87 is in the top 5%. We want a good balance between milk and fertility, so we screen bulls for fertility first and then look at milk. We want a high PTA for yield, solids, butterfat and protein percentages. We think PTA for yield is important because in previous experiments, we found a greater response to supplementary feed when milk PTA is higher,” Alan says.

Given that the EBI hasn’t changed dramatically over the five years of the experiment, yet fertility performance has, Finbar explains that management changes were made.

“We are no longer relying just on automated heat detection aids. We started using scratch cards three years ago. If the scratch card isn’t activated, the cow isn’t served. We use the automated heat detection system only for pre-breeding heats, but we don’t rely on it alone for breeding. It needs to be used in conjunction with something else.

“The other big change we made is to the timing of AI. We had a long-standing arrangement whereby the AI technician used to call in the middle of the day. This has since been changed to twice a day and we feel it has made a big difference.

If the scratch card isn’t activated, the cow isn’t served

“It’s important to recognise that we have a great team of really committed people working here. Research and farm staff take turns to monitor the cows for signs of heat an hour before morning and evening milking. Nobody lives on the farm, so the only time the cows are seen is before and during milking.”

In terms of mineral status, no changes have been made in recent years. The concentrate is a standard ration with added yeast and acid buff. A mineral analysis of pasture, bloods and liver biopsy was last conducted in 2017 and there are plans to do one again next year.


Nitrogen fertiliser applied on the milking platform was 235kg N/ha in 2020, down from 250kg N/ha in previous years. The silage ground received 160kg N/ha, so the amount of chemical nitrogen across the whole farm is 214kg N/ha. Slurry is only spread on the silage ground, with no slurry applied to the milking platform.

Current experiments in the research herd are looking at different crude protein contents in the meal and at including native ingredients in rations. Finbar says the purpose of this is to reduce the amount of nitrogen being excreted and the carbon footprint of the system’s herd. A crude protein content of 14% is now standard while cows are grazing.

A carbon lifecycle analysis has been completed and a new analysis is set to begin shortly using more up-to-date data. Finbar says that the carbon footprint of the milk produced in the Lyons system herd compares favourably to other systems.


UCD lecturer Michael Wallace has modelled the financial performance of the system’s herd at Lyons and compared this to a low-cost grass-based herd at similar scale. His findings suggest that after accounting for all economic costs, including a value on all labour, land and interest, the systems herd is more profitable at €566/ha compared to €507/ha for the low-cost system.

Actual performance from Lyons was used in the analysis, but the poor performance year of 2018 was excluded from the analysis.


The researchers at UCD have proven that a profitable high-input high-output grass-based system is achievable. But that doesn’t mean it’s the right system for everyone. In fairness, UCD has stripped out most of the complexities from high-input systems, but at the same time, it’s still not as simple to operate as a low-cost system.

The researchers recognise this and almost an hour a day of extra labour is factored into the economic analysis. This is an important consideration for farm families looking to push up yield per cow – more work and more skills required with little or no extra profit. Remember, it took UCD five years to come close to achieving its target output – even with top 1% cows and top management. It would be nice to compare performance of these elite cows in Lyons with average cows under the same system.

Then there is the question of the environmental footprint of high-input systems. Just 53% of the diet in Lyons is grazed grass, while it is 74% in low-cost systems. What impact does this have on the environment and on the end product? We heard at Dairy Day about the importance of our grass-based system to the taste and quality of our dairy exports. It is our unique selling point. UCD says that the carbon footprint of milk produced at Lyons is similar to Teagasc research herds.

In brief

  • The systems research herd at UCD is on track to produce 610kg MS/cow in 2020 from 1.5t of meal per cow.
  • The herd consists of 57 cows in 2020 and their EBI is €206, putting them in the top 1% nationally.
  • They aim for a balanced EBI between milk and fertility sub-index.
  • 54% of the diet is grazed grass, the remainder is meal and grass silage on an as fed basis – this is 93 to 94% grass and silage.
  • The project is funded by the Dairy Levy.