Cows in Milk

Stocking rate was lower than initially planned on both sides due to various reasons unrelated to the trial such as consolidating all A2 cattle on another farm for the premium payout. We dried off the regenerative herd 10 days earlier than the conventional side and 10 day earlier than expected, due to lack of autumn feed and not wanting to impact spring growth by overgrazing the diverse paddock. This is something we will aim to avoid next year.

Daily per Cow Production

This graph is quite interesting to us as it shows that when pasture quality wasn’t an issue in spring and autumn the regen outcompetes the conventional in per cow production, but once we lost quality in the diverse pastures over summer, the per cow production visibly dropped.

We have come to the realisation that we put too much focus on maintaining a higher cover rather than driving quality and this had a detrimental impact on milk production. If you compare our Average Cover graph to this graph you will see that production drops once we exceeded 2500 kgDM/ha cover on October 24th. At the time we thought this was to be expected with regenerative farming, but we now realise we can address this with management.

The regen herd was exceeding conventional in the start of the season and we want to try to extend this period of excellent per cow production this coming season.

YTD Production per Cow

The difference between the per cow production between the two herds grew over the summer months when pasture quality on the regenerative side was not optimal.

Weekly Milk Production per Ha

Same trend, we are impressed with the regenerative herd performance in the spring and autumn but lost it over summer due to the dip in low cow production discussed above. We identified in the summer that the quality of feed wasn’t good enough and in hindsight we probably accepted that rather than addressing it and that is a key learning from this season. We are now questioning why we can’t have equal per ha production on both sides, so expect changes in this space this coming season.

YTD Milk Production per Ha

This graph did what we anticipated. We are aiming to lift the per ha production on the regen side by 200 kgMS/ha next season, and then lift another 200 the following year.

Milk Composition

Not much to mention in this graph, relatively consistent throughout the season with the regenerative milk typically being higher in fat and the conventional milk higher in protein.

Milk urea was fairly consistent between the two herds throughout the season

Average Cover

This graph demonstrates the issues we had with milk production in relation to pasture quality. Our production started to reduce from the 24th of October which was when our covers started to exceed 2500 kgDM/ha. We were too focused on cover and we lost quality and milk production as a result. It is easy to blame the autumn growth (which is no doubt an issue) but the greater problem was that we didn’t get enough per cow production over the summer months. This force our hand in autumn to push harder and import supplements. If we had an extra 40 kgMS/cow through the 100 days of summer we would not have had to push so hard in autumn and our financial results would have improved. Next season we will focus more on quality.

Daily Pasture Growth

Pasture growth was very similar throughout the season, but the diverse pasture did out compete the conventional a few times throughout the season. We had an unusually cloudy summer for Canterbury and this impacted growth on both sides.

YTD Pasture Growth

The regenerative side grew 16,988 kgDM/ha while the conventional side grew 16,561 kgDM/ha, which is an additional 427 kgDM/ha grown on the regenerative side, something we did not anticipate. Some of this could be due to the longer grazing rounds. We went into the season concerned we would not grow enough feed without the longer round lengths, but now we are confident we are able to grow the feed so will focus on quality this coming season. On reflection we realised we had a preconception that we couldn’t grow grass without urea, but then when we did, we did not react quickly enough to address quality issues.

Round Length

Through most of the season there was roughly a 10 day gap between the round length of each side of the farm. At the time we thought having that 10 day buffer was the right thing to do, but on reflection we realise we should have been more adaptive. We left a buffer to allow recovery time of the pasture without urea but have learned that is not necessary and we should be less fixated on higher residuals. Going forward we are going to focus on following leaf stages rather than consistent grazing round. Next year will jump around a bit more as we are also intending to do some deferred grazing and be less regimented on how we need to graze the feed in general.

Supplements Fed (YTD)

We fed more supplements on both sides than expected, especially on the regen side. We planned to stop grain on the regenerative side in October but we ended up continuing to feed it throughout the season after noticing a drop in BCS of the heifers. Typically we run a heavier herd and a lighter herd but because of the trial they are all in one, resulting in some heifers being outcompeted and underfed so we supplemented these cows with grain. We also lost autumn growth on the regenerative side so imported supplement to make up for this deficit due to the attractive payout. Next year we will improve by focusing on quality feed and deferred grazing

N Applied (YTD)

This is probably the most compelling graph to illustrate the difference between the two systems. We applied less than 6 kgN/ha on the regenerative side, most of which went on the 4 paddocks that have not been transitioned to diverse pastures yet. The diverse pastures received 0.9 kgN/ha. We are very proud of this, not too many farms with N at that level would be growing 17 TDM/ha, so now we need to focus on turning it into quality feed which is the plan for next season. Also worth noting that we are also proud of achieving 160 kgN/ha on our Conventional side as this is well below the 190 kgN/ha limit.

Somatic Cell Count

Somatic cell count has consistently been higher on the regenerative side from the beginning, and we have had plenty of conversations about the cause of this. We have various theories of what could be causing this including the way we split the herd (on BW and PW, not previous SCC counts), the low-quality feed, or even the possibility of the higher cover causing issues. This is a focus for next year, we have invested heavily in dry cow for both herds so hopefully we will have a re-set over winter and then we can have confidence that the results are indicative of the farm system and not some other factor. If there is still a considerable difference by Christmas, we will consider it a trend and do some digging with vets and other experts to determine the cause.

YTD Mastitis Treatment

Mastitis rates were higher on the regenerative side which was to be expected due to the higher SCC rates.

YTD Lameness Treatment

This result was interesting as lameness was considerably higher on the conventional side. Kiri thinks it could because they are the second herd to be milked and the quality of the race the conventional girls mainly walk on. We also considered it could be walking distance but did some analysis and it is negligible between the two herds. We believe it has something to do with the feed, the track conditions or being the second herd with more stones on the yard. This is quite interesting to us and a focus for this year to try to determine why

YTD Animal Health Treatments

The regenerative herd had more animal health treatments compared to the conventional side throughout most of the season, but both herds ended up with 34% of the herd being treated at the end of the season. We find it interesting the numbers lined up in the end so are curious to see if this happens again next year.

Irrigation Applied (YTD)

This was an unusual year for irrigation, didn’t apply much this year. The difference between the two sides is not a reflection of the systems, there are other aspects at play such as running a pump for a pump test on the regenerative side. In a dryer season we may start to see a difference, but not yet.

Per Paddock Growth Graphs

These graphs are a new addition because they are so compelling. We graphed the cumulative grass growth of each paddock on the farm, for the full year and in 4-month intervals. Our conventional paddocks are green, our regenerative paddocks with diversity are black, and our regenerative paddocks that are still in ryegrass/white clover are blue.
We did put a small amount of synthetic N on these blue paddocks as growth was so dismal in the first few rounds.

Paddock 3 : 8 kgN/ha
Paddock 4: 16 kgN/ha
Paddock 5: 31 kgN/ha
Paddock 6: 16 kgN/ha

This equated to 5.5 kg/ha of synthetic N applied to the entire regenerative side. The diverse paddocks received 0.9 kgN/ha in the form of fish hydrolysate.

The conventional paddocks received 161 kgN/ha.

From the full season perspective, 3 of the bottoms 5 are blue paddocks, and all four are in the bottom quarter. We had very poor performance in these paddocks and that makes sense due to the lack of diversity or conventional amounts of urea to make them perform. This graph gives us additional confidence in diverse pastures without synthetic N, as they are 2 of the top 5 so there is definitely some interesting possibilities in this space.

By splitting the graphs into spring, summer and autumn it is easy to see the variation that occurs throughout the season and why the diverse pastures have increased our pasture curve in the summer months.

The diverse paddocks perform adequately in the spring, superbly in the summer and under performed in the autumn. The regenerative paddocks that were in ryegrass white clover were in the bottom quarter consistently. We intended to keep 1 or 2 paddocks like this for the remainder of the trial just to observe what happens for interest’s sake, but currently we would not recommend farmers drop out their nitrogen without increasing diversity first.


To accurately compare the two systems, they were modelled with just the predominate soil type on farm (Mayf_2a.1) rather than using all 5, as the soil types are not equally distributed between the regenerative and conventional side. Each side has an effective area of 146 ha. Stocking rate, production, fertiliser, crops and supplements were modelled based on actuals from our season. We have pivots and sprinklers on the farm and they were equally represented on both sides.