NZ regenerative-agriculture trial results 2021–2026
Updates the 2021 "anecdotal" verdict — NZ veg trials confirm the crop-specific pattern, but absolute kg/m² and nutrient density remain unmeasured
A Perplexity Deep Research synthesis (2026-06-11) of published results from NZ regenerative-agriculture controlled trials launched 2021–2024. It materially updates the Grelet & Lang (2021) “anecdotal/no controlled evidence” characterisation (LIT_013): ≥5 funded trials now in controlled phases, the vegetable sector strongest. For Neobiome it is the NZ localisation the UK
veg_yield(LIT_058) lacked — NZ veg trials confirm the crop-specific pattern, though they report relative gains, not absolute marketable kg/m². Well-cited but AI-prepared, mostly industry/non-peer-reviewed → data_quality: medium.
Summary
A Perplexity Deep Research synthesis answering RT_242: have NZ regenerative-agriculture trials launched 2021–2024 published quantitative results that update the 2021 “anecdotal” characterisation? It reviews five trials (two vegetable/cropping, three pastoral), with inline citations to industry whitepapers, conference manuscripts and openly published farm-system data. The headline for Neobiome: NZ vegetable trials (LeaderBrand/Woolworths/Plant & Food, Gisborne) now provide controlled quantitative evidence that confirms the crop-specific regenerative pattern seen in the UK AMOS data, but report relative/biomass gains rather than absolute marketable yield, and nutrient density remains unmeasured across all NZ trials.
Key claims
- claim: "Updates the 2021 Grelet & Lang baseline: at least five substantively funded NZ regenerative-agriculture trials launched 2021–2024 have moved into controlled/semi-controlled field phases, and four have published quantitative outcomes (yield, soil carbon). The vegetable sector has the strongest emerging evidence; the pastoral sector has credible production data but is mostly mid-trial; nutrient-density data are effectively absent across all trials. Only one peer-reviewed controlled NZ RA dataset exists to date (Lincoln pasture farmlet, J. NZ Grasslands 2024)."
source_location: "Executive Summary; Synthesis table; Key Limitations"
- claim: "Vegetable trial — LeaderBrand / Woolworths NZ / Plant & Food Research (Gisborne, 2 commercial sites, regen vs standard zones, MPI SFF Futures co-investment, 2022–2025): 4 of 6 cash crops showed increased yield in the regenerative zone. Site 1 (sandy loam, winter salad): ~45% more biomass in the regen zone, mesclun 'nearly double'; Site 2 (clay, sweetcorn/brassica): Year-2 sweetcorn comparable to standard DESPITE a 34% reduction in N fertiliser (hairy-vetch cover crop fixing ~150 kg N/ha). Topsoil total C gains modest (+0.1 pp Site 1, +0.4 pp Site 2 over 2 yr); soil biological respiration and mineralizable N rose. Nutrient density not measured."
source_location: "Trial 1 (LeaderBrand/Woolworths/PFR)"
- claim: "Process-crop & pastoral trials (mixed/early): Carbon Positive (LandWISE/HBFFT, Hawke's Bay, 6-yr 3-arm) — process sweetcorn 2022/23 conv 18.9 / hybrid 17.4 / regen 16.7 t/ha; process tomatoes 2023/24 conv 140.6 / hybrid 148.9 / regen 93.4 t/ha (regen N/moisture-stressed, p<0.001); green beans ~26 t/ha with regen highest at <½ the N. Align Farms Clareview (dairy, split-farm, 3 seasons) — regen ~18–22% lower milk solids/ha (1,329 vs 1,714 kg MS/ha 2023/24) but lower fertiliser cost and lower somatic cell count. Whenua Haumanu (Massey/Lincoln, NZ$26.1M 7-yr, most rigorous) early — diverse pasture higher mass but lower annual yield in sheep farmlets; dairy season-total milk inconclusive."
source_location: "Trials 2 (Carbon Positive), 4 (Align Farms), 5 (Whenua Haumanu)"
- claim: "Soil carbon is hard to move in short trials — detectable changes in TOTAL soil carbon require 5+ years; labile (hot-water-extractable) carbon is more responsive but reversible (Carbon Positive: labile C peaked 29% above baseline after cover crops, then returned). Crucially, in the Carbon Positive trial all three management systems showed the SAME carbon-cycling pattern in years 1–3 — crop sequence, not management approach, dominated. Nutrient density of produce (vitamins/minerals/fatty acids) is unmeasured in every NZ RA vegetable trial to date — the explicit RT_242 nutrient-density gap remains open."
source_location: "Trial 2 soil carbon; Key Limitations and Gaps"
- claim: "The largest, most rigorous trials (AgResearch/Synlait/Danone dairy, NZ$3.85M 5-yr; Whenua Haumanu NZ$26.1M 7-yr) have not yet reported outcomes — the highest-quality NZ RA data are expected 2026–2028. Trial durations to date (2–4 yr) remain short relative to the soil-carbon/ecosystem timescales RA advocates invoke."
source_location: "Trials 3 & 5; Key Limitations"Retrieval provenance
- Upstream source: multiple NZ trial sources — Our Land & Water (Grelet & Lang 2021), Woolworths/Plant & Food Research whitepaper (2025), LandWISE Carbon Positive, Align Farms Clareview, Massey/Lincoln Whenua Haumanu, MPI SFF Futures.
- Prepared by: Perplexity Deep Research, 2026-06-11.
- Prompt: RT_242 prompt — “Search for published RESULTS (yield, soil-carbon, nutrient-density) from the NZ regenerative-agriculture controlled trials launched 2021–2024 (Our Land & Water; MPI SFF Futures vegetable trials; AgResearch/Synlait/Danone; Align Farms Clareview). Have any reported quantitative outcomes that would update the 2021 ‘anecdotal/no controlled evidence’ characterisation (Grelet & Lang)? Report each trial, status, and any yield/quality numbers with units.” (with
@Neobiome Research Projectattached) - Verification: well-cited (footnotes + source URLs) but AI-extracted from mostly industry/non-peer-reviewed, ongoing trials — not independently verified against each primary.
data_quality: medium. (Convention:wiki/CLAUDE.md→ Retrieval provenance.)
Neobiome Intelligence relevance
Resolves RT_242 — the NZ localisation the UK AMOS veg_yield (LIT_058) was missing. The NZ vegetable trial (LeaderBrand/Woolworths/Plant & Food) confirms the crop-specific reversal that AMOS and CR_021 identified: regenerative wins on intricate-husbandry crops (salad/mesclun/spinach — +45% biomass, mesclun ~2×) while bulk/process crops (sweetcorn, tomatoes) are comparable or lower, often N/moisture-limited in transition. So the UK→NZ transfer of the yield pattern is now directionally supported by NZ controlled data. ⚠ But the NZ trials report relative/biomass gains and management contrasts, not absolute marketable kg/m² — so they do not replace the conservative veg_yield ≈ 21 t/ha (LIT_058); that figure stands, now with NZ directional backing rather than UK-only. This updates LIT_013’s “anecdotal as of 2021” to “controlled quantitative evidence emerging 2022–2026, strongest in vegetables.” Two caveats persist for the model: nutrient density is still unmeasured in every NZ veg trial (the diet-quality argument in D_001 remains NZ-unevidenced), and soil-carbon gains are modest/slow (<5-yr trials). ⚠ Mostly industry-run, non-peer-reviewed, short and ongoing.
Research targets
Research gaps
- RT_242 (RESOLVED → this page): NZ RA trial results 2021–2026 characterised; the yield pattern is NZ-confirmed, though absolute kg/m² and nutrient density remain unmeasured.
- Watch (no new RT): the high-rigour trials (AgResearch/Synlait/Danone; Whenua Haumanu) report 2026–2028 — a future refresh would firm the NZ pastoral/soil-carbon picture and possibly the first NZ nutrient-density comparison.
Connections
Links to
Referenced by