LIT_089: Townshend & Boleyn (2010) — measured Canterbury field trial of hemp cv. Fasamo (seed yield + oil)

LIT_089: Townshend & Boleyn (2010) — measured Canterbury field trial of hemp cv. Fasamo (seed yield + oil)

Source

doi:10.33584/rps.14.2008.3180 — original publication (opens in a new tab; the file is not redistributed)

Townshend & Boleyn (2010) — Plant density effect on oil seed yield and quality of industrial hemp cv. Fasamo in Canterbury

The MEASURED NZ hempseed-yield primary — upgrades OT_112's COMPILED hemp range, the way OT_088 (faba) anchors legume protein

A peer-reviewed Canterbury field trial (Midlands Seed Ltd) of industrial hemp cv. Fasamo, published in the Agronomy Society of New Zealand Special Publication No. 13 / Grassland Research and Practice Series No. 14. Two seasons, two sites, six plant-population treatments: machine-dressed seed yield 804–931 kg/ha (2006-07, Ashburton) and 1,640–1,849 kg/ha (2007-08, Methven). Where OT_112 gives a compiled hemp seed range (800–1,000 kg/ha, from seed companies / FAR / overseas), this is a measured NZ two-season trial — the hemp analogue of what OT_088’s 55-paddock survey did for faba. ⚠ Load-bearing scope: the QUALITY axis measured is OIL (content 30.9–32.5 g/100g + fatty-acid composition, AOAC 991.36) — NOT seed protein. This paper does not measure hempseed protein %; do not read a protein figure from it. Retrieved open-access from the NZGA journal (nzgajournal.org.nz), verified via pdftotext -layout.

⚠ Authorship correction: the RT_362 retrieval note filed this as “Nott & McKenzie” — the raw’s own title page reads J.M. Townshend and J.M. Boleyn (Midlands Seed Ltd). The verbatim source is authoritative.

Summary

A two-year, two-site agronomic field trial of the German-bred, dual-purpose, monoecious industrial-hemp cultivar ‘Fasamo’ grown for oilseed in Canterbury, run by Midlands Seed Ltd to determine the optimum plant population for hempseed (oilseed) production and economic return. Target populations of 125–250 plants/m² were laid out in randomised complete blocks at Ashburton (2006-07, Lismore stony silt loam, border-dyke irrigation) and Methven (2007-08, Mayfield/Hororata silt loam, Rotorainer irrigation). Its load-bearing output for the Neobiome food model is a measured NZ hempseed yield: machine-dressed and corrected to 8% moisture, 804–931 kg/ha in the marginal-site 2006-07 season and 1,640–1,849 kg/ha in the well-prepared-arable 2007-08 season — the between-site/season difference dwarfing the effect of plant population. Seed from the 2006-07 trial was analysed for oil content (30.9–32.5 g/100g) and fatty-acid composition (linoleic ~53%, α-linolenic ~20%, oleic ~10%, γ-linolenic 3.8%; AOAC 991.36) — the hemp oil approaching the reported ~3:1 linoleic:α-linolenic essential-fatty-acid balance. Quality was measured on the OIL, not on seed protein. The paper recommends a target established plant population of 150–225 plants/m² depending on site yield potential. Economic returns (before fixed production costs) ran 3,482–3,838/ha (2006-07) and 7,183–8,092/ha (2007-08).

Key claims

- claim: "MEASURED hempseed yield, 2006-07 (Ashburton). Machine-dressed seed yields, corrected to 8% moisture content, ranged 804–931 kg/ha across the six plant-population treatments (Table 1: established population 124→809, 140→804, 176→880, 201→887, 224→928, 260→931 kg/ha; LSD(5%) 133). 'In 2006-07, yields ranged from 804 kg/ha to 931 kg/ha, with a positive linear trend of 0.823 kg/ha yield gain for every extra plant/m2 established.' Trend equation: yield (kg/ha) = 719 + 0.823 × (plants/m2), regression slope SE 0.339 (P<0.05). [measured NZ field trial, verbatim, verified]"
  source_location: "Abstract; Results and Discussion — Yield (p.88); Table 1 (p.89)"
- claim: "MEASURED hempseed yield, 2007-08 (Methven). Machine-dressed seed yields, corrected to 8% moisture content, ranged 1,640–1,849 kg/ha (Table 2: established population 185→1671, 218→1823, 259→1640, 273→1849, 287→1724, 346→1696 kg/ha; LSD(5%) 189). 'In 2007-08 yields were much higher, ranging from 1,640 to 1,849 kg/ha.' A small non-significant negative trend with population: yield (kg/ha) = 1751 − 0.069 × (plants/m2), slope SE 0.513. 'the greatest difference in yield was between sites and not between plant populations' — agronomic management and climatic conditions had a greater impact than plant population. [measured NZ field trial, verbatim, verified]"
  source_location: "Abstract; Results and Discussion — Yield (p.88); Table 2 (p.89)"
- claim: "MEASURED seed OIL content and fatty-acid composition (2006-07 seed only; Table 3, AOAC 991.36 by AgriQuality/AsureQuality; samples pooled across plots per treatment so no statistical analysis). Oil content 30.9–32.5 g/100g. Fatty acids as % of total fatty acids: linoleic 53.0–53.1%, α-linolenic 20.2–20.4%, oleic 9.7–9.8%, γ-linolenic 3.8%, palmitic 6.2–6.3%, steridonic 1.5–1.6%. 'There were small differences in oil content across treatments, which ranged from 30.9-32.5 g/100g, but no consistent trend was evident. Plant population had a negligible effect on fatty acid composition.' Abstract: hemp oil fatty-acid content 'approaches the optimal 3:1 balance of essential fatty acids linoleic acid: α-linolenic acid'. ⚠ THIS IS OIL, NOT SEED PROTEIN — no seed protein % is measured or reported anywhere in this paper. [measured NZ field trial, verbatim, verified]"
  source_location: "Abstract; Methods (oil analysis, p.87); Results and Discussion — Oil Quality (p.89); Table 3 (p.89)"
- claim: "MEASURED economic return (before fixed production costs). 2006-07: $3,482–$3,838/ha; 2007-08: $7,183–$8,092/ha (Tables 1 & 2). No statistically significant relationship between established plant population and return in either season. Return formula: Economic Return ($/ha) = {($4.55/kg × yield kg/ha) − ($8.00/kg × seeding rate kg/ha)}, where $4.55 is the net price to the grower (assumed contract price $5.00/kg machine-dressed seed dried to 8% MC, less $0.45/kg cartage+drying+dressing), stock seed cost $8.00/kg, and the seeding rate for each target population was 25, 30, 35, 40, 45, 50 kg/ha respectively. [measured NZ field trial, verbatim, verified]"
  source_location: "Abstract; Methods — economic-return formula (p.87); Results and Discussion — Economic Return (p.90); Tables 1 & 2 (p.89)"
- claim: "Agronomic context and recommendation. Fasamo is a spring-planted, frost-sensitive annual; the 2006-07 crop was sown 10 Oct 2006 / harvested 6 Mar 2007 and the 2007-08 crop sown 2 Nov 2007 / harvested 28 Feb 2008 (a grain crop of roughly this length). Both sites were irrigated Canterbury arable and received fertiliser (2006-07 Ashburton: 250 kg/ha Cropmaster 15 + effluent in irrigation water; 2007-08 Methven: 150 kg/ha Cropmaster 15 pre-plant + 125 kg/ha urea). Recommendation: 'a target plant population of 150-225 plants/m2 is recommended for Fasamo oilseed production in Canterbury depending on site yield potential.' In-paper comparators: McPartland et al. (2004) reported average NZ yields of 950–1,800 kg/ha depending on cultivar; commercial Fasamo in Canterbury has ranged 400–1,000 kg/ha (unpublished, Midlands Seed Ltd). NZ regulation: only cultivars <0.3% THC on the National Approved List may be grown, and only the oil extracted from seed may be sold for human consumption (rest of plant → livestock feed). [measured NZ field trial, verbatim, verified]"
  source_location: "Abstract; Introduction (p.86, THC/regulation); Methods — trial sites, sowing/harvest dates, fertiliser (pp.86–88); General/Recommendation (p.90)"

Neobiome Intelligence relevance

This is the measured NZ hempseed-yield primary that firms OT_112’s compiled hemp figure into a field-measured range — exactly the compiled→measured upgrade that OT_088 performed for the legume-protein cell.

1. Firms the hemp oilseed yield term. OT_112’s Table 4 gives hemp as “800–1,000 kg/ha seed” — a compiled range from seed companies / FAR / overseas information (Table 6), not a NZ measurement. This trial measures it directly: 804–931 kg/ha on a marginal, border-dyke-irrigated pastoral/effluent site (2006-07) and 1,640–1,849 kg/ha on a well-prepared arable site (2007-08). So OT_112’s compiled 0.8–1.0 t/ha matches the low (challenging-site) end of the measured NZ evidence, while a well-managed Canterbury arable site roughly doubles it — and the dominant driver is site/season, not plant population. For the engine’s food/oilseed layer, cite the measured NZ range ~0.8–1.0 t/ha (conservative, marginal-site) to ~1.6–1.85 t/ha (good arable) rather than the single compiled point. In-paper NZ comparators corroborate the spread (McPartland et al. 2004: 950–1,800 kg/ha; commercial Fasamo Canterbury 400–1,000 kg/ha).

2. The measured QUALITY axis is OIL — it firms the dietary-fat/oil-adequacy dimension, NOT the protein cell. Table 3 gives measured seed oil content 30.9–32.5 g/100g and a fatty-acid composition (linoleic ~53%, α-linolenic ~20%, oleic ~10%, γ-linolenic 3.8%) approaching the reported ~3:1 linoleic:α-linolenic essential-fatty-acid balance. This is directly relevant to the dietary-fat-adequacy dimension the food model does not yet represent — the FNP 91 / OT_167 point that a self-sufficient community must produce enough dietary fat/oil (oilseeds, nuts) to clear the 15–20%E fat floor, an adequacy axis beyond veg + protein volume. Hemp is a credible NZ oilseed for that floor, with measured composition. Explicit boundary: this paper measures NO seed protein %. The plant-protein cell (LEGUME_PROTEIN_KG_HA, faba-anchored via OT_088) gets no hemp protein figure here; hempseed protein remains a borrowed/generic constant (RT_363’s hemp-protein leg is not closed by this source). Do not apply, imply, or back-derive a hempseed protein % from this trial.

3. Land-class and input placement corroborate OT_112. Both trials sit on irrigated Canterbury arable (broadly LUC 1–4) — consistent with OT_112’s finding that hemp belongs in NZ’s established arable regions, not on marginal grazing land, and with the model’s arable gate (CR_052, AF-41). On inputs, both sites were fertilised (Cropmaster 15 + urea/effluent) — reinforcing OT_112’s note that hemp needs ~100–130 kg N/ha, a poorer fit for the regenerative/low-input food scope (D_001) than the N-fixing pulses. Do not overstate hemp as a low-input crop.

Caveats for the model:

1. Single cultivar, two sites, two seasons. A robust NZ measurement, but Fasamo-specific and Canterbury-specific; the near-2× year-to-year swing shows how sensitive the figure is to site preparation, soil aeration and irrigation quality — use the range, not a point.
2. Oil, not protein. As above — the quality data firms the oilseed/fat axis only.
3. 8% MC, machine-dressed basis. Yields are corrected to 8% moisture content and machine-dressed (cleaned) — the tradeable-seed basis, not field-fresh.
4. Economic returns are pre-fixed-cost and use 2006–08 contract prices (5.00/kg contract, 8.00/kg stock seed) — indicative, de-date before any cost use.

Notes

PDF (7 pp, pp.85–91) read in full via pdftotext -layout; every key-claim figure grep/table-verified against the extracted text. Downloaded open-access from the NZ Grassland Association journal (nzgajournal.org.nz/index.php/rps/article/download/3180/2804) — an authoritative first-party download, not AI-prepared, so no retrieval-provenance block. The identical paper exists as an Agronomy Society of NZ copy (agronomysociety.org.nz/files/SP13_11…) but that host returns curl-404; the NZGA open copy is the same work. Authorship: the RT_362 retrieval note mis-cited “Nott & McKenzie”; the raw title page reads J.M. Townshend and J.M. Boleyn, Midlands Seed Ltd — the verbatim authors are used here. Publication year 2010 per the retrieval record and the ASNZ Special Publication 13 / R&P Series 14 volume (internal references run to 2008; AOAC 2002 methods). data_quality: high — a measured NZ field-trial primary read verbatim (matches the treatment of the sibling measured primary OT_088); the model-relevant yield is independently corroborated (McPartland et al. 2004; the NZHIA multi-site paper in the same RT_362 drop; OT_112’s compiled range), but held at high rather than verified given it is a single-cultivar two-site trial.

Research targets

Documents to retrieve

• None new — this source is itself the measured NZ hemp-yield primary. It partially advances RT_362 (measured NZ crop yield × composition for the OT_112 specialty-grain/pulse set): the hemp measured-yield leg is now in. RT_362 stays OPEN — soybean and quinoa have no retrievable NZ field-trial primary (soybean’s only NZ trial is paywalled Blair 1966; no NZ quinoa trial exists), and the chickpea measured-yield doc (McNeil, Agronomy Society SP7_23) is curl-blocked and must be fetched via Comet.

Research gaps

• RT_363 (hemp-protein leg) NOT closed. This trial measures hemp seed oil, not protein; a NZ-measured hempseed protein % was not found in any open trial. The hemp-protein cell stays a borrowed/generic constant. No new RT minted (RT_363 already carries the soybean/quinoa protein gap; hemp protein folded into it here rather than opening a speculative placeholder — backlog is closing).

Connections

Links to

Sources (4): CR_052 · OT_088 · OT_112 · OT_167

Decisions (1): D_001: Neobiome food production — regenerative…

Referenced by

Sources (2): LIT_098 · OT_112

SSI indicators (1): I02: Food Security & Sustainable Agriculture

EDT domains (1): D02: Smart Food Systems & Agriculture