# Amaranth (Amaranthus spp.) — Literature Survey

**ResearchForge sweep** · topic workspace: `research/amaranth-seeds/`
**Date:** 2026-08-19 · **Depth:** comprehensive (24 query variants, 4 sources)
**Status:** metadata-level survey only — screening, full-text acquisition, and claim extraction are **not yet done** (human gates).

## 1. Method and limits

- 24 query variants covering the PRD research areas for amaranth (grain production, germination, planting/density, photoperiod/phenology, irrigation/water, soil/salinity, fertilization, pests/weeds, harvest/maturity, threshing/cleaning, drying/storage, popping/processing, nutrition, Mexico relevance).
- Sources: **OpenAlex, arXiv, Crossref, Semantic Scholar** (the full `all` preset stalled on chia and was not re-run here; see §7).
- 1,899 raw records → 1,560 deduped **unique DOIs**.
- **Limits:** metadata-level only (titles/abstracts/venues/years); no PDFs acquired, no full-text claims, no numeric values. Screening decisions and full-text acquisition are human-required per the rforge automation policy. `web_search` unavailable (insufficient balance), so no supplementary web pass. Semantic Scholar rate-limited (HTTP 429) on 20/24 queries.

## 2. Bottom line

1. The literature is **extensive and nutrition/biology-heavy** but the **agronomy is thinner and more scattered** than chia's. Biology/genome/chemistry dominates the corpus; field agronomy (density, fertilization, water) is present but spread across many single-site studies.
2. Amaranth agronomy in the academic set is **regionally concentrated** (Kenya/Africa, Argentina, India, Egypt, China, Italy) — precisely reflecting that amaranth is grown as a smallholder/alternative crop. **Mexico origin/genepool** is under-represented in this sweep despite the crop being Mesoamerican; the strongest Mexican-agnostic signal is `Amaranthus hypochondriacus`/`cruentus` breeding and the two grain-amaranth genomes.
3. **Human-scale postharvest knowledge (manual harvest, threshing, winnowing, cleaning of small plots) is essentially absent from the academic literature** — the milling/flour/popping literature is food-science, not manual small-scale processing. Confirms the project premise that practical/grower sources and our own experiments will carry the human-scale core, exactly as it did for chia.

## 3. Key themes and exact papers (initial candidates, DOIs from the sweep)

### 3.1 Biology, taxonomy, genomes
- **Draft genome of *A. hypochondriacus*** (C4, high-lysine pseudocereal) — `10.1093/dnares/dsu021` (2014)
- **Chromosome-level *A. cruentus* genome** — `10.1111/tpj.15298` (2021)
- Transcriptomic analysis of grain amaranth (*A. hypochondriacus*) — `10.1186/1471-2164-12-363` (2011)

### 3.2 Germination
- Temperature effects on germination of **nine Amaranthus species** — `10.1614/ws-03-012r` (2004)
- Germination of grain amaranth (*A. hypochondriacus* × *A. hybridus*): seed quality × temperature — `10.1016/s1161-0301(97)00049-x` (1998)
- Soaking/germination and anti-nutritional/bioactive changes — `10.1016/j.crfs.2021.11.019` (2021)

### 3.3 Planting date and density (core small-scale questions)
- **Row spacing × plant population × cultivar** for grain amaranth, Northern Great Plains — `10.2134/agronj2000.922329x` (2000)
- **Optimal sowing dates of three grain-bearing amaranth species**, semi-arid Argentina — `10.5424/sjar/2004023-93` (2004)
- Planting density × poultry-manure effects on *A. cruentus* — `10.15835/nbha3723458` (2009)

### 3.4 Photoperiod / phenology / flowering
- Genetic variability in phenology, *A. tricolor* — `10.47815/apsr.2021.10033` (2021)
- Grain yield × related-trait associations, *A. cruentus* — `10.1515/johr-2016-0025` (2016)

### 3.5 Water / drought (Mexico-relevant)
- Betacyanin photoprotection in *A. cruentus* leaves under **water stress** — `10.1007/s11099-011-0062-7` (2011)
- Irrigation suitability evaluation, *A. cruentus* (S. Africa) — `10.5897/ajps.9000154` (2008)
- Physiological drought-tolerance screening in vegetable amaranth (*A. tricolor*) — `10.3390/agriculture11100994` (2021)

### 3.6 Soil / salinity
- Growth and organic/inorganic solutes in **amaranth under salt stress** — `10.1590/1983-40632016v4742580` (2017)
- Strategies of underexploited Amaranthus species to confront **salt stress** — `10.3389/fpls.2021.604481` (2021)

### 3.7 Fertilization
- Vegetable and grain amaranth yield/quality affected by **applied N** — `10.21273/hortsci.25.9.1078` (1990)
- **Burkholderia** growth-promoters increase grain-amaranth yield — `10.1371/journal.pone.0088094` (2014)
- NPK × poultry manure on three Amaranthus — `10.1155/2014/828750` (2014)
- Maize-stover compost × N on growth/yield/nutrients — `10.1016/s0304-4238(01)00305-3` (2002)

### 3.8 Pests / weeds (small-scale relevant)
- Farmers' knowledge & practices: insect pests of **leafy amaranth**, Kenya — `10.1093/jipm/pmz029` (2019)
- Weeding frequency effects on grain amaranth yield — `10.1016/s0261-2194(97)00017-3` (1997)
- Screening *A. hypochondriacus* cultivars for disease tolerance — `10.1007/s13237-020-00335-y` (2020)

### 3.9 Harvest / maturity / seed moisture
- Effect of seed moisture at hand-harvest/machine-threshing on seed quality — `10.15258/sst.2004.32.1.15` (2004)
- Evaluation of accessions for early vegetative yield — `10.24297/jaa.v5i3.5041` (2016)

### 3.10 Threshing / winnowing / seed cleaning (research gap)
- **Small-scale and portable seed cleaning** (book chapter) — `10.1201/b10312-68` (2016)
- Adaptation of grain-cleaning equipment for small seeds (kalonji/sesame analogy) — `10.3390/agriengineering7060179` (2025)
- Roller milling potential of amaranth grains — `10.1016/j.jcs.2016.11.006` (2016)
- > Academic coverage of **manual hand-scale threshing/winnowing of amaranth** is effectively absent — flagged for grower/YouTube + original experiments.

### 3.11 Drying / storage / longevity
- Microwave drying & seed viability — `10.1016/j.biosystemseng.2021.12.019` (2022)
- Priming & storage effects on amaranth germination — `10.15258/sst.2006.34.1.18` (2006)

### 3.12 Popping / grain utilisation
- Germination and popping effects on anti-nutritional compounds & digestibility — `10.3390/foods11142075` (2022)
- Whole-amaranth flour bread properties & nutrition — `10.1016/j.lwt.2012.07.031` (2012)

### 3.13 Mexico relevance (user-directed)
- Entomofaunistic study of amaranth (*A. hypochondriacus*) cultivation (Puebla, MX) — title-only record in sweep (no DOI captured)
- The crop is **native to Mexico** (huautli/alegría); Mexican-origin agronomy appears thin in this sweep — flagged for extension/grower + Spanish-language sources.

> ⚠️ **No numeric claims are asserted in this report.** All values must be extracted from the papers themselves during screening/reading. The enrichment from this sweep is knowing *which* papers exist per topic.

## 4. Coverage by research area (PRD §8.2 map)

| Research area (amaranth) | Found in sweep? | Notes |
|---|---|---|
| Grain production / biology / genome | **Strong** | two genomes; breeding, nutrition |
| Germination | **Strong** | temperature light salinity |
| Planting date / density | **Strong** | N. Plains row-spacing; Argentina sowing dates |
| Photoperiod / flowering | Moderate | phenology papers exist |
| Water / drought | Moderate | *A. cruentus* water-stress papers |
| Soil / salinity | Moderate–Weak | some salt-stress studies |
| Fertilization | **Strong** | N, NPK×manure, compost, microbial |
| Pests / weeds | Moderate | Kenya leafy-amaranth; weeding trials |
| Harvest / maturity | Weak | little captured (role of seed-moisture paper) |
| Manual threshing / winnowing / cleaning | **None found** | confirmed research gap |
| Drying / storage | **Strong** | drying, viability, priming |
| Popping / grain processing | Moderate | food-science, not small-scale |
| Mexico genepool | Weak in this sweep | native crop; extension/Spanish sources needed |

## 5. Claim hygiene (no numbers asserted in this pass)

- This sweep proves **existence and coverage**, not truth of any specific number.
- All numeric claims enter topic pages with evidence levels and DOIs once full text is read, per `docs/methodology/evidence.md` and `source-quality.md`.
- The yield page must record context per figure (cultivar, location, date, density, water, N, harvest method); conflicts are expected and welcome — amaranth yield reports are strongly context-dependent.

## 6. Evidence gaps (what the sweep did **not** find)

1. **Human-scale manual processing** (hand harvest, thresh, winnow, clean several kg): essentially absent — the core gap, to be filled by growers/YouTube + experiments.
2. **Mexico-specific agronomy at home-garden scale**: thin in this academic sweep despite Mesoamerican origin — needs extension/Spanish-language/grower sources.
3. **Sowing at home-garden scale** (beds, not ha): quantitative gap — papers report kg/ha and plants/m²; conversion to g/m² per `units.md` needed.
4. **Small-scale pest/bird/weed pressure** in grain amaranth specifically: weak; weeding reviewed, insect pests mostly on leafy amaranth.
5. **Long-term home storage** of amaranth seed: partial only.

## 7. Failures and limitations of this run

- Semantic Scholar: 20/24 queries rate-limited (HTTP 429) — recorded in `failures.jsonl`; no `search resume` re-run performed this pass.
- OpenAlex + Crossref dominate the useful set; arXiv is noise-heavy for amaranth (food-science preprints).
- The `all` multi-source preset was not re-run (it stalled on the chia pass; see `research/chia-seeds/report.md`). OpenAlex/Crossref coverage is adequate for a metadata survey.
- `rforge evidence grid/gaps` not available in installed v0.1.20 (only `evidence audit|risk-bias-*`), so synthesis was done manually from metadata.

## 8. Next steps (human gates marked)

1. **[HUMAN] Screening.** Review `queue.csv` (pending papers) — include/exclude/uncertain per paper. (Required before any full-text work.)
2. **[HUMAN] Full-text acquisition approval** — then `rforge oa fetch` for accepted OA papers (automation policy: `full-text-acquisition-approval` = human-required).
3. **[AGENT after gates]** read accepted papers, register per source note (S-####) using `templates/academic-source.md`, extract claims with evidence tags inline into `docs/crops/amaranth/*.md`.
4. **[AGENT]** update `data/crops/amaranth.yaml` progressively where evidence warrants.

## 9. Provenance

See `provenance.json` in this directory. Query logs: `queries.txt`; raw per-source: `raw/`; manifest: `manifest.json`; failures: `failures.jsonl`. Coverage analysis script: `scripts/analyze-amaranth.mjs`.