Chepil greens (Crotalaria pumila)

Accepted name and local-name boundaries

Kew accepts Crotalaria pumila Ortega in Fabaceae. In the Central Valleys of Oaxaca, a cultivation study identifies the chepil it describes as C. pumila. A Guerrero taxonomic thesis reports “chipil” applied across C. mollicula, C. pumila and C. longirostrata, sometimes indistinctly. That regional name overlap does not make the species synonyms or make their food and safety evidence interchangeable.

Oaxaca food reports

A market study in Zaachila and Zimatlán records “Chepiles” as C. pumila and lists leaves and flowers. Vendors reported four food modes: cooked with black beans, cooked with rice, tamales, and espesado with corn dough. The authors describe chepil tamales in religious-festival contexts and report them sold in the studied markets. The paper does not provide a complete quantity-based method for these dishes.

Regional growing context

The 2021 Oaxaca study describes spontaneous chepil in fields and disturbed vegetation during the rains and deliberate planting in irrigated plots to supply the dry-season market. Its two study municipalities are San Antonino Castillo Velasco and Santiago Apóstol, around 1,479–1,480 m. The paper gives a semi-warm subhumid summer-rain setting, 828.1 mm reported annual rainfall and a 20.7°C mean temperature. These describe those localities, not species-wide tolerances.

Field practices reported by growers

The same Oaxaca paper describes flat irrigation tables about 120–150 cm wide and up to 50 m or longer, rows about 60–70 cm apart, and broadcast sowing of about 100 g chepil seed per 20 m table. It records manure or ammonium sulfate use and local weeding practices. These are reported regional field methods, not household instructions or a controlled comparison. No soil pH or container dimensions are given.

Season and harvest observations

The Oaxaca study reports spontaneous rainy-season growth from May to September and irrigated production mainly during the dry-season market period, roughly September to April or May. It describes harvest from about two months. Some growers uproot plants; others cut and leave the root and about 5 cm of stem, reporting that plants can regrow for additional cuts. These are local producer accounts, not guaranteed yields or validated harvest intervals.

Separate research is not edible-crop guidance

A 2004 paper abstract reports laboratory and greenhouse germination treatments, but its full text was unavailable in the recorded review. A 2024 greenhouse study used C. pumila for phytoremediation research in metal-contaminated soil. A Chapingo thesis measured spontaneous forage biomass in one Valley of Mexico site and season. None is an edible greenhouse growing guide or a household crop trial.

Food and safety limits

A 1989 UNAM thesis analyzed dried, ground plant samples from Alcozauca and reports qualitative alkaloid screening and hemagglutination activity in an extract. The thesis interpreted its alkaloid-screen results as not indicating considerable risk in view of the other reagents, but this was a historical laboratory interpretation, not a human exposure assessment. The authors also said fresh or steamed preparation effects needed further study. These tests do not establish a particular toxin, safe serving amount or validated cooking endpoint, and cannot be transferred to other Crotalaria species named as chipil.

What remains unresolved

No complete recipe batch, voucher-confirmed match for every vernacular food report, safe preparation endpoint, or validated food storage method was established. The reviewed sources do not provide productive container or edible greenhouse guidance, a named hardiness scheme, frost/heat limits or soil pH. The evidence covers selected Oaxaca, Guerrero and Valley of Mexico settings and does not represent all Mexican quelite practice.

Research gaps

Market food modes do not provide a full recipe batch or validated preparation endpoint.

The broad chipil/chepil name applies to more than one Crotalaria species in reported Guerrero usage; not every food report is voucher-linked.

The 1989 lab screening does not identify a specific toxin or establish human risk, safe serving or preparation control.

No species-specific soil pH, household container design or edible protected-crop protocol was found; greenhouse phytoremediation and forage trials are different purposes.

No species-wide frost, heat or climate tolerance and no Mexico-wide coverage were established.

Research status

The underlying plant account received a partial source review. Its review scope is described in the original plant article.