Spider plant greens

Name and regional context

Kew’s Plants of the World Online accepts Cleome gynandra L. and lists Gynandropsis gynandra as a homotypic synonym. ITIS retains the genus placement Gynandropsis under a different taxonomic treatment, so both names appear in research and older crop materials for this subject. South Africa’s Agricultural Research Council lists Cleome gynandra as spider plant; the stable subject here follows Kew’s accepted placement.

The BvAT guide records names including Mwangani (Swahili), Thageti (Kikuyu), Tsisaka (Luhya), Alot-dek (Luo), Saget (Kalenjin), Chinsaga (Kisii) and Mwianzo (Kamba). The HORTINLEA manual records several of the same names with different spellings and additional forms. These are community- and source-specific names; a common name alone does not identify a gathered plant.

Zambian sources use several names for this species. Nguni and Mwila’s 2007 survey records Shungwa and Luyuni-yuni as Tonga names, Lubanga as Lenje, and Suntha as Nyanja in three named community sites. The 2007 National Food and Nutrition Commission report instead assigns Lubanga to Bemba; the difference is left unresolved. The 2023 Zambia row lists lubanga and suntha without a language or district. A shared spelling does not establish one national or community-wide mapping.

The central Malawi study lists C. gynandra among indigenous vegetables grown and names “Cats’ whiskers (Luni)” in its stakeholder-promotion text. A separate Table 2 row gives Luni without a scientific name, and the authors say some vernacular-only identities remain under follow-up. A UNDP Small Grants Programme summary separately associates C. gynandra with Luni but does not name a language. An indexed excerpt from the B4FN report also associates Luni with C. gynandra and gives “Chechewa” and herb leaves, but the linked report could not be inspected directly. These records do not establish one crosswalk for every Luni entry. [S13, S17, S18, S19]

In Zimbabwe, Flora of Zimbabwe lists Bangara, Nyevhe, Rudhe, Runi and Tsuna as Shona names and Ulude as Ndebele. The title of a Zimbabwe horticulture paper spells Nyeve and assigns it to Shona; Honde Valley Table 2 records a Cleome leaf-and-relish entry as RuniC, with M and voucher ITM044. These source spellings and study labels are not interchangeable name mappings. [S21, S22, S23]

Food uses described in sources

The HORTINLEA manual describes leaves, young shoots and flowers boiled or stewed, including with amaranth, milk, groundnut paste, or tomato and oil. The BvAT guide says fresh leaves and shoots are cooked alone or mixed with other vegetables, and notes that leaves may be mixed with milk. These accounts document uses; they do not give a single standardized recipe.

PROSEA describes the leaves as bitter and reports that cooking and fermentation reduce bitterness; it also records pot-herb and sauce-flavouring uses in East and West Africa and salted leaves used as pickles. Its entry does not provide household process controls. The BvAT guide’s reported sour-milk marination for two to three days omits quantities, temperature, acidity, vessel and endpoint, so it is not a preservation instruction. Medicinal claims in source material are outside this article’s food-use scope.

The 2007 survey interviewed 280 households in Hoffmeyer (Nyimba District), Subcentre (Siavonga District), and Chikumbi (Chibombo District). In those communities, respondents most often reported harvesting leaves (64%); 21% reported leaves with stems and 6% leaves with flowers. The paper places leafy vegetables in relishes eaten with nshima, but its general relish description is not a Cleome-only formula. The NFNC report’s Cat’s whiskers entry names Cleome gynandra and says to boil until cooked, then add groundnuts; it gives no quantities or measured endpoint. The CTDT/Oxfam Zambia row identifies leaves and lists boiled and dried preparation categories. These are records of reported practice, not a standardized method.

Jinazali and colleagues sampled young tip leaves from uncultivated C. gynandra plants growing around homes and in the wild in Dedza, Lilongwe and Salima during February–March 2014. They collected 1,500 g of fresh material per district for food-composition analysis. The paper’s introduction mentions relish use through earlier references; its study was not a new household-use survey or cooking report. Maseko and colleagues’ four-village Zomba study lists C. gynandra and C. monophylla as separate wild-food entries, without assigning either row a plant part, preparation, local name or linked voucher. [S14, S20]

Zimbabwe food records describe leaves rather than a complete recipe. Flora of Zimbabwe says the leaves are locally eaten. In a two-village Honde Valley study, the table lists C. gynandra leaves as relish. A separate Shurugwi study across seven communities reports the species among edible weeds and records local rainy/harvest-period and dry-season availability. These observations describe those study settings, not national prevalence or a standardized dish. [S21, S23, S24]

Growing conditions in the available guides

The BvAT guide describes a warm-season crop in East and Southern Africa, commonly found during rainy seasons. It reports sensitivity to cold, poor performance below 15°C, better growth in sunlight than shade, and preference for sandy loam over wet, marshy or heavy clay soils. It advises sowing with the rains and watering during vegetative growth when rainfall is inadequate. These are guide statements, not universal thresholds for every cultivar or location.

The same guide reports short-term drought tolerance but also says drought can hasten flowering and lower yield. PROSEA similarly describes full-sun growth, sensitivity to waterlogging and relative drought resistance, while noting its account is Southeast Asian in geographic framing. Its agronomy describes seed propagation, nursery or direct sowing, thinning, and leaf harvest from the second month onward. Different regions and production systems are represented; the sources do not establish one global calendar or a container/greenhouse regimen.

Zambia evidence describes practice rather than a measured crop response. In Nguni and Mwila’s survey, direct broadcasting was the most commonly reported planting method for Cleome (76%); priority leafy vegetables were generally planted in the warm, wet season from late November to early April. Those figures describe the sampled communities and are not a Cleome-specific sowing rate or climate limit. A later Eastern and Southern Province household study recorded Cleome collection, preservation, and cultivation in its current-season table; its authors identify voucher 22265 for Cleome and report 3% cultivated among 157 respondents. The CTDT/Oxfam country table lists agricultural fields and home gardens among settings where the species occurs. None of these is a controlled yield trial.

The 2026 participatory study covered 84 farming households in Dedza and Nkhotakota and lists C. gynandra among vegetables grown. Across indigenous vegetables, it reports that 66.6% of farmers grew them in the rainy season without additional irrigation and 11.0% used irrigation; these are aggregate practices, not a Cleome calendar or yield result. A UNDP project summary for central Malawi reports outreach, seed multiplication and one season of farmer growing and selling, but no plot-level design or species-specific yield. Jinazali’s paper gives altitude, rainfall and temperature descriptions for its three sampling zones, attributed to a 2003 ministry guide; these describe collection locations, not plant tolerance. The study sampled uncultivated weeds and did not test production, propagation, water treatments, containers or protected growing. [S13, S17, S20]

A Zimbabwe field paper reports a single outdoor randomized-block trial at the Horticultural Research Centre near Marondera using a green-stem landrace, seven animal-manure treatments and three replicates. The paper reports 32.68 tonnes per hectare of fresh leaves under its 50-tonnes-per-hectare goat-manure treatment and 4.94 tonnes per hectare in the control. Those are results from one site and trial management, not home rates or universal yield expectations. The institute page describes the HRC at 1,630 m, with mean annual rainfall of 873 mm and monthly mean maximum temperatures of 19.5–24.6°C; these are station descriptors, not crop limits. [S25, S26]

The trial reports 0.3 × 0.15 m spacing and a 1.62 m² net plot, then states a thinning density of 22,000 plants per hectare. Uniformly applying that density to 1.62 m² gives about 3.56 plants; the stated spacing implies about 222,222 plants per hectare. The paper does not explain timing or layout, so the inconsistency remains unresolved and neither figure is selected as the error. The article abstract for an older Zimbabwe study says production of both C. gynandra and C. tridens was examined in two natural regions under fertilizer treatments, but provides no methods or results; its linked full text could not be retrieved in this pass. [S25, S22]

Harvest, handling and open questions

The BvAT guide describes repeated shoot harvests influenced by soil fertility and moisture, followed by allowing plants to flower and set seed. Its overnight bagging and next-morning water-dipping passage is market-handling advice from that guide, not a validated household storage life. PROSEA says leaves may be dried, a practice it describes as common in Africa, but does not specify a safe moisture target or storage conditions.

The ARC page places Cleome gynandra within its South African indigenous-crop research programme and describes ongoing work on planting dates, planting methods, water and fertilizer; this is programme context rather than a complete home-growing manual. The sources reviewed leave unresolved locally calibrated frost and heat responses, container and protected-culture performance, food storage limits, and safe household fermentation controls.

The 2007 survey reports that about 60% of respondents who preserved Cleome leaves blanched them before sun-drying; the paper also describes varied drying and storage practices. The Hohenheim study separately reports some respondents blanching edible weeds, including Cleome, before drying. These survey accounts do not validate a preservation control or storage duration. A 2024 COLEAD notice says the Zambia Agriculture Research Institute was setting up field trials involving Cleome in crop systems, but publishes no results. Zambia-specific soil response, crop yield, container performance, protected growing, and food-storage endpoints remain open.

Jinazali and colleagues oven-dried remaining samples for one day at 60°C after laboratory moisture testing. That was analytical sample preparation, not a household drying or preservation method. The Malawi material therefore adds no validated drying endpoint, storage duration, or food-safety control. The 2017 Maseko correction page remains a subscription preview; the full 2018 correction states corrected funding information and the article’s later open-access licence, and does not report a change to its food table. [S15, S16, S20]

Zimbabwe sources add no measured frost-survival threshold, species response curve, irrigation treatment, mature-container crop or greenhouse production method. The HRI page mentions a production manual for Nyevhe, but that manual was not inspected here. No household food-storage or preparation endpoint is established by the field trial or visible Acta abstract. [S25, S26, S22]

Research gaps

No reviewed source establishes a universal frost/heat threshold, container scale, or greenhouse growing regime.

The reported sour-milk marination and fermentation accounts lack process parameters and safety endpoints; they are descriptive rather than executable.

No household fresh-leaf storage duration or safe preservation endpoint was established.

The broad Kew use category 'poison' lacks part, dose and context in the inspected profile and is not interpreted here as a food hazard determination.

The 2022 final journal article could not be reopened in full; detailed table and method statements here rely on the full 2021 author-uploaded preprint.

The Zambian household percentages describe sampled respondents, not yield, national prevalence, or a recommended cultivation method.

The NFNC report and Nguni–Mwila survey assign Lubanga to different languages; the sources do not resolve the difference.

No Zambia-specific controlled soil, irrigation, container, protected-crop, or yield trial was found in the inspected sources.

The reported blanching and drying practices do not establish validated household storage or food-safety limits.

Malawi’s Luni and Denje vernacular rows are not binomial-linked in the central Malawi chapter; the B4FN table remains indexed-only and does not resolve every local-name use.

The Malawi crop survey reports aggregate indigenous-vegetable practice, not a Cleome-specific growing calendar, yield, irrigation response or container/protected-culture trial.

The Jinazali article reports a food-composition sample, not a household preparation or cultivation study; its cited agro-ecological ranges are site context, not plant tolerances, and laboratory drying is not a preservation endpoint.

The 2017 Maseko correction text remains inaccessible; the full 2018 correction concerns funding and licensing and does not report a change to the food table.

Zimbabwe food accounts describe leaves and local dishes but do not provide a complete source-specific recipe with quantities and an endpoint.

The Marondera trial reports spacing, plot area and thinning density that do not reconcile; its timing and layout are not explained.

The Acta Horticulturae abstract names two Zimbabwe natural regions but not their locations, trial methods or results; the linked PDF was blocked by a JavaScript challenge.

No Zimbabwe source in this amendment establishes a frost limit, irrigation response, mature productive container, greenhouse method, or household storage and food-safety endpoint.

Research status

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