A scoping review of interventions for crop postharvest loss reduction in sub-Saharan Africa and South Asia

The study addresses the question of which interventions small-scale producers and associated value-chain actors in sub-Saharan Africa (SSA) and South Asia can adopt or adapt to reduce postharvest losses (PHLs) along food crop value chains, and the barriers and facilitators to adoption. Motivated by rising global food demand, persistent undernourishment in SSA and South Asia, and SDG 12.3 targets to reduce losses, the paper highlights that PHLs occur across all postharvest stages and are often preventable through appropriate tools, handling practices, training, policy, market and infrastructure improvements. Despite renewed attention after past food crises, impacts of investments have been limited due to narrow focus on technologies, short time frames, and lack of institutional support. This scoping review synthesizes evidence across 22 staple and horticultural crops to inform policy and investment decisions and to identify evidence gaps.

The paper situates its contribution within a context of limited synthesized evidence on PHL reduction interventions, despite numerous trials since the 1970s. Prior work has emphasized PHL magnitudes and opportunities (for example, World Bank/FAO/NRI 2011; Affognon et al. 2015) and highlighted shortcomings such as poor coordination, focus on technology over systems, and lack of training, finance, and policy support. SDG 12.3 and the Malabo Declaration provide policy impetus. Existing reviews and grey literature often lack standardized metrics and comprehensive, cross-crop, cross-country comparisons. This study expands the evidence base through systematic searching, screening, and a structured database enabling cross-tabulations by crop, country, postharvest stage, and intervention type, while also noting persistent gaps (training, policy, finance, market interventions; socio-economic and environmental outcomes).

Design: Scoping systematic review and meta-analysis of intervention efficacy. Scope: Interventions applicable to small-scale producers and associated value-chain actors in SSA and South Asia for 22 focal crops across five crop groups (cereals; legumes; roots/tubers; fruits; vegetables), spanning 1970s–2019. Search: Comprehensive searches (May 2019) in CAB Abstracts, Web of Science, and Scopus plus 47 additional electronic databases/grey literature sources. Initial records: 14,576; deduplicated to 12,786. Update search (Oct 30, 2019) added 121 deduplicated records. Screening: Titles/abstracts auto-coded by semantic ML to assist filtering; all 12,907 records double-screened by researchers; 1,906 full texts assessed; inclusion of 334 studies after exclusions (PRISMA provided). Inclusion required original research in SSA/South Asia on a focal crop; field/real-world scale; effect of an intervention on PHL (with comparator: other intervention, adopters vs non-adopters, or pre/post adoption). Language limited to English/French. Minimum replicate sizes set by commodity (e.g., ≥50 kg/replicate for maize; ≥25 kg for rice/wheat/sorghum; ≥10 kg for legumes; ≥20 kg roots/tubers; ≥10 kg fruits/vegetables). Laboratory-only, artificially infested, or insufficiently sized studies were excluded. Data extraction: Coding framework with four-tier intervention classification—Tier 1: type (technology/tool/equipment; handling practice change; training/extension; finance; policy; markets; support/organization; infrastructure). Tier 2: postharvest stage (harvest, drying, storage, packaging, transport, etc.). Tier 3: specific intervention (e.g., hermetic bag, traditional granary + synthetic chemical). Tier 4: detailed description (e.g., chemical name/rate, container size). Part II captured PHL metrics (quantity and quality), adoption barriers/facilitators, study design/duration/scale, intervention costs, and any socio-economic/environmental outcomes. Standardization and synthesis: Quantitative and qualitative loss metrics aggregated into groups. For durable crops, storage losses standardized to 6 months (cereals) and 4.5 months (legumes). Perishables not time-standardized due to varying temperatures and non-linear effects. Meta-analyses pooled means at intervention level (Tier 3); 95% CIs calculated when n>2; for durable crops, ANOVA with Holm-corrected LSD (R agricolae) to group means (P<0.05). Data visualized via an open-access SQL database (https://PHCeres2030.net/).

Evidence profile: Of 12,907 screened records, 334 studies (2.6%) met inclusion criteria. Most were journal articles (85.9%); earliest 1971; 42.2% published in the last decade. India accounted for 32.2% of studies; 25 countries had none. Maize was most studied (24.9%). By crop group: cereals 43.3%, roots/tubers 19.9%, fruits 19.2%, vegetables 10.7%, legumes 6.8%. By postharvest stage, storage dominated (storage-dry 42.5%; storage-fresh 40.1%). Interventions were overwhelmingly technologies/tools/equipment (88.3% of studies; 89.0% of interventions); far fewer on handling practices (14.1%; 10.5%), training (0.6%; 0.3%) and infrastructure (0.3%; 0.1%); none on policy, finance, markets or organizational support. Cereals: Storage interventions (pesticides—synthetic/botanicals, modified atmospheres, hermetic containers) predominated. In maize, across 78 storage studies (74 SSA, 4 South Asia), hermetic storage (bags, metal/plastic silos), admixture with diatomaceous earth (DE) or cooking oils, and fumigated/insecticide-sprayed bag stacks kept mean 6-month weight loss below ~2%. Highly variable outcomes with synthetic protectants (e.g., polypropylene bag + synthetic chemical, n=21: 7.2% ±11.2% weight loss). Handling improvements (selecting tight husks, proper drying, hygiene, timely harvest) reduced infestation, weight loss, and aflatoxin. For wheat/rice/sorghum, hermetic storage, sealed containers, improved granaries, pits, and chemical/botanical/DE treatments often kept 6-month weight loss <2% and damage <5–6%; traditional sacks/granaries without treatment had much higher losses (e.g., rice: 2.8–21.8% weight loss; 16.4–20.3% damage). Legumes: Research volume lower; 86.9% focused on storage of dried legumes; cowpea dominated (53.2%). Untreated storage in jute/poly sacks resulted in severe damage (46–70%) and high weight loss (cowpea ~18.9% over 4.5 months). Hermetic bags consistently reduced quantity and quality losses across cowpeas, groundnuts, and beans; synthetic chemicals, botanicals, and DEs admixed with grain also lowered damage by ≥20 percentage points versus controls. Simple handling changes (e.g., weekly sunning/sieving) reduced bean damage to ~3.6–4.1% vs 37.7% control. Sorting/drying groundnuts markedly lowered aflatoxin B1 (e.g., from 55 ppb to 17 ppb; Gambian training study reduced average to 0.28 ppb with 1.9% weight loss). Roots and tubers: 70.7% of interventions were storage protectants/structures (ventilated, evaporatively cooled, cold rooms, improved pits). Potato: improved pits, evaporative cooling, cold rooms, and sprout suppressants (e.g., chlorpropham) reduced quantity (<15.5%) and quality losses (<8.5%). Yams: botanicals, essential oils, biocontrol, heat, irradiation reduced quantity losses (<20%); irradiation lowered both quantity and quality losses and sprouting. Handling practices (16% of interventions) such as curing, piecemeal harvest, gentle handling, sorting damaged tubers reduced losses; infrastructure (better roads) associated with reduced potato losses in Ethiopia. Fruits: Interventions included storage protectants (waxes/coatings with/without fungicides/botanicals, heat), packaging (crates, liners, padding, MAP, shrink-wrap), and storage structures (evaporative/cold rooms). Packaging reduced quantity loss in citrus; however, liners often increased decay due to humidity (e.g., wooden boxes: 6.6% quality loss; with liner: 22.6%). Shrink-wrapping reduced both quantity and quality loss. Without protectants, citrus suffered high losses (~34% quantity and quality). Mango quality loss was reduced by hot-water treatment (~13%) and pesticides (~16.6%) over 9–14 days; improved handling cut quantity loss from 25% to 5% and damage from 68.3% to 22.5%. Vegetables: Onions stored in shaded/traditional structures had high losses (quantity 22.1–50.5%; quality 5.5–73.0%); structures with airflow/RH/temperature control reduced losses (quantity 6.1–16.6%; quality 2.9–5.7%). Curing reduced onion losses (e.g., from 47.0% to 31.0%). Tomatoes: traditional packaging (wooden boxes, rough baskets) had higher losses (quantity 23.8–48.3%; quality ~17%); plastic crates and improved baskets reduced losses (quantity 7.9–17.5%; quality 3.2–6.9%); MAP showed low losses (quantity 6.2%; quality 5.7%). Socio-economic/environmental outcomes: Only 13.1% mentioned such outcomes; 12.5% economic, 3.0% social, 1.2% environmental. Costs reported in 11.4% of studies: from <US$1 (tools, packaging) to ~US$2,000 (evap-cooled cold room, 8–10 t modified AC) to US$4,000 (20 t hermetic cocoon) to US$36,000 (combine). Some storage interventions linked to reduced aflatoxin risk and improved consumption smoothing; mechanization reduced drudgery but could displace labor. No studies reported gendered outcomes; only two covered economic, social, and environmental outcomes together. Adoption: Five studies analyzed adoption drivers (mostly maize). Adoption positively related to efficacy, lifespan/durability, favorable cost–benefit, household size, literacy, land size, financial services, and off-farm income; negatively to road distance and female primary decision maker for metal silos. Barriers included high upfront costs, weak distribution, limited participatory development/testing, low awareness, lack of credit/subsidies/input markets, and trade-offs (e.g., bulkier packaging, reduced seed viability). Facilitators included cost-effective, time-saving, effective, maintainable interventions, integration with existing practices, quality-sensitive markets, participatory learning approaches, and training/awareness.

The review shows that effective PHL reduction solutions exist, particularly hermetic storage and appropriate protectants for cereals and legumes, improved storage environments and sprout suppression for roots/tubers, and better packaging, coatings, and temperature/humidity control for fruits/vegetables. However, the evidence base is narrow—dominated by on-farm storage technologies for maize—and thin for non-storage stages, legumes, vegetables, and for actors beyond farmers. Findings address the research question by cataloging interventions, quantifying their technical efficacy, and identifying adoption factors. The study underscores that PHLs accumulate along value chains and that one-size-fits-all technological fixes are insufficient. There is urgent need for studies across multiple stages (harvest, handling, cooling, packaging, transport, processing), multiple seasons/sites, and involving traders, transporters, processors and service providers. Moreover, integrating technical interventions with training, finance, infrastructure, policy, and market incentives is essential to realize sustained PHL reductions and to deliver socio-economic and environmental benefits aligned with the SDGs.

This scoping review synthesizes 334 studies across 22 crops in SSA and South Asia, offering an open-access database of interventions by crop, country, postharvest stage, and type. It highlights technically effective interventions (e.g., hermetic storage; DE/botanical/synthetic protectants; improved storage structures; wax/coatings and heat treatments; improved packaging and curing) and major evidence gaps in non-storage stages, training/finance/policy/market interventions, standardized metrics, multi-season/site validation, and socio-economic/environmental outcomes. Future research should broaden crop and actor coverage; rigorously evaluate combinations of technologies with training, finance, infrastructure, and policy; standardize loss measurement; and embed participatory, real-world trials. Policy and investment recommendations (Box 2) emphasize awareness-raising, context-appropriate solutions, coupling technical solutions with training/management, studying policy/finance/infrastructure impacts, supporting quality-sensitive markets, and fostering multistakeholder platforms for co-learning and multi-location/multi-season studies.

The evidence base is limited in size and skewed geographically (India overrepresented) and by crop (maize) and stage (storage). Many interventions have few studies per crop/context, with heterogeneity in design, duration, metrics, and environments, constraining robust cross-study comparisons. Exclusion of small-scale/lab-only/artificial infestation studies reduced volume but improved decision relevance; however, short storage durations in some studies limit inference for longer farmer storage periods. Non-publication bias may overstate efficacy. Details on treatment rates, costs, and preceding fumigation were often missing. Few studies involved farmer/actor participation, potentially limiting real-world effectiveness and acceptability. Socio-economic and environmental outcomes, gendered impacts, and adoption drivers were rarely measured. Standardized, comprehensive loss metrics were inconsistently applied.