A research brief for Singapore's vegetable resilience strategy
Why storable roots, fruiting vegetables and leafy greens require different production models
Executive conclusion Singapore should avoid assuming that higher-tech farming is automatically the best response to land scarcity. The better approach is to match each crop with the production model that delivers the strongest overall outcome across cost, energy, resilience, freshness and resource use. The evidence supports a crop-specific architecture: i) use diversified sourcing and storage for storable roots and tubers; ii) centralised protected cultivation that exploits natural light for selected fruiting vegetables; and iii) distributed, near-consumer production for highly perishable leafy greens where Harvest-to-Consumption Time matters most. This is a stronger basis for policy than ranking technologies by farm-gate yield or price per kilogram alone.
1. The better suited methods should be crop-specific, not technology-specific
Singapore Food Story 2 has already moved towards a more targeted local-production strategy. SFA's 2035 goal is for local farms to build the capacity to supply 20% of local fibre consumption, covering fresh leafy and fruited vegetables, beansprouts and mushrooms. [1][2] This policy direction now needs to be translated more carefully into execution. Selecting crops and technologies primarily on the basis of high yield, market price or technological sophistication can produce misleading conclusions. A more robust approach is to match each crop category with the production and resilience model best suited to its biology and supply-chain characteristics - whether that means strategic storage, protected cultivation or production close to the point of consumption.
Established controlled-environment research also cautions against using a single productivity measure. Recent benchmarking work argues that yield should be assessed together with energy, water, cost and carbon, while crop-selection studies show that plant architecture, crop duration, harvest index and lighting needs materially affect viability. [3][4] A high fresh-weight yield can also favour water-rich crops without necessarily representing better nutrition or lower system cost. For Singapore, the more useful unit of analysis is therefore the delivered edible output at the point of consumption, including the resources and post-harvest steps required to get it there.
Recommended crop-production framework
2. Storable roots and tubers: resilience through sourcing and storage
Roots and tubers need to be treated differently. Potatoes, sweet potatoes and yams can be stored for months under appropriate curing, temperature and humidity conditions, but cassava can deteriorate rapidly after harvest. [5][6] FAO guidance, for example, lists potential storage periods of around 10 months for fresh-market potatoes and around six months for sweet potatoes under suitable controlled conditions, while separately warning that cassava is much more perishable.
The policy implication: where a crop is demonstrably storable, Singapore can create resilience through diversified source countries, contractual supply access and appropriately designed storage rather than high cost urban cultivation. Root and tuber crops also sit outside the specific fresh leafy/fruited vegetable scope highlighted in SFA's current 20% fibre target, so they are best treated as a comparator in this paper rather than the core local-production case.
3. Fruiting vegetables: the strongest case for centralised protected cultivation
Fruiting vegetables such as tomatoes, cucumbers, capsicums and eggplants are biologically different from leafy greens. They require a longer reproductive cycle, larger plant architecture and, for many crops, active pollination or fruit-set management. These characteristics make them less suited to dense stacked plant factories, but potentially well suited to centralised protected cultivation that captures free solar radiation.
Wageningen University & Research provides useful evidence of the production potential of this architecture. Its next-generation semi-closed tomato greenhouse research targeted approximately 60 kg/m²/year while reducing energy input by about 40% relative to the reference greenhouse concept. Separate Winterlight greenhouse trials reported cucumber yields of 112-118 kg/m²/year, with 40-50% lower energy use than commercial reference growers during the trial periods. [7][8]. The Winterlight trials demonstrate that increasing the effective use of natural sunlight can raise cucumber productivity while reducing dependence on purchased energy, provided the greenhouse and climate-management system are designed around that
objective.
These figures should not be transplanted directly to Singapore. Tropical protected cultivation has its own heat and humidity constraints. A recent Samoa case study reported annual protected-cultivation yields of about 16.35 kg/m² for tomato and 10.35 kg/m² for bell pepper, substantially below advanced temperate greenhouse benchmarks. [9] This is precisely why Singapore should test crop and greenhouse architecture under local conditions instead of assuming that either a European greenhouse or a fully enclosed plant factory will transfer efficiently.
Harvest maturity, quality and crop-specific food safety
Proximity to consumption provides another potential advantage for selected fruiting vegetables. It can allow the fruit to be harvested at a later stage of maturity because less shelf life and transport durability are required.
This is particularly relevant to tomatoes, which are commonly harvested at the mature-green stage for long distance distribution and subsequently ripened after harvest.
Tomato ripening also involves an important biochemical change. Immature green tomatoes contain higher concentrations of the steroidal glycoalkaloid α-tomatine, which declines substantially as the fruit ripens. This is a crop-specific consideration rather than a general rule for all fruiting vegetables: eggplants contain different glycoalkaloids, while green bell peppers are routinely and safely consumed before full colour maturity.
Long-distance distribution may require some climacteric produce to be harvested before full eating ripeness and subsequently ripened under controlled conditions, including ethylene exposure. While this is an established and safe commercial practice, production closer to consumers can allow selected crops to remain on the plant to a later stage of maturity, potentially improving flavour, eating quality and reducing post harvest handling.
Critical distinction: Leafy greens are among the easiest crops to grow efficiently in vertical farms, but the farming system itself remains expensive because it relies heavily on artificial lighting, cooling and infrastructure. Fruiting vegetables require more space and longer growing cycles, but can be better suited to greenhouse production where sunlight provides much of the energy required for growth.
4. Leafy greens: the strongest case for proximity to consumption
Leafy greens create a different problem because the harvested product remains metabolically active. Respiration and water loss continue after harvest, and quality deteriorates even when refrigeration slows the process.
Harvard T.H. Chan School of Public Health notes that produce picked at peak maturity generally begins losing nutrients soon after harvest and that the journey from farm to supermarket can take one to two weeks before additional household storage time is added. [12] Harvard Health similarly notes that prolonged storage and exposure to light degrade vitamin C. [13]
Peer-reviewed evidence provides a more specific example. A 2017 Food Chemistry study of spinach and other salad leaves found losses of 35-86% of ascorbate over ten days of storage at 4°C, with substantial variation between species. [14] The critical point is not that imported vegetables are nutritionally inadequate; it is that Harvest-to-Consumption Time is a scientifically relevant variable for highly perishable leafy greens and should be measured alongside yield.
This changes the system boundary. A central plant factory may perform well at the farm gate but still require packing, cooling, transport, distribution and retail holding after harvest. A distributed system can shorten or remove several of those stages by producing close to the point of consumption. Its own electricity use must still be counted, so the correct comparison is total resource use and total delivered cost from cultivation to consumption. Nutritional quality and not output quantity should be an important policy goal for a developed economy like Singapore.
5. Why price-per-kilogram and farm-gate yield can give the wrong policy signal Technology screening often gravitates towards simple indicators such as kg/m², selling price/kg or farm-gate cost/kg. These are useful, but they are insufficient for public food-resilience decisions.
• Fresh-weight yield can be misleading: Water-rich fruiting crops can generate very high kg output, while a kilogram does not measure nutritional contribution.
• Farm-gate energy is only part of the system: Cold storage, transport, packaging, post-harvest losses and HVAC should be inside the comparison boundary where they are required.
• Crop cycle matters: Short-cycle leafy greens can turn production space rapidly; fruiting crops can deliver repeated harvests over a longer cycle. Both advantages need to be annualised.
• The energy source matters: A greenhouse can exploit sunlight; a fully enclosed plant factory must create photons electrically. This architectural difference often dominates energy economics.
• Resilience is not identical to lowest cost: Distributed capacity sacrifices some economies of scale but reduces concentration risk and places production close to demand. Singapore’s push for crisis-ready food security needs a reality check. We cannot afford to fall into a trap of diminishing returns, blindly throwing everything we have into local production just to harvest an ounce of gold. Recent research explicitly supports this broader approach: a 2025 study argues that vertical-farming viability should be benchmarked beyond yield by integrating energy, water, cost and carbon, while a 2024 meta-analysis found that controlled-environment agriculture generally raises yield and reduces water use but also increases energy use and global-warming impact relative to conventional systems. [3][15]
6. Recommended evaluation framework for Singapore
For local fibre resilience, Singapore can consider evaluating candidate production models using a common delivered-system scorecard rather than selecting a technology principally because a crop grows quickly or achieves a high farm-gate yield.
7. Policy recommendations
Do not make fully enclosed, artificial-light vertical farming the default architecture for Singapore's vegetable strategy. It is technically capable, but energy intensity remains a structural constraint and varies substantially by crop.
Prioritise centralised protected cultivation for selected fruiting vegetables where natural light can carry most of the photosynthetic load. Singapore should validate greenhouse designs specifically for tropical heat, humidity and disease pressure rather than importing temperate assumptions.
Treat highly perishable leafy greens as a distinct proximity problem. Evaluate distributed urban production as a complementary pathway in homes, offices, schools, hospitals, hotels and public buildings, supported by central propagation, food-safety protocols and technical management.
Use imports and storage where biology allows it. For storable crops, resilience may be achieved more cheaply through diversified sourcing and storage capacity than through local high-tech cultivation.
Measure delivered nutrition and resilience, not only farm-gate kilograms. SFA's pilot framework should compare total-system energy, cost, waste, freshness, nutrition and concentration risk across competing production architectures.
The land trade-off must also be recognised. Fruiting vegetables generally require larger, single-layer growing areas and therefore cannot achieve the extreme footprint efficiency possible with multi-tier leafy-green systems. For land-scarce Singapore, the objective should therefore not be to minimise either land or energy in isolation, but to optimise both across the food-production system. Selected fruiting vegetables may justify scarce agricultural land where protected cultivation can achieve high annual output using predominantly natural sunlight. Highly perishable leafy vegetables, meanwhile, offer an alternative route to land efficiency:
distributing production within existing buildings and urban spaces rather than requiring additional dedicated farmland.
Recommended policy proposition
Store what stores well. Use protected, solar-assisted cultivation for fruiting vegetables where it is locally resource-efficient. Grow highly perishable leafy greens as close as practical to where they are consumed. The pilot proposed to SFA should test whether this crop-specific architecture delivers better food resilience per unit of land, energy, capital and post-harvest infrastructure.
References
[1] Singapore Food Agency. “Our Singapore Food Story 2.” 2026. https://www.sfa.gov.sg/about-us/our
singapore-food-story-2
[2] Singapore Food Agency. Singapore Food Statistics 2025. 2026, p. 10.
https://www.sfa.gov.sg/docs/default-source/publication/sg-food-statistics/sgfs-2025
publication_060526-fa.pdf
[3] “Beyond yield: integrating energy, water, cost, and carbon to benchmark indoor vertical farming viability.” Energy Conversion and Management, 2025.
[4] “Specific aspects of leafy greens biology and cultivation technology in controlled environment.” Advances in Botanical Research, 2025.
[5] FAO. Basic Harvest and Post-Harvest Handling Considerations for Fresh Fruits and Vegetables.
https://www.fao.org/4/y4358e/y4358e05.htm
[6] FAO. Handling and Storage Methods for Fresh Roots and Tubers. https://www.fao.org/4/x5415e/x5415e04.htm
[7] de Gelder, A., et al. “A concept for reduced energy demand of greenhouses: the next generation
greenhouse cultivation in the Netherlands.” Wageningen UR / Greensys, 2012.
[8] Wageningen Plant Research. Komkommer in de winterlichtkas: Teeltresultaten 2017-2019. 2022. DOI: 10.18174/566962.
[9] “Protected Cultivation with Drip Fertigation Is a Feasible Option for Growing High-Value Vegetables in Samoa: A Case Study.” Sustainability 17 (2025): 7208.
[10] “Unveiling the Productivity Potential of Vertical Farming Through Diverse Crop Cultivation: An Empirical Analysis.” Agronomy 16 (2026): 1700.
[11] “Vertical farming economics: crop performance targets for cost-competitive vertical farming.” Frontiers in Sustainable Food Systems, 2026.
[12] Harvard T.H. Chan School of Public Health, The Nutrition Source. “Common Questions About Fruits and Vegetables.” https://nutritionsource.hsph.harvard.edu/common-questions-fruits-vegetables/
[13] Harvard Health Publishing. “Myths and truths about vitamin C.” https://www.health.harvard.edu/diet and-nutrition/myths-and-truths-about-vitamin-c
[14] Dewhirst, R. A., et al. “Novel insights into ascorbate retention and degradation during the washing and post-harvest storage of spinach and other salad leaves.” Food Chemistry 233 (2017): 237-246. DOI:
10.1016/j.foodchem.2017.04.082.
[15] Verteramo Chiu, L. J., et al. “A meta-analysis of yields and environmental performance of controlled environment production systems for tomatoes, lettuce and strawberries.” Journal of Cleaner Production (2024): 143142. DOI: 10.1016/j.jclepro.2024.143142.