Editorial: Crop suitability and food security: the evolution of new technologies

Introduction

Global food security stands at the intersection of an expanding population, a changing climate, and increasingly fragile natural resource bases. In 2024, 638 to 720 million people were hungry according to the latest joint assessment by FAO, IFAD, UNICEF, WFP and WHO, and it is estimated that nearly 28% of humanity, 2.3 billion people, suffered from moderate or severe food insecurity (FAO et al., 2025). These burdens fall unevenly across regions: the prevalence of undernourishment in Africa has surpassed 20%, more than double the global average, and hunger has continued to rise across Western Asia, even as parts of Southern and South-eastern Asia and South America have registered gains. Projections suggest that, absent a change in trajectory, roughly 512 million people could remain chronically undernourished in 2030, nearly 60% of them in Africa. Against this backdrop, matching crops to the land, water, and climatic conditions in which they are actually grown is not a peripheral technical exercise; it is a first-order determinant of whether the world can expand food production without exhausting the resource base, and of whether the most vulnerable regions can adapt as rainfall patterns and growing seasons continue to shift. The Food and Agriculture Organization estimates that meeting future food demand while safeguarding ecosystems will require not only more land and water, but smarter, evidence-based decisions about where, how, and what to cultivate.

This Research Topic, “Crop suitability and food security: the evolution of new technologies,” was launched to capture this transition. It brings together six contributions that, collectively, illustrate how new technologies are reshaping the way researchers and policymakers diagnose agro-ecological risk, evaluate land and soil potential, and design interventions that protect livelihoods under climatic and economic stress. They offer a timely snapshot of where the crop suitability and food security literature now stands, and where its open challenges lie. Geographically, the Research Topic reflects the uneven but instructive distribution of crop suitability research itself. Two contributions are anchored in the drylands of the Middle East and North Africa, in Najran, Saudi Arabia, and the Nile Delta of Egypt, regions where water scarcity and soil salinity make suitability assessment a precondition for any expansion of cultivated area. Two further studies draw on Chinese provincial and national data, illustrating how large, centrally coordinated agricultural systems translate suitability and vulnerability science into infrastructure investment and land-management policy. A fifth study spans Sub-Saharan Africa, the region in which, as the statistics above indicate, food insecurity is most acute and where the economic dimension of suitability cannot be separated from the ecological one; and the sixth synthesizes decades of evidence on stress-tolerant crop adoption drawn largely from smallholder settings in Asia, including Vietnam, and other tropical regions. Although the specific climatic, institutional, and market conditions vary across these settings, the underlying lessons travel well: the value of low-cost, remotely sensed indicators where ground survey capacity is limited; the need to couple ecological suitability with economic and institutional variables; and the recognition that infrastructure and policy can materially alter a region's exposure to climate risk. Researchers, practitioners, and policymakers working in other data-scarce or climate-exposed settings, from the Sahel to South Asia to Latin America, should find in these six studies a transferable toolkit rather than a set of purely local findings.

Diagnosing land and soil potential with geospatial technologies

Two contributions to this Research Topic advance the geospatial toolkit used to characterize land and soil suitability in resource-constrained drylands. Alqurashi et al. develop a GIS-based geometric approach to dryland soil quality modeling in Najran, Saudi Arabia, demonstrating how freely available spatial layers can be combined into a defensible quality index even where dense ground-truthing is costly. In a complementary contribution from arid Egypt, Ganzour et al. integrate Google Earth Engine data with field observations and laboratory analysis into a suitability-resources quality index, coupling crop water requirements and fertilizer recommendations with geospatial suitability mapping to support irrigation and nutrient management decisions at the field scale. The two studies show that credible suitability diagnostics do not require costly ground campaigns: by anchoring index construction in openly available satellite, elevation, and reanalysis products, both approaches offer a template that can be replicated in other dryland regions where survey infrastructure and laboratory capacity are limited.

Machine learning and data fusion for risk and economic assessment

A second cluster of papers applies machine learning to problems of risk prediction and economic viability. Zhou and Sun propose a Multi-Source Data Fusion-Based Risk Assessment framework built around an Agricultural Risk Fusion Network and an Adaptive Fusion Strategy, which combine graph-based representation learning with temporal modeling to integrate climate, soil, remote-sensing, and crop-yield data into interpretable, spatially and temporally resolved ecological risk scores. Evaluated against several state-of-the-art baselines across four agricultural datasets, the framework illustrates how deep learning architectures can move beyond single-indicator risk assessment toward system-level diagnostics that explicitly model interdependencies among climatic, edaphic, and anthropogenic stressors. Benchmarked against several established baselines across four independent agricultural datasets, the proposed framework is reported to outperform single-indicator and simpler fusion approaches, underscoring the practical value of modeling spatial and temporal dependencies jointly rather than treating risk factors in isolation.

Kárpáti et al. adopt a complementary economics-oriented view, using Random Forest and multilayer perceptron models to predict a suitability proxy based on yield and price across Sub-Saharan Africa, using two decades of FAO data. This study shows that suitability for food security purposes cannot be divorced from the market and livelihood conditions under which farmers can maintain a particular crop choice over time, by recasting crop suitability not just in terms of ecological fit, but in terms of economic viability, and by combining prediction with profitability-based clustering across representative countries. Their clustering based on profitability also shows that economically viable crop choices are not necessarily the same as ecologically optimal crop choices, a difference with direct relevance to the targeting of national agricultural investment and extension strategies across Sub-Saharan Africa.

Infrastructure, policy, and climate vulnerability

Land and soil potential are necessary but not sufficient conditions for food security; the resilience of production systems also depends on built infrastructure and policy design. Li and Zhang examine how high-standard farmland construction shapes the climate disaster vulnerability of grain production using a 2003–2020 panel of 30 Chinese provinces, applying an improved vulnerability framework to identify both the direct mitigation effects of farmland infrastructure investment and its spatial spillovers into neighboring regions. This work situates infrastructure policy firmly within the crop suitability and food security agenda, showing that engineered improvements to land quality can durably reduce the climate sensitivity of staple grain production, with implications that transcend administrative boundaries. Their estimates of spatial spillovers suggest that the benefits of vulnerability reduction from high-standard farmland construction spill over to neighboring provinces, which conventional, single-region evaluations may systematically understate the true returns to such infrastructure investment.

Synthesizing decades of evidence on adoption and livelihoods

The Research Topic closes with a systematic synthesis that looks beyond any single technology to ask what the cumulative evidence base says about outcomes. Duong et al. meta-analyze 56 empirical studies (2000–2025) and extract 247 effect-size estimates to quantify the association between the adoption of stress-tolerant alternative crops—drought-, heat- and salt-tolerant varieties—and income, food security, resilience, and adaptive capacity. Their meta-regression shows that these relationships are robust, and more prominently observed for outcomes related to food security and resilience than short-term income gains, and that the magnitude of these relationships across climate stressors is moderated by institutional support, credit access, and market linkages. This contribution is a timely reminder that the ultimate test of any suitability technology is whether it leads to sustainable improvements in farmer livelihoods and that this translation is as much mediated by the institutional environment as by the underlying agronomy. This suggests a clear policy lever: significant livelihood benefits from adoption can be achieved not through technology alone but through increased credit access and market linkages in tandem with stress-tolerant crop promotion.

Looking ahead, the studies gathered here point to several open priorities for the field: extending data-fusion and graph-based risk models to regions and crops where labeled, high-resolution data remain scarce; testing whether economically defined suitability proxies remain stable under volatile global commodity prices; and building stronger evidence on how infrastructure and institutional interventions interact with, rather than substitute for, farmer-level adaptation strategies. As new sensing platforms, open data repositories, and machine learning architectures continue to mature, the central challenge for the crop suitability and food security research community will be to ensure that these technologies remain grounded in the livelihood realities that ultimately determine whether food security gains are realized and sustained.

Realizing this potential will depend as much on institutional design as on scientific advance. Suitability indices, risk-fusion models, and vulnerability frameworks of the kind presented in this Research Topic can only translate into food security gains if they are embedded in functioning science-policy interfaces that carry evidence from the research community into extension services, land-use planning, and agricultural investment decisions. This requires sustained interdisciplinary collaboration among soil scientists, agronomists, economists, and data scientists, together with genuine engagement of the farmers, extension agents, and local institutions who ultimately implement suitability-informed recommendations on the ground. Strengthening these channels for evidence-informed decision-making, rather than developing suitability tools in isolation from the policy process, will determine how quickly, and how equitably, the innovations described in this Research Topic are translated into resilient food security outcomes.

References

• 1

  FAO IFAD, UNICEF, WFP, and WHO. (2025). The State of Food Security and Nutrition in the World 2025 – Addressing High Food Price Inflation for Food Security and Nutrition.Rome: FAO. doi: 10.4060/cd6008en

  • CrossRef
  • Google Scholar
  • View reference in article