Emerging Food Safety Risks to Watch in 2026
The hazards moving into 2026 do not arrive on a fixed schedule. Climate volatility, price-driven fraud, shifting residue limits and disrupted supply chains are reshaping where risk concentrates — and periodic reviews tend to catch them late. Here is what quality and food-safety teams should be watching, and why continuous monitoring is now the baseline.
The emerging food safety risks to watch in 2026 share a common trait: they move faster than the annual review cycles most teams still rely on. A hazard that was a southern-European concern can migrate north within a few growing seasons; a commodity that was cheap and low-risk can become a fraud target within a single price spike. The through-line is volatility, and it rewards teams that watch signals continuously rather than reassessing risk once a year.
This article walks through five risk themes for 2026 — climate-driven contamination, fraud in high-priced commodities, pesticide and residue shifts, supply-chain and geopolitical disruption, and why static annual reviews miss all of them — then closes with a practical view of how to stay ahead. Every figure is drawn from a named, verifiable source, and several findings carry important nuance that headline numbers tend to strip out.
| Signal | What drives it | Where it shows first | Buyer-side control |
|---|---|---|---|
| Climate-driven contamination | Temperature and moisture shifting mycotoxin geography | Harvest-year data and biomonitoring, not yet in specifications | Re-map origin risk annually rather than trusting a fixed approved-origin list |
| Fraud in high-priced commodities | Price spikes and harvest shortfalls widening the incentive gap | Price and yield signals months before alert volume rises | Tie verification intensity to price movement, not to a fixed audit calendar |
| Moving residue limits | MRL revisions taking effect between purchase and arrival | Official journals, ahead of any test result | Monitor the limit as a live value; a compliant origin can become non-compliant in transit |
| Supply-chain disruption | Substitution toward less-vetted suppliers under shortage | Sourcing changes inside your own approved-supplier records | Treat a new supplier under time pressure as the highest-risk case, not the routine one |
| The annual review cycle itself | Assessment cadence slower than the rate at which new signals arrive | In the gap between two reviews, never inside a single one | Keep the annual reassessment, but let continuous monitoring trigger out-of-cycle reviews |
How is climate change reshaping food safety hazard geography?
Climate-driven contamination is the shift in where and how strongly biological and chemical hazards appear in food as temperature, humidity and extreme-weather patterns change. It is no longer a modelling hypothesis. The European Environment Agency's 2025 briefing states plainly that a warmer, more humid climate across European regions is promoting a higher prevalence of mycotoxins in staple crops, and identifies developed countries with moderate climates as facing the highest risk of climate-change-induced mycotoxin contamination [1]. For a buyer, the practical consequence is that an approved-origin list built on historical incidence starts to age: an origin that was low-risk for aflatoxin or DON five harvests ago is not automatically low-risk today, and the change arrives one growing season at a time rather than in a regulation. Two hazards carry most of that movement in Europe — aflatoxin B1 in maize and deoxynivalenol in wheat — and both are driven by weather in the field rather than by anything a supplier controls inside the plant.
Which mycotoxins are moving north in Europe?
Aflatoxin in maize and deoxynivalenol in wheat are the two mycotoxins whose European geography is moving most visibly, though not in a simple straight line north. Under a +2C scenario, the EEA finds aflatoxin contamination in maize rising particularly in southern Europe, including Spain, Italy and the Balkans. Under a more extreme +5C scenario, southern risk can actually decline as heat suppresses the fungus, while risk migrates into more northern European countries [1]. The direction is scenario-dependent, which is exactly why teams should treat northward migration as a projection with nuance rather than a fixed trajectory. The species mix is shifting too: the EEA reports that the types of Fusarium fungi on European wheat are constantly changing, with particular concern about the increasing presence of F. graminearum — a strong producer of deoxynivalenol (DON, or vomitoxin) — in central and northern Europe [1].
How much does a drought year change mycotoxin contamination?
A drought year can multiply mycotoxin contamination several times over within a single harvest. A 2025 review in Toxins documented that during drought years in Serbia, aflatoxin-contaminated maize samples reached 72% in 2012 and 37% in 2015, against 0% in the wetter 2014 season [3]. Volatility, not just gradual warming, is the acute driver. The exposure that follows is measurable in people, not only in crops: human biomonitoring under HBM4EU, cited in the EEA work, found that DON was detected in 14% of adult participants in studied European countries at levels that may affect health, with populations in Poland and, to a lesser extent, Luxembourg, France and Portugal very likely exposed at risky levels [1][2]. One bad season therefore shows up twice — in the commodity, and in the population that eats it.
What do climate models project for mycotoxins by 2050?
Peer-reviewed modelling projects DON in European wheat reaching its highest contamination class more frequently by 2050. A 2026 machine-learning study from Wageningen University, using CMIP6 climate scenarios, concentrates that increase in coastal regions of northwest Europe. The authors themselves flag limited sensitivity for elevated-contamination cases [4] — so it belongs in a watchlist as a peer-reviewed projection, not as settled fact. The uncertainty is worth stating precisely, because it determines how the finding should be used: a model that is accurate overall but weaker on the extreme class is better at telling you which regions to watch than at telling you which lots will fail. For a buyer, that means putting northwest-European wheat origins on a DON watch-list now rather than reclassifying them today, and revisiting the list as each harvest adds real measurements to the projection.
Why is food fraud rising in high-priced commodities?
Food fraud is the intentional substitution, dilution, mislabelling or adulteration of food for economic gain, and its incentive scales directly with price. When a commodity's price spikes, the margin available to a fraudster who cuts, extends or misrepresents it grows with it. The FoodAkai Global Food Fraud Index projected sharp 2025 increases for exactly the commodities under the steepest price and supply pressure: +66% for cocoa, +25% for herbs and spices, +20% for fats and oils, +74% for fish and seafood and +358% for nuts, seeds and nut products [5]. These are index projections rather than confirmed year-end counts, but the pattern is consistent — reported fraud cases have risen roughly tenfold in four years, at an estimated $40bn annual global cost [5]. Understanding how iComplai predicts food fraud risk across these commodity categories is central to catching the shift early.
Why is olive oil the leading food fraud target?
Olive oil is the flagship case because record-low harvests met record-high prices in a product that is hard to authenticate by sight or taste. Enforcement seizures continued into 2025: Portuguese officials seized over 16,000 litres of cooking oil falsely labelled as olive oil, plus 82,000 counterfeit labels, while Italian police dismantled a ring blending low-grade oils with chemicals and selling them as extra virgin [6]. The pattern tracks the price curve directly — enforcement activity rose alongside record prices and has begun to ease as prices retreat. The practical read for a buyer is that verification intensity on an oil should follow its price chart rather than the audit calendar.
How much imported honey is adulterated with sugar syrup?
Close to half of the honey consignments the EU sampled at its borders were suspected of sugar-syrup adulteration. In the EU "From the Hives" coordinated action, 147 of 320 imported honey samples (46%) were suspected of adulteration with sugar syrups, with suspicion rates reaching 74% for consignments from China, 93% from Turkey and 100% from the UK [7]. Because sugar-syrup adulteration is hard for routine labs to detect, the true rate is likely higher. Cocoa follows the same price-driven logic under record 2024–25 prices.
Which spices are most often adulterated?
High-value spices carry the highest baseline adulteration rates of any food category. A 2025 systematic review estimates 20–30% of commercial saffron is adulterated globally, ranging from about 3.5% in regulated EU markets to roughly 60% in India, with adulterants including hazardous synthetic dyes such as Sudan compounds and auramine-O [8]. That bridges fraud into a genuine safety hazard rather than a purely economic one. Coffee, by contrast, is a real structural substitution risk — cheap robusta passed off as arabica, which can be flagged through NMR profiling — but the same 2025 index projected coffee fraud down, so it is a standing watch-item rather than a 2025 surge.
- Cocoa: record prices raise the incentive to substitute cocoa fat or misrepresent product [5]
- Olive oil: counterfeit labelling and chemical blending; enforcement tracks the price curve [6]
- Honey: ~46% of sampled EU imports suspected of sugar-syrup adulteration, hard to detect [7]
- Saffron and paprika: 20–30% global saffron adulteration, some adulterants hazardous [8]
- Coffee: ongoing robusta-for-arabica substitution risk, though not a 2025 spike
Are EU pesticide residue limits changing in 2026?
Maximum residue levels (MRLs) are the legal ceilings for pesticide residues in food, and they are actively changing in 2026 rather than sitting fixed. Commission Regulation (EU) 2026/140, dated 22 January 2026 and applying from 12 February 2026, amends MRLs for six active substances — acequinocyl, chlormequat, metalaxyl-M, pyraclostrobin, sulfoxaflor and trifloxystrobin — under Regulation (EC) No 396/2005 [9]. A supplier compliant one month can find a threshold has shifted the next. That matters more than it sounds, because an MRL is not a property of the supplier: it is a legal value attached to a substance-and-crop pair that the EU revises on its own timetable, and where no specific limit is set, Regulation (EC) No 396/2005 applies a default of 0.01 mg/kg. The specification agreed at order and the limit in force at arrival can therefore be two different numbers, with nothing having changed at the farm.
Why do acetamiprid MRLs keep moving?
Acetamiprid limits keep moving because EFSA reassessed the substance's toxicology. EFSA lowered its acceptable daily intake and acute reference dose from 0.025 to 0.005 mg/kg body weight per day, and MRLs were subsequently revised across 2025 — Commission Regulation (EU) 2025/1212 amended acetamiprid MRLs with effect from 20 August 2025 (revising several levels, some upward) [10]. Acetamiprid remains approved.
How often do EU checks find unauthorised pesticides in imports?
Residues of pesticides no longer authorised in the EU keep turning up, most often in imports. EFSA's 2023 EU pesticide residue report, published in 2025, analysed 132,793 samples and found 96.3% within legal limits; unauthorised substances including chlorpyrifos and imidacloprid were among those exceeding limits in specific imported commodities, and non-compliant consignments concentrated in a small set of third countries [11]. A single unapproved substance can trigger border rejections: ethylene oxide, not approved in the EU, was screened for in 3,651 samples, exceeded MRLs in 40 and drove 24 non-compliant findings, 13 from India and 4 from Türkiye [12].
Why do single-substance MRL checks miss the real exposure?
Checking one MRL at a time misses the residues that arrive together. Advocacy analysis of EFSA data reports that 41.6% of EU fruit and vegetables contain pesticide residues and 25.5% contain multiple residues [13]. Combined "cocktail" exposures are not routinely assessed against single-substance limits, which is why residue risk warrants continuous pesticide risk prediction rather than one-off, single-compound checks. For a testing programme the more useful question is therefore not only whether a lot passes, but which residues this commodity and this origin usually carry together.
How does supply-chain disruption create food safety risk?
Geopolitical and trade disruption becomes a food-safety issue when it forces buyers off established, vetted supply chains and onto unfamiliar ones under time and cost pressure. The 25% US tariff on products from Mexico and Canada implemented in March 2025 pushed food companies to shift purchasing abruptly to domestic and alternative suppliers, creating demand surges where production pressures rise faster than food-safety infrastructure. Industry guidance now urges enhanced documentation verification, supplier approval and lab testing to confirm authenticity and safety of rapidly onboarded product [14]. The risk is not the tariff itself but the speed of the response: a supplier onboarded in days carries none of the audit history, none of the residue record and none of the origin knowledge that the supplier it replaced had accumulated.
How do tariffs increase economically motivated adulteration?
The mechanism linking trade pressure to safety is economically motivated adulteration. When legitimate import channels become prohibitively expensive, suppliers face a stronger incentive to substitute lower-value alternatives without disclosure — conventional produce sold as organic, or extra-virgin olive oil diluted with cheaper oils under premium labelling. Tariff pressure can also spawn black-market diversion, where product moves without temperature verification and with falsified traceability, eliminating the ability to conduct effective recalls [15].
Do EU fraud notification counts confirm the trend?
This is not hypothetical, and it shows up in authority signals. The same drivers — price spikes, opaque and fragmented supply chains, climate and conflict shocks — are named as the core reasons behind the 2025 rise in food fraud, with the estimated $40bn annual global cost attributed to Authenticate [16]. EU enforcement data reflects it directly: in 2024, 1,959 RASFF notifications were flagged as potentially resulting from intentional behaviour, a 21% increase over 2023, while total Alert and Cooperation Network notifications reached 9,460 (up 8%) and RASFF notifications 5,250 (up 12%) [17].
Why does supply-chain complexity multiply fraud risk?
Supply-chain complexity is itself a fraud multiplier. As tariffs, shortages and trade realignment lengthen and fragment sourcing networks, visibility drops and every additional tier becomes another point at which product can be introduced or mislabelled before it reaches the buyer [18]. That is precisely the kind of shift continuous horizon scanning is built to surface early: a change two tiers upstream rarely announces itself in a certificate, but it does show up in where authority notifications start appearing, and in which origins begin repeating on the same hazard. The practical test is whether you can name the tier-two supplier for your highest-value raw materials. Where you cannot, the approved-supplier record describes the party you contract with rather than the party who handles the product, and the distance between those two is where undisclosed substitution becomes possible without anyone breaching a contract.
Why do annual reviews miss these emerging food safety risks?
Annual reviews miss emerging risks because emerging risks, by definition, do not respect an annual cadence. EFSA defines an emerging risk as one arising from a newly identified hazard with potential for significant exposure, or from an unexpected new or increased exposure or susceptibility to a known hazard — and it stresses that identifying such risks requires operational processes for continuous monitoring and structured assessment of new signals so policymakers can anticipate and prevent, rather than react [19]. The mismatch is structural rather than a question of effort. An annual review is a sampling method with a sampling interval of twelve months, applied to a process that produces new evidence daily, so anything that appears and is acted on between two reviews is invisible to both of them. It is also backward-looking by construction: it summarises the year that has closed, at the moment next year's sourcing decisions are already being made.
How many food safety notifications arrive each day?
The volume of incoming signals is measurable: iComplai recorded 21,166 authority notifications in 2025, about fifty-eight a day — 13,575 US import refusals, 5,328 RASFF notifications and roughly 2,263 from other authorities. A review held once a year is a single reading of a feed that moved every day for the preceding twelve months, and what it reads is the state of that feed on one arbitrary date rather than everything that crossed it. Volume alone is not the argument; the shape of it is. Notifications do not arrive evenly across commodities, and a single origin or hazard can move from absent to repeated within a few weeks — exactly the pattern a twelve-month interval cannot resolve. Most of them will never concern a given buyer, which is the other half of the problem: the signal that matters is the small subset touching your own materials and origins.
How many emerging risks does EFSA confirm in a year?
EFSA confirms only a handful of emerging risks each year, out of a much larger pipeline of candidate signals. In 2024 alone, EFSA collected 65 emerging issues, further characterised 38, and confirmed 7 as actual emerging risks, most needing follow-up in contaminants and biological hazards [20]. A static review captures a fraction of that pipeline and, crucially, captures it late. EFSA also names climate change explicitly as a driver, noting that rising temperatures and humidity likely contributed to aflatoxins appearing in southern Europe in the early 2000s and spreading northward since, alongside the northward spread of Vibrio and ciguatera [21]. It has even commissioned predictive mapping of future aflatoxin B1 in EU cereals precisely because warming can introduce contamination in regions previously considered lower-risk [22].
Does the early-warning infrastructure already exist?
The infrastructure for early warning already exists, and has for decades. The EU's Rapid Alert System for Food and Feed (RASFF), established in 1979 and operating around the clock under Article 50 of Regulation (EC) 178/2002, exists to exchange information between authorities fast enough for immediate action [23]. National authorities, official journals and market data add further continuous feeds on top of it. The gap is not data availability — it is whether a team is watching that data continuously or only at review time. That is a tooling question, not a data question. The same feeds that support a border authority's decision on a Tuesday are available to a quality team on that Tuesday, but only if something is reading them then rather than in December. This is the article's synthesis, supported by EFSA's emphasis on continuous anticipation rather than a direct EFSA quotation.
What does iComplai's own notification data show about these risks?
Two of these signals are visible in iComplai's own analysis rather than only in the literature. In the hazelnut case, the platform's models began flagging elevated fraud-related signals in February 2025, coinciding with reports of poor harvest yields in Turkey, the largest producing origin — the price-and-shortfall pattern in signal 2, showing up months before alert volume did. And in our analysis of honey notifications, iComplai data identified 53 distinct substances across honey-related records since 2010, which is the practical reason a single-substance test plan under-detects: the residue you screen for is rarely the only one present. The hazelnut work is set out in full in the hazelnut early-signal case.
An emerging risk is only useful while it is still emerging
By the time a hazard reaches a recall notice it is no longer a forecast. iComplai tracks the signals ahead of that point across notifications, enforcement and world media, and routes the ones that touch your categories to the people who can act on them.
See how the model worksHow can food safety teams stay ahead of emerging risks in 2026?
Staying ahead of emerging food safety risks in 2026 means replacing the annual review with continuous horizon scanning — continuous monitoring of regulatory, enforcement, recall, market and environmental signals — as the baseline, not an occasional exercise. Across all five themes above, risk is moving faster than periodic assessment can track. That is the shift from reactive to predictive food safety: catching a residue-limit change, a fraud pattern in a spiking commodity, or a supplier shift under tariff pressure while there is still time to act. In practice it pairs horizon scanning with material-level raw-material and supplier risk assessment, so a new signal lands against a named commodity and origin rather than in a general newsfeed.
What does continuous horizon scanning look like in practice?
In practice it means a system that reads the authority and market feeds every day and reports only what touches your own materials, origins and suppliers. This is what iComplai is built to do. The platform continuously monitors global signals to surface emerging raw-material and supplier risks before they escalate, drawing on authority sources including RASFF, FDA and FSA alongside 50+ regulatory and news sources, across more than 7.5 million records, updated daily. Rather than compiling one annual view, it watches the signals covered throughout this article — shifting MRLs, fraud enforcement, climate-linked contamination and supply-chain disruption — as they emerge.
If your 2026 risk assessment still runs on an annual cycle, the practical next step is to see continuous horizon scanning applied to your own commodities and suppliers. You can request a demo to walk through how iComplai turns these signals into early, actionable warnings for your team.
Frequently asked questions
What are the biggest emerging food safety risks to watch in 2026?
Five themes dominate the 2026 outlook. Climate-driven contamination is shifting where mycotoxins appear, with aflatoxin risk in maize rising in southern Europe under a +2C scenario and DON-producing Fusarium species spreading in central and northern Europe. Fraud is concentrating in high-priced commodities such as cocoa, olive oil, honey and saffron, because the incentive to adulterate scales with price. EU pesticide residue limits are being revised mid-year, so an origin that met the specification at order can miss it on arrival. Supply-chain and tariff disruption pushes buyers onto less-vetted suppliers under time pressure. The fifth risk is the review cycle itself: all four move faster than an annual assessment can track.
Which mycotoxins are increasing in Europe because of climate change?
Aflatoxin B1 in maize and deoxynivalenol (DON) in wheat are the two mycotoxins whose European geography is moving most clearly. The European Environment Agency’s 2025 briefing finds aflatoxin contamination in maize rising particularly in southern Europe — Spain, Italy and the Balkans — under a +2C scenario, while under a more extreme +5C scenario southern risk can decline as heat suppresses the fungus and risk migrates north. The EEA also reports that Fusarium species on European wheat are changing, with growing concern about F. graminearum, a strong DON producer, in central and northern Europe. Exposure is already measurable: HBM4EU biomonitoring detected DON in 14% of adult participants at levels that may affect health. Drought years amplify the effect sharply rather than gradually.
Which foods are most at risk of food fraud in 2026?
High-priced and hard-to-authenticate commodities carry the highest fraud risk. A 2025 fraud index projected steep increases for cocoa, nuts and seeds, fish and seafood, herbs and spices, and fats and oils, and 2025 enforcement produced confirmed olive-oil cases in Portugal and Italy, including more than 16,000 litres of mislabelled cooking oil and 82,000 counterfeit labels. Cocoa and honey remain high-incentive categories on price and past sampling evidence rather than on 2025 enforcement outcomes: the EU’s 2021–22 From the Hives exercise found 46% of sampled honey consignments suspected of sugar-syrup adulteration. Saffron has the highest baseline, with 20–30% of commercial product estimated adulterated globally and some adulterants presenting a genuine safety hazard.
What happens if an EU pesticide MRL changes after a supplier is approved?
An MRL change applies from the date set in the amending regulation, so a supplier and a specification agreed under the old limit can fall out of step within weeks. Commission Regulation (EU) 2026/140, dated 22 January 2026 and applying from 12 February 2026, amended MRLs for six active substances: acequinocyl, chlormequat, metalaxyl-M, pyraclostrobin, sulfoxaflor and trifloxystrobin. Acetamiprid limits were revised across 2025 after EFSA cut its acceptable daily intake and acute reference dose from 0.025 to 0.005 mg/kg body weight per day. The practical control is to monitor the limit as a live value rather than a fixed line in a specification.
How often should a food safety risk assessment be updated?
Often enough to keep pace with the signals it is meant to track, which for most raw materials means continuously rather than once a year. iComplai recorded 21,166 authority notifications in 2025 — about fifty-eight a day, made up of 13,575 US import refusals, 5,328 RASFF notifications and roughly 2,263 from other authorities. An annual review is a single reading of a feed that moved every day for the preceding twelve months. EFSA takes the same position for emerging risks, stressing that identification requires operational processes for continuous monitoring and structured assessment of new signals; in 2024 it collected 65 emerging issues, characterised 38 and confirmed 7 as actual emerging risks. A workable compromise is a fixed annual reassessment plus continuous monitoring that can trigger an out-of-cycle review.
What is horizon scanning in food safety?
Horizon scanning in food safety is the continuous monitoring of regulatory, enforcement, recall, market and environmental signals to identify hazards before they reach a specification, a test result or a recall. It differs from routine compliance checking in what it looks at: not whether a shipment meets today’s limit, but whether the limit, the origin risk or the fraud incentive is about to change. The EU’s Rapid Alert System for Food and Feed, established in 1979 and operating around the clock under Article 50 of Regulation (EC) 178/2002, is one of its primary feeds, alongside national authority notifications, official journals, and price and harvest data. EFSA runs a formal horizon-scanning programme on the same principle.
See the signals before they become findings
iComplai monitors authority notifications, recalls and adulteration signals for your own materials and suppliers, daily.
Talk to iComplaiWhere these claims come from
- European Environment Agency (EEA). "Mycotoxin exposure in a changing European climate" (Briefing 02/2025). eea.europa.eu
- Food Safety News. "EU agency says rising temperatures increase mycotoxin risk" (reporting HBM4EU biomonitoring on DON exposure). foodsafetynews.com
- Bereziartua et al. "Pre-Harvest Aflatoxin Contamination in Crops and Climate Change Factors: A European Overview," Toxins (Basel), July 2025. pmc.ncbi.nlm.nih.gov
- Wageningen-based modelling of climate-driven mycotoxin risk in European wheat, as summarised in the EEA briefing cited above — the primary finding is the EEA/HBM4EU material rather than the trade write-up.
- FoodNavigator. "Global food fraud surges in 2025" (citing FoodAkai, Digicomply, Authenticate). foodnavigator.com
- FoodNavigator. "Olive oil a major target for food fraud" / Food Fraud Advisors. foodnavigator.com
- eucrim. "From the Hives: Results of the EU Action Against Honey Adulteration" (DG SANTE / OLAF / JRC). eucrim.eu
- Critical Reviews in Food Science and Nutrition (2025). "Combating saffron fraud: a systematic review." pubmed.ncbi.nlm.nih.gov
- EUR-Lex. Commission Regulation (EU) 2026/140. eur-lex.europa.eu
- EUR-Lex. Commission Regulation (EU) 2025/1212 (acetamiprid MRLs). eur-lex.europa.eu
- EFSA Journal. "The 2023 European Union report on pesticide residues in food." efsa.onlinelibrary.wiley.com
- Food Safety News. "EFSA report details pesticide residue situation." foodsafetynews.com
- PAN Europe. "EFSA annual pesticide residue report" press release (May 2026). pan-europe.info
- New Food Magazine. "The cascading food safety impacts of tariffs on the food industry." newfoodmagazine.com
- Food Processing. "Tariffs on Food: A Cascading Threat to Food Safety, Quality and Legitimacy." foodprocessing.com
- FoodNavigator. "Global food fraud surges in 2025" (drivers and $40bn cost estimate). foodnavigator.com
- FoodTimes. "EU Alert and Cooperation Network: 2024 Annual Report Analysis." foodtimes.eu
- AIB International. "How Supply Chain Complexity Is Impacting Food Fraud." blog.aibinternational.com
- EFSA. "Emerging risks and horizon scanning" (topic page). efsa.europa.eu
- EFSA. "2024 Annual Report on Emerging Risks and Horizon Scanning Activities." efsa.onlinelibrary.wiley.com
- EFSA. "Climate change and food safety" (topic page). efsa.europa.eu
- EFSA. "EFSA launches project to predict the effect of climate change on aflatoxin B1 in cereals." efsa.europa.eu
- European Commission. "Rapid Alert System for Food and Feed (RASFF)." food.ec.europa.eu