The European Union is currently in the process of regulating mycotoxins in food products. Mycotoxins are a group of toxins, which have in common that they can be produced by specific molds. Coffee is included in this process, together with a series of otherfood products, because of a mycotoxin called Ochratoxin A (0 TA).

What Are Molds and What is Ochratoxin A?

Molds can and do grow on any agricultural food product. They just need adequate levels of moisture, temperature and time. Different molds have different optimal levels for these three factors. In addition, molds do differ in their preference for different food products (so called substarate specificity). These factors together make that molds usually grow in very localized spots. These factors also make that usually each type of food raw material harbours its own specific group of molds. On green coffee, normally spores of about ten different types of mold can be found.

Presence of a mold in a given food does not per se mean that the food is unfit for consumption. Some molds are integral part of the production of a specific food, e.g. some cheeses. Also, parts of some molds are directly consumed, e.g. different fungi. However, a small number of molds can produce toxins, as a group called mycotoxins. As a consequence of the very localized spotwise growing of molds can produce toxins, and the fact that there mycotoxins are unevenly distributed over the agricultural commodity, sampling and analysis of these mycotoxins is a complicated exercise with limited accuracy. Well-known mycotoxins are Aflatoxin, Ochratoxin A and DON (Deoxynivalenone). The first one is known to occur on several types of nuts and the latter two on cereals. Besides on cereals, Ochratoxin A is also found to occur in wine, coffee, beer, cocoa, dried fruits (like raisins and currants), meat, spices and fruit juices.

What Harm Does Ochratoxin A Do?

The most sensitive toxic effect of OTA is its renal toxicity. This is known from the damage which OTA causes in the kidneys of pigs. At higher doses it was found to cause liver and kidney cancer in rodents. Whether or not OTA is a genotoxic carcinogen is still unclear. The currently available studies show conflicting results in this respect and the experts in the field still have deviating views on it.

OTA is also frequently linked to a kidney disease mainly occurring in the Balkan countries, so called Balkan Endemic Nephropathy (BEN). However, a clear link to this type of kidney disease is not established and other factors are likely to be involved. At doses in animals, which are considered widely beyond the levels of human consumption, OTA can cause neurotoxic, immunotoxic and teratogenic effects.

The International Agency on Research on Cancer classified OTA as a "possible human carcinogen." The EU Scientific Committee for Foods and the joint expert committee of WHO and FAO recommended tolerable daily and weekly intake levels, respectively. The mean total intakes from food in Europe do not exceed these recommended levels. Coffee contributes on average just a few percent of these recommended tolerable intake levels.

How Effective is the ICO/FAO Project for Mold Prevention in Coffee?

The ICO/FAO mold prevention program started in 1999 with a project in Uganda. The program is now funded mainly by the Common Fund for Commodities, the Dutch government, FAO's Technical Cooperation Program and the European Coffee sector. The first couple of years of the program were devoted to build up local expertise in a number of coffee producing countries around the globe.

Just a few of the remarkable findings are mentioned here. Cherries fallen on the soil show very frequently high OTA contamination levels. In general, OTA levels in the flesh of the cherry can be more than ten times higher than in the contained beans. These observations stress the importance of appropriate harvesting and drying practices and adequate separation of beans from cherries/husks plus disposal of the latter.

IMPROVE YOUR COFFEE QUALITY BY THE PREVENTION OF MOLD GROWTH

Good quality coffee receives bigger payments. In Order to protect your revenues from coffee, it is essential that you provide only the best quality commodity available. Use Good Agricultural Practices to Achieve Results: During harvest

The soil under the tree should be covered with a clean sheet of plastic during picking to avoid cherries getting contaminated by dirt or mixed up with moldy cherries ("gleanings") from previous harvests.

Cherries that have fallen to the ground are known to be susceptible to mold growth and therefore should not be used.

Process fresh cherries as quickly as possible. Avoid storage of cherries, especially ripe and over-ripe ones, as any period of storage (in a bag or in a pile) increases the likelihood of mold growth.

Do not dry on bare soil: Use trays or tarpaulins. Mold spores from previous lots are known to remain on the ground and this could result in clean cherries being contaminated during drying.

The layer of drying cherries should be not more than 4 cm thick after the first 2-3 days of drying.

Drying cherries must be regularly raked (5-10 times per day). Protect cherries during drying from rain and night dew. Avoid all re-wetting of partially dried and dried cherries, protecting them from moisture and rain.

During primary processing, place site processing plant in a dry area, not in a swamp. Dispose of pulp from wet processing away from clean dry coffee. Compost it before using it as mulch in the field.

Keep equipment and facilities clean separating residual partially processed material and accumulation of dust and discarded material.

Clean coffee from all husk material - more than 90% of mold comes from husks in sun dried cherries.

Remove as many defects (husks, un-hulled cherries or moldy beans) as possible. Use clean bags for storing and transporting cleaned dried beans. Keep separate cleaned dried beans from discarded material. Prevent recontamination by avoiding contact of clean green beans with dust, husks and dirty bags.

Do not store cleaned, dry green coffee near rejects and husks. Processing should achieve a uniform green bean moisture content that is as low as feasible, but certainly not higher than 12.5%.

During transport and storage cover bags during transport and storage to prevent re-wetting.

Load and unload trucks or containers only in dry weather or under cover. Do not use damaged containers and prevent water leaks. Make sure that pallets or wooden floors of trucks and container are dry.

Store coffee is well-ventilated and leak-proof warehouses. Store away from the walls. Cover bags/loose beans in container with waterproof or water-absorbent cover to prevent re-wetting of the top layer of coffee from condensation.

Provide good quality control tests (including cupping) and ensure that they are adhered to, especially to check for moisture and defects.

Mold (mainly Aspergillus Ochraceous) apparently grows much faster and produces much more OTA in the flesh of the cherry than in the bean itself. The ICO/FAO project is now in its dissemination phase. Training tools have been developed and training of trainers courses are being given to spread all gathered know-how to the surrounding countries aiming to reach all coffee producing countries. It is expected that 2 to 3 complete harvest cycles will be needed to spread this know-how in the countries to all relevant levels in the coffee chain and to become fully effective.

In addition to the ICO/FAO project, the OTA-taskforce of the European coffee sector has over the recent years raised the awareness about OTA in the whole coffee chain, via spreading of do's & don'ts (recommended practices in handling of coffee throughout the production chain), international workshops, communications to/by coffee buyers, publications etc. (A copy of these joint ICO/FAO/OT. A Taskforce recommendations are listed in the Table).

Since 1995 the mean OTA level of roasted coffees sold in Europe has gone down by 30%. This is confirmed by both German Food Inspection data and coffee industry own data. It demonstrates that the preventive effort was indeed effective. What Risks are there During Coffee Transport?

A series of transport trails in which containers were equipped with monitoring devices for temperature and relative humidity have been executed and some more are planned. In these trials, actual conditions in the container in between and on top of the coffee were registered. These observations showed that the temperature within the stack of coffee changes only quite slowly without big fluctuations. It went gradually down from the higher temperature at stuffing the container in the producing country to the lower temperature in the consuming country. At the same time, the relative humidity between the coffee rose in a complementary way. However, on top of the coffee stack and also at the ceiling of the container very substantial fluctuations both in temperature and relative humidity were seen. These fluctuations clearly followed a day/ night pattern, with highest temperatures registered towards the end of the day and lowest temperatures early in the morning.

The relative humidity was lowest at the end of the day and condensation (100% relative humidity) was easily reached early in the morning, both in producing and in consuming countries.

The fluctuations are most pronounced during periods of terrestrial transport or storage. The exposure of the container to direct sunlight is evidently a major factor. Apparently this is hardly the case during maritime transport. Such is plausible from the limited fraction of containers on a ship, which are exposed to direct sunlight, and the preferred below deck loading of coffee containers.

The learning from these observations is that stuffing of containers should be done protected from rain and that the moisture content of the coffee should be below 12.5%. These are the factors which set the starting humidity conditions at closing of the doors of the container. Proper lining of the walls with water absorbent paper and a cardboard layer over the coffee stack will reduce the risk of condensation and water dripping from the ceiling directly on the coffee.

Otherwise such would create spot wise conditions favorable for mold growth and OTA production. Appropriate positioning of bags with a drying agent in the headspace of the container over the cover on the coffee will reduce the risk further.

Duration of terrestrial transport and storage of the green coffee within the container should be minimized to reduce the risk of condensation and local rewetting of the coffee.

What is the Fate of OTA during Further Processing?

Several studies have been done about the fate of OTA during further coffee processing and the results of several of them have been communicated via scientific publications. It has been shown that the usual decaffeination processes remove about three quarters of the OTA present before decaffeination. This is plausible, as OTA is quite soluble in the extraction media normally used in decaffeination. If appropriately dried, there will be no mold growth and new OTA formation in the decaffeinated green coffee.

Ten different studies on OTA and roasting are now available. Two studies applied conditions fully outside the commercial roasting practice. All other 8 studies have clearly shown substantial reduction of OTA during roasting. For green coffees with relevant OTA contamination the reduction ranged between 69 and 96%.

A Practical Way to Deal with the Risk of Mold and OTA in Green Coffee

The ICO/FAO mold prevention project has gathered basic understanding in the coffee producing countries and has now embarked on a comprehensive dissemination program to spread this know-how and make it effective in the coffee producing countries. The Code of Practice, as adopted in June 2002 by the European Coffee Federation, gives a comprehensive listing of practical recommendations for transport, stuffing, shipping, receipt, stripping of container and storage of green coffee.

Downstream in the coffee chain there are a number of important points to control OTA levels. Most international coffee transport now goes by container. The European Coffee Contract requires now the moisture content of the green coffee should be below 12.5% to reduce any risk of water condensation during transport (a lower moisture content is even better). At stuffing of the container this moisture level should not be exceeded. The walls of the container should be properly lined and particularly coffee in bags should be covered with cardboard or another water-absorbent sheet to avoid dripping water re-wetting the coffee. Use of an appropriate drying agent in the headspace of the container can further reduce the risk of condensation.

The OTA-taskforce of the European coffee sector initiated a statistical analysis of the traditional quality characteristics, as commonly used in coffee business, and the OTA levels in over 700 more or less random containers of green coffee arriving in Europe during 1998.This exercise confirmed by some origins have more frequently problems than others and that these countries are spread around the globe. Interestingly it showed also correlations with some commonly used quality characteristics. Although the statistical correlations were not very strong, both coffees, which had a visual damage, and coffees with an earthy/ moldy smell had clearly higher frequencies of OTA contamination.

Within this collection of the coffee lots, those without an earthy/moldy smell showed less than 1 % risk to contain more than 10 ppb OTA. The advantage of using the smell is that it spreads much more homogeneous than the actual mold and its OTA do. Similarly visually clean coffees without visible damage showed less than 1 % risk of containing more than 10 ppb OTA.

The strong advantage of these traditional characteristics is that they are much more easily checked and are less dependable on the strongly inhomogeneous distribution of the molds. They allow limiting any checking for OTA by complex sampling plus actual measuring to the cases of doubt only.

There are not yet any formally validated procedures for checking coffee lots for OTA. There is only a CEN validated analytical procedure for OTA in roasted coffee (which does not include sampling). However, particularly in green coffee, the sampling is the most crucial step in checking OTA because of the strong inhomogeneity of the mold and the OTA. At this moment the only thing one can do is to use a procedure analogous to another mycotoxin in another food product.

As an indication for the inevitable inaccuracy, 50 samples of 300g each taken from 50 different bags of the same lot showed a mean OTA level of 7.3 ppb with a variation coeffecient of 39%, when analyzed in one and the same lab. For OTA homogeneous samples of instant coffee the same lab showed analytical variation coefficients up to 13%. This clearly shows that the sampling is the main source of variation. This can partly be compensated by taking multiple samples from different spots spread over the lot, up to an amount of several kg's. The total combined sample should be ground and then intensively homogenized. Only thereafter a smaller sample can go into the actual analysis.

The variation between (certified) labs is an additional story. Even between selected labs the analytical variation coefficient on the same homogeneous sample amounts another 25%.

In order to control mold and OTA risks in the downstream coffee chain it is crucial that containers are properly lined and properly stuffed with green coffee with a moisture content below 12.5% and that bags are covered with a cardboard sheet. Putting bags with appropriate drying agent over the cardboard sheet in the headspace of the container will add extra safety. The duration of terrestrial transport and storage of the green coffee within the container should be minimized to reduce the risk of condensation and local rewetting of the coffee.

At the downstream decision points, the practicable criteria are little or no beans in cherry, clean visual appearance without visible damage/ condensation and absence of an earthy/mouthy smell/note.

In a practicable way, the downstream coffee chain can control the OTA risk in coffee by a system of three control levels: Random checking of green coffees from different origins for OTA to establish their level of performance (can be at a low frequency).

Checking at the downstream decision points for "cherries", "visual damage" and "earthy/ moldy" and in case of doubt or suspicion in combination with OTA checking.

Random checking of final coffee products for actual levels being acceptable. If not so, stricter application of criteria in preceding steps is required.