Info

1 (no subdivision)

Risk assessors usually are involved in the generation and scientific analysis of most data sets on mycotoxins. Risk assessors also regularly perform and promote research activities whose goal is to evaluate the impact of mycotoxins on human and animal health. Such data are crucial to risk managers who must develop the necessary responses to various levels of contamination. These objectives are amenable to many risk assessment strategies, e.g., point estimates, probabilistic approaches, total diet, duplicate diet, etc., all of which require reliable concentration data on mycotoxins. High quality concentration data is a prerequisite for all further decisions and analyses and should use a statistical sampling plan that focuses on which samples, where and when they should be taken and the number of samples required to ensure a representative analysis.

More recently [Pan-European proactive identification of emerging risks in the field of food production (PERIAPT): http://www.periapt.net/], risk assessors were assigned roles in identifying and evaluating emerging risks of mycotoxin contamination, including re-emerging mycotoxins, less investigated compounds, new sources of contamination attributable to changes in climate or diet and contamination due to reduced hygienic conditions or to calamities, e.g., war, in the country of origin. In all cases, samples taken to evaluate toxin levels should be selected only after the additional information on possible contamination problems has been incorporated into the sampling plan.

Risk communication remains a very broad and misunderstood concept, and involves all stakeholders. Stakeholders need information on mycotoxin levels that are consistent with their needs. For example, producers of raw products need data from samples collected before harvesting and at various stages of processing to identify an appropriate destination and to ensure that any necessary treatment/remediation has been successfully accomplished. Consumer organizations and food processors also need to know toxin levels for materials that enter a particular process as well as levels in the final product on the shelf.

Representativeness of laboratory samples

The soundness of residue data for mycotoxins relies both on "fit for purpose" sampling and on the strategy employed for taking a sample that is quantitatively representative of the entire lot. The primary concern in this process is that mycotoxins often have a peculiar, non-homogeneous distribution both in the bulk lot and in the product as it is packaged for retail sale. The sampling step of the process has by far the largest source of error. Many exhaustive reviews have been published since the 1960s, most of them based on studies by Whi-taker and colleagues (Whitaker and Wiser, 1969; Whitaker et al., 1974, 1976, 1979, 1994, 1998, 2000; Whitaker, 2004; Vargas et al, 2006; Miraglia et al, 2006).

Many studies are available that evaluate the distribution of contamination in raw agricultural commodities, that quantify the variance of each sampling step, and develop the Operative Characteristics curves required for an operative sampling plan. An Operative Characteristics curve provides an estimate of the performance of a given sampling plan. This curve is based on a well-defined sampling plan, has an accept/reject limit, and estimates the risk to the seller/exporter and to the buyer/importer that the lot will be misclassified.

The characteristics of a sampling plan generally are chosen based on the final objective of the test. Thus a sampling plan adopted by industry may differ from one used for official regulatory control. Both sampling theory and the pragmatic limitations of sampling have been taken into consideration in developing sampling plans that depend on the crop, the mycotoxin and the global perspective of the organization that will use the results acquired.

We describe the sampling procedures in force in the European Union in some detail and provide cross references to sampling plans in Codex Alimentarius. The U.S. Department of Agriculture and the U.S. Pistachio Industry have adopted specific sampling plans for raw shelled peanuts destined for further processing and for pistachio nuts, respectively.

European Union

Since 1998, the European Union has enacted a package of legislative provisions (European Commission 1998, 2001, 2002a,b, 2003, 2004, 2005a,b, 2006) that define the sampling procedures for the most prominent mycotoxins, i.e., aflatoxins, ochratoxin A, Fusarium toxins and patulin, in many raw and processed products. This package represents one of the more comprehensive tools for sampling mycotoxins worldwide. These procedures were developed for the regulatory control process and are coupled with the maximum tolerable levels for mycotoxins set at European levels. Most of these sampling plans account for theoretical sampling issues, with some also considering the practicalities of the sampling process.

Regulation 401/2006/EC (European Commission, 2006) is a unified provision that contains sampling procedures for all regulated mycotoxins in food, both in bulk and in packages, and that harmonizes sampling protocols for regulatory purposes amongst all EU members (Tables 1-3). Regulated mycotoxins are: aflatoxins in cereals, milk, dried fruits and figs, spices, baby food, peanuts, pistachios and other nuts; ochratoxin A in cereals, roasted coffee, dried fruits, spices, wine, dried vine fruits and baby food; and Fusarium toxins in cereals and baby foods.

Sampling in bulk

A procedure common to all of the EU Directives/Regulations is the subdivision of the lots into a number of sublots. The number of sublots depends on the size of the lot, the matrix being examined, the manner in which incremental samples are taken from a sublot and the way in which the samples from the sublots are pooled to form an aggregate sample, from which the test sam-ple(s) is derived. The procedures can be used in both static and dynamic sampling protocols.

For aflatoxins, ochratoxin A and Fusarium toxins in cereals and cereal products, and for aflatoxins and ochratoxin A in dried fruit and derived products, a lot is accepted if the test sample derived from the aggregate sample is less than the maximum limit once corrections for recovery and measurement uncertainty have been applied. A lot is rejected if the test sample exceeds the maximum limit beyond reasonable doubt once corrections for recovery and measurement uncertainty have been applied.

In samples of dried figs, peanuts and nuts for aflatoxin B1 and for total aflatoxins (Table 1), the size of an aggregate sample and of incremental samples (up to 30 kg and 300 g respectively) is specified that it is higher than that for the food products listed in Table 2. The number of incremental samples to be taken depends on the weight of the lot, with a minimum of 10 and a maximum of 100.

For aflatoxins in food products (Table 1), and for lots > 15 tons, before grinding, the aggregate sample (= 30 kg) from each sublot must be mixed together and then divided into three 10 kg laboratory samples from which test samples are taken. Subdivision of the aggregate sample is not required for dried figs, peanuts or other nuts subjected to further sorting or to other physical treatment, or if equipment capable of homogenizing the samples is not available. If a lot weighs < 15 tons, then the number of incremental samples, the weight of aggregate samples and the number of laboratory samples to be taken from aggregate samples, as well as the acceptance criteria on the basis of the analysis of the aggregate sample/laboratory samples, depends on the size of the lot and on the final destination of the product. The rationale for sampling procedures and acceptance criteria for aflatoxins in spices (Table 1) is similar to that applied to cereals and dried fruits. The aggregate sample for milk and milk products, infant formulae and follow-on formulae (Table 2), including infant milk and follow-on milk, must be at least 1 kg or 1 l, unless such samples are not available, e.g., if the sample consists of a single bottle.

Table 2. Minimum number of incremental samples to be taken from a lot for aflatoxin Mi in milk and milk products, patulin in apple products, and ochratoxin A in wine, grape juice and must.

Form of commercialization

Volume (l) or

Minimum number Minimum volume

51 Ways to Reduce Allergies

51 Ways to Reduce Allergies

Do you hate the spring? Do you run at the site of a dog or cat? Do you carry around tissues wherever you go? Youre not alone. 51 Ways to Reduce Allergies can help. Find all these tips and more Start putting those tissues away. Get Your Copy Of 51 Ways to Reduce Allergies Today.

Get My Free Ebook


Post a comment