Monte carlo technique to simulate aflatoxin testing programs for peanuts

A computer model that accounts for sampling, subsampling, and analytical variability was developed to simulate aflatoxin testing programs. Monte Carlo solution techniques were employed to account for conditional probabilities that arise from multiple samples, subsamples, and/or analyses being used in testing programs. The aflatoxin testing program to be used on the 1974 peanut crop was evaluated by use of the described model. ARS, USDA, and Biological and Agricultural Engineering Department, North Carolina State University, Raleigh, NC 27607.     

Variability associated with testing cottonseed for aflatoxin

The sampling, subsampling, and analytical variance associated with testing cottonseed for aflatoxin were estimated by use of 4.54 kg samples, 100 g subsamples, and the Velasco method of analysis. Regression analysis indicated that each of the above variance components is a function of the concentration of aflatoxin in the populations tested. Functional relationships are presented to determine the sampling, subsampling, and analytical variance for any size sample, subsample, and number of analyses. Paper 4821 of the Journal Series of the North Carolina Agricultural Experiment Station, Raleigh, NC 27607.     

Variability of aflatoxin test results

Using 12 lb samples, 280 g subsamples, the Waltking method of analysis, and densitometric procedures, the sampling, subsampling, and analytical variances associated with aflatoxin test procedures were estimated. Regression analysis indicated that each of the above variance components is a function of the concentration of aflatoxin in the population being tested. Results, for the test procedures given above, showed that sampling constitutes the greatest single source of error, followed by subsampling and analysis. Functional relationships are presented to determine the sampling, subsampling, and analytical variance for any size sample, subsample, and number of analyses.     

Evaluation of cottonseed aflatoxin testing programs

A computer model was developed to simulate cottonseed aflatoxin testing programs. By use of the model, probabilities of obtaining certain aflatoxin test results for various lot concentrations and sample sizes were determined. Also, the effects of sample size and the definition of good and bad sample quality on the probability of lot acceptance were demonstrated. Paper number 5247 of the Journal Series of the North Carolina Agricultural Experiment Station, Raleigh, NC 27607.     

Sampling cottonseed lots for aflatoxin contamination

Large samples called “sublots” were drawn from 41 commercial lots of contaminated cottonseed. Each sublot was subdivided into twenty 5 lb samples which were analyzed for aflatoxin. The mean, median, variance, coefficient of variation, and the estimated range among the sample concentrations were computed. The results indicated that: (A) the variance among sample concentrations was large and was found to be a function of sample concentration and (B) the distribution of sample concentrations was skewed; the density of sample values was greater below the sublot concentration.     

Preparation of lot samples of nut meats for mycotoxin assay

A procedure has been devised for preparing lot samples of mycotoxin-contaminated nut meats so that a representative analytical sample may be removed. The sample is rapidly reduced to coarse size. A relatively large portion (about 1/10 of total sample) of subsample is then split out and further comminuted to a fine particle size with the aid of a fat solvent (meat-solvent, w/v, 3:2). The analytical sample is removed from this mixture. The procedure was tested with shelled almonds and shelled walnuts using radioactive nuts to simulate the mycotoxin contamination and provide a simple, precise measure of the contaminated nut meat distribution. The pooled coefficient of variation was 18% for the subsamples and 4.4% for the analytical samples. Considering the dilution factors used (1.50 and 2.14 contaminated nuts/10⁴ nuts) and the low degree of reliability of the lot sample, the sample preparation methods tested appear to be practical and reliable.     

Variability associated with testing corn for aflatoxin

The sampling, subsampling (both coarse and fine ground meal), and analytical variances associated with testing shelled corn for aflatoxin were estimated by the use of 500 g samples, 50 g subsamples, and the CB method of analysis. The magnitudes of the variance components increased with an increase in the aflatoxin concentration. Functional relationships were developed to predict the variance for a given aflatoxin concentration and any size sample, subsample, and number of analyses. At 20 ppb total aflatoxin, the coefficient of variantion associated with a 4.54 kg sample, 1 kg subsample of coarsely ground meal (passes a #14 screen), a 50 g subsample of finely ground meal (passes a #20 screen) and one analysis were 21, 8, 11, and 26%, respectively.     

Design and analysis of sampling plans to estimate aflatoxin concentrations in shelled peanuts

Methodology for use in the design and evaluation of sampling plans to estimate aflatoxin concentrations in lots of shelled peanuts is presented. Use of the operating characteristic curve for comparing and evaluating processor and consumer risks related to various sampling plans and application of the negative binomial distribution to estimate probabilities associated with sampling lots of shelled peanuts for aflatoxin concentration are discussed. Operating characteristic curves are developed for two different single-sample plans, an attribute multiple sample plan, and the plan presently used by the peanut industry to estimate aflatoxin concentrations in commercial lots of shelled peanuts. An estimated prior distribution of lots according to aflatoxin concentration is used to predict, among others, such values as the per cent of all lots tested that will be accepted by the sampling plans and the average aflatoxin concentration in the accepted lots. All four of the sampling plans described in the paper are compared on the basis of values such as these. Other factors to be considered in the critical evaluation and selection of sampling plans for estimating aflatoxin concentrations in commercial lots of shelled peanuts are discussed. Paper number 3197 of the Journal of Series of the North Carolina State University Agricultural Experiment Station, Raleigh, N.C.     

Comparison of the observed distribution of aflatoxin in shelled peanuts to the negative binomial distribution

Suitability of the negative binomial distribution for use in estimating the probabilities associated with sampling lots of shelled peanuts for aflatoxin analysis has been studied. Large samples, called “minilots,” were drawn from 29 lots of shelled peanuts contaminated with aflatoxin. These minilots were subdivided into ca. 12 lb samples which were analyzed for aflatoxin. The mean and variance of these aflatoxin determinations for each minilot were determined. The shape parameterk and the mean aflatoxin concentrationm were estimated for each minilot. A regression analysis indicated the functional relationship betweenk andm to be:k=(2.0866+2.3898m) × 10⁻⁶. The observed distribution of sample concentrations from each of the 29 minilots was compared to the negative binomial distribution by means of the Kolmogorov-Smirnov test. The null hypothesis that each of the true unknown distribution functions was negative binomial was not rejected at the 5% significance level for all 29 comparisons. Journal Series Paper of the North Carolina State University Agricultural Experiment Station, Raleigh, N.C.     

Measurements of High Oleic Purity in Peanut Lots Using Rapid,Single Kernel Near-Infrared Reflectance Spectroscopy

High oleic peanuts have improved shelf life vs. conventional peanuts. Purity (percentage of high oleic peanuts within a lot) is critical to ingredient performance and final lot value. Contamination can result from unintentional mix-ups at the breeder/seed level, improper production handling, or due to physiologically immature high oleic kernels. Therefore, industry groups have established unofficial sampling plans to monitor purity. Assuming equivalent measurement performance and simple random sampling, increasing the sample size decreases variance among replicated sample test results and increases the precision of estimated lot purity. A novel instrument (QSorter Explorer by QualySense AG) using near-infrared reflectance spectroscopy was evaluated for high speed (20 kernels per second) high oleic purity measurements. The study objectives were to assess instrument performance in: (1) measuring oleic acid (%) in runner peanuts and (2) estimating the true high oleic purity of artificially mixed peanut lots. Three grades (Jumbo, Medium, and No 1) of US Runner mini-lots each at seven different contamination levels (0, 5, 10, 20, 30, 50, and 100%) were prepared. Oleic acid (%) of individual kernels was measured by scanning replicated samples of 10, 50, 100, and 500 kernels using the QSorter Explorer. The variance associated with each sample size and lot contamination level on returned purity values is discussed in the context of binomial sampling. Overall, the demonstrated measurement performance and capacity of the QSorter Explorer to process much larger sample sizes suggest this instrument can better identify true high oleic peanut lot purity vs. other currently available technologies.     

Sampling diced almonds for aflatoxin

To ensure that diced almonds meet the current FDA guideline limit for total aflatoxin, it is necessary to have a sampling plant that will allow representative sampling with defined precision-i.e., with confidence limits on the average aflatoxin found. A sequential sampling plan using 4.54-kg samples of diced almonds or 150-g samples of meal by-product (fines screened from diced nuts during production) was constructed with data from a study of aflatoxin distribution among samples of 2 selected lots of almonds. These 2 lots of whole nuts, estimated to have 400 and 25 ppb aflatoxin, were diced and boxed with normal processing equipment and procedures to approximate the distribution of aflatoxin in the product during commercial production. With a square root trans-formation of the data from 4.54-kg samples of diced nuts, the aflatoxin in samples of both lots approximated a normal distribu-tion and the within-lot variances were not significantly different, which allowed the statistical plan described. A supplemental study was made of aflatoxin distribution in the meal by-product. The lack of a significant difference between the results for diced nuts and those for the corresponding meal suggests that diced almonds can be monitored for aflatoxin indirectly by sampling the meal, which will allow the use of fewer analyses of 150-g samples of less expen-sive product to reach a decision.     

Aflatoxin content of peanut hulls

The degree of aflatoxin contamination in peanut hulls was determined by analyzing inoculated hand-shelled hulls and hulls from peanuts known to contain aflatoxin. Hulls adjusted to 20% moisture, inoculated withAspergillus flavus, and incubated 7 days at 25 C supported growth ofA. flavus but not aflatoxin production. Peanuts from 20 selected Segregation III (visibleA. flavus) lots contained 13–353 ppb of aflatoxin. The machine-shelled hulls from these lots were analyzed and 3 lots contained no detectable aflatoxin, 13 lots contained 4–88 ppb and 4 lots contained >116 ppb. Aflatoxin concentrations of 53–87 ppb were detected in hulls when peanuts containing relatively high levels of aflatoxin (up to 26.8 ppm in damaged kernels) were carefully machine-shelled. Hulls from the same samples obtained by hand-shelling contained no detectable aflatoxin. When machine-shelled hulls were screened through successively smaller screens, the aflatoxin concentration of the smallest fraction (<3.18 mm) was always highest and indicated that small peanut kernels and peanut parts in the hulls actually contained the aflatoxin. Separating hulls over a 4.76 mm round-hole screen appeared to provide a means of removal of most aflatoxin in peanut hulls. No aflatoxin was found in hulls from uncontaminated peanuts.     

Bright greenish-yellow fluorescence and aflatoxin in recently harvested yellow corn marketed in north carolina

Corn kernels that exhibited bright greenish-yellow fluorescence (BGYF) under long-wave ultraviolet light were hand-picked from samples of yellow corn produced in eastern North Carolina. The BGYF kernels from 113 4-kg samples contained an average of 8665 parts per billion (ppb) aflatoxin compared to an average of 46 ppb in the non-BGYF kernels. A regression analysis between the ppb aflatoxin concentration and the wt % BGYF kernels in 2,304 4.5-kg samples produced the regression equation: ppb in sample =197 (wt % BGYF). The correlation coefficient for the analysis was 0.90. Testing programs to reduce aflatoxin concentrations in purchased lots of corn based on either the BGYF method or the AOAC chemical assay method were compared. The average aflatoxin concentration in lots accepted by the AOAC method was 4 ppb, 10 ppb or 18 ppb when an acceptance level of < 20 ppb, < 50 ppb or < 100 ppb, respectively, was used. For the BGYF method, the average aflatoxin concentration in accepted lots was 10 ppb, 16 ppb or 22 ppb when an acceptance level of < 0.10% BGYF, < 0.25% BGYF or < 0.50% BGYF, respectively, was used. Approximately the same percentage of lots were accepted by both methods when either the low, medium or high acceptance level was used. Paper no. 6930 of the Journal Series of the North Carolina Agriculture Research Service (NCARS), Raleigh, NC.     

Variability in corn hybrid resistance to preharvest aflatoxin contamination

Preliminary field studies suggested evidence for resistance of certain corn hybrids to the preharvest infection of kernels byA. flavus and contamination of the kernels with aflatoxin before harvest. A major constraint in evaluating corn hybrids for resistance to the contamination has been the unusual heterogeneity associated with the toxin distribution. A few kernels containing high levels of toxin are routinely responsible for contamination of large sample lots. Extraordinary heterogeneity is also observed in toxin occurrence among fields within a region and among large geographic areas. Edaphic and climatic differences appear to render immature kernels susceptible to aflatoxin accumulation in a discontinuous manner. To reduce intrinsic variability and acquire definitive information on hybrid differences in susceptibility to contamination, several techniques have been developed including: (a) an increase in the number of regional test sites, (b) expansion of the sample sizes, (c) an increase in replication numbers, and (d) elevation of toxin levels in kernels by experimental treatments. Reduction of test variability has allowed for delineation of hybrid differences in aflatoxin resistance. In a diallel set study, genotypes have been identified with heritable qualities of reduced aflatoxin levels in developing kernels. These results provide a basis for further characterization of a genetic facility for resistance to the toxin-producing fungi; these factors have the potential for incorporation into commercial hybrids.     

The relationship of peanut maturity to 2-methylpropanal in headspace volatiles

Headspace volatiles from peanuts of five consecutive maturity classes from three soil temperatures were determined to evaluate the relationship of peanut maturity and 2-methylpropanal. Cured, ground peanuts were held in closed vials for 30 min at 150 C before headspace sampling. Quantitation as percent of total volatiles and ppm in the peanuts revealed a decrease in 2-methylpropanal as peanuts matured regardless of seed size within maturity class. From peanut lots in which maturity-size relationships had been determined, headspace analysis of commercial sized lots indicated that the concentration of 2-methylpropanal was related to the maturity distribution (percent of each maturity class) within each sized lot. These data/techniques may have application in estimating lot-to-lot maturity-related peanut flavor/quality potential. Presented at the AOCS meeting in Honolulu, HI in May 1986.     

Mycoflora,aflatoxins and free fatty acids in California cottonseed during 1967–1968

In central California, neither fungal infections nor aflatoxins are significant problems in cottonseed during the receiving and storage seasons. However, in southern California, the 1967 harvest contained a relatively high percentage of seed which were invaded before harvest by fungi, includingAspergillus flavus. Seed infection and concentrations of aflatoxins in seed increased significantly during the time between harvest and storage in southern California. For a short time during storage, seed infection byA. flavus increased because of the moisture the seed received late in the season; however, aflatoxin concentrations in seed did not increase in storage. The aflatoxin content of the seed removed from storage was a reflection of the relative amount of aflatoxins the seed contained when they were received for storage. In 1967, the conditions that existed in the large, densely packed seed pile did not favor accummulation of aflatoxins in seed, even thoughA. flavus was active.     

电镀产品质量的抽样检验- GB/T 12609新版标准的应用

结合GB／T1 2609新版标准的内容,介绍了电镀产品的抽样检验程序、检验的实施以及检验后的处置。对新标准中检验批和批量,缺陷和不合格品类型,批质量水平 的表示。接收质量限,检验水平等指标进行了解释。重点阐述了抽样方案的确定方法和抽样方式的选择。同时,深入讨论了关于孤立批和连续批的理解和应用。     

Theoretical investigations into the accuracy of sampling shelled peanuts for aflatoxin

Within a population of shelled peanuts, aflatoxin may be concentrated in less than 0.5% of the peanuts. Those peanuts containing aflatoxin might have concentrations up to 1,000,000 μg of aflatoxin per kilogram of peanuts. Because of the distribution pattern, sample means vary widely, and the true average level of aflatoxin in the population is difficult to estimate. The objective of this study was to determine the effect of sample size, N, on sampling accuracy. The negative binomial distribution of aflatoxin since it allowed for a high probability of zero counts along with small probabilities of large counts. Using both the Monte Carlo technique and a direct computation method, the effect of sample size on sampling accuracy was quantitatively described. Journal Paper No. 2775, North Carolina State University Agricultural Experiment Station, Raleigh, N.C.     

Aflatoxin in freshly harvested 1979 georgia corn and formation after collection

In the crop year 1979, freshly harvested dent corn was coUected at maturity in 57 sets of 2 equivalent samples/set. One set was dried the day of collection in Georgia and the other set was shipped to Peoria in corrugated cardboard boxes before drying, The set that was not dried in Georgia was shelled and dried as soon as possible after arrival in Peoria to prevent further aflatoxin formation. Both sets of samples were analyzed randomly at the Northern Regional Research Center, Peoria. In 22 Peoria-dried samples, aflatoxin was detected in levels ranging from 2 to 449 ng/g total toxin but was not detected in the matching samples dried die same day of collection in Georgia. It took an average of 7 days to ship samples from Georgia. Of the 57 samples dried in Georgia, 63% were negative for aflatoxin; aflatoxin was below violative levels (>20 ng/g) in 82%; the average aflatoxin level in all samples was 36 ng/g. In the matching 57 samples dried in Peoria after shipment, aflatoxin was detected in all but 37%; aflatoxin was below violative levels in 70% of the samples; the average aflatoxin level in all samples that had been dried later was 78 ng/g. There was a significant increase in aflatoxin-positive samples associated with shipment prior to drying. These results indicate that aflatoxin formed during shipment of the 1979 freshly harvested corn samples from Georgia.     

Sample preparation for aflatoxin assay: The nature of the problem and approaches to a solution

Cases have been reported of individual peanuts, cottonseeds or Brazil nuts so highly contaminated with aflatoxin that, for a 50 g portion to be representative of the whole, the sample preparation procedures should grind each unit to a large number of particles and distribute them uniformly throughout the sample. Assuming uniform contamination of the individual kernel, each 50 g sample should contain 1/100 of that kernel. Even though these extreme cases may be encountered only infrequently, the more usual situation still presents difficulties because of great variability in individual kernel contamination. However, if the extreme can be handled, one can expect to handle the more usual situation. Equipment and procedures to achieve this distribution goal are described. The equipment studied includes a food chopper (Hobart), a nut mill (Thomas Mills), a disc mill (Bauer), a hammer mill (Fitzpatrick Model D comminuting machine), a hammer mill designed specifically for peanut samples (Dicken’s subsampling mill), a Polytron homogenizer (Bronwill Scientific), a vertical cutter-mixer (Hobart), and a sample splitter (Jones riffle). Commodities examined were shelled peanuts and in-shell Brazil nuts, walnuts, pecans and almonds. Comminution and mixing effectiveness were determined by particle size analysis, by distribution of kernels made radioactive by neutron activation and by aflatoxin analysis of naturally contaminated products. From the results we conclude that the ultimate in sample uniformity can be achieved with a disc mill, solvent addition to obtain a fluid system and mixing and grinding of the fluid with a dispersion mixer-grinder. A practical uniformity can be achieved in a vertical cutter-mixer with less expenditure of time and effort for the commodities studied. Presented in part at the AOCS-AACC Joint Meeting, Washington, D.C., April 1968.