Aflatoxin contamination of Virginia peanuts for the crop-years 1982–1986

International Biodewrioration 24 (1988) 399--407

Aflatoxin Contamination of Virginia Peanuts for the Crop-Years 1 9 8 2 - 1 9 8 6

G. C. Llewellyn", C. E. O'Rear b, P. C. Sherertz", G. Ananaba c, C. G. McWright b & W. V. DasheF "Bu reau of Toxic Substances, Virginia Department of Health, Richmond, Virginia 23219, USA bDepartment of Forensic Sciences, The George Washington University, Washington, DC 20052, USA "Department of Biology. Atlanta University. Atlanta. Georgia 30314, USA

ABSTRACT To establish those environmental conditions which promote the growth of aflatoxin (AFT)-producing Aspergillus spp. on peanuts, a four-year (1982-86) investigation was undertaken to examine possible relationships between air temperature (AT). predpitation (P). and AFT contamination of stored nuts. The mean percentages of nuts that possessed various AFT levels for the years 1982-86 (July-June) rangedfrom 74.2 to 88.0forO-4 ppb. 6.3 to 14. 9for 5-15 ppb. 2"4 to 5"9for 16-25 ppb. 2. 3 to 6 4for 26-100 ppb, and 0 to 4" 7 for > lOO ppb. The mean percentages for the years which exceeded USDA/FDA regulations were 7.1 (1982-83), 7"6 (1983-84), 11.6 (1984-85). and 17"0 (1985-86). Examination of the mean percentage > 15 ppb for each month during thesefour years revealed that the following months fell within that range: September. November. December, January. February, and May (I 982-83); July, October. April. and June (1983-84): August. and June, (I 98485): and July, April. and May (1985-86). Comparisons of pooled-AFT levels, rainfall, and temperature overfour years suggested a 'betterfit' between mean monthly P and mean percentage AFTs > 15 ppb, than between the latter and mean monthly AT. However, application of a predictor equation suggested a correlation between AFT levels and monthly AT. 399 hm,rnational Biocleterioration 0265-3036/88/$03.50© 1988 Elsevier Science Publishers Ltd. England. Printed in Great Britain.

G. C. Llewellyn et al.

400

INTRODUCTION Periodically, stored peanuts have been reported (Scheele et al., 1972; Sanders et al., 1981; Llewellyn et al., 1983, 1986; Pierra et al., 1985) to be contaminated with bifuranocoumarins, i.e. AFTs. AFTs are reported to be hepatocarcinogens, mutagens, teratogens, and toxins (Wogan & Busby, 1980), and they have created possible health hazards to both humans and their domesticated animals following consumption of AFT-tainted nuts. For the past few years, Llewellyn et al. (1983, 1986) have attempted to establish the environmental conditions responsible for AFT contamination of peanuts. Here, are reported mean monthly AFT levels, rainfall, and AT data for the years 1982-1986. An attempt has been made to assess any statistically significant relationships between mean monthly AFT levels and these parameters.

MATERIALS AND METHODS Peanut Collection sites

The peanuts used in this investigation were gathered from sites in Southeastern Virginia (Fig. 1). AFF data The AFT data were obtained from monthly reports prepared by the USDA/State of Virginia Peanut Laboratory (Suffolk, Virginia). To determine AFT levels, bagged samples (1.36 kg) were selected at random, when delivered to either warehouses or shelling plants, prior to unloading. Both an AFT analysis and a peanut grading sample were removed from at least 25% of these samples. Each was homogenized using a Dickens' subsampling mill and yielded approximately 1100 g of sample which was divided into four subsamples. Two of these were stored (Suffolk Office, Virginia Department of Agriculture and Commerce, Division of Markets) and the other two were used in this investigation. The stored subsamples were retained for up to 30 days. The remaining two subsamples were sent to the Mycotoxin Laboratory for AFT analysis. Prior to AFT quantification by thin-layer chromatography (TLC), a modification of the Best Foods' rapid-celite method (sensitive to 2 ppb) served to extract the samples (Nesheim et al., 1964; Horwitz et al., 1970).

Aflatoxin contamination of Virginia peanuts 1982-1986

,(f - . ?---'--

.

-l..L--"---~ ,

----'~7-'-

.---___

~

~

.... *':S ....

,

V

401

I

R

__

(3

-':-.

I

i_-".;

N

I

A

Fig. !. Map of USA and Virginia showing peanut sample collection locations. Peanut samples were collected in the counties of Southampton (1), Isle of Wight (2), Sussex (3), Greensville (4). Surry (5). Dinwiddie (6). Prince George (7). Brunswick (8). Northampton (9). James City (10), Chesterfield ( 11), and New Kent ( 12): also in the cities of Chesapeake (13) and Suffolk (14).

An AFT negative certificate was issued if the first subsample possessed < 16 ppb AFTs. Initial subsamples found to contain >75 ppb were issued a certificate of restriction. Ifa subsample contained 17-74 ppb, a second was requested. The first and second AFT values were averaged and those with a value of < 22 ppb received AFT negative certificates, whereas those with average values of >38 ppb were restricted. In addition, combined subsamples with AFT contamination values of 22-38 ppb required a third subsample. After averaging the values for all three lots those with an average value of < 25 ppb were certified negative and those with > 26 ppb were restricted. Furthermore, use of these lots in commercial processing was restricted.

Environmental parameters Temperature and rainfall data were provided by the USA Department of Commerce, National Oceanic and Atmospheric Administration Data Service (NOAA) (Asheville, North Carolina, NC). The data used in this investigation were recorded from July, 1982 to June, 1986 in" the Norfolk, (Virginia) area. Relative humidity data were not available.

Data expression and analysis Mean percentages and standard deviations for peanuts with various AFT levels (0--4, 5-15, 16-25, 26-100, and > 100 ppb) were determined for

402

G. C. Mewellyn ct al.

N ~

,.0

+I+I+I+I

+I

+I+I+I+I

+I

+I+I+I+I

+I

+I+I+I+I

+I

+I+I+I+I

+I

m

0 "I::

> I

,.~

I

-~

.<

E 0

e, ¢..

.o e,

+,

Aflatoxin contamination of Virginia peanuts 1982-1986

e.

A

>. e~

E e~

e~

o2

<

>"

e-. ¢. e~

e.. o

~^

E

+1

+1

+t

+1

t~ e., 0 e.,

0

4< E

k

r.~

,)

Go,

ol0

oo

0~

! C~4

!

!

I

403

404

G. C LleweUyn et al.

four years (July through June), as well as the four-year sum for 1982-86 years. In addition, the peak AFT contamination months for the eight years were determined to be those that exhibited a value > 10% of the mean for a given year. Both linear (regression analysis) and non-linear (a predictor equation) statistical analyses were used to assess any possible relationships between either monthly P or AT and AFT levels. The non-linear analysis involved the predictor equation AFT contamination level (Ai) = 16.85 + 8.69 sin (Qi), where Qi = n/3 × ( M i - 3.75) and Mi is the monthly AT.

RESULTS AND DISCUSSION The mean percentages for different AFT levels in peanuts over four years were: 83.7 + 6.4; 9.1 + 4.0; 4.0 + 1.7; 4.9 + 1.9; and 1.9 + 2.1 for 0-4, 5-15, 16-25, 26-100, and > 100 ppb, respectively (Table 1).

Monthly AFT levels Table 2 presents the year-dependent peak months for AFT-contaminated peanut samples expressed as 10% or more above the mean contamination for a given year. When the data for the four years were summed the ranking was June, July, April, August, September, October, November, December, January, February, and May.

Monthly AFT levels, AT, and P For the years 1982-1986 (Fig. 2) only the August values exceeded the USDA/FDA regulation of > 15 ppb AFT contamination. The low levels suggested that the bulk ofthe Virginia peanuts stored for that period were relatively safe to consume. Linear regression analysis revealed: 1) there was no linear relationship between either monthly P or AT and the percentage of AFT, 2) an inverse linear relationship existed between annual P, AT, and the percentage of AFq" contamination. In contrast, the non-linear predictor equation indicated that a correlation existed between Ai and the sine of the Mi. Comparison of these results with those from previous investigations (Sledd et al., 1976; Llewellyn et al., 1983, 1986) provides useful information. Sledd et al. (1976) observed that the percentage of shelled peanuts for the crop-years 1964-1968 contained AFTs > 3 0 p p b and exhibited two annual peaks, June and December + one month. These

Aflatoxin contamination of lOtginiapeanuts 1982-1986 I

1

I

1

l

I

1

!

l

!

!

405

I

E

u v Z O In

I0

Xl

Z 0

/i

e

W~.. >_3 '*c

!0

I

6

d

"4

t

V

-

8

xm

t6

.<

z

4

]~ ~ I--

"1" r"<

~

"~ m

i

l

L2

0

o

I

'o

4

1

~.

/)::f

AFT,

6

IX

4 2 0

,

*

I

,

I.

I

J

A

S

0

N

O

.

"r

i

i

I

i

1

d

F

M

A

M

J

o.

v

30

~

20

.-t c

m •

0

I0

MONTHS

Fig. Z. Mean monthly AFT levels, P, and AT for four crop-years.

peaks appeared to correlate with both above average rainfall and AT in the month preceding an elevated AFT-contamination month. Subsequently, Llewellyn et al. (1983) reported a statistically significant correlation between elevated AFT levels and rainfall in the proceeding month for the crop years 1969- ! 974. This finding supports the contention of Sledd et al. (1976). In contrast, Llewellyn et al. (1986) applied both linear and non-linear data analyses to determine the relationship between environmental parameters and AFT contamination for the crop-years 1974-1982. They indicated that: 1) the percentage of AFTs > 15 ppb was not necessarily related to either AT or P; 2) total AFT levels were related non-linearly to AT; 3) the environmental variable which most influenced AFT levels appeared to be AT in the month that the sample was obtained; 4) as AT increased, there was an associated decrease in total AFT levels, and vice versa. Linear regression analyses of the data in the present investigation (Table 3), indicated that: 1) a linear relationship between either monthly P or AT and the percentage of AFT contamination was not apparent; 2) an inverse linear relationship between annual P and the percentage of AFT contamination was possible. Non-linear analyses which involved the predictor equation [Ai --- 16-85 + 8-69 sin (Qi)] indicated that there was possibly a relationship between AFT levels and mean monthly AT in the month the peanuts were sampled, thus supporting the interpretations of Llewellyn et al. (1986).

406

G. C. Llewellyn et al.

TABLE 3 Summary of Statistical AnalysesApplied to AssessWhether Relationships Exist Between AFT Levels. AT. and Rainfall Statistical test employed

Linear regression relationships and correlation co-efficient analysis

Non-linear relationships. predictor equation Ai = 16.85 + 8.69 sin (Qi); Where Qi = n/3 X (Mi - 3.75)

Rest

Inference

(a) r = 0-36 for monthly P and % AFT contamination; r = 0-46 for monthly AT and % AFT contamination

A linear relationship betweeneither monthly P or AT and % AFT contamination does not exist

(b) r = -0-92 for annual and % AFT contamination r = -0-80 for annual AT and % AFT contamination

An inverse linear relationship exists between annual P, AT and % AFT contamination

Ai was correlated with the sine of the monthly AT Mi

Correlation between mean monthly AFT levels and temperature the month preceding peanut sampling

Investigators other than Llewellyn a n d his associates have provided data which indicate that environmental parameters other than AT and P affect c o n t a m i n a t i o n of peanuts with AFTs. Certain o f these parameters are: l) kernel moisture between 23% a n d 46% (Ashworth et al., 1965); 2) drought (Sanders, 1981); 3) soil moisture a n d temperature (Hill et al., 1983); 4) soil temperature and drought (Sanders et al., 1984).

ACKNOWLEDGEMENTS We t h a n k Mrs Ruby Wright a n d Ms Barbara Spain for thoughtful clerical assistance a n d Ms Rosaline Odom, (Atlanta University, Science Librarian for N I H - R C M I Biomedical G r a n t No. GI2RR03062) for verifying the accuracy o f the citations. The authors wish to express their appreciation to R a y m o n d E. Oliver o f the Virginia Department o f Agriculture and C o n s u m e r Services, Agricultural Marketing Services for providing the peanut data, a n d to their Suffolk, Virginia Peanut Laboratory for the analysis o f the peanut samples for AFTs. Also, we wish to express o u r appreciation to the USA D e p a r t m e n t of Commerce, National Oceanic and Atmospheric Administration (NOAA) for

Aflatoxin contamination of Virginiapeanuts 1982-1986

407

providing us with the e n v i r o n m e n t a l data used in this study. Ms Vickie O'Dell also assisted with the project.

REFERENCES Ashworth, L. J., Jr., Schroeder, H. W., Langley, B. C. (1965). Aflatoxins: Environmental factors governing occurrence in Spanish peanuts. Science 148, 1228-9, Hill, F. P., Blankenship, P. D., Cole, R. J., Sanders, T. H. (1983). Effects of soil moisture and temperature on preharvest invasion of peanuts by the Aspergillusflavus group and subsequent aflatoxin development.Appliedand Environmental Microbiology 45, 628-33. Horwitz, W., Chichilo, P. O., Reynolds, H. (eds) (1970). Aflatoxins. In: Official Methods of Analysis of the Association of Official Analytical Chemists, Section 26, 020-26, 061:429, Association of Official Analytical Chemists, Washington, DC. LleweUyn, G. C., Johnson, R. H., O'Rear, C. E. (1983). A model for evaluating aflatoxin occurrence in shelled peanuts. In: Biodeterioration 5. ed. T. A. Oxley & S. Barry, John Wiley and Sons, Chichester, pp. 638-51. Liewellyn, G. C., O'Rear, C. E., Moll, M. B., Dashek, W. V. (1986). Aflatoxin contamination of peanuts grown in Virginia for the crop years 1974 through 1982. In: Biodeterioration 6, ed, S. Barry, D. Houghton, G. Llewellyn, & C. O'Rear, Commonwealth Agricultural Bureau, International Mycological Institute, London, pp. 254-61. Pierra, K., lchinoe, M., Maeda, K., Itch, Y., Makano, M. (1985). Storage conditions of imported raw shelled peanuts and production of aflatoxins. Journal of the Food tlygienic Society of Japan 26, 234--42. Sanders, T. H., Hill, R. A., Cole, R. J., Blankenship, P. D. ( 1981). Effect ofdrought and occurrence of Aspergillusflavus in maturing peanuts. Journal of the American Oil Chemists' Society 58 (A), 966-70. Sanders, T. H., Blankenship, P. D., Cole, R. J., Hill, R. A. (1984). Effect of soil temperature and drought on peanut pod and stem temperatures relative to Aspergillusflavus invasion and aflatoxin contamination. Mycopathologia 86, 51-4. Scheele, R. A., Llewellyn, G. C., O'Rear, C. E., Eadie, T. (1972). The occurrence of aflatoxin in samples of peanuts from Virginia. Virginia Journal of Science 23, 116. Sledd, C. A., Eadie, T., O'Rear, C. E., Lleweilyn, G. C., (1976). An evaluation of environmental factors and occurrence of aflatoxin in Virginia shelled peanuts: 1964 to 1968. In: Proceedings of the Third International Biodegradation Symposium. ed. J. M. Sharpley & A. M. Kaplan, Applied Science Publishers, London pp. 625-634. Wogan, G. N. & Busby, W. F. (1980). Naturally occurring carcinogens. In: Toxic Constituents of Plant Foodstuff. ed. I. E. Liener, Academic Press, New York, pp. 329-69. An extended bibliography for this paper m a y be obtained from the senior author.