APPLIED MICROBIOLOGY, OCt. 1970, p. 644-645 Vol. 20, No.

4
Copyright © 1970 American Society for Microbiology Printed in U.S.A.

Microbiological Examination of Cocoa Powder1

D. A. GABIS,'2 B. E. LANGLOIS, AND A. W. RUDNICK
Food Science Section, Department of Animal Sciences, University of Kentucky, Lexington, Kentucky 40506
Received for publication 6 May 1970

Total aerobic plate counts of 36 cocoa powders ranged from 100 to 27,000 per g,
with 86% having counts of less than 9,300 per g. Bacillus and Micrococcus were the
only genera identified.

Downloaded from http://aem.asm.org/ on May 23, 2018 by guest

Limited information available on the bacteri- TABLE. 1. Tests used in characterization of isolates
ology of cocoa powders used for the manufacture from cocoa powder
of chocolate-flavored milk indicates that the General Morphology
microflora is mainly aerobic sporeforming bacilli Gram stain Rod or coccus
and micrococci (2). The microflora of the cocoa Catalase Motile, not motile
powders apparently results from contamination Lecithinase Spores present
of the cocoa beans after fermentation (4). This Methyl red Chromogenesis
Voges-Proskauer Colony margin
note confirms the earlier reports on the bacteria Salt tolerance (7 and 10%) Agar stroke growth
isolated from cocoa powders. Oxygen relationship
Thirty-six cocoa powders were obtained from Citrate utilization
10 different manufacturers and were analyzed for Carbohydrates Nitrogen metabolism
Acid and/or gas from Gelatin liquefaction
coliforms and total aerobic plate counts by using Dextrose Peptonization of casein
Violet Red Bile Agar and Standard Plate Count Sucrose Indole
Agar, respectively. In general, plating procedures Lactose Urease
outlined in Standard Methods for the Examination Maltose Nitrate-nitrite reduction
Mannitol cleavage with pepton4e
of Dairy Products were used (1). as nitrogen source
Isolates were obtained randomly from the total pH of hydrolyzed starch brotl
aerobic count plates by the method of Harrison pH of dextrose broth
(3). The isolates were cultured routinely in Brain Nutrient broth Growth at
Surface growth 7C
Heart Infusion broth and agar slants. The tests Subsurface growth 21 C
carried out to aid in the identification of each iso- Amount of growth 32 C
late are presented in Table 1. A total of 37 features Sediment 40 C
were recorded, coded, and transferred to IBM SOC
data cards. The computer used was the IBM
360-50 with a program of a simple ratio determi- counts of over 15,000 per g included two over
nation prepared to determine the similarity be- 22,000 per g and three between 15,000 and 17,000
tween a pair of cultures (6). Characteristics of per g. No coliforms were detected in any samples.
each species from the genera Micrococcus and Study of the isolates indicated that Bacillus and
Bacillus were taken from Bergey's Manual and Micrococcus were the only genera comprising the
placed on IBM cards. Similarity values were then microflora of the samples. All samples contained
determined between these known species, and the bacilli but only nine contained micrococci. Cocoa
unknown isolates having the greater similarity of powders may be a source of contamination of the
characteristics to those of the known species were finished product, but the numbers added are
classified as being the same species. generally small. Normal storage conditions after
The total aerobic plate counts ranged from 100 processing would be sufficient to prevent growth
to 27,000 per g and can be divided into three of these organisms as far as spoilage is concerned.
distinct groups as shown in Table 2. The results A total of 519 isolates were obtained from 36
show that 31 samples (86.1 %) had counts of less cocoa powder samples. Of the isolates from the
than 9,300 per g. Of these, 12 (39%) had counts cocoa powder samples plated and incubated at
of less than 1,500 per g. The five samples with 32 C, only 10 were of the genus Micrococcus and
I Published with the approval of the Director of the Kentucky
the rest were of the genus Bacillus. Computer
Agricultural Experiment Station as journal article no 70-547. analysis determined that there were 4 Micrococcus
2 Present address: Department of Food Science, North Caro- species and 20 Bacillus species represented in the
lina State University at Raleigh, N.C. 519 isolates. The identified species and the num-
644
VOL. 20, 19 70 NOTES 645
TABLE 2. Grouping of cocoa powder samples by TABLE 3. Species and number of isolates from cocoa
total aerobic plate countsa powders identified by numerical taxonomy
Counts per g Species No. of isolates
Company
100-1,500 2,200-9,300 14,000-27,000 Bacillus licheniformis 233
B. cereus ....................... 102
B. megaterium .................. 54
A 0 2 1 B. subtilis ....................... 40
B 0 1 0 B. alvei ......................... 11
C 2 1 0 B. badius .................1...... 11
D 0 2 1 B. larvae ......... 11
E 0 1 0 B. pumilus ...................... 8
F 1 4 0 B. pantothenticus ................ 7

Downloaded from http://aem.asm.org/ on May 23, 2018 by guest

G 1 2 1
H 2 3 0
I 4 3 2
J 2 0 0
Total 12 19 5 B. laterosporus ..................
a Incubated at 32 C.
ber of isolates for each are given in Table 3.
M. freudenreichii represented one-half of the
Micrococcus species, and the remaining micro-
cocci were identified as M. luteus, M. colpogenes,
and M. agilis. The small number of micrococci
isolates confirms the report of Fuller et al. (2).
B. licheniformis comprised 45% of the total
bacilli population. This species plus B. cereus,
B. megateriun, and B. subtilis made up about 85 %
of the bacilli isolates. Fuller et al. (2) reported
that B. subtilis was the predominant species in the
cocoa powders they examined. Sneath (5) states
that B. subtilis and B. licheniformis are closely
related and subject to misclassification on a
monothetic basis. It is also quite possible that the
microflora of cocoa powders may change slightly
over the years.
In conclusion, total numbers of aerobic bac-
teria added to chocolate-flavored milk by cocoa
powders would be insignificant, cocoa powders
are not generally a source of coliform contamina-
tion, the genera found in cocoa powders plated
at 32 C are Micrococcus and Bacillus as deter-
mined by numerical taxonomy, and bacilli are
B. lentus ........................ 6
B. circulans ..................... 5
B. firmus ....................... 5
B. polymyxa .................... 4
3
B. brevis ........................ 2
B. coagulans .................... 2
B. macerans .................... 2
B. pulvifaciens .................. I
B. sphaericus ................... 1
B. stearothermophilus ........... 1
Micrococcus freudenreichii 5
M. luteus ....................... 2
M. colpogenes ................... 2
M. agilis ........................ 1
the predominant type of bacteria found in cocoa
powders.
LITERATURE CITED
1. American Public Health Association. 1967. Standard methods
for the examination of dairy products, 11th ed. American
Public Health Association, Inc., New York.
2. Fuller, J. E., W. S. Mueller, and R. W. Swanson. 1942. Bac-
teriological study of chocolate milk. J. Dairy Sci. 25:883-894.
3. Harrison, J. 1938. Numbers and types of bacteria in cheese
Proc. Soc. Agr. Bacteriol. 1:12-14.
4. Haynak, S. T., S. Polansky, and R. Stone. 1941. Microbio-
logical studies of cocoa fermentation. Food Res. 6:471-479.
5. Sneath, P. H. A. 1962. The construction of taxonomic groups,
p. 289-332. In G. C. Ainsworth and P. H. A. Sneath (ed.),
Microbial classification, 12th Symposium of the Society for
General Microbiology. Cambridge University Press, Cam-
bridge.
6. Sokal, R. R., and P. H. A. Sneath. 1963. Prnciples of nu-
merical taxonomy. W. H. Freeman and Co., San Francisco