765
Journal of Food Protection, Vol. 47, No. 10, Pages 765-769 (October 1984)
Copyright®, International Assooiaiton of Milk, Food, and Environmental Sanitarians
Microbiology of Hydroponically-Grown Lettuce
EVE C. RISER 1 , JOSEPH GRABOWSKI, and EDWARD P. GLENN*
Environmental Research Laboratory, University of Arizona, Tucson International Airport, Tucson, Arizona 85706
(Received for publication January 9, 1984)
ABSTRACT
The microbial quality of lettuce (Lactuca sativa L. var. " O s -
tinata") cultivated in a hydroponic system was evaluated. Over
a 3-month study period, samples of lettuce, nutrient solution,
and peat-vermiculite growing mixture from the greenhouse were
analyzed for total aerobic bacteria, yeasts, molds, and col-
iforms. There was a consistent amount of each type of organism
occurring within each sample group for a given month, and the
numbers of aerobic bacteria and coliforms present were gener-
ally similar to those reported for lettuce and leafy vegetable
crops propagated by the conventional method of agriculture.
Over the study period, the modal values for each type of or-
ganism in lettuce were: aerobic bacteria, 7.9 x 10 6 CFU/g
(range of 3.8 x l O 4 to 2.3 x 10 8 ); coliforms, 1.5 x 10 4 CFU/g
(range of none detected to greater than or equal to 5.3 x 10 7 );
molds, 2 . 9 x l 0 3 C F U / g (range of 1.2 X l O 2 to 5.3 XlO 4 ); and
yeasts, 2.4 X l O 4 CFU/g (range of 6.9 x l O 2 to 2.3 XlO 6 ). The
primary organisms associated with the growing system were
Citrobacter freundii, Enterobacter cloacae, and Enterobacter
agglomerans. No organisms of human health concern (i.e. Sal-
monella spp., Clostrium botulinum, Escherichia coli, or
Staphylococcus aureus) were detected in the samples. The bac-
teriology of lettuce produced for market by this type of hydro-
ponic farming and packaging appears to be generally compara-
ble to that of field-grown lettuce and to present no unique
microbiological hazards to consumers.
Hydroponic farming methods have been adapted to
large-scale production of lettuce and other salad greens
(6,8). Recycling of non-sterile nutrient solution poses
microbiological hazards unique to hydroponic culture and
raises the question of the safety of the products from this
agricultural system.
Lettuce and other leafy vegetable field crops are a suit-
able environment for many types of bacteria
(1,2,4,5,9,10,11). Organisms of public health signifi-
cance, such as Salmonella typhimurium, Escherichia coli,
and Staphylococcus aureus, were able to survive for at
least 6 h when inoculated onto lettuce (9); Salmonella
species have been isolated from lettuce grown in fields
irrigated with sewage effluent (7).
'Present address: Department of Geosciences, University of Arizona,
Tucson, AZ 85721.
JOURNAL OF FOOD PROTECTION, VOL. 47, OCTOBER 1984
Downloaded from http://meridian.allenpress.com/jfp/article-pdf/47/10/765/1655515/0362-028x-47_10_765.pdf by guest on 05 February 2020
The microbial quality of field-grown produce is not
regulated by those agencies (FDA and USDA) that are
charged with the responsibility of establishing microbial
standards for other agricultural products. Likewise, the
microbial quality of produce grown by hydroponic farm-
ing is not controlled by these agencies. There are no
specifications for the microbiological quality of produce
grown by this method, and there is no information in the
literature on the microbial hazards that could be encoun-
tered with this type of farming.
We conducted specific assays to determine the extent
of microbial contamination of the growing mix, nutrient
solution, and lettuce issuing from a hydroponic lettuce
greenhouse. The intent was to evaluate the microbiologi-
cal quality of the produce and to determine if organisms
of human health concern could pose a serious problem
in hydroponic farming.
MATERIALS AND METHODS
Microbial monitoring
The lettuce (Lactuca sativa L. var. "Ostinata", De Ruiter Seeds,
Inc.) was grown in a greenhouse containing 12 long troughs filled with
aerated, recycled nutrient solution as described elsewhere (6). Each
trough was 3.7 m wide by 69.5 m long and approximately 30 cm deep,
holding 76,000 L of solution. Individual lettuce plants were grown in
plastic cups filled with a peat-vermiculite mixture. The cups were in-
serted into evenly spaced holes in floating styrofoam boards allowing
contact of the water and the roots in the open-bottomed cups. Each
month, for 3 months, samples of lettuce, peat-vermiculite, and nutrient
solution (water) were taken from each raceway for microbial analysis.
Microbiological methods employed were selected from the Bac-
teriological Analytical Manual of the Food and Drug Administration
(3) and the Compendium of Methods for the Microbiological Examina-
tion of Foods (7). The analytical procedure was designed to enable a
general monitoring of the entire 12-trough system once every month
for 3 months, as samples were available (i.e., lettuce of harvest size).
To facilitate processing of the large number of samples (one each of
lettuce, peat-vermiculite, and water per trough) plate counts were deter-
mined from single platings of the appropriate dilutions. Readings for
each of the available 12 samples were then averaged to obtain a single
value to represent that sample group (Table 1).
Aerobic plate count, yeasts, molds and coliforms
Lettuce of harvestable size was collected with the use of sterile plas-
tic gloves. The heads were cut aseptically from the roots, transported
to the laboratory in sterile plastic bags, and refrigerated until processed
(usually within 1 h). Processing began by grinding the entire head in
a sterile Waring blender. A 25-g portion was then homogenized with
225 ml of lactose broth (Difco).
766 RISER, GRABOWSKI AND GLENN
Peat-vermiculite was aseptically removed from six plastic growing
cups per trough and separated from the root system. Twenty-five g of
the mixed peat was blended with 225 ml of lactose broth.
Water was withdrawn from different depths of the trough by sterile
pipet and transferred to a sterile bottle. One ml of this was inoculated
into 10 ml of double-strength lactose broth.
Serial 10-fold dilutions were prepared of each of the above lactose
broth suspensions using 0.1% peptone water (Difco). Pour plates were
made of each dilution in Plate Count Agar (Difco) and incubated at
35°C for 48 h for total aerobic bacteria counts. Each dilution was also
incorporated into pour plates of potato dextrose agar (Gibco) containing
chloramphenicol (Argent Chemical Laboratory, Inc., Redmond, WA).
These plates were incubated at 21°C for 5 d to determine total yeast
and mold counts.
For the presumptive coliform counts, serial dilutions of lettuce and
peat-vermiculite were added to tryptic soy agar (Difco) pour plates and
overlayed with violet red bile agar (Gibco). The confirmed coliform
counts were calculated by determining the percentage of presumptive
colonies from the plates that were positive in brilliant green bile broth
(BGB, Gibco).
Serial dilutions of trough water were inoculated into lauryl sulfate
tryptose broth tubes in triplicate and incubated at 35°C for 48 h. Sam-
ples from positive tubes were inoculated into tubes of BGB which were
incubated at 35°C for 48 h; most probable number (MPN) estimates of
coliforms per ml were calculated.
Organisms of human health concern
To detect E. coli, positive BGB tubes were streaked onto eosin
methylene blue agar (EMB, Gibco). The plates were incubated at 35°C
for 24 h. Presumptive colonies were confirmed as E. coli by use of
API 20E strips (Analytab Products, Inc., New York).
Spread plates for S. aureus were prepared by the addition of 0.1-ml
volumes of the 1:10 dilutions to the surface of Baird-Parker medium
(Gibco). The plates were incubated at 35°C for 48 h. Typical colonies
that produced coagulase-positive tests were considered S. aureus.
One-ml portions of the 1:10 dilutions of lettuce or of peat-vermiculite
were inoculated into cooked meat medium (Difco) and incubated at
35°C for 5 d to promote growth of any Clostridia present. Companion
inocula were transferred into trypticase-peptone-glucose-yeast extract
broth with trypsin (TPGYT) and incubated at 24°C for 5 d. Two ml
of 100% ethanol were added to 2 ml of each culture. The alcohol-cul-
ture mixture was held at room temperature for 1 h. Inocula from the
mixtures were streaked onto anaerobic egg agar and incubated anaerobi-
cally at 35°C for 48 h. Typical Clostridium colonies were identified by
using anaerobic API 20A strips (Analytab Products, Inc., N.Y.).
The remainder of the 1:10 dilutions of lettuce and peat in lactose
broth and the water in the double-strength lactose broth were incubated
at 35°C for 24 h for detection of salmonellae. One ml of each was
inoculated into selenite cystine broth (Gibco) and tetrathionate broth and
incubated at 35°C for 24 h. The broths were streaked onto xylose lysine
deoxycholate (XLD) agar and Hektoen agar and incubated at 35°C for
24 h. Suspect Salmonella colonies were identified by inoculation into
R/B I Enteric Differential Tubes (Flow Laboratories, Inc., Inglewood,
CA) and API 20E strips.
Other bacteria
Oxidase-negative colonies and oxidase-positive, glucose-positive, fer-
menting gram-negative rods which appeared on the EMB were also
identified by the R/B I tubes and the API 20E strips.
Spoiled lettuce
A head of our hydroponically-grown lettuce of indeterminate age, but
poor appearance, was purchased from a local supermarket to be assayed
for total aerobic bacteria, yeast, and mold counts.
RESULTS
Aerobic plate count, yeasts, molds, and coliforms
The distributions of counts for aerobic bacteria, yeasts,
molds, and coliforms are recorded in Fig. 1 and a statisti-
JOURNAL OF FOOD PROTECTION, VOL. 47, OCTOBER 1984
cal summary of the data is given in Table 1. For a given
month, there was a consistent amount of each type of
organism occurring within each sample group for all the
troughs tested. In addition, the group counts were basi-
cally stable over the 3-month period, with a suggestion
of a slight reduction in yeast and possibly mold counts
during this time (Fig. 2). Aerobic bacteria counts aver-
aged 4.5 xlO 5 CFU/ml in the water samples, 2.2X10 7
CFU/g in the lettuce, and 1.9 x 108 CFU/g in the peat.
Average yeast counts were 4 x 102 CFU/ml in water,
2.4 x 105 CFU/g in lettuce, and 2.6 x 106 CFU/g in peat.
Molds averaged 1.3 x 102CFU/ml in water, 1.1 xlO 4
CFU/g in lettuce, and 2.6 x 105 CFU/g in peat. Aerobic
bacteria constituted the major class of organisms present
Downloaded from http://meridian.allenpress.com/jfp/article-pdf/47/10/765/1655515/0362-028x-47_10_765.pdf by guest on 05 February 2020
in all sample groups, followed by yeasts, then molds.
The greatest numbers of all types of organisms were
found in the peat-vermiculite followed by the lettuce,
then the water.
The highest coliform counts were also in the peat-ver-
miculite and lettuce, with the culture water typically hav-
ing low coliform counts of approximately 102 per ml or
less in 35 of 35 samples tested. Lettuce typically had col-
iforms counts of less than 103 to less than 105 per gram
(mode of about 1.5 x 104), but in a number of samples
(5 of 35) exceeded 105 and ranged to as high as 5.3 x 107
coliforms per gram in one sample (Fig. Id; Table 1).
Organisms of human health concern
No E. coli, S. aureus, salmonellae, or C. botulinum
was detected. C. novyi, C. sporogenes, C. tetani, and
C. bifermentans were isolated infrequently from lettuce
and peat-vermiculite.
Other bacteria
Table 2 lists the sample group and the number of isola-
tions of each gram-negative rod (all aerobic bacteria) re-
covered from the available 13 specimens each month.
The most frequently isolated organisms from all sub-
strates were Citrobacter freundii and Enterobacter
cloacae followed by Klebsiella pneumoniae, E. agglome-
rans, and Providencia stuartii. E. cloacae appeared
equally on the three substrates and the isolations in-
creased over the three months. K. pneumoniae disap-
peared from the lettuce during this time, although the or-
ganism was still present to the same extent in the water.
Spoiled lettuce
The total aerobic bacteria count was 1.4xlO 8 CFU/g,
the total yeast count was 2.1 x 104/g, and the total mold
count was 1.2 X 103 CFU/g. The coliform count was not
determined.
DISCUSSION
The total counts from the lettuce hydroponic culture
system over the 3-month investigation indicated a very
stable level of aerobic bacteria, yeasts, and molds, which
was characteristic of the site of isolation. The consistent
values suggested that each glass of microbes would attain
 MICROBIOLOGY OF HYDROPONICALLY-GROWN LETTUCE 767

35
a) Total aerobic count ^ Yeast count
| = " Water
30 EZZZZJ Lettuce
25 — Peat

20
15
> is
10
CO
o> 5

Downloaded from http://meridian.allenpress.com/jfp/article-pdf/47/10/765/1655515/0362-028x-47_10_765.pdf by guest on 05 February 2020

o.
E
o
0 J 1 i_n i_l_H 11 H i_LS i £«-i J B«J • '

CO C) Mold count d) Coliform count

30
CD

E 25
20
15
10
5
0 d^ J 1 L
slOMO^Itf-KflO3- O^IOMWlOMyiO'-IO 8 >10" <10' 101-K)2102-K)31(?-104104-105105-K)610'i-1()r1()r-1{)8 >10"
Organ isms/gm Organisms/gm

Figure 1. Distribution of aerobic bacteria (a), yeasts (b), molds (c) and coliforms (d) in hydroponically-grown lettuce, peat-vermiculite
growing mix, and nutrient water. The data represent available samples taken from 12 troughs over the three-month study period.

optimum numbers in each of the substrates. Available nu-
trients and other growth regulating factors restricted the
degree of proliferation beyond that point.
These microbial levels appeared to act as a natural sat-
uration point with only moderate fluctuations within a
particular trough or over time. Ercolani (2) also reported
stable aerobic counts averaging about 107 CFU/g, with
an insignificant variation, (P = 0.05), in lettuce sampled
from retail outlets over a 2-year period. Splittstoesser
(11) noted that aerobic counts rose on various field-grown
vegetables after harvest as a result of equipment contami-
nation and multiplication of bacteria on the harvested
crops. However, the aerobic counts did not increase
beyond the level of 107 CFU/g.
When the hydroponic lettuce was sealed in a plastic
bag for market, the fresh appearance was maintained for
at least 2 weeks. A head of obviously spoiled market let-
tuce of undetermined age had yeast and mold counts that
were lower than these normally observed in fresh lettuce
while the aerobic bacteria count was only one log higher.
The similarity of counts obtained from spoiled and fresh
lettuce suggested a threshold level of organisms for this
substrate, with decay of the lettuce being due to the ac-
tion of the microbes and lettuce enzymes over time. The
stability of the microbe populations measured by the
routine monitoring, further supported the view that
aerobic bacteria, yeasts, and molds would not multiply
to levels that would result in rapid deterioration of the
crop or that would present a human health hazard.
No Salmonella, E. coli, S. aureus, or C. botulinum
was isolated form any of the samples. These bacteria
were not present in the peat-vermiculite mixture used for
growing the lettuce, and the sanitary practices used by
the hydroponic farm and packing line workers were ap-
parently adequate to prevent their introduction into the
system.
Total coliform counts (done by MPN) in the water
ranged from less than 1 to 2.4 X 103 CFU/ml of water.
In lettuce, the counts were higher and ranged from none
detected to 5.3 x l O 7 CFU/g of lettuce. Coliforms are
common in the microflora of raw vegetables (77) and
salad greens (4). Fowler and Foster (4) reported similar

JOURNAL OF FOOD PROTECTION, VOL. 47, OCTOBER 1984

768 RISER, GRABOWSKI AND GLENN

TABLE 1. Summary of microbial count data" from a hydroponic lettuce-growing facility.

Statistical data
Sample type nb Range Mode Mean Std. Dev.
I. Aerobic bacteria
A. Nutrient water 34 8.0 X 10 3 -2.0X 10 6 2.9X105 4.5 XlO 5 5.0 X l O 5
B. Lettuce 34 3.8X104-2.3X108 7.9X106 2.2 XlO 7 4.4 XlO 7
C. Peat Mix 33 4.4xl07-1.9xl08 1.9X10 8 1.9X10 8 7.1 x l O 7
II. Yeasts
A. Nutrient water 34 nd-4.2xl03 7X101 4X 10 2 9.2 x l O 2
B. Lettuce 34 6.9X102-2.3X106 2.4 X 10 4 2.4X105 4.5 XlO 5
C. Peat Mix 33 3.9xl04-1.6xl07 7.9X105 2.6 XlO 6 5.3 X l O 6
III. Molds
A. Nutrient water 34 nd-2.3 X 10 3 2X101 1.3 X l O 2 3.9 x l O 2
B. Lettuce 34 1.2X10 2 -5.3X10 4 2.9 X10 3 1.1 XlO 4 1.4 x l O 4

Downloaded from http://meridian.allenpress.com/jfp/article-pdf/47/10/765/1655515/0362-028x-47_10_765.pdf by guest on 05 February 2020

C. Peat Mix 33 9.5 x 1 0 3 - 8 . 7 x l 0 5 2.2X105 2.6 XlO 5 2.1 x l O 5
IV. Coliforms
A. Nutrient water 34 9 x l 0 " 1 - ^ 2 . 4 x 10 3 4.6X101 4.1 X l O 2 7.7 XlO 2
B. Lettuce 33 n d c « 5 . 3 X 107 1.5 x l O 4 1.7 XlO 6 9.2 x l O 6
C. Peat Mix 33 n d - < l . l X 107 S105 8.9X105 2.1 x l O 6

"Counts are total plate count in colony forming units (CFU for aerobic bacteria, yeasts, and molds, and MPN for coliforms.
b
Number of samples processed.
c
None detected.
coliform counts in three types of salads (two of which 5.5 x 103 per gram and 3.6 x 103 per gram, respectively,
contained lettuce and the third contained only cabbage) which were values below those in this study. However,
sampled from the kitchen of a military hospital. Using by the MPN method, Fowler and Foster (4) found mixed
the total plate count method for coliforms, they found salad and green salad to contain 1.4 X 104 and 7.9 X 104
the range of coliforms for mixed salad (lettuce, radishes, coliforms per gram, respectively. These values conform
and tomatoes) was 1.0 x 101 to 2.1 x 105 per gram, and closely to the modal coliform value found on lettuce in
for green salad (lettuce only) the range was 1.0 x 101 to this study. The high mean value of MPN coliforms in
1 . 6 x l 0 5 per gram. The mean plate count values were the present study is due to a single atypical value of
5.3 X 107 MPN coliform per gram of lettuce, which was
10 nearly two orders of magnitude higher than the next high-
-. b) est coliform count from lettuce.
8 Although the test for fecal coliforms was not included
in the present analyses, fecal contamination by animals
or humans did not appear to be a factor since the stan-
dard coliform tests did not reveal the presence of E. coli
in any of the samples. Several authors (7,5) reported that
4h °-Peot fecal contamination often occurred on crops either man-
o-Lettuce ured or irrigated with untreated sewage but was seldom
•-Water found on other crops. The fecal contamination encoun-
tered under those conditions would not be a problem with
the controlled nutrient water source of the hydroponic
system.
Nelson and Oebker (70) reported that high bacterial
counts on lettuce were almost exclusively due to gram-
negative rods of no probable sanitary significance.
Likewise, several species of gram-negative rods were
found to be associated with hydroponic farming. The pur-
pose of the present study was not to do a comprehensive
evaluation of all gram-negative bacteria present, but
rather to determine if the gram-negative species as-
sociated with the hydroponic environment might be of
August September October August September October
human health concern. Since no quantitative estimations
were made on the gram-negative bacteria, the number of
Figure 2. Average values of aerobic bacteria (a), yeasts (b),
isolations (Table 2) provided a useful indication of the
molds (c) and coliforms (d) by month in lettuce (open circles),
degree to which an organism could grow in each sub-
peat-vermiculite growing mix (squares), and nutrient water
(closed circles). Data points are mean values of available sam- strate, and whether it might be present in significant
ples from the 12 troughs. numbers.

JOURNAL OF FOOD PROTECTION, VOL. 47, OCTOBER 1984

 MICROBIOLOGY OF HYDROPONICALLY-GROWN LETTUCE 769

TABLE 2. Number of isolations of specific gram-negative rods recovered from different sample groups over time. L = Lettuce;
W = Water; P = Peat-Vermiculite Growing Medium.
Total #
August September October Isolations
Organism L w P L w P L W P L W P Total
Aeromonas hydrophila 0 0 0 0 1 0 2 0 0 2 1 0 3
Citrobacter amalonaticus 0 0 0 0 2 2 0 0 0 0 2 2 4
Citrobacter diversus-levinea group 0 0 0 0 0 0 1 0 1 1 0 1 2
Citrobacter freundii 2 9 7 7 10 9 6 9 8 15 28 24 67
Enterobacter aerogenes 0 0 0 0 0 0 1 0 3 1 0 3 4
Enterobacter agglomerans 3 1 2 3 2 4 4 0 1 10 3 7 20
Enterobacter cloacae 4 4 4 10 9 6 11 10 12 25 23 22 70
Enterobacter spp. 0 2 0 0 0 0 0 0 0 0 2 0 2

Downloaded from http://meridian.allenpress.com/jfp/article-pdf/47/10/765/1655515/0362-028x-47_10_765.pdf by guest on 05 February 2020

Klebsiella pneumoniae 5 8 2 1 5 1 0 8 4 6 21 7 34
Proteus morganii 0 0 0 2 3 1 1 1 1 3 4 2 9
Proteus rettgeri 0 3 0 0 0 0 0 1 1 0 4 1 5
Proteus vulgaris 0 0 1 0 0 0 0 0 0 0 0 1 1
Providencia stuartii 1 1 1 2 1 5 2 2 1 5 4 7 16
Pseudomonas alcalifaciens 1 0 0 0 2 0 0 0 1 1 2 1 4

Certain bacteria were common to all sources, while
others tended to be found predominantly in a particular
site. The primary organisms generally associated with the
hydroponic system were C. freundii and E. cloacae. K.
pneumoniae was a common isolate but appeared to have
a better survival rate in the water and was not frequently
transmitted to the lettuce. Proteus rettgeri also occurred
in the water and peat-vermiculite but not on the lettuce.
On the other hand, E. agglomerans may have found let-
tuce the most favorable substrate.
Splittstoesser (77) proposed that pH, inhibitors, and
specific nutrients in different vegetables could influence
the constitution of their microflora. Isolates which might
be termed lettuce bacteria in the hydroponic system were
E. cloacae, C. freundii, and E. agglomerans. These or-
ganisms are considered opportunistic human pathogens
encountered infrequently in human disease; they are com-
mon inhabitants of soil, water, the surface of plants, and
the human intestine. The lettuce that was grown hydro-
ponically was therefore normally colonized by relatively
harmless bacteria, and the more threatening opportunist
pathogens which were occasionally isolated did not ap-
pear to be able to persist on this substrate.
The results from this study indicate that bacteria of
health significance in the food industry such as 5. aureus,
C. botulinum, E. coli, and Salmonella spp. did not nor-
mally occur in the hydroponic system. The total numbers
of microorganisms were not a problem and the presence
of coliforms and gram-negative rods were not of concern.
Use of similar greenhouse conditions and materials as
monitored during this study, accompanied by reasonable
observation of sanitary practices, should result in a hy-
droponic method of farming which is a microbiologically
safe means of growing crops. Microbial monitoring re-
quires facilities that few hydroponic operations would be
able to maintain. Fowler and Foster (4) observed that
bacterial guidelines for field-grown lettuce were both un-
feasible and unnecessary. The present data support this
conclusion in relation to total microbial monitoring of the
hydroponic system.
ACKNOWLEDGMENTS
We thank Dr. D. V. Lightner for generously providing the facilities
and equipment for this study, and for helpful comments on the manu-
script. This work was partially supported by a grant from Kraft Foods,
Inc.
REFERENCES
1. Dunlop, S. G., and W.-L. L. Wang. 1961. Studies on the use
of sewage effluent for irrigation of truck crops. J. Milk Food Tech-
nol. 24:44-47.
2. Ercolani, G. L. 1976. Bacteriological quality assessment of fresh
marketed lettuce and fennel. Appl. Environ. Microbiol. 31:847-
852.
3. Food and Drug Administration, Bureau of Foods, Division of
Microbiology. 1978. Bacteriological analytical manual. Association
of Official Analytical Chemists, Washington, DC.
4. Fowler, J. L., and J. F. Foster. 1976. A microbiological survey
of three fresh green salads - can guidelines be recommended for
these foods? J. Milk Food Technol. 39:111-113.
5. Geldreich, E. E., and R. H. Bordner. 1976. Fecal contamination
of fruits and vegetables during cultivation and processing for mar-
ket. A review. J. Milk Food Technol. 34:184-195.
6. Glenn, E. P. 1984. Seasonal effects of radiation and temperature
on the growth of greenhouse lettuce in a high insolation desert
environment. Hort. Sci. 22:9-21.
7. Intersociety/Agency Committee on Methods for the Microbiological
Examination of Foods. 1976. Compendium of Methods for the
Microbiological Examination of Foods. M. L. Speck (ed.), Ameri-
can Public Health Association, Washington, DC.
8. Jensen, M. 1980. Exciting future for sand culture. Am. Veg.
Grower Nov.:33-35.
9. Maxcy, R. B. 1978. Lettuce salad as a carrier of microorganisms
of public health significance. J. Food Prot. 41:435-438.
10. Nelson, F. E., and N. F. Oebker. 1974. Microbiology of head let-
tuce. Abstr. Ann. Mtg. Amer. Soc. Microbiol. 1974:18 (Abstr.
E103).
11. Splittstoesser, D. F. 1970. Predominant microorganisms on raw
plant foods. J. Milk Food Technol. 33:500-505.

JOURNAL OF FOOD PROTECTION, VOL. 47, OCTOBER 1984