-------
DISCLAIM
The 1nfonnat1on 1n this document has been funded wholly or 1n part by
the United States Environmental Protection Agency under Interagency
Agreement No. DH21930587-01-3 to the United States Department of
Agriculture, Beltsvllle, Maryland. It has been subjected to the Agency'.s
peer and administrative review, and 1t has been approved for publication as
an EPA document. Mention of trade names or consnerclal products does not
constitute endorsement or recommendation for use.
11
-------
FOREWORD
The U.S. Environmental Protection Agency 1s charged by Congress with
protecting the Nation's land, air, and water systems. Under a mandate of
national environmental laws, the agency strives to formulate and Implement
actions leading to a compatible balance between human activities and the
ability of natural systems to support and nurture life. The Clean Water
Act, the Safe Drinking Water Act, and the Toxic Substances Control Act are
three of the major congressional laws that provide the framework for
restoring and maintaining the Integrity of our Nation's water, for
preserving and enhancing the water we drink, and for protcting the
environment from toxic substances. These laws direct the EPA to perform
research to define our environmental problems, measure the Impacts, and
search for solutions.
The Water Engineering Research Laboratory 1s that component of EPAls
Research and Development program concerned with preventing, treating, »nd
managing municipal and industrial wastewater discharges; establishing
practices to control and remove contaminants from drinking water and to
prevent Its deterioration during storage and distribution; and assessing
the nature and controllability of releases of toxic substances to the air,
water, and land from manufacturing processes and subsequent product uses.
This publication 1s one of the products of that research and provides a
vital communication link between the researcher and the user community.
Composting 1s capable of stabilizing municipal sewage sludge and
eliminating pathogenic organisms from it. It 1s thus one of the methods
capable of converting a potential waste that has become a massive national
problem Into a useful resource. However, salmonella bacteria, serious
pathogens of humans and other animals, have been reported to repopulate
sewage-sludge composts producing a health hazard for some uses of compost.
The goal of this study 1s to gain an understanding of the factors Involved
1n the growth of these organisms 1n composts. An understanding of these
factors 1s essential to controlling this potential health hazard.
Francis T. Mayo, Director
Water Engineering Research Laboratory
111
-------
ABSTRACT
Research was conducted to Investigate the regrowth of salmonellae 1n
composted sewage sludge. Though composting effectively stabilizes and
disinfects sewage sludges, the decrease 1n salmonellae may be only
temporary, because this pathogen can survive and grow without a human or
animal host.
Modification of an agar medium Improved our ability to detect
salmonellae 1n composts. Salmonellae were detected 1n four composts from
30 composting sites across the United States. However, all composts
supported salmonella growth when sterilized by radiation. These results
and results by others suggest that the mlcroflora 1n composts can suppress
salmonella growth.
To determine the nature of salmonella suppression 1n composts, we
Investigated the effects of groups of the compost mlcroflora, and the
characteristics of the substrates used by salmonellae 1n composts. Results
Indicated that suppression of salmonella regrowth 1s mainly a result of
bacterial competition for a limited number of substrates that these
organisms use 1n common with salmonellae.
This report was submitted by the U. S. Department of Agriculture 1n
fulfillment of Interagency Agreement Nos: EPA No. DW21930587-01-3 and USDA
No. AD-12-F-4-A029 under the partial sponsorship of the U.S. Environmental
Protection Agency. This report covers the period February 1982 to February
1986, and the work was completed as of August 1986.
1v
-------
CONTENTS
Foreword 111
Abstract 1v
Contents v
Figures vl
Tables v11
Abbreviations 1x
Acknowledgments . ....... x
1. Introduction 1
2. Conclusions ..... 3
3. Recommendations 4
4. Modified agar method for detecting environmental salmonellae
by the MPN method .......
Introduction . ........ 5
Materials and methods .................. 6
Results . 9
Discussion ..... 11
5. Growth of salmonellae 1n 30 composted sewage sludges
Introduction ....... 13
Materials and methods 14
Results 17
Discussion . 22
6. M1crob1al suppression of salmonella regrowth
Introduction 25
Materials and methods 25
Results 29
Discussion ...... 36
7. Salmonella regrowth as Influenced by substrate
Introduction 38
Materials and methods ....... 38
Results ...... 39
Discussion .............. 44
References ........ 50
Appendix
A 54
B . 55
C ................ 56
-------
FIGURES
Number Page
1 Growth and death of S. typh1mur1um (circles) and
S.* newport (trlangTes) In compost at two
moisture contents 19
2 Growth and death of S. typh1mur1um (circles) and
J>. newport (tr1 angles) 1 ri 3 composts 20
3 Growth and death of js. typhlmurlum (triangles)
and S. newport (circles) 1n composts at reduced moisture
content 20
4 Growth of salmonellae 1n a mineral-salts medium as
Influenced by the amount of compost #6175 added ... .40
5 Growth of salmonellae on extracts as Influenced by
extraction time 41
6 Correlation of the maximum amount of salmonella growth
with amount of glucose added to a mineral-salts medium, 0 ... .43
7 Growth of salmonellae 1n a mineral-salts extract of
compost #6175 as Influenced by the amount of compost
extracted .44
8 Growth of salmonellae 1n a mineral-salts medium as
Influenced by added amount of extract from compost
#6252 45
9 Growth-rate constants (k, h-1) for salmonellae
as plotted against: A. amount of compost
extract added, and B. total amount of salmonellae
grown .46
10 A plot of the population and rate-constant data
of F1g. 9B according to the linear form of Monodls
equation (see F1g. 9 for meaning of symbols) 48
v1
-------
TABLES
1 Bacterial strains used 1n study 7
2 Effect of BG dye on plate counts of cultures as shown
by percent Inhibition and tlter 10
3 Characteristics of plating media used for
selective-differentiation of salmonellae 12
4 Colony counts on various media for distilled water
suspensions of ^. newport 413 using XL,
Nutrient (N), and PC Agar bases 16
5 Influence of compost mlcroflora on salmonella growth 18
6 Regression analysis and k(j for total salmonellae 1n
a1r-dr1ed, unlrradlated composts 21
7 Regression analysis and k
-------
Number
14 Regression analyses comparing growth rate constants (kg)
for growth of salmonellae 1n extracts of compost #6175
as Influenced by extraction time 42
15 Intercepts (PIA,,,), slopes (l/km), and correlation
coefficients (CC) of equations using amount of compost
or total growth as tha Independent variable for data
of Individual composts and the combined (COMB) data
of the three composts 47
16 Determination of the relative fit of multiple versus
single parameters 1r use of the transformation of
Monodjs equation to describe the dependence of the
growth-rate coefficient on maximum population as a
measure of substrate concentration. ... ... 48
-------
LIST OF ABBREVIATIONS
BG « Brilliant Green
BGS — Brilliant Green Sulfa
BPW — Buffered Peptone Water
BS — Bismuth Sulflte
CPU — Colony Forming Un1t(s)
GNS — Gram Negative Selection
kd — death rate constant
kg — growth rate constant
km — one-half the maximum growth rate constant achievable
KPN -- Most Probable Number
mXLBG — modified Xylose Lyslne Brilliant Green
OD — Optical Density
PBW — Phosphate Buffered Water
PC — Plate Count
TS — Tryptlc Soy
TSI — Triple Sugar Iron
TTBG -- Tetrathlonate with Brilliant Green
XL -- Xylose Lyslne
XL+ — Xylose Lyslne agar base + hydrogen sulfide Indicators, amp1c1H1n,
tetracycllne and kana.5iyc1n
XLBG — Xylose Lyslne agar with 12.5 ppm Brilliant Green
XLD — Xylose Lyslne Deoxycholate
1x
-------
ACKNOWLEDGMENTS
The authors thank A. V. Gibson of our laboratory for her technical
assistance throughout the work and Kevin Poff (student, University of
Maryland) for his help 1n the work of Sections 1 and 2. We tha^k »;. E.
Powers of our laboratory for conducting much of the technical work of
Section 6, and J. R. Surge (Statistical Consulting and Analysis, ARS, USDA)
for his guidance and work relative to the regression analyses conducted 1n
Section 7.
The authors are particularly appreciative of the helpful suggestions
concerning the research given by Joseph Parrel 1 and Gerald Stern, both of
the USEPA.
Special acknowledgment 1s made to those composting facility operators
and municipalities who graciously provided samples and technical
Information. The assistance of Wendy Aaronson (USFDA) for characterizing
the plasmlds 1n our Salmonella cultures 1s deeply appreciated. We thank
Norman Stern (USDA, ARS) for graciously providing the antibiotic-resistant
salmonellae. Helpful criticisms were accepted froro John M. Damare (USDA,
FSIS), Edward Dougherty (USDA, ARS) and Norman Stern.
We thank the staff at the Blue Plains Wastewater Treatment Plant,
Washington, D.C. for assistance with obtaining compost. We thank
Lawrence Slkora (USDA, ARS) for assistance with the adiabatlc composter,
and Lawrence Bromery of NASA, Greenbelt, Maryland for Irradiation of
compost samples.
-------
SECTION 1
INTRODUCTION
Composting 1s a very effective process for stabilizing and disinfecting
sewage sludge. The high temperatures achieved 1n the composting process
Inactivate pathogenic organisms and result 1n population densities that
approach or are below analytical detection limits. For viruses, bacteria,
and parasites requiring specific hosts for survival, 1nact1vat1on results
1n a permanent decrease 1n their densities. For salmonellae, which can
propagate 1n the absence of specific hosts, the reduction in numbers may be
only temporary. Repopulatlon of compost by salmonellae may occur through
regrowth of the organisms existing 1n the compost at an undetectable
concentration or through the growth of organisms Introduced from an outside
source. A likely source may be feces from salmonella-Infected birds,
reptiles, or other animals. Salmonellae Infecting these animals are also
Infectious to humans. Thus even though the composting process achieves
treatment conditions that meet the further pathogen reduction criteria set
forth 1n 40 CFR 257 '.'Criteria for Classification of Solid Waste Disposal
Facilities and Practices; Interim, Final, and Proposed Regulations" as
corrected 1n the Federal Register of September 21, 1979, there may still
be a potential for repopulatlon of composted sewage sludge by salmonellae.
Anecdotal reports of salmonellae 1n composted sewage sludge have been
made. Russ and Yanko reported that salmonellae grew 1n a compost stored 1n
their laboratory. The few studies that have been done Indicated that
salmonellae can grow extensively only 1f the compost has been sterilized,
Indicating that the mlcroflora present 1n composts prevent salmonella
regrowth through antagonistic effects that are not understood.
To aid 1n our evaluation of the ootentlal for salmonellae to grow 1n
sewage-sludge compost, we modified an agar medium to adapt 1t to detecting
salmonella In compost by the most-probable-number (MP.N) method. This
method wa> used to assay the salmonella content and salmonella growth
potential of sewage-sludge composts collected from 30 compost sites across
the United States.
Methods designed to segregate the mlcroblal populations of the compost
on the basis of temperature growth range and other physiological and
biochemical properties were used to select Individual and groups of
organisms from compost. These organisms could then be tested for their
antagonistic capabilities to elucidate factors Involved in the Inhibition
of salmonella growth.
-------
To gain Information on the number of soluble substrates used 1n
salmonella regrowth, kinetic studies of salmonella growth 1n composts were
conducted and analyzed according to mlcroblal growth equations.
-------
SECTION 2
CONCLUSIONS
Selecting salmonella colonies when enumerating low numbers of
salmonellae in sewage sludge and compost samples can be difficult because
of the growth of organisms that mimic salmonellae. This difficulty can be
greatly alleviated by modification of the standard xylose lyslne brilliant
green (XLBG) agar medium. The modification Involves using high
concentrations of brilliant green (BG) dye that has not been heated beyond
50QC.
Studies of composts collected from 30 sites throughout the United
States Indicated that the Inhibition of the growth of salmonellae by the
Indigenous mlcroflora 1s a general phenomenon.
When complete mlcroflora of compost (bacteria, actlnomycetes, fungi,
and protozoans) are present or Introduced Into sterile compost, they fully
suppress the regrowth of salmonellae. A major proportion of suppression
comes from the col 1 forms, with complementing activity from other
gram-negative bacteria. ThermophlUc and mesophlUc actlnomycetes also
supplement the suppresslve activity, but the effect of fungi 1s
negligible. The specific proportion attributable to protozoans was not
defined.
Three composts from widely separated composting sites 1n the United
States contained water-extractable substrates that supported the growth of
Salmonella typh1mur1um. Kinetic studies of salmonella growth Indicate that
these substrates In the different composts are very similar 1f not
Identical, and that total salmonella growth 1s a sensitive assay for their
concentration 1n composts.
-------
SECTION 3
RECOMMENDATIONS
Modification of the xylose lyslne (XL) agar base (agar Increased to 2%
and 6-7 ppm BG dye added to the autoclaved medium after cooling to 50°C)
provides a useful alternative to other plating medium for salmonella assay
of sewage sludges and sewage-sludge composts. Increasing the BG dye
content of the modified XLBG to 9 ppm was found to Increase the
effectiveness 1n discriminating for salmonella colony growth. We suggest,
however, that each user run a study to determine what concentration works
best for salmonella measurement. Additional studies are recommended to
compare the recovery of salmonellae from similar samples (I.e., sludges,
composts) with other procedures. Also, comparison of the modified medium
and other media to recover Indigenous salmonellae 1s recommended.
The resident mlcroflora 1n the composts apparently provide a safety
factor preventing the colonization of sewage-sludge composts by
salmonellae. It has been suggested that composts be sterilized by
Irradiation. We suggest that complete sterilization may result 1n
unchecked growth of salmonellae 1f the composts become Inoculated. The
possibility that partial sterilization may destroy pathogens and yet
Inhibit salmonella growth needs evaluation.
The fungi play essentially no role 1n suppressing the growth of
salmonellae Introduced into composts. Schemes to prevent or control fungal
growth can be used 1f they do not eliminate gram-negative bacteria,
particularly conforms, from the compost.
The finding that bacteria most closely related to salmonellae play the
major role 1n suppressing salmonella growth and that similar water-soluble
substrates 1n the three composts studied support salmonella growth suggest
that a study to determine the identity of these substrates may furnish the
key to understanding and perhaps controlling the regrowth of salmonellae 1n
composts.
Studies are recommended to determine the contribution of protozoans and
other parasites for suppressing salmonellae regrowth 1n composted sewage
sludge.
-------
SECTION 4
MODIFIED AGAR MEDIUM FOR DETECTING ENVIRONMENTAL
SALMONELLAE BY THE MPN METHOD
INTRODUCTION
The presence of salmonellae 1n the environment presents a potential
health hazard that has been studied for nearly 90 years (IK Technical
Improvements 1n the detection of salmonellae have been made over the years,
but the basic question of the significance of the presence of salmonellae
remains unanswered (2). Recent evidence shows an Increase 1n the Isolation
of salmonellae from humans since 1977, and age-associated attack rates have
been shifting (3). However, during the period 1966-1975, nearly 30% of the
outbreaks of salmonellosls were of unknown origin (2). It would seem that
many of these outbreaks were due to environmental exposure. Sewage sludge
remains a source for potential exposure to salmonella. When properly
monitored, sewage sludge composting can be effective in reducing numbers of
pathogenic bacteria to acceptable levels (4), leaving a safe humus-11ke
material for soil Improvement 1n gardens and lawns.
While examining samples of compost, we have also evaluated methods of
detecting and enumerating salmonellae. While no standard procedure exists
for enumerating or even detecting salmonellae 1n the environment (5),
several methods have been offered with qualifications (5, 6, 7). Since
salmonellae are typically found 1n very low numbers, if at all, in compost
(4), the most probable number (MPN) method 1s the one of choice.
Additionally, partlculate matter Interferes with membrane filtration of
large samples. In conformance with the concepts of stress-Induced Injury
(8) and the need to avoid selective agents 1n primary enrichment media (9),
the peptone-water-enrichment procedure 1s currently used (10, 11). This
procedure 1s followed by elevated temperature Incubation in tetrathionate
broth or 1i. tetrathionate broth with brilliant green (TTBG; 11). Plating
on brilliant green (BG) agar, bismuth sulflte (BS) agar, xylose lysine
deoxycholate (XLD) agar or brilliant green sulfa (BGS) agar (5, 6, 12)
offers an opportunity to select those colonies that are salmonella-like.
The suspect colonies can be presumptively tested serologically and, once
purified, biochemically identified and serotyped. Theoretically, 1f only
one salmonella 1s planted 1n an MPN tube, it can be detected on the agar
plates following growth 1n the selective enrichment media.
BS agar 1s routinely used for plate isolation of Salmonella typhi (6),
a pathogen rarely Isolated froni humans 1n the United States (37 and not yet
reported (confirmed) from aerated-pile compost. BG agar uses the dye
brilliant green (13) to select for gram-negative enteric bacteria and
-------
lactose fermentation to Indicate various non-salmonellae. XLD agar uses
the ability of salmonellae to ferment xylose, decarboxylate lyslne and
produce hydrogen sulflde 1n addition to the selective activity of the
bile-salt (detergent), deoxycholate. On XLD agar, conforms and Proteus
sp. are differentiated by lactose and sucrose fermentation, respectively
(6). While each of these media offers specific advantages., they tend to
suffer specific drawbacks and have required modification. 'For example,
there exists some controversy on the need to age BS before use (6). On BG
agar many Proteus and Pseudomonas colonies appear very similar to
salmonellae unless sulfadlazine 1s added (6) and the medium is prepared
without excessive heating (14). For environmental samples such as compost,
Proteus and Pseudomonas organisms are expected to be 1n great numbers
(unlike clinical samples), and from BGA plates of compost MPN tests, the
vast majority of colonies are these genera and not salmonellae. XLD agar
shows excellent differential properties and 1s selective for gram-negative
organisms. However, when diluted samples of compost were spread plated on
XLD agar, motile organisms spread too easily across the detergent-wetted
agar surface.
We examined xylose lyslne agar with 12.5 ppm brilliant green (XLBG; 12)
as a plating medium for the salmonella MPN procedure and also for
direct-plate colony counts of laboratory manipulated composts. When this
amount of BG dye was added to autoclaved and cooled XL agar base (Difco),
the medium became very dark and was extremely toxic to all Inocula. Edel
and Kampelmacher (10) have reported Increased selectivity of BG agar when
heating was minimized. Moats £t£l_. (14) have discussed the destruction of
BG upon heating 1n medium. We report the effect of varying concentrations
of BG dye that has not been heated beyond 5Qoc and propose the use of a
BG dye concentration in XL agar for detecting H2S positive salmonellae
from environmental samples with low pathogen concentrations.
MATERIALS AND METHODS
Bacterial strains
Reference cultures used 1n this study are listed 1n Table 1.
Environmental isolates Proteus mirabilis AB2-la and Salmonella AB2-3a were
recovered during this study from compost. Cultures were maintained, from
-720C stock cultures in duplicate at 4 and 200C on tightly capped agar
slants containing 0.5% Bacto-peptone (Difco), 0.1% yeast extract (Difco),
0.5% sodium chloride, 0.21% dlsodium phosphate, 0.08% sodium dihydrogen
phosphate, and 1.5% agar (pH 7.2).
Culture preparation
For direct plating assay using static salmonellae, ^. typhlmurium was
grown 1n tryptic soy (TS) broth (Difco), and S. newport 4lT(antibiotic
resistant) was grown in TS broth, with ampiclTl1n "(25 ug/ml) and
tetracycllne (75 ug/ml) at 36 !C overnight 1n 150 ml flasks. These
bacteria were pelleted by centrlfugatlon at 17,000 x g for 20 min, and
resuspended 1n sterile, chilled 0.5% saline. The salt was removed by
repeating the centrlfugatlon and resuspending the bacteria in sterile
6
-------
TABLE 1. BACTERIAL STRAINS USED IN STUDY
Organism
Esci.
-------
distilled water. Optical density (OD) was measured at 420 nm and the
suspensions were diluted to 6 x 10? S. typh1mur1um per ml and 3 x 10?
S. newport per ml. It was determine!! by standard curves that the number of
salmoheTlae per ml of water equaled 2.2 x 108 salmonellae per ml per OD
unit + 5 x loo (for dilutions with OD readings between 0.1 and 0.4).
AHquots of each culture suspension were combined equally and stored at
4oc for 21 days. Plate count agar with and without antibiotics confirmed
the total salmonella count as 5.6 x 10
-------
agar, and/or XLBG agar) by streaking a loopful of broth. Salmonella-like
colonies were picked to triple sugar Iron (TSI) agar slants and confirmed
with polyvalent 0 and H antisera (D1fco). Positive tubes were used to
estimate the salmonella MPN, using the tables 1n Standard Methods (5).
Salmonella sp. recovery from the jar (but not the MPN tubes) was used as a
qualitative Index only. Various cultures were biochemically characterized
by the M1n1tek rapid Identification system (BBL Microbiology Systems).
Direct plate enumeration of salmonellae
Composts manipulated 1n the laboratory contained various levels of
.S. typhlmurlum and J>. newport up to 10?/g. Direct plate counts were
possible when salmonelIae exceeded 300 per g because 1n these situations
more uninjured salmonellae existed and the ratio of salmonellae to
competitors wus lower. From the 10-1 suspension used for the MPN,
decimal dilutions 1n BPW were prepared, and 0.1 ml was Inoculated by spread
plate to XLBG 1n triplicate. These plates were Incubated 18-24 h, and
dark-centered colonies were counted as salmonellae. S.. newport was counted
using XL agar base with amp1c1H1n (25 ug/ml) plus kanamycin and
tetracyc'JIne (each 75 ug/ml). Direct plate counting was much preferred to
the labor Intensive MPN test, when salmonella densities permitted*
RESULTS
A mixed culture of S. typhlrnurium and j>. newport was stressed by
storage 1n distilled waTer at 40C for 3 weeks and used to test the
recovery efficiency of BG-conta1n1ng media. Autoclavlng BG dye greatly
reduced Its toxldty. In XL agar, the manufacturers recommended
concentration of BG dye could be doubled without loss of salmonellae 1f the
dye was autoclaved with the medium. In contrast, the manufacturer's
suggested dye content (12.5 ppm) was toxic 1f added after autoclavlng, as
was 10 ppm. After 18 h Incubation of the mixed culture, 10 ppm XLBG plates
supported colonies of S. newport only. S. typhlmurlum colonies did not
emerge until after 24 W. However, XL w1"fh 6 through 9 ppm BG supported the
full salmonella population. On all XL agars up to 10 ppm BG,
_S. typhlmurlum colonies were visibly smaller than £. newport which
maintained an orange colony color through the 24 h observation. Hydrogen
sulflde production was evident for all salmonella colonies examined.
Pure cultures of salmonellae and Proteus mirabllis grown 1n TTBG were
spread-plated on XL supplemented with 0, 6, 7, 8, 9 or 10 ppm BG dye
(unheated), and on BG agar. Each plating was repeated on media aged an
additional 24 h. Very little difference was noted between counts from BG
agar and the XL agar base controls. P. mlrabllls AB2-la grew very sparsely
on XL with greater than 6 ppm BG. _S. typhlmurlum and Salmonella AB2-3a
produced colonies on all XL media supplemented with unheated dye
(Table 2). Salmonella AB2-3a appeared sensitive to the freshly poured agar
as compared to day-old media (Table 2), but this was not observed for
S. typhlmurlum which showed reduced colony size at 24 h as BG content
Increased toward 10 ppm. At 48 h, £. typhlmurlum colonies on 8, 9 and
10 ppm XLBG agars were of equal size to 24-h colonies on XL agar.
-------
TABLE 2: EFFECT OF BG DYE ON PLATE COUNTS OF CULTURES AS
SHOHN BY PERCENT INHIBITION AMD TITER
Proteus AB2-U S. typhlEiur1um Salmonella
Agar/BG, ppm*
BG agar/12.5
XL/0#
XL/6
XL/7
XL/8
XL/9
XL/10
Aget
(h)
6
30
6
30
6
30
6
30
6
30
6
30
6
30
Recovered
(% of Control)
63
N.D.t
100
100
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Recovered
(% of Control)
143
79
100
100
96
93
65
79
61
69
78
63
31
56
Recovered
(% of Control )
105
109
100
100
47
75
51
91
47
70
47
72
40
79
* BG added to XL agar base after sterilization; BG agar (D1fco) prepared by
manufacturer;s directions.
t Hours between agar hardening and Inoculation.
£ No data.
# Control.
XLBG plates containing 6 to 10 ppm unheated dye were heavily
streak-Inoculated with P. vulgarls, Pseudomonas fluorescens. S.
typhlmurlum, S. newport, and Escherlchla colTT" These grew well on XL agar
"basil ETcolT was completely Inhibited by 8 ppm BG, and P. vulgarls and
Ps. f 1 uorelcens colonies were barely perceptible at 9 and~")u ppm after 24
"and WfcSTTionellae were easily detected as black centered and red to
orange after 24 h.
When fifteen cultures (Table 1) were streaked on plates of XL with
6 ppm unheated BG, the growth of Shlgella boydH, _E. colj_, Enterobacter
10
-------
hafnlae, Serratla llquefaclens. Serratla marcescens, Ps. fluorescens and £.
vmgaris was greatly inhibited. '~Th"e"coTl forms Enterd5a"cter"cToacaeT
Enterobacter aerogenes, Klebslella pneumonlae, K. oxytocaTnci
utroDacte"r"treundll were bright yellow without bTack centers. The
Klebs1elTa~spp. and Enterobacter aerogenes produced mucold colonies three
iran or more 1n diameter^The saTmonelfaeTtested showed black-centered red
or orange colonies approximately 2 mm diameter &t 18-24 h. No difference
was noted when these Inoculations were repeated onto XLBG (6 ppm)
containing the alternate dye lot.
The Increased sensitivity of salmonellae recovery from compost as a
result of the suppression of competing organisms when BG was added after
autoclavlng was strikingly evident. Previous MPN assays of nineteen
composts find nine sludges detected salmonellae (0.3/g) 1n three sludges
when only BS agar and BG agar were used. From these, four hundred
salmonella-like colonies were picked for study but only twenty-eight (7%)
were confirmed as Salmonella sp. In assays using commercial XLBG agar
(either alone or 1n combination with BG agar or BS agar), 78 colonies were
picked and 21 (27%) were confirmed as Sajjnpnel 1 a sp. Using only BG agar
and BS agar, the MPN',s from two other 'samples yielded salmonellae. One
sample was sludge. The other was an atypical '.ompost containing 10?
conforms and 17,000 salmonellae per g, as well as P. m1rabH1s 1n
unusually high numbers. Colony picks from this compost were more
frequently confirmed from BG agar (49 of 52 picks) and BS agar (36 of 40),
but not commercial XLBG agar (20 of 52). The overall efficiencies for BG
agar and BS agar were 26% and 20% respectively, from MPN analysis of 28
composts (one was salmonella-positive) and 10 sludges (four were
salmonella-positive). In recent studies using modified XLBG agar (BGS
7 ppm) as the sole plating medium of 15 composts, three samples (20%)
yielded salmonellae (two were detected but less than 0.3 per g, one was
21 per g). From these 15 tests, 26 salmonella-like colonies were picked
and 21 (81%) were serologlcally and biochemically confirmed.
DISCUSSION
BG has been used as a selective agent for more than 70 years. In their
early studies, Browning ejt al. (13) obtained nearly pure cultures of
salmonellae from feces IncuEated 1n peptone water with 3 to 5 ppm BG.
A1tl/,ugh they did not specify, we must assume the dye was not heated. The
recent observations of Edel and Kampelmacher (10) and Moats et al_. (14)
that commercial BG agar suffers from autoclavlng, demonstrated* the need to
study the selectivity of BG for salmonellae 1n environmental samples.
Prepared by manufacturers', Instructions, neither XLBG agar nor BG agar
were adequately selective for Salmonella sp. However, when unheated BG
(7 ppm) was added to XL agar, only salmonellae grew with black-centered
colonies. This was particularly Impressive since C. freundU often mimics
Salmonella sp. on commercial BG agar and XLBG agar (unpublished data). The
nearly complete Inhibition of P. yulgarls and £. m1rab111s on the modified
XLBG agar greatly helped colony picks for transfer to TSI agar. The
black-centered colonies contrasted clearly enough to allow rapid location
of suspect colonies after Incubation for 18-24 h. This contrast allowed
11
-------
use of areater amounts of sample which Increased the sensitivity of the
qualitative assay relative to the ^sntUatlve assay. Samples positive to
the qualitative assay did not always have high enough levels for
quantltatlon.
MPN methods for salmonella enumeration use selective enrichment with
TTBG. Thus, plated organisms would be presumed to be adapted to BG. In
fact, the BG content of the tetrathionate medium used (6) exceeds the
amount In the modified XLBG agar (7 ppm vs 10 ppm). Interestingly, Proteus
AB2-la grew to a tlter equal to or greater than S. typhlmurlujn In the TTBG,
but was completely Inhibited on the modified XLB5 (Table 2).Clearly the
activity of BG dye was reduced In the TTBG broth, perhaps by heat or iodine
addition. Additional studies will be required.
Based on the data from Table 2, little difference results from changing
the BG content from o to 9 ppm. However, Salmonella AB2-3a seemed more
sensitive to fresh XLBG than day-old media, me maximum effect was only
twofold (at 10 ppm) and would make little difference for MPN methods where
colony presence, not colony count, 1s important. However, 6 and 7 ppm BG
In day-old agar gave more consistent recovery. For convenience, we propose
preparation of XLBG agar plates a day 1n advance.
We suggest that our modification to XL agar base (2% agar, and 6 to
7 ppm unheated BG added to the autoe laved medium at 50°C) provides a
useful alternative to other plating media for Salmonella sp. (Table 3).
However, each laboratory should evaluate Its needs, and determine what
medium formulations are required. Other than caution (9), we offer no
reason not to Increase the BG content of the modified XLBG agar to 9 ppm 1n
the MPN procedure when TTBG broth enrichment Is used.
TABLE 3. CHARACTERISTICS OF PLATING MEDIA USED FOR
SELECTIVE-DIFFERENTIATION OF SALMONELLAE
Medium
Bismuth Sulflte Agar
Brilliant Green Agar
Brilliant Green Agar
XLBG
XLBG
XLBG
BG Dye
Content
{ppm)
25
12.5
5.0
12.5
12.5
6-7
Maximum
Dye Temp.
(°C)
100
121
50
121
121
50
Agar
Content
(»)
2.0
2.0
1.5 or 1.35
1.5
1.35
2.0
Source
Dlfco, BBL
Dlfco, BBL
Reference 14
Dlfco
BBL
This study
12
-------
SECTION 5
GROWTH OF SALMONELLAE IN 30 COMPOSTED
SEWAGE SLUDGES
INTRODUCTION
Municipal sewage sludge contains many microorganisms Including
Salmonel1 a sp. which may present health risks to the general population
U/J. To provide a m1crob1olog1cally safe product, sewage sludge can be
composted by the Beltsvllle aerated-pile method to achieve temperatures
over time that kill most pathogens (4). A temperature of 55<>c In the
portion of the pile exhibiting the minimum pile temperature for a period of
2.5 days will provide adequate destruction of pathogenic microorganisms for
release of the compost for use by the general public (4).
Although proper composting kills salmonellae, sterilized composted
sludge supports their growth (18, 19), and there have been anecdotal
reports of salmonellae 1n unster111zed marketed composts and one published
report of salmonella growing 1n an unsterlllzed compost (19). Since many
vectors (such as Infected birds and other animals, and heavy equipment used
In the .composting process) might serve to Introduce salmonellae Into
finished compost, 1t was necessary to determine If relnoculatlon of
finished compost with salmonellae would result In a hazardous product.
Data from studies using raw sludge may have Implications for regrowth 1n
compost. Salmonellae do not grow extensively when Inoculated Into raw
sludge containing autochthonous mlcroflora. Conforms, once established 1n
a sterilized sludge, will greatly Inhibit the growth of Inoculated
salmonellae (20). Studies using compost from a single site Indicated that
a moisture content of 20% and a carbon-nitrogen ratio In excess of 15:1 was
necessary for repopulatlon (19). However, the effect of moisture may vary
with composts, and the carbon-nitrogen ratio will be greatly Influenced by
the relatively Inert carbon of woodchlps that can be expected to vary
greatly from compost to compost.
In this study we have attempted to establish the relative Importance of
the factors known to Influence the growth of salmonellae 1n other media.
These factors Include pH, water activity, temperature, available nutrients
and antagonistic effects due to the presence of other organisms. We
present data showing growth and death of Salmonella sp. In composts 1n the
laboratory, and levels of salmonellae 1n various composts throughout the
United States.
13
-------
MATERIALS AND METHODS
Compost sample evaluation
Sewage sludges composted by the Beltsville aerated-pile method were
shipped to our laboratory 1n sealed 20-liter containers. Thirty facilities
provided 31 composts. Each compost was assigned an Identification number
which 1$ used In this report to observe compost site anonymity. Coarse
material (woodchlps) was removed by a 0.6 cm mesh screen. Water-logged
composts were drained overnight at room temperature to permit screening.
Moisture content was then determined by drying prewelghed subsamples at
95oc overnight, and rewelghlng. All results were expressed on a
dry-weight basis. In addition, subsamples were assayed for salmonellae.
Remaining portions of the composts were stored for seven or fewer days at
AOC 1n loosely sealed plastic bags. Storage beyond one week was at
-200C.
Sample pH was determined by suspending 1 part of the sample 1n 2 parts
(w/v) of distilled water. Water activity measurements were attempted on
six samples of fresh compost using a dew point mlcrovoltmeter (Wescor,
Inc., Logan, Utah) connected to a thermocouple psychrometer (Decagon
Devices, Pullman, Washington) calibrated against saline of known molallty.
However, after adjustment of the water contents of the composts, water
activity data were not linear and this method was discontinued.
The HPN assay for salmonellae 1n compost has been previously described
In Section 4 of this report and 1n Hussong, et al. (21). Briefly, 1t
Involves enrichment 1n buffered peptone broth, selective enrichment In
tetrathlonate broth with added brilliant green, selective differentiation
on modified xylose lyslne brilliant green agar (mXLBG), presumptive
screening on triple sugar Iron agar and confirmation with slide
agglutination assays. Our 5 tube assay detects 0.2 or greater salmonellae
per gram compost, and the 3 tube assay detects at least 0.3 salmonellae per
gram (5). The qualitative assay detects salmonellae 1n at least 14 g of
compost (21). The first 17 composts were tested by the MPN using brilliant
green agar and bismuth sulflte agar (10). Subsequent tests employed mXLBG
agar (21). The first 17 samples were subsequently retested by the MPN
assay using mXLBG agar, but after freezer storage. Fecal conform MPN
assays were performed as described elsewhere (5).
Sample preparation for regrowth and die-off assay
For compost Incubation studies, samples were adjusted to 451 water
content by air drying or atomizing sterile distilled water over the compost
and aseptlcally mixing as necessary. Twenty g (dry weight) of compost were
placed In 100-nl plastic (snap cap) vials wrapped 1n foil. Preliminary
studies showed that the foil prevented condensation of water vapor on the
Interior vial walls (presumably by reflection of radiant energy that would
causa differential heating of the dark compost mass resulting In
evaporation of water and condensation on the cooler Interior vial wall).
Sterilization of composts was achieved by 3 megarads of radiation ($OCo),
which presumably has little effect upon growth factors other than thorough
elimination of competitive micronora. Air exchange was provided through a
14
-------
sterile, cotton-plugged 18-20 gauge syringe needle that was pushed through
the bottle cap (just after Inoculation). Composts were packaged,
Irradiated, inoculated and Incubated within one week. For the preliminary
trial and for long-term Incubation, air-dried composts (101 moisture) were
also prepared.
Salmonella cultures for regrowth and dleoff assay
Salmonella typhlmurlurn ATCC 14028 and antibiotic-resistant strain
Salmonella newport 413 were grown separately for 24 h at 36<>C In 150 ml
Tryptic soy broth (D1fco) 1n a water bath shaker. S. newport was routinely
subcultured on media containing 25 ug/ml amp1c1H1n~and 75 ug/ml
tetracycllne before broth Inoculation (21).
Each fully grown culture was pelleted by centrlfugatlon at 17,000 x g
for 10-15 mln. The pellet was resuspended In chilled 0.51 saline, pelleted
again by centrlfugatlon and resuspended 1n chilled distilled water. The
cell concentrations of the suspensions were estimated by determining 0.0.Is
(420 nn). Bacterial concentrations were adjusted by diluting with
distilled water to about 6 x 10' S. typhlmurlurn per ml and 2.5 x 10'
S. newport per ml. Equal parts oT the adjusted suspensions were combined
"and 10 ui were aseptlcally planted on the top-center surface of each test
compost, and the vial was recapped and Incubated. Previous attempts at
distributing the Inoculum within the sample were too time consuming and
risked contamination of the samples. Spread plate Inoculation on plate
count (PC) agar (D1fco), nutrient agar (D1fco) or mXLBQ agar was used to
enumerate salmonellae when composts were Inoculated and to confirm
0.0.-derived concentrations.
Regrowth and die-off assays
In preliminary study, Irradiated and nonlrradlated samples of compost
16175 were Inoculated and Incubated at 36QC. One vial was taken from
each Irradiation and non1rrad1at1on treatment and tested for salmonellae
after 7, 15, 24 and 36 h. Subsequent short-term Incubations were sampled
at 7 d. Long-term Incubations were conducted at room temperature and were
sampled at one, three, and six months.
For each Inoculation series, plate counts were prepared of each
strain. Also, S. newport 413 was plated on XL agar (01fco) with hydrogen
sulflde Indicators, ampk1111n (25 ug/ml), tetracycllne (80 ug/ml) and
kanaraycln (80 ug/ml) added which will be referred to as XL+ agar. This
agar was used to estimate plasmld (R factor) loss. Loss was normally less
than 20%, and numerically within the standard error of the mean of the
plate count. We had observed variability In plate counts of the resistant
strain on nutrient agar amended with antibiotics; consistently higher
counts were obtained from XL+ agar, which was used thereafter 1n the
preliminary studies (Table 4).
At each sample time, the compost was removed from Its vial aseptlcally
and diluted 1/10 1n chilled peptone buffer. Serial decimal dilution? were
prepared and 0.1 ml Inocula were spread onto mXLBG and XL+ agar
15
-------
TABLE 4. COLONY COUNTS ON VARIOUS MEDIA FOR DISTILLED
HATER SUSPENSIONS OF S. NEWPORT 413
USING XL, NUTRIENT (N),"MJ PC AtiAR BASES
Pour Plates Spread Plates
Trial Medium Cfu/ml (+S.D.)* Plasmld Cfu/ml (^5,0.)* Plasmld
Frequency* Frequency
A
B
XL+J
XL
PC+
PC
XL+
XL
N+
N
4.7E7 (2.616)
5.9E7 (2.5E6)
6.0E7 (9.8E6)
7.2E7 (5.7E6)
1.0E6 (1.1E4)
8.8E5 (7.0E4)
1.1E4 (8.7E3)
9.6E5 (4.4E4)
5.0E7 (2.1E6)
0.79
6.1E7 (n.d.)#
4.8E7 (1.0E7)
0.83
5.4E7 (3.5E6)
6.9E5 (3.65E4)
1.00
7.7E5 (6.4E4)
3.3E3 (2.3E3)
0.011
7.2E5 (4.7E4)
0.82
0.89
0.90
0.005
* Colony forming units per ml, and standard deviation from three plates.
t Plasmld frequency as determined by the proportion of colonies counted on
antibiotic-containing media.
| •»•, media containing antibiotics; see text.
§ No data, two countable plates were averaged.
plates. Additionally, buffered peptone enrichment MPN broths were Inoculated
and til media were Incubated at 36°c for 18 to 24 h. Plate counts showing
too few salmonellae (less than 25 salmonellae on plates representing a 10-2
final dilution) were discarded and their MPN analyses continued to estimate
densities of 0.2/g to 240/g (In a 3-d11ut1on MPN assay). £. newport was
•numerated on XL+ agar 1n both plate counts and MPN assays. Since S. newport
waslnoculated as 25* of the total population, the difference between tnT
total salmonellae counted on mXLBG and the resistant salmonellae counted on
XL+ easily permitted monitoring both populations. Since each strain Is a
different serogroup and showed minor colonial morphology distinctions,
additional cross checks were available to confirm these data.
Soeclflc growth rate constants (kg) were calculated for salmonellae 1n
Irradiated composts by first-order klnStlcs and are therefore expressed as
unit tlme-1 (usually hours). Similarly, specific death rate constants
16
-------
(Ifd) were calculated for salmonellae In nonlrradlated composts, by
first-order kinetics and therefore are expressed as unit tlme-i 1n hours,
days or weeks as suited to a particular Inactlvatlon rate.
RESULTS
Salmonellae were Initially detected In four of the 30 composts
received. Of these four, two contained less than the detection limit (0.3/g)
of the MPN assay. The other two had levels of 21/g (sample #6262) and 1.7 X
104/g (sample #6252). A follow-up Inquiry of sample #6252 left doubt
concerning the sample collection method. A second sample (#6263) from the
same source was free of salmonellae. Fecal conform densities for samples
#6252 and #6263 were 1.3 x 107/g and 1.1 x I05/g, respectively. The pH|s
for the composts upon receipt ranged from 4.2 to 7.6 (Table 5). Water
activity averaged 0.9953 (^ 0.0062) for 6 samples with average water content
of 511 (jf 51), and there was no correlation (0.0454) between the tfcc.
Of the 17 samples processed by the first MPN method, only sampie #6252
yielded salmonellae. After -20 HC storage for 1 yr and retestlng with the
mXLBG MPN method (21), again only sample #6252 yielded salmonellae and at a
density of approximately one log less (2.4 x lOo/g). Fecal conforms were
similarly reduced (6.2 x 106/g) 1n sample #6252 after freezer storage. The
serogroups of the salmonellae recovered from all of the composts were groups
B (samples #6252 and #6262), C] (samples #6252, #6256, and #6265) and E
(sample #6262).
In the preliminary study (F1g. 1), the growth rate for salmonellae at
45% moisture was double that at 11% moisture 1n Irradiated compost. In
nonlrradlated compost, salmonella counts declined. As shown 1n Table 5, the
seeded salmonellae grew well 1n all but two of the Irradiated composts,
reaching densities of about 107/g or more after 7 days. Of the
nonlrradlated 451 moisture composts, about half did not yield salmonellae
after 7 days at 36oc.
Over the 6-month Incubation, salmonella densities In nonlrradlated
composts declined several logs 1n the first month, and 1n the 451 moisture
subset, salmonellae became undetectable thereafter (F1g. 2). Although growth
occurred during the first 36 h 1n Irradiated compost #6175 at 111 H20 (F1g.
1), all three air-dried composts (10, 6, and 71 H20) showed several log
declines after 4 weeks, whether Irradiated or not (F1g. 3).
Antibiotic-resistant salmonellae were no longer detected 1n dried composts
after 4 weeks.
In Irradiated, moist (451 moisture) composts, both strains of
salmonellae reached densities of about 10
-------
TABLE 5. INFLUENCE OF COMPOST MICROFLORA ON SALMONELLA GROWTH
Sample
No.
61/5
6236
6237
6238
6239
6240
6241
C242
6243
6244
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6556
6257
6258
6259
6260
6261
6262
6263
6265
6266
6267
Salmonel
pH
6.5
6.3
7.2
5.8
4.8
6.8
7.2
7.2
6.6
7.6
4.2
7.0
5.9
7.1
7.0
7.2
n.t.
6.6
7.3
5.5
7.5
7.2
7.3
7.2
7.4
6.9
7.6
6.8
n.t.
5.3
7.7
la counts
Organlsms/g*
0 Time
(Inoculum Level)
t>.t>E3 / 3.3E3
5.3E4 / 2.2E4
5.3E4 / 2.2E4
5.3E4 / 2.2E4
4.3E4 / 2.0E4
5.3E4 / 2.2E4
5.3E4 / 2.2E4
5.3E4 / 2.2E4
5.3E4 / 2.2E4
5.3E4 / 2.2E4
2.4E4 / 6.5E3
2.4E4 / 6.5E3
2.4E4 / 6.5E3
2.4E4 / 6.5E3
2.4E4 / 6.5E3
2.4E4 / 6.5E3
6.0E3 / 5E2
6.0E3 / 5E2
6.0E3 / 5E2
5.2E3 / 2.0E3
2.6E4 / 1.6E3
5.2E3 / 2.0E3
2.6E4 / 1.6E3
5.2E3 / 2.0E3
2.6E4 / 1.6E3
6.4E3 / 4.2E2
2.6E4 / 1.6E3
2.6E4 / 1.6E3
6.4E3 / 4.2E2
6.4E3 / 4.2E2
6.4E3 / 4.2E2
are presented as
/ Day
Irradiated
n.t. T
6.2E8 / 1.9E8
3.4E7 / 1.3E7
6.7E7 / 4.1E7
4.5E3 / n.d.f
5.2E8 / 3.3E8
2.1E9 / 1.7E8
3.9E8 / 2.1E8
4.1E7 / 2.8E7
1.1E8 / 4.1E7
- / -
1.0E8 / 1.1E4
8.2E8 / 1.1 E7
6.5E6 / 1.0E5
5.1E7 / 6.8E5
4.4E7 / 3.5E5
7.6E7 / 7.2E7
6.3E7 / 4.3E6
2.1E7 / 5.2E5
1.1E7 / 4.7E4
2.4E8 / 4.9E5
1.7E7 / 3.1E4
9.9E8 / 3.9E4
3.2E5 / 6.7E2
3.2E8 / n.t.
1.6E9 / 2.8E5
2.2E7 / 1.3E5
9.4E6 / 3.8E5
1.6E7 / 1.2E5
2.8E8 / 8.9E6
8.4E8 / 1.0E5
direct plate counts
k4 MS* +A+ *1 * »1 nwtnAl 1 i
Non- Irradiated
n.t.
1.1 / 0.7
1.7 / 1.2
-/-*
1.6 / 1.5
110 / 3.4
0.3 / -
1.5 / 0.4
9.3 / 1.1
9.3 / 0.4
- / -
110 / 4.3
- / -
4.3 / 1.5
4.3 / -
- / -
920 / -
1.7 / 0.2
- / -
- / -
- / -
2.3 / -
- / -
- / -
- / -
- / -
- / -
- / -
- / -
- / -
- / -
(greater than
>A /*n + 4K4n+4/»_
resistant S. newport.
t n.t., not tested.
i -, MPN value of less than 0.3 (first ten samples) or less than 0.2
(remaining samples).
I n.d., no data (no colonies on 10~2 plates).
18
-------
I7
o
at
(A
O
O
COMPOST 6175
COMPOST 6175
45% H20 / A
12 18 24 30 6
HOURS
12
18
24
30
F1g. 1. Growth and death of S. typhlmurlum (triangles) and £. newport
(circles) 1n compost at two moisture contents. Open symbols
Indicate nonsterlUzed compost and filled symbols Indicate
Irradiated (sterile) compost. Curves are for the sums of the two
species', populations (See Table A-l for equation parameters).
Rearesslon analyses of data for a1r-dr1ed, unlrradlated composts and
moist, irradiated composts up to 26 weeks after Inoculation with
salmonellae are shown 1n Tables 6 and 7, respectively. Salmonella
densities 1n air-dried, Irradiated composts during the 26-week period were
too irregular to analyze on an Individual basis. The kd for these data
combined is shown 1n Table 8. The change 1n relative rates of 1nact1vat1on
(k/i's) of the moist and dry composts with Irradiation Indicate the
possible effect of mlcroflora. For sterile compost, dryness Increased the
rate of salmonella 1nact1vat1on over that at the moist condition. With
nonsterlle compost the higher rate occurred 1n the moist compost possibly
because of the presence of other mlcroflora.
19
-------
i
m 7
0*
Ks
.
o
3,
COMPOST (243 F
49« HjO
COMPOST 1240
45% HjO
COMPOST 8238
45* H20
12
2( • 4
12
WEEKS
2«
12
28
Fig. 2. Growth and death of JS. typhlmurlum (triangles) and _S. newport
(circles) 1n 3 composts. Open symbols Indicate nonsterlllzed
compost and filled symbols Indicate Irradiated (sterile) compost.
Symbols on the X-ax1s Indicate "less than" values for undetected
salmonellae. Curves are for the sums of the two species,
populations (See Table A-l for equation parameters).
2»
COMPOST 8243
10 * HjO
COMPOST 8240
COMPOST 6238
7% HO
12
WEEKS
12
F1g. 3. Growth and death of £. typhlmurlum (triangles) and J5. newport
(circles) In composts at reduced moisture content. Open symbols
Indicate nonsterlllzed compost and filled symbols Indicate
Irradiated (sterile) compost. Symbols on the X-axis Indicate
"less than" values for undetected salmonellae. Curves for total
salmonellae are for moist and dry results combined (See Table A-1
for equation parameters).
20
-------
TABLE 6. REGRESSION ANALYSIS AND kd FOR TOTAL SALMONELLAE IN
AIR-DRIED, UNIRRADIATED COMPOSTS
Compost Incubation* kHt
No. Time d
6175 -0 - 30 h
6243 0 - 12 wk
6240 0 - 26 wk
6238 0 - 12 wk
6243, 6240, 6238
(Combined)
(h-i)
0.0373
0.00429
0.00246
0.00484
0. 00291
y-
Inter-
cept
6.0E3
2.0E4
4.3E3
no test
8.4E3
In y
Inter-
cept
8.70
9.91
8.34
no test
9.04
1 * *
Significance*
-0.7747
no test
no test
no test
-0.8794
no test
no test
no test
99.9
n
4
3
4
3
10
* Incubation time for points used In linear regression analysis.
t Death, rate constant.
| Correlation coefficient.
I Significant value of r, exceeded at n - 2 degrees of freedom, (80, 90, 95,
99 or 99.9%). -: less than 80%.
TABLE 7. REGRESSION ANALYSIS AND kd FOR TOTAL
SALMONELLAE IN MOIST, STERILE COMPOSTS FROM
4 THROUGH 26 WEEKS OF INCUBATION
Compost
No.
6243
6240
6238
(Combined)
(h-1)
0.0018
0.0030
0.0015
0.0021
y-
Inter-
cept
1.9E9
1.7E10
7.3E8
2.9E9
In y
Inter-
cept
21.4
23.6
20.4
21.8
rt %
Significance^
-0.8308 no test
-0.9736 no test
-0.9986 no test
-0.8949 99
n
3
3
3
9
* Death rate constant.
t Correlation coefficient.
t Significant value of r, exceeded at n - 2 degrees of freedom,
+ (80, 90, 95, 99 or 99.9%).
21
-------
TABLE 8. DEATH RATE (kd. H) FOR TOTAL
SALMONELLAE IN EACH TEST CONDITION AT ROOM
TEMPERATURE FOR EXTENDED INCUBATION TIMES
Sterile Nonsterlle
Moist -0.0021 -0.0143
Dry -0.0049 -0.0029
DISCUSSION
Of the 30 composts received for study, four (12%) were
positive, but only two (6%) provided an MPN value for salmonellae.
Examination of compost #6263, from the same source as 16252, makes
the high salmonella count of lOVg appear adventitious. Raw
sludge from this sample site was also tested and salmonellae were
undetected. In the presence of 10? fecal conforms per g of
sample 16252, competition should have prevented salmonellae (1000:1
minority) from repopulatlng. Thus, we suspect these high numbers
of salmonellae were due to Improper sampling which Introduced
unusually contaminated material.
Water activity was measured to determine the fraction of water
content available for mlcroblal growth. Although this Is not an
unusual procedure for foods 1t does not seem to be readily
applicable to composts. Our data did not follow theoretical trends
and could not be utilized.
Growth of salmonellae 1n laboratory-sterilized compost has
been previously reported (18). The capacity of 93% of this studyjs
composts to provide nutrients for growth of salmonellae Indicated a
great potential for regrowth under these conditions. However,
salmonellae died In nonlrradlated compost. Those Irradiated
samples where salmonellae did not grow were found to be unusually
add, with pH values of less than 5.0.
The six-month Incubation trials used moist (Fig. 2) and dry
(F1g. 3) composts, and Irradiated and nonsterlle subsets. All
samples were Incubated at room temperature (Instead of 36°C) to
simulate packaged compost under storage, and also to save Incubator
space. Thus, slightly different kinetics were not unexpected. In
2 of 3 moist nonsterlle composts, salmonellae persisted through
four weeks with loss of viability occurring before 12 weeks. In
the remaining compost, salmonellae were not vial be at 4 weeks (F1g.
2). However, 1n low moisture composts, the salmonellae densities
22
-------
declined In Irradiated composts as well as nonsterlle composts.
This decline Is 1n apparent contradiction to the Increase In
population of the preliminary trials 1n which salmonellae grew to
10//g 1n dry Irradiated compost. Had these levels been achieved,
higher 4-week densities would have been expected.
If the death rates of salmonellae In the various composts are
first order, then the kinetics of the short term studies should
apply to the longer term studies. Applying the KH of -0.0307h-1
for air-dry nonsterlle compost 16175 (Fig. 1) to the Initial
salmonella concentration {8 x 104/g) of the nonsterlle compost
data of Fig. 3, predicts that these organisms should be below
detectable concentrations by 5 days. Since there are still
survivors In excess of 12 weeks, compost 16175 must be very
atypical or first-order kinetics do not apply. The latter seems
most likely. Although we have used a straight line for
convenience, a hyperbola might better fit the data than the
straight lines used 1n F1g. 3.
From Inspection of F1gs. 2 and 3, 1t 1s apparent that
salmonella 1nact1vat1on was more rapid 1n moist nonsterlle compost
than 1n moist sterile compost. Apparently active mlcroflora were
more Important than dry conditions and the presence of weakly
active mlcroflora 1n salmonella 1nact1vat1on. This observation may
account for an earlier observation (20) of the relative stability
of salmonellae In air-dried raw sludge.
An advantage of using Si. newport 413 as well as S. typhlmurlum
arose from the presence of the easy markers (serogroup and
antibiotic resistance) to separate them. We serotested
antibiotic-resistant Isolates from laboratory Incubations and found
no serogroup B salmonellae. This suggests that resistance factor
transfer did not occur to a level detectable by our methods and
great hazards due to generation of resistant salmonellae (22) may
not occur 1n compost. While our methods would have detected
transfer on a large scale, more sensitive methods exist to test
this more precisely. More Importantly, had antibiotic resistance
been a factor 1n regrowth suppression, group B salmonellae would
have been overrun by resistant strains 1f the specific antibiotics
were present. Since both strains grew rather well In Irradiated,
sterile composts, growth- suppressing antibiotic or toxic compounds
were not present.
Salmonellae grew well In moist, Irradiated composts (except
under conditions of add pH), and died 1n non-Irradiated composts
when Incubated at 36<>c or at room temperature. Since the
difference between these composts was the presence of other
microorganisms, we conclude that competitive Influences suppress
salmonellae 1n spite of available nutrients. Although low pH
composts prevented regrowth, these are atypical composts and may be
23
-------
less desirable for geochemlcal reasons. Suggested alternative
final treatments for composts have Included sterilization by
Irradiation (18) or other methods. Although this eliminates
salmonellae In composts, where the potential for relntroductlon of
the pathogen exists, we feel the resident mlcroflora provide a
safety factor. In the absence of these normal flora, relntroduced
salmonellae may grow unchecked, thus creating a greater hazard.
Two Immediate questions are presented by these findings. They
Involve the nature of the specific nutrients used by the
salmonellae and the Identification of the competitive resident
mlcroflora that suppress salmonellae. Resolving these additional
problems will provide a better understanding of mlcroblal control
In the environment.
24
-------
SECTION 6
ROLE OF MICROFLORA IN SUPPRESSION OF SALMONELLA
REGROWTH IN COMPOSTED SEWAGE SLUDGE
INTRODUCTION
Composting 1s used by municipalities to treat sludge generated at
wastewater treatment plants and transform It Into an agriculturally useful
product (23). Sewage sludge, when composted according to recommended
criteria (23, 24), Is considered free of human pathogens and parasites.
However, a question arises concerning the possible regrowth of salmonellae
as a result of Inoculation of finished compost by animals Infected or
machinery contaminated with these organisms.
Previous studies (19, 25, 26) show that sterilized compost can support
abundant growth of salmonellae. However, results from a recent study of
more than 30 composts (Section 5 of this report) show that when the
Indigenous mlcroflora are present In finished compost (unsterlllzed
compost), any relntroduced salmonellae will die rapidly (26). The specific
nature of the mlcroblal competition responsible for the suppression of
salmonella regrowth 1s the subject of the data reported here. The studies
undertaken were based on two hypotheses: 1) actlnomycetes and fungi present
In compost produce secondary metabolites that are Inhibitory to
salmonellae, and 2) bacteria, especially enterobacterla, are better adapted
to saprophytlc colonization of compost than salmonellae. Understanding the
nature of this suppression would provide a basis for use of Indicator
microbes or mlcroblal activity to possibly control the growth of
salmonellae In composts.
MATERIALS AND METHODS
Compost samples
Sewage sludge from the Blue Plains Wastewater Treatment Plant,
Washington, D.C. composted by the Beltsvllle aerated-pile method (23) was
used for this study. Samples were collected using sterile tools and
equipment, and processed at our laboratory within 4-6 h of collection.
Temperatures 1n the compost-pile zones were measured Immediately before
sampling and recorded as the temperature for the sample. Large pieces of
wood were removed aseptlcally from the compost. Moisture content was
determined In duplicate on 10 g subsamples dried to constant weight at
95<>C. When required, moisture contents of the compost samples were
adjusted to 45-50* prior to sterilization.
25
-------
SECTION 6
ROLE OF MICROFLORA IN SUPPRESSION OF SALMONELLA
REGROWTH IN COMPOSTED SEWAGE SLUDGE
INTRODUCTION
Composting 1s used by municipalities to treat sludge generated at
wastewater treatment plants and transform It Into an agriculturally useful
product (23). Sewage sludge, when composted according to recommended
criteria (23, 24), Is considered free of human pathogens and parasites.
However, a question arises concerning the possible regrowth of salmonellae
as a result of Inoculation of finished compost by animals Infected or
machinery contaminated with these organisms.
Previous studies (19, 25, 26) show that sterilized compost can support
abundant growth of salmonellae. However, results from a recent study of
more than 30 composts (Section 5 of this report) show that when the
Indigenous mlcroflora are present 1n finished compost (unsterillzed
compost), any relntroduced salmonellae will die rapidly (26). The specific
nature of the microblal competition responsible for the suppression of
salmonella regrowth 1s the subject of the data reported here. The studies
undertaken were based on two hypotheses: 1) actlnomycetes and fungi present
In compost produce secondary metabolites that are Inhibitory to
salmonellae, and 2) bacteria, especially enterobacterla, are better adapted
to saprophytlc colonization of compost than salmonellae. Understanding the
nature of this suppression would provide a bar1s for use of Indicator
microbes or microblal activity to possibly control the growth of
salmonellae 1n composts.
MATERIALS AND METHODS
Compost samples
Sewage sludge from the Blue Plains Wastewater Treatment Plant,
Washington, D.C. composted by the Beltsvllle aerated-pile method (23) was
used for this study. Samples were collected using sterile tools and
equipment, and processed at our laboratory within 4-6 h of collection.
Temperatures In the compost-pile zones were measured Immediately before
sampling and recorded as the temperature for the sample. Large pieces of
wood were removed aseptlcally from the compost. Moisture content was
determined In duplicate on 10 g subsamples dried to constant weight at
950C. When required, moisture contents of the compost samples were
adjusted to 45-50% prior to sterilization.
25
-------
Actlnomycetes were Isolated from dilutions of compost plated onto
tryptlc soy (TS) agar. or Czapekjs agar containing 100 ug/ml cyclohexlmlde
Incubated at 25oc and at 50<>c. Bacteria were Isolated on TS agar with
cyclohexlnrtde at 25oc and at 50QC. Fungi were Isolated on oxgall (27)
gentamlcln (100 ug/ml) agar Incubated at 25°c and at 45°c.
Identifications of actlnomycetes and fungi were based on key
morphological and physiological characteristics (27, 28, 29). Reference
cultures of some actlnomycetes were Included for comparison of key
characteristics and of response by salmonellae 1n agar-plate assays. These
cultures were: Streptomyces eurythermus NRRL 2539, St. grlseoflavus
NRRL 5312, £t. nygroscopi'cus NRRL Z3B/. St. macrosporeus ATCC 197H3.
° leous NKRL 3009, St. thermodlasTatlcus NRRL 5316. St. thermoflavus
, . . .
NRRL I25U3. st. thermonltrTTlcans NRRL 12517. St. thermophltus ATCC 19282.
St. thermotoTerans NRRL Z345. and St. thermovloTaceus NRRL 12374. Bacteria
were identified by gram-stain reaction and nutritional and biochemical
characteristics determined 1n the Enterotube and Oxl-Ferra systems
(Hoffman-LaRoche, Nutley, N.J.).
Plate assay for antagonism
Isolates of actlnomycetes, bacteria, and fungi were screened
Individually In agar plate assays for their potential to Inhibit salmonella
growth. Six salmonella Isolates were used: S. t>ph1mur1um ATCC 14028,
S. typh1mur1um APHIS, S. enterltldls 6176, tHVee Isolates of salmonella
"From sewage sludge, Including 2 isolates of serogroup C1 (designated AOC1
and BOC1) and 1 Isolate of serogroup B (designated BOB). Assay plates were
arranged so that the potential antagonist was streaked along the agar over
a grid mark 70 mm long and 25 mm from the edge of the plate and three
salmonella streaks were made perpendicular to the antagonist. Spores and
conldla of actlnomycetes and fungi were obtained from Czapek's agar and
yeast malt extract agar cultures of the Isolates. Bacteria were obtained
from 24 h TS broth cultures. With actlnomycetes and fungi, antagonists
were allowed to grow on the test plates at temperatures appropriate to
their optimal growth (as determined 1n preliminary tests 1n our laboratory)
and for up to 4 d prior to Inoculation of the salmonellae. Antagonism was
determined by measuring the zone of Inhibition extending from the
antagonist streak to the start of growth of the salmonella streaks.
Controls consisted of salmonella streaks on plates without antagonist
streaks. Isolates that Inhibited salmonellae were retested on the plate
assay before being used In compost assays.
Responses of salmonellae to compost actlnomycetes and bacteria were
checked by an agar overlay procedure. In the latter, dilution plates of
compost, as described above for bacteria and actlnomycetes, were
replica-plated onto fresh TS agar when colonies appeared. The parent
dilution plates were overlaid with TS agar In which salmonellae were seeded
from a 24 h culture. Overlays were Incubated at 30°C and any zones of
Inhibition In the salmonella layer would correspond to colonies in the
original layer that were antagonists. These antagonists could be picked
and *ubcyltured from the replica plates,
-------
Compost Assays for Antagonism
Two approaches were used to test antagonism by specific microbes or
groups of nonspecific Microbes. For both approaches, plastic vials with
plastic caps were filled with 20 g (dry weight) of test compost and either
Incubated at the temperature designated for the particular test or
Irradiated with a 60c0 source to a final sterilizing dose of 3 Megarads
as described previously (26). Duplicate or triplicate vials of each test
were prepared. Vial caps were fitted with a sterile, cotton-plugged,
20-gauge hypodermic needle to allow for gas exchange.
In the first series of tests, Isolates from the plate assays that
showed Inhibition of salaonellae were Inoculated Into sterile compost
simultaneously or 7 d prior to Inoculation with salmonellae (Isolate
BOC1). A Bacillus sp. Isolated from compost was also Included 1n this
series of tests and Inoculum preparation and enumeration were performed as
for salmonellae. Fungal Inoculum consisted of 1 ml conldlal suspensions In
sterile distilled water. Salmonella Inoculum consisted of 1 ml of cells
from a 24 h TS agar slant In autoclaved compost extract (compostrwater,
1:200). Populations of fungi and salmonella were enumerated at Inoculation
time and at 2 d after salmonella Inoculation. Fungi were enumerated on
Plate Count (PC) agar containing 100 ug/ml gentamldn Incubated at 25oc.
Salmonellae were enumerated by a mlcrotlter MPN method (30) using TS broth
with 11 glucose and 50 ug/ml cyclohexlmlde Incubated at 36°C. In tests
with Bacillus as antagonist, salmonella enumeration media also contained
50 ug/ml vancomycln (Sigma, St. Louis, Mo.) and enumeration media for
Bacillus sp. contained 25 ug/ml polymlxln B (Sigma, St. Louis, Mo.).
incubation of Inoculated compost vials was at 36<>c. Control vials
consisted of salmonellae only, fungus only, or Bacillus sp. only.
In the second series of tests the total compost mlcroflora was
considered In terms of subcomponents with minimal selection of Individual
Isolates as test antagonists. This strategy was an attempt to avoid the
potential pitfalls of arbitrarily selecting Isolates that would be
relatively Ineffective antagonists. The complex mlcroblal population
present In compost at 36°C (the Incubation temperature of compost at
which consistent suppression of salmonella has occurred) was considered 1n
terns of the following determinate components: thermophlllc spore-forming
and nonspore-forming bacteria, mesoph111c spore-forming and
nonspore-forming bacteria, fungi, actlnomycetes, protozoa, and nematodes.
Two additional groups were selected, enterobacterla and nonenterobacterlal
gram negatives. The separate components were obtained by using
temperature, particle-size separation, and selective media culture to
prepare composts that contained only the subcomponents desired.
Compost Inoculation tests were designated as cases and they differed 1n
terms of the microbes present. Compost Inoculation tests were designed to
relate salmonella growth to the growth state of the saprophytlc population,
I.e., Inactive or active, and the temperature zone from which the compost
was obtained. Case I was compost obtained from a 70<>c zone 1n a 21 d
aerated pile and was expected to contain Initially Inactive thermophlles,
I.e., spores of thermophlllc actlnomycetes and Bacillus spp. that germinate
27
-------
and grow during the Incubation period. Case IA was the same as case I but
was incubated at 55oc for 1 week prior to Inoculation with salmonellae
and was expected to contain active thermophlles at the time salmonellae
were Introduced. Case II was compost from an adlabatlc composter (31)
operated at 55<>c for 25 d, and It was expected to contain active
thermophlles at the time of salmonellae Inoculation and mesophlles that
survived the previous 55oc and attained active growth during the
Incubation period. Case III was ambient compost, 60 d cured, and contained
all components of all classes of microbes represented In active state at
the time of salmonellae Inoculation.
Case IV was compost recolonlzed with microbes present 1n three
different filter-sized fractions of liquid suspensions of compost. Filter
sizing facilitated separation of different groups of microbes. Several
fractlonatlon procedures (32) were evaluated by assaying for the presence
and type of microbes present 1n the sequentially obtained supernatants and
filtrates. The simplest and most discriminating procedures are described
below. Fraction 1 contained bacteria and a Nocardla sp., fraction 2
contained bacteria and actlnomycetes, fraction 3 contained bacteria,
actlnomycetes, fungi, and any other microbes present 1n mature
ambient-temperature compost.
In the first compost Inoculation trials, fractions were obtained by the
following procedure: Ten g of screened (2 mm), cured, ambient compost were
suspended 1n 90 ml 0.05 M phosphate-buffered water pH 7.2 (PBW) and blended
3 times for 30 sec each at 4<>C; the blended material was diluted to 500
ml with PBW, and centrffuged 15 m1n at 2000 X g at 4°c. The sediment was
resuspended In 50 ml PBW, and 0.01 ml of the suspension was Inoculated Into
each of 2 vials for fraction 3. Supernatant was filtered through a 10 urn
membrane filter (MIlHpore), diluted 1:10 1n PBW and 1.0 ml used as
fraction 2 Inoculum. The remaining undiluted 10 urn filtrate was passed
through a series of sterile membrane filters (Nuclepore) with pore sizes
2.0, 1.0, and 0.8 urn with the final filtrate diluted 1:10 1n PBW serving as
Inoculum for fraction 1, at the rate of 1.00 ml/v1al.
Subsamples of all fractions were plated for bacteria, actlnomycetes and
fungi to determine Inoculum tlter and quality. Bacteria and actlnomycetes
were enumerated on PC agar Incubated at 36<>c and 55°C for mesophlles
and thermophlles respectively. Fungal colonies were counted on oxgall
gentamlcln agar Incubated at 36oc. Colonies of gram-negative bacteria
were counted on gram negative selection (GNS) agar Incubated at 36<>c
containing the following 1n g/Hter of distilled water: peptone, 5.0;
glucose, 1.0; K2HPOA. 0.03; cydohex1m1de 0.1; and agar, 20.0. For
tests with each of the fractions, compost was Incubated with the respective
Inocula at 36oc for 7 d prior to Inoculation with salmonellae (S.
typhlmurlum ATCC 14028), at which time populations of the expected groups
of bacteria, actlnomycetes and fungi were enumerated as described above.
Populations of all expected groups were also determined by the same methods
at the end of the 7 d salmonella incubation. Salmonella Inoculum was
prepared and populations were enumerated by either a quantitative MPN
procedure or by a qualitative procedure previously described (26).
Control* for each different trial consisted of vials Inoculated with
28
-------
salmonellae only and vials not Inoculated with salmonellae. Counts
reported for wesophlllc and gram-negative bacteria were from the
unlnoculated vials.
Results from the first tests of case IV suggested the need for slight
changes In the testing of the effect of the populations 1n the fractions on
salmonella regrowth. Thus, 1n the retestlng of case IV, fractions were
prepared as described above except that fraction 1 Inoculum for each vial
of compost consisted of 0.01 ml of a filtrate passing a 0.6 urn sterile
filter (Nuclepore). Fraction 2 was 0.01 ml of filtrate passing a 2.0 urn
filter (Nuclepore). Further, one set of fraction 2 vials was supplemented
with fungal Inoculum (0.01 ml) which consisted of equal parts of conldlal
suspensions prepared 1n PBW of A. fumlgatus. H. grlsea. Paecllomyces sp.,
and Penlcinium sp. No fungal Tnoculum was aulded to a second set of
fraction 2 vials. All Inocula were stirred Into compost after addition.
Controls for these tests were the same type as those for the first case IV
tests.
Case V was sterile Irradiated compost recolonlzed by a mixture of
compost-derived conforms prior to Inoculation with salmonellae. Conforms
used were E. coll. Enterobacter hafnlae, Enterobacter sp., and dtrobacter
freund11.
Case VI was sterile Irradiated compost recolonlzed by a mixture of
compost-derived nonenterobacterlal Isolates, mostly pseudomonads. Bacteria
used were Identified as Pseudomonas aeruglnosa. £s. veslcularls, and a
pseudomonad-Hke sp.
RESULTS
Plate assays for antagonists
Several hundred microorganisms wero Isolated from compost and from this
collection 126 were selected as representative of the range of
morphologically and taxonomlcally different microbes present. From the
collection of 126, 23 were bacteria, 61 were actlnomycetes, and 42 were
fungi. The Identities of the organisms tested are listed 1n Table 9 1n so
far as Identifications were made. None of the actlnomycetes or bacteria
Inhibited salmonellae 1n either the Individual Isolate assays or 1n the
dilution plate overlays. In contrast, six fungal Isolates Inhibited
salmonellae In Initial tests, and three showed continued strong Inhibition
upon retestlng. The latter three were used 1n compost regrowth assays.
Compost assays with Individual antagonists
A. fumlgatus and Penlcl 11-1 urn 1096-15 did not suppress salmonella growth
when grown prior to or simultaneously with salmonellae 1n gamma-Irradiated
compost. However, a weak amount of Inhibition occurred 1n tests with
Penlcllllum 1098-11 (Table 10). The Bacillus-Inoculated composts, the
previously Inoculated as well as the simultaneously Inoculated, suppressed
salmonella regrowth about 10 times Us uncontrolled level, I.e., about the
29
-------
TABLE 9. MICROOGANISMS ISOLATED FROM COMPOST AND TESTED FOR
ANTAGONISM TOWARD SALMONELLAE IN AGAR PLATE ASSAYS*
Actlnomycetes
Streptomyces spp., St. eurythennus, St. grlseus. St. thermofuscus, St. thermo-
Vlo1aceus7 St. thermovulgarls
Micropoiyspora fa?hT
Nocardlaspl
Tnermoac'tlnomyces vulgarls
_. Bacteria
AT call genes facealls
BacniusTpp.
Cltrobacter freundjl
fenterobacter sp.. E. cloacae, E. hafnla
Escher1chlT"co11 ""
navobacterlum sp.
Mlcrococcus sp.
Pseudomonas spp., Ps. aeruglnosa. Ps. cepacta, Ps. veslcularis
Serratla sp.. !• marcescens
Fungi
Acrophlalophora sp. 220A-2
Aspergl 11 us"Tum1 gatus 267A, 335A
A. nlger. f098-6. 1098-11
~K. nldulans 850-13
foperglllus sp. 256-4, 856-4,
1098-13, 1)15-8,
1115-11, 1256, 1641-45
Chaetomlum thermophUe 856-3
Chaetomlum
sp. 856-3
pur pure us
1 096-1 2
Iplcoccum nlgrum 582-67
Humlcola grisea 541 ( B ) 1
H. Insolens 748
F. lanuglnosa 619-2, 1160-3
ffalbranchea pulchella 557
paeci lotnyces sp.
Papulaspora sp. 1641-7
Penlcimum dupontH 1332-13,
1098-9
P. ochraceum 1098-7
1083, TenlcllHum spp. 1083-13,
1096-13, 106-15, 1098-4,
1098-10, 1098-11,
1115-13, 1641-5, 1641-6
Phanerochaete chrysosporlum 1002
Plectosphaerella sp. 1159-n"
Rhlzopus 1097-13
Scopuiarlppsls brevlcaulls 1096-2
Sporotrlchum thermophn 1s "582-77
Toruia therniophiia
Trlchoderma sp. 1098-4
TrichurusTpl rails 1115-13
* Isolate numbers are those used 1n our laboratory.
same as Pcn1c1111um 1098-11 (Table 10). However with Bacillus as a
competitor, salmonella numbers Increased by a factor of 10*. Thus, none
of the Isolate* that showed antagonism on agar plate assay were very
effective antagonists 1n compost.
30
-------
TABLE 10. GROWTH OF SALMONELLAE IN GAMMA-IRRADIATED COMPOST
COLONIZED BY MICROBIAL ANTAGONISTS
Name
A. fumljatus
Pen1c11l1um
Pen1c1ll1um
1098-11
Bacillus
Antagonist
inoculation
Status |
Prior
Simultaneous
None
Prior
Simultaneous
None
Prior
Simultaneous
None
Prior
Simultaneous
None
Salmonellae*
TO
5.51
5.53
5.52
5.26
d
5.32
4.87
d
5.07
2.17
d
> on^s g
6. 93+. 17
6.877.25
or
7. 52+. 19
7.317.21
d
7. 02+. 16
7.137.09
7.3S+.67
5.407.18
LogioMPN/g
TO T2
6.15
6.15
6.15
6.86
6.86
6.86
5.04
4.00
4.00
3.13
3.34
2.64
S.99+.23
9.42+. 15
9.067.13
9. 03+. 07
9.467.09
9.027.10
6. 60+. 13
6.01+.26
7. 19+;. 35
6.52+. 20
6. 56+. 46
7.647.05
* Salmonella counts are expressed 1n Logio Most Probable Number per gram
dry weight (g) of compost and are t.ie means of two tests. Inocula counts
are shown under TO and counts after 2 days are shown under T2.
f Values are the means of two tests. Counts at the time of Initial
Inoculation are shown under the TO column and counts after 2 days
Incubation with salmonellae are shown under the T2 column.
t Refers to the sequence of Inoculation. Prior means compost was Inoculated
with antagonist and Incubated 7 days before Inoculation with salmonellae.
# d Indicates less than detectable, I.e., 100 colon1es/g.
Compost assays with mixtures of antagonists
In general, the amount of salmonellae growth depended on the types of
microbes present 1n compost. Results of cases I-III described above are
shown 1n Table 11. For case I, I.e., compost obtained from the 70&C zone
of a 21-d aerated compost pile and Incubated at 36<>C with salmonellae,
salmonellae grew to levels comparable to those seen 1n Inoculated sterile
compost. Populations of actlnomycetes, bacteria, and fungi were Initially
low to absent 1n the first trial tr.d high In the second trial.
ThermophlHc and mesoph111c populations of actlnomycetes and bacteria were
as great as or greater than those of salmonellae after Incubation at 36oc
for 7 d. Thus, case I was the first example of raw compost that did not
suppress salmonella growtn. Results from the second test of case I were
31
-------
TABLE 11. SALMONELLA POPULATION CHANGES AS RELATED TO
COMPOST-TEMPERATURE-ASSOCIATED MICROFLORA *
Case
I
I
IA
IA
II
II
III
Tr1 al Day
Assaye
1 0
7
2 0
7
1 0
7
2 0
7
1 0
7
2 0
7
1 0
7
Compost-Temperature-Associated
T Mlcroflora {Logio Colon1es/g)
id
TA J
4.
7.
7.
7.
9.
7.
34
43
25
38
43
09
7.25
8.59
9.
7.
7.
7.
73
08
34
68
7.76
8.17
TB
3.47
7.84
6.01
8.22
8.52
9.29
7.78
8.04
8.95
9.32
7.69
8.15
6.90
7.96
MA
ND 1
ND
5.67
7.64
6.45
5.90
7.22
8.85
5.50
6.68
8.54
9.40
8.79
9.28
MB
3.47
8.53
6.67
8.98
8.08
9.28
8.08
9.99
7.69
9.09
9.07
9.89
9.23
9.43
GN
0
0
0
0
ND
ND
6.90
8.42
ND
7.80
8.00
8.82
6.93
7.16
F
0
1.5
0
0
7.43
6.41
2.60
4.57
6.29
4.98
6.24
7.67
4.70
4.82
Salmonellae
Log10MPN/g
1.
7.
2.
6.
2.
4.
2.
3.
1.
3.
2.
2.
60
98 +
96
33 +
10
99 +
66
61 +
88
14±
66
80 +
.06
.37
.01
.02
1.2
.83
2.66
-0.52
* Values are means + standard error of the mean expressed as Logio
for duplicate vlaTs, except for salmonellae 0 days, which were
values calculated from optical density measurements and a standard
curve. 0 Indicates no colonies were observed.
t 0 refers to the time of salmonella Inoculation and 7 Is 7 days later.
i TA and TB: thermophlUc actlnomycetes and bacteria; MA and MB:
mesophlllc actlnomycetes and bacteria; GN: gram-negative bacteria;
F: fungi.
t Not determined.
similar to those from the first. Results from the two tests of case I and
other cases are reported separately because the composts used were from
different collections and thereby probably differed at least qualitatively
In their mlcroblal flora and nutrients.
For case IA. I.e., compost from the same 70°C zones as case I but
Incubated at 55°C for 7 d prior to Inoculation with salmonellae,
substantial suppression of salmonella regrowth occurred. The average
32
-------
salmonella population was about 105/g compared with about 108/g for the
same compost containing a more limited mlcroblal flora, I.e., case I.
Results from the second test of case IA also were similar to those from the
first test. Notable differences between cases I and IA are: 1) the
amounts of the different microbes present at the time of salmonella
Inoculation, and 2) the presence of gram-negative bacteria and of fungi In
Case IA.
Results for case II, I.e., compost from a 55°c adlabatlc composter,
were comparable to those from case IA 1n terms of salmonella suppression
and population levels of other microbes. The notable difference between
composts of cases IA and II Is that the former was previously heated to
temperatures capable of killing many spore-forming fungi and many different
kinds of bacteria. In contrast, the maximum temperature of case II compost
was about 55<>c and therefore less destructive than the 70°C
temperatures of case IA compost. However, numbers of different groups of
microbes, as shown In Table 11, did not differ markedly and the relative
amounts of suppression of salmonellae were similar for cases IA and II.
Results for case III, 60 d cured, screened, ambient compost, were
comparable to those from other recent tests by the same procedures with 30
different composts (26). Salmonellae were not only Inhibited but also
reduced to levels below the Inoculation amounts. Case III was not repeated
again because the results confirmed other recent results for salmonellae
(26).
For case IV, I.e., recolonlzatlon of compost by groups of microbes
present 1n different size-selective fractions of compost extract, results
are shown 1n Table 12. Filtration failed to produce a fraction completely
dominated by bacteria and free of actlnomycetes. No attempts were made to
segregate nematodes and protozoa from the other mlcroflora present In
fraction 3. Multiple flltratlons through 0.8 urn and 0.6 urn filters did not
reduce the actlnomycetes In fraction 1. The fraction 1 microbes, which
Included no fungi and Initially undetectable levels of thermophlUc
actlnomycetes and a single type of mesophlUc actlnomycete, Nocardla.
substantially Inhibited salmonella growth 1n comparison with the compost
Inoculated with salmonellae only. However, fraction 2 microbes, with and
without fungi, but with Initially high numbers of thermophlUc
actlnomycetes and a wide diversity of mesophlllc actlnomycetes, Including
Streptomyces thermofuscus and SJt. thendovulgarls. reduced the salmonella
population to levels below those of the inocula. Ultimately, only fraction
3 microbes, which Included the total complement of microbes found In
compost, were able to reduce salmonellae to undetectable levels comparable
to those encountered In Incubations of salmonellae with ambient, cured
compost. Case IV showed that fungi have a relatively minor role 1n the
suppression. The results with fungi supported the results from the
previous series of compost Inoculation tests with A. fumlgatus and
PenlcllHa (Table 10).
Comparison of results for cases I, IA, and II Indicated that the
presence of either fungi, gram-negative bacteria or Initially high levels
of thermophlUc and/or mesophlllc actlnomycetes might lead to Increased
suppression of salmonellae. The relatively minor effect that the Initial
33
-------
TABLE 12. SALMONELLA GROWTH IN COMPOST RELATIVE TO POPULATIONS
OF THERMOPHILIC AND MESOPHILIC BACTERIA, ACTINOMYCETES,
AND FUNGI: CASE IV
Compost Day
Extract Assayed
Inoculum
Trial 1
Fraction 1
(0.8 urn)
Fraction 2
(10 urn)
Fraction 3f
(sediment)
None
0
7
0
7
0
7
0
7
Compost-Tempera!
Mlcroflora (Logir
TA*
0
5.5
7.4
7.9
7.1
7.3
0
0
TB
7.6
7.9
7.1
7.7
7.6
7.6
0
0
MA
7.4
7.8
7.9
8.7
7.8
8.7
0
0
:ure-Assoc1at
) Colon1es/g)
MB F
8.4
8.7
8.4
8.8
8.6
8.9
0
0
0
0
0
0
3.6
5.7
0
0
ed
GN
2.0
4.5
6.9
6.7
6.5
6.7
0
0
Salmonellae
(LogioMPN/g)
2.20
4.74 +
2.20
1.76 +
2.20
-0.52
2.20
6.74 +
.97
1.3
.65
Trial 2
Fraction 2
(2.0 urn
filtrate)
No Fungi
With Fungi
Fraction 3 f
(sediment)
None
0
7
0
7
0
7
0
7
6.2
6.6
6.0
6.6
7.8
7.6
0
0
6.8
7.5
7.1
7.8
7.8
7.8
0
0
7.3
7.6
6.8
7.3
8.7
8.6
0
0
9.6
9.8
9.6
9.8
9.5
9.5
0
0
0
0
6.4
7.1
3.0
2.0
0
0
7.8
7.8
7.8
7.8
8.2
7.5
0
0
1.17
0.06 i
1.17
0.28 +_
1.17
-0.52
1.17
6.57 +.
.30
.23
.53
* Abbreviations are the same as those for Table 11.
t Not replicated.
levels of thermophlUc actlnomycetes or mesophlllc bacteria (determination
Includes gram-positive and gram-negative, but gram-negatives were 100/g) had
on salmonellae was shown by the results for the two trials of case I.
However, the relatively suppresslve effect that mesophlllc actlnomycetes and
gram-negative bacteria had was shown by comparison of case I with case IA.
Major amounts of suppression were associated with high levels of gram-negative
bacteria as was shown by comparison of results from trials 1 and 2, fraction
2, case IV.
34
-------
Conforms and gram-negative bacteria as antagonists to salmonellae
Results of cases V and VI shown In Table 13 Indicated that the
collfonns were more active In suppression of salmonellae than were the
pseudomonads. The latter were the most abundant types of noncollform
gram-negative bacteria obtained from the screened, cured, compost. The
collform mixture used In the Inoculation tests Included E. col 1,
Enterobacter hafnlae, Enterobacter sp., and Cltrobacter TreUndTl. Conform
mixture inoculated
TABLE 13. SUPPRESSION OF SALMONELLAE REGROWTH AS RELATED TO GROWTH
OF COLIFORMS AND PSEUDOMONADS: CASES V AND VI
Antagonist Inocu-
Mlxture latlon
Status
Conform Simul-
taneous
Conform Prior |
Pseudo- Prior
monads
None added
dSalmonellae*
LogioMPN/g Assay
Day t
4.6 + .09 0
7
-0.61 + .60 # 0
7
1.41 + .97 0
7
4.9 + .45 0
7
Antagonists
Logio colonles/g
4.3 +
8.5 T
7.9 +
7.7 +
8.3 +
8.4 +
none
.18
.03
.25
.18
.23
.10
* dSalmonellae 1s the difference between the day 0 and day 7 counts. Values
are the means + standard error of the mean for four tests with conforms
and two tests with pseudomonads, with 2 or 4 vials per test.
t 0 1s the time of salmonellae Inoculation and 7 1s 7 days later.
t Prior means that the antagonist was Inoculated and Incubated with
gamma-Irradiated compost 7 days prior to Inoculation with salmonellae;
simultaneous means that the antagonist and salmonellae were Inoculated on
the same day.
# Significantly different from
p • .10, df » 10, Studentjs t
the mean of the pseudomonad treatment at
test.
35
-------
simultaneously with salmonellae did not suppress salmonella growth 1n the
sterile compost (Table 13). In contrast, conforms Inoculated Into compost
7 d prior to the salmonella Inoculation suppressed salmonella growth strongly
(Table 13).
Tests with the gram-negative bacteria antagonists (noncollforms) Included
Pseudomonas aeruglnosa. Ps_. veslcularls. and a pseudomonad-llke sp. No tests
were maae or simultaneous Inoculation because of the lack of Inhibition 1n
the previous tests of such with the Bacillus sp. and with conforms. The
gram-negative antagonist mixture, Inoculated 7 d prior to salmonella, did not
suppress salmonellae to the extent that the conforms did (Table 13).
DISCUSSION
Several reports provide support for the concept that composted sewage
sludge suppresses the growth of salmonellae (19, 25, 26, 33, 34). These
reports suggest that the suppression Is mlcroblally mediated. In.our effort
to understand the mlcroblal nature of the suppression, It became evident that
composts from different temperature zones 1n a compost pile differ In their
suppresslve characteristics depending on the types and relative amounts of
different microorganisms present. Active and complete suppression was
obtained only 1n compost thoroughly colonized by actively metabolizing
members of all major groups of microbes, I.e., bacteria, actlnomycetes, and
fungi. Such compost Is repeatedly encountered 1n the cured compost pile that
1s not forcibly aerated. Were It possible that a self-heated patch of
compost would be Inoculated with salmonellae Inadvertently, 1t seems that
salmonellae could grow for a limited time. Such a situation represents a
very localized occurrence. It requires that the superheated compost not be
mixed with ambient or thermophH1c compost, which could contain a more
diverse mlcroflora. The occurrence of localized sites of relnoculated but
unmixed compost would be limited because of the nature of the screening and
compost-handling processes. However, mlcroblal diversity per se would be
Insufficient, since 1t appears that certain groups of microbes are more
competitive with salmonellae than others. The suppression produced by
bacteria, especially conforms and then noncollform gram-negatives, tended to
dominate that produced by actlnomycetes. However, by themselves mixtures of
conforms or compost-.jlonlzlng pseudomonads could not suppress salmonella
growth to the extent that the total complement of compost microbes could.
Fungi appeared to have a negligible Influence on suppression. Our data
Indicated that thermophlllc and especially mesophlllc actlnomycetes
supplemented the activity of the conforms and other gram-negative bacteria
In the suppression. Accounts of antagonism toward salmonellae by
thermophlllc actlnomycetes are known (33, 34, 35) and Involve some of the
same actlnomycetes found 1n the compost used for this study, I.e.,
St. thermofuscus, St. thermovlolaceus, St. thermovulgarls, although 1n our
"agar plate assays antagonism between actTnomycetes and salmonellae was not
observed. The lack of antagonism on plate assays 1s a problem commonly
encountered by soil m1crob1olog1sts 1n their Investigation of mlcroblal
antagonisms Important 1n biological control. Plate assays have come to be
regarded as almost useless 1n assessing blocontrol phenomena and mlcroblal
competition. Therefore, we place more Importance on the compost
assay-Inoculation test results.
36
-------
Given the diversity of the mlcroblal population of ambient compost and
the response of salmonellae to this Indigenous population, we conclude that
salmonella regrowth would be negligible for cured compost. Because total
inhibition 1s not related to the activities of single groups of microbes,
there presently 1s no practical way of assessing the Inhibitory character of
the compost by assaying for the presence of specific competitive microbes.
Although our data support a major role for gram-negative bacteria, especially
conforms, 1n the suppression, the use of this group of microbes even coupled
with determinations of populations of actlnomycetes could result 1n false
assessments. Therefore, we do not recommend the use of any of the groups or
combinations of groups of microbes as Indicators to assure a
salmonella-suppresslve compost. With additional study, specific
recommendations about precise numbers or thresholds of each mlcroblal group
needed to assure suppresslveness 1n compost might become available.
In this work we have not specifically addressed the role of protozoans.
We can only say that they may play a role 1n the suppression produced by a
mixture of microorganisms that our data Indicate 1s dominated by the conform
bacteria.
Another approach for assessing the suppresslveness of compost might be
based on an Index of the mlcroblal activity of the compost. The
dehydrogenase assay might be used to assess the general level of mlcroblal
activity but techniques would need to be devised to eliminate Interference
from the colored compounds that are present 1n compost extracts used 1n the
assay (L. Slkora, pers. comm.). However, measurements of mlcroblal
respiration on screened, cured compost could be superfluous. Screening
compost results 1n the mixing of various mlcrosltes and thereby the
relnoculatlon of previously pasteurized mlcrosltes with a multitude of
compost microbes. With curing for 30 d, any sites would have ample time for
Intense competition which would lead to Inhibition and kill of salmonellae.
Only sites from the very hot cores of a 21-d pile could be expected to
support any substantive regrowth of accidentally Introduced salmonellae
during disassembly of a pile. However, hot spots, even If Inoculated with
salmonellae, would be Inoculated with other compost microbes and would be
mixed during screening. During storage of screened compost any salmonellae
that might have been Introduced would have to compete with the Inhibitory
flora found In the ambient compost. Our data and those of others (26)
Indicate that successful regrowth of salmonellae In competition with the
mixed compost flora present 1n ambient compost 1s extremely unlikely.
37
-------
SECTION 7
SALMONELLA REGROWTH IN COMPOST AS INFLUENCED BY SUBSTRATE
INTRODUCTION
Composting Is capable of destroying the primary pathogenic organism?
Including salmonellae that may be present 1n sewage sludge (4). However,
there have been anecdotal reports of salmonellae In marketed composts. If
these reports are valid, possible explanations Include growth of
organisms: 1) that may have survived composting because of failure to
obtain a lethal time-by-temperature regime; or 2) that may have been
Inoculated Into the compost by Infected birds or other animals, or by
equipment contaminated with salmonella-containing sewage sludge.
Available data Indicate that salmonellae do not grow extensively when
Inoculated Into compost (18), and sewage sludge unless they have been
sterilized. Further, repopulatlon of the sterilized sludge with conforms
will Inhibit the growth of salmonellae (20). A study of compost from a
single site Indicated that a water content of 20% or greater and a
carbon/nitrogen ratio of greater than 15:1 were necessary to support
growth. Also, as with sludge, the native flora Inhibited growth (19). But
data from studies with sewage sludge and a compost from a single site can
not be considered definitive. To understand better the regrowth potential
of salmonellae, knowledge of the substrates Involved and the nature of the
Inhibiting mlcroflora 1s needed.
In this study, the kinetics of salmonella growth 1n suspensions and
extracts of radiation-sterilized compost were studied to determine the
number and relative amounts of substrates utilized.
MATERIALS AND METHODS
Compost
The compost collection, processing, and storage methods have been
previously described (26). Briefly, sewage sludges composted by the
Beltsvllle Aerated-P1le Method (36), were collected from storage piles
containing finished composts (composts that had been through the complete
process and were ready for utilization), and shipped to our laboratory 1n
sealed five-gallon containers. Composts were Identified by a composting
site number. Woodchlps were removed by sieving with a 0.6 cm screen. The
remaining compost was sieved through a 0.147-mm pore-sized screen and
stored In plastic bags for periods of less than a week at 4 C . Storage
beyond a week was at -20 C. For utilization as substrate, the compost was
38
-------
sterilized by Irradiation (3 megarads. 60c0).
Compost extract
Compost extracts were prepared by shaking overnight weighed amounts of
Irradiated compost In 100 ml of a minimal medium (37) modified by reducing
the NH4C1 concentration from 1.0 to 0.5 g and omission of the glucose.
The mixtures were then centrlfuged (19,600 x g, 20 m1n) and decanted to
obtain a partlculate-free extract. Initially filtration through a Gelrnan
0.4' urn pore-sized, 45 mm d1a. membrane filter was used to remove the
partlculate material, but results from the control {membrane-filtered
medium without compost) showed that the filter was contributing significant
amounts of substrate.
Experimental Inoculum
Salmonella typhlmurlum ATCC 14028 was used for the experiments. The
Inoculum was prepared by Introducing a loopful of the organism as grown
overnight on a nutrient agar slant Into the minimal medium described
above. At first, a filter-sterilized glucose solution was used as the
carbon source (final concentration 1n the medium, 2.0 g/Mter). However,
use of the glucose-adapted cells for the Inoculum resulted 1n an
appreciable lag phase upon Inoculation Into compost extract (compare F1g. 4
to F1g. 5). Therefore, an extract of compost was substituted for the
glucose as a carbon source for growth of the Inoculum to reduce this lag
phase. After overnight growth, the culture was diluted to produce a zero
time concentration 1n the experimental flasks of between 100 and 1,000
colony-forming units (CFU)/m1.
Experimental Procedure
In the first study, weighed amounts of Irradiated compost were added to
100 ml of the minimal medium 1n a 250 ml screw-capped flask. After adding
the salmonella Inoculum, the flask was placed 1n a water bath (36 +_ 1°C
with shaking) and sampled with time. In all other studies, compost extracts
were Inoculated, Incubated, and sampled as above. The specific amounts of
compost used In each study will be given 1n the results section. The
amounts used ranged from 0.025 to 20.0 mg/ml. The sample volume taken was
either 0.1 or 1.1 ml depending upon the dilution needed for counting. For
counting, ten-fold serial dilutions were made and plated by spreading on
xylose lyslne brilliant green (XLBG) agar. Flasks were sampled at either
1- or 2-hour Intervals for eight hours (1st study, 5 h) to determine growth
rates, and a sampling at 24 h was taken (1st study, not done) to determine
total growth potential of the substrate.
RESULTS
Influence of quantity of compost added on growth rate
If growth proceeds by first-order kinetics Indicating a readily .
available single substrate, total growth can be expected to be related to
substrate concentration, but the rate of growth should not be except at
39
-------
very low concentrations (38). In the first study, 0. 25, and 250 mg of
Irradiated compost were added to flasks containing 100 ml of the minimal
medium lacking the addition of a carbon source. Each treatment was
replicated. After Inoculation with salmonellae grown overnight In the
minimal medium containing glucose, the flasks were Incubated with shaking
at 37 + loc and sampled at hourly Intervals for five hours. As shown 1n
Fig. 47 growth occurred with the addition of compost and the rate of growth
Increased with the Increase 1n the amount of compost added to the flasks.
Five hours was not enough time to achieve maximum growth, and the growth
was not Indisputably first order.
4,0
3,5
3,0
2,5
2,50 MG/ML
0,0 MG/ML
JL
012345
HOURS
Fig. 4. Growth of salmonellae 1n a mineral-salts medium as Influenced by
the amount of compost 16175 added.
40
-------
variations In growth rate that occurred with different amounts of
compost suggested that substrate concentration was controlling the growth
rate. However. It was possible that the rate of growth was controlled not
X! ^JSL^*^™1?1^0' substrate furnished In proportion with the addition
of compost but also by the rate of diffusion of the substrate from the
•atrix of the compost particles, and/or the rate of solublllzatlon of the
substrate. To test for these two possibilities, a series of time extracts
of the compost was prepared by shaking 500 mg of compost In 100 ml of the
carbon-source-free minimal medium and removing the particulate material by
centrlfugatlon. Shaking times for extraction were 4. 8. 16, and 24 h. The
inoculum was grown 1n compost extract to reduce the lag phase.
The plots of the data (Fig. 5) appeared to show first-order growth
because the effect of change in substrate concentration that should produce
second-order kinetics was too small to be evident. The plots also showed a
reduction In the lag phase from use of the compost-condltloneu Inoculum
(Fig. 5). In addition there wasj^s^ljht trend of Increased growth rate
6 •
5 •
12
HOURS
16
20
2«
Fig. 5,
Growth of salmonellae on extracts as Influenced by
extraction time (4h, right half filled; 8h, filled;
16h, open; and 24h, left half filled) of compost 06175
using a mineral-salts medium.
41
-------
and total population size with Increase 1n extraction time from 4 through
16 n. Against this trend, the 24-h extract supported a slower growth rate
than the other extracts and a slightly lower total population. These
results were difficult to rationalize solely on the basis of release of
substrate with time through diffusion and/or slow solub111zat1on. A
regression analysis of the data (Table 14) showed that the 16- and 24-h
compost extract slopes were not the same as the 4-h slope (p - 0.01 and
0.05 respectively). The slope of the 8-h extract was not significantly
different from the 4-h slope (p > 0.05). We concluded that these
TABLE 14. REGRESSION ANALYSES COMPARING GROWTH RATE CONSTANTS (kq)
FOR GROWTH OF SALMONELLAE IN EXTRACTS OF COMPOST 16175 AS
INFLUENCED BY EXTRACTION TIME
Extraction ka
Time
(h)
4
8
16
24
*
*«
(h-1)
1.381
1.372
1.493
1.211
Significant at 0.05%
Significant at 0.01%
In y
Inter-
cept
3.545
3.752
3.393
3.395
t for
H0:
Paramccer » 0
47.23
47.02
49.93
44.06
PR> [tj
0.6228
0.8342
0.0129*
0.0054**
differences, although significant, could be tolerated 1n kinetic studies
for determining the effect of amount of compost added 1f the extraction
period was held within one hour'.s deviation from 16 h.
Although the data cannot be used to determine the mode of release of
substrate, It does appear that the lack of first-order kinetics 1n the
first study could have been the result of some slow release process.
Therefore, we decided to use extracts Instead of compost for the rest of
the studies.
Glucose equivalent
Tests were conducted to determine the amount of glucose used by the
organisms. The results were used to calculate the glucose equivalent of
the substrate In the compost. Glucose was added to flasks of minimal
medium to form a glucose concentration series as follows: 0.004, 0.40,
4.00, 40.0, and 400.0 ug/ml. The growth rates for all concentrations were
not first order (not shown), but after 48 h, the total growth was
correlated with the amount of glucose used (F1g. 6). Each CFU required
4.82 x 10'8 mg of glucose. Lack of first-order growth on glucose could
be explained by the fact that bacteria require appreciable C02 tension
for first-order growth on simple carbohydrates, but the CO? requirement
can be ameliorated by sucdnate, purlnes, and pyr1m1d1nes (39).
42
-------
-5
-3 -2 -1
LOG GLUCOSE MG/ML
F1g. 6. Correlation of the maximum amount of salmonella growth with amount
of glucose added to a mineral-salts medium.
Influence of water soluble compost-substrate concentration on growth
To determine the effect of concentration, extracts were made from
different quantities of the compost #6175 and Inoculated with the test
strain grown on extract of compost. The quantities of compost extracted
per ml of basal medium were 1.25, 20, 40, 60, and 100 mg. The results
showed only small differences 1n the rates of growth at concentrations of
20 mg/ml and above. Only the lowest concentration (1.25 mg/ml) produced a
large difference 1n growth rate (F1g. 7). Total growth, however, appeared
to be proportional to the amount of compost extracted. Similar studies
with composts 16252 and 16266 were conducted. The results for some of the
data for compost 16252 are shown 1n F1g. 8.
Regression analyses of the data from the three studies showed that the
growth-rate equations were first order (p • 0.01). Also there appeared to
be a relationship between growth rate, total growth, and amount of compost
extracted, but some of the curves were out of place as was that for two of
the 200 mg/ml curves (data pofnts are filled circles) 1n F1g. 8.
43
-------
£
o
100 MG/ML
---o
1.25 MG/Ml
60 MG/flL < V—°
' V-o
6/, L--0
r a '
140 MG/ML
20 MG/ML
8
8
F1g.
21 0 4
HOURS
7. Growth of salmonellae 1n a mineral-salts extract of compost
16175 as Influenced by the amount of compost extracted.
DISCUSSION
With the exception of the results shown 1n F1g. 4, analysis of the
growth-rate data for all three composts showed first-order regressions, but
the rate coefficients were not constant for the lower concentrations. The
dependency of growth rate on concentration was evaluated by plotting the
growth-rate constants against the substrate concentrations for the three
composts (F1g. 9A). The data appeared to describe a single curve
Indicating that the substrate could have been the same or very similar 1n
all three composts. However, there were outlying points that did not seem
to fit Into the general pattern.
The quantity of extractable substrate should have been directly related
to the amount of compost mass unless the compost mass from which the
samples were taken was not uniform or some change 1n the available
substrate had occurred between samplings. If either was true, then maximum
growth might have been a better measure of substrate quantity than compost
mass, and should have been better correlated with the growth-rate constant
(k) than mass was. It also should have compensated for differences 1n
substrate concentrations among the composts. Plotting k versus total
population grown (P) produced fewer outlying data points possibly
44
-------
6
o
o
"^OOJ^G/ML __
200 MG/ML
T-
20 MG/I.
200 MG/ML
1
2.0 MG/ML
— _ _ . ^ _. _ _ x\
- ~— 2 N.r_ _r_ _x
- - - - -*J "~rz.--.-Q
O
8 12
HOURS
24
F1g. 8. Growth of salmonellae 1n a mlneral-satts medium
as Influenced by added amount of extract from
compost 16252 (See Table B-l for data used).
Indicating the Instability of the substrate with time despite refrigeration
and/or Its uneven distribution 1n the refrigerated compost samples when
they were subsampled (F1g. 96).
If the substrates were qualitatively different, then the maximum rates
at which they decompose should not be the same. Linearization of the data
for the three composts would make 1t possible to compare the fit of the
data to a common regression line.
45
-------
7 «
1.6
I "
M
0.8
O.U
B
0 0.1 0.5 1.2 1.6 2.0 2.
-------
produced a larger correlation coefficient for the combined data of the
three composts (Table 15). Both Intercepts and slopes were more similar.
TABLE 15. INTERCEPTS (Pi/km), SLOPES (I/km). AND CORRELATION
COEFFICIENTS (CO OF EQUATIONS USING AMOUNT OF COMPOST
OR TOTAL GROWTH AS THE INDEPENDENT VARIABLE FOR DATA
OF INDIVIDUAL COMPOSTS AND THE COMBINED (COMB)
DATA OF THE THREE COMPOSTS
Compost
No.
#6175
#6263
#6266
COMB
x »
Pi Am
0.883
1.834
7.877
1.132
Amount of
lAm
0.563
0.714
0.568
0.673
Compost
CC
0.996
0.872
0.901
0.907
x =
Pi Am
3.228
6.126
4.796
3.420
Total Growth
lAm
0.571
0.594
0.533
0.589
CC
0.995
0.999
0.999
0.997
Also, the values for the correlation coefficients were Increased for composts
#6252 and #6266 showing that using total growth was more representative of the
amount of the substrate available than was compost mass. Therefore, we
plotted the transformation for the total growth versus k data of F1g. 9B 1n
Fig. 10.
The constant km can be used to compare the efficiency with wh.ch
substrates are utilized (38). According to Ostle (40), the correlation
coefficient can be used to compare regression parameters for goodness of
fit. In Table 16 four variations of a model equation are compared as to
their effect on the magnitude of their resulting correlation coefficients.
Model 1 utilized an Intercept (a) and a regression coefficient (b) for each
of the sets of data for the three composts. Models 2 and 3 eliminated
respectively two regression coefficients and two Intercepts. Model 4
utilized only one regression coefficient and one Intercept. There was a
decrease 1n the value of the correlation coefficient with each decrease 1n
the number of parameters used, but the total decrease from model 1 to model
4 was only 0.0011 (0.1U). This change was small, and dictated on the
basis of parsimony that model 4 was preferred. There was no good reason to
believe that more than one kind of substrate among the composts was needed
to produce these results.
The water-soluble substrate 1n the compost was decomposed 1n accord
with first-order kinetics, but glucose was not. For minimal media with
slnale simple carbon sources, high concentrations of C02 are required to
produce first-order growth (39). Although growth may not be first order,
alven enough time the substrate will be exhausted producing growth
proportional to the amount of substrate added. The C02 requirement can
be met by providing additional metabolites. These results Indicated that
there probably was a water-extractable single energy source 1n each compost
extract producing the observed first-order growth. If It was a simple
compound such as glucose or some other sugar, there must also have
47
-------
F1g. 10. A plot of the population and rate-constant data of F1g. 9B
according to the linear form of Monodjs equation (see F1g. 9
for meaning of symbols).
TABLE 16. DETERMINATION OF THE RELATIVE FIT OF MULTIPLE VERSUS
SINGLE PARAMETERS IN USE OF THE TRANSFORMATION OF MONODlS EQUATION
TO DESCRIBE THE DEPENDENCE OF THE GROWTH-RATE COEFFICIENT ON
MAXIMUM POPULATION AS A MEASURE OF SUBSTRATE CONCENTRATION
Model
1
2
3
4
No.
a
3
3
1
1
of Parameters
b
3
1
3
1
Correlation Coefficient
0.9988
0. 9983
0.9987
0.9977
been metabolites available to compensate for the lack of an adequate C02
concentration.
46
-------
The results of this study showed that It was possible to extract a
water-soluble substrate from compost that would support first-order growth of
!• typhlmurlum. The first-order nature of the kinetics and the high degree of
correlation for the combined data using the linear form of Monodls equation
suggested that there was a single substrate among the composts. The
Identification of this substrate, and the testing for Its presence 1n other
composts might possibly furnish valuable Information as to the factors
Involved 1n the regrowth of salmonellae 1n composts.
49
-------
REFERENCES
1. Foote, C.J. 1895. A bacterlologlc study of oysters, with special
reference to them as a source of typhoid Infection. Med. News
66:320-324.
2. Sllllker, J.H. 1980. Status of Salmonellae - Ten years later. J.
Food Protection 43(4):301-313.
3. Centers for Disease Control. 1983. Human Salmonella Isolates - United
States, 1982. Morbidity and Mortality Weekly Report J32(45):598-600.
4. Burge, W.D., D. Coladcco, and W.N. Cramer. 1981. Criteria for
achieving pathogen destruction during composting. J. Uater Pollut.
Control Fed. 53:1683-1690.
5. American Public Health Association. 1976. Standard methods for the
examination of water and wastewater, 14th ed. American Public Health
Association, Washington, D.C.
6. Galton, M.M., G.K. Morris, and W.T. Martin. 1968. Salmonellae 1n
foods and feeds. Communicable Disease Center, Atlanta, GA.
7. Bordner, R.H., J.A. Winter, and P. Scarplno. 1978. Microbiological
Methods for Monitoring the Environment. EPA-600/8-78-017. 338 pp.
8. Blssonette, G.K., .J.J. Jezeskl, G.A. McFeters, and D.G. Stuart. 1975.
Influence of environmental stress on enumeration of Indicator bacteria
from natural waters. Appl. M1crob1ol. j29(2):186-194.
9. Mossel, D.A.A., G.A. Harrewljn, and C.F.M. Nesselrooy-van Zadelhoff.
1974. Standardization of the selective Inhibitory effect of surface
active compounds used 1n media for the detection of Enterobacterlaceae
In foods and water. Hlth. Lab. Sc1. Jl_:260-267.
10. Edel, W. and E.H. Kampelmacher. 1973. Comparative studies on the
Isolation of "sublethally Injured" Salmonellae 1n nine European
laboratories. Bull. W.H.O. 48:167-174.
11. Thomason, B.M., D.J. Dodd, and W.B. Cherry. 1977. Increased recovery
of salmonellae from environmental samples enriched with buffered
peptone water. Appl. Environ. Mlcroblol. 34(3):270-273.
12. Taylor, W.I. 1965. Isolation of -shlgellae. I. Xylose lyslne agars;
new media for Isolation of enteric pathogens. Amer. J. CUn. Pathol.
44:471-475.
13. Browning, C.H., W. Gllmour, and T.J. Mackle. 1913. Isolation of
ryphold bacilli from faeces by means of brilliant green 1n fluid
medium. Hygiene 21:335-342.
50
-------
14. Moats, W.A., J.A. Klnner, and S.E. Maddox. 1974. Effect of heat on
the antimicrobial activity of brilliant green dye. Appl. Mlcroblol.
27:844-847.
15. Hussong, D., J. M. Damare, R. M. Welner, and R. R. Col well. 1981.
Bacteria associated with false-positive most-probable-number coll form
test results for shellfish and estuaries. Appl. Environ. Mlcroblol.
41.: 35-45.
16. Kaper, J. B., G. S. Sayler, M. M. Ba1d1n1, and R. R. Colwell. 1977.
Ambient- temperature primary nonselectlve enrichment for Isolation of
Salmonella spp. from an estuarlne environment. Appl. Environ.
Micro t)loi7 33:829-835. ~~^ -
17. Geldrelch, E.E. 1972. Water-borne pathogens, p. 207-241. In
R. Mitchell (ed.), Water Pollution Microbiology. John Wiley Ind Sons,
Now York.
18. Brandon, J.R., W.D. Burge, and N.K. Enk1r1. 1977. Inact1vat1on by
Ionizing radiation of Salmonella enter1t1d1s serotype montevldeo grown
1n composted sewage slUJge^ A"ppl. Environ. Mlcroblol. ^3:1011-1012.
19. Russ, C.F. and W.A. Yanko. 1981. Factors affecting salmonellae
repopulatlon 1n composted sludges. Appl. Environ. Mlcroblol.
41^:597-602.
20. Yeager, J.G., and R.L. Ward. 1981. Effects of moisture content on
long-term survival and regrowth of bacteria 1n wastewater sludge.
Appl. Environ. Mlcroblol. 41_:1117-1122.
21. Hussong, D. , N.K. Enk1r1, and W.D. Burge. 1984. Modified agar medium
for detecting environmental salmonellae by the most-probable-number
method. Appl. Environ. Mlcroblol. 48:1026-1030.
22. Holmberg, S.D., J.G. Wells, and M.L. Cohen. 1984. Animal-to-man
transmission of antimicrobial resistant Salmonella; Investigations of
U.S. Outbreaks, 1971-1983. Science 225:833-835.
23. Wlllson, G.B., J.F. Parr, E. Epstein, P.B. Marsh, R.L. Chaney,
D. Coladcco, W.D. Burge, L.J. Slkora, C.F. Tester, and S.B. Hornlck.
1980. Manual for composting sewage sludge by the Beltsvllle
aerated-pile method. Joint EPA-USDA report EPA-600/8-80-022-65 pp.
24. Composting processes to stabilize and disinfect municipal sewage
sludge. Office of Water Program Operations. WH-541. Washington,
D.C. EPA 430/9-81-011, 1981. 38 pp.
51
-------
25. Brandon, J.R. and K.S. Neuhauser. 1978. Moisture effects on
1nact1vat1on and growth of bacteria and fungi In sludge. Publ. No.
SAND 78-1304. Sandla Laboratories, Albuquerque, N.M.
26. Hussong, D., W.D. Burge, and N.K. Enklrl. 1985. Occurrence, growth,
and suppression of salmonella growth 1n composted sewage sludge. Appl.
Environ. Mlcroblol. 50:887-893.
27. Mlllner, P.O., P.B. Marsh, R.B. Snowden, and J.F. Parr. 1977.
Occurrence of Asperglllus fumlgatus during composting of sewag
sludge. Appl. Environ. Mlcroblol. 34:765-772.
28. Cross, T. 1968. ThermophlUc actlnomycetes. J. Appl. Bacterlol.
31_: 36-53.
29. Mlllner, P.O. 1982. ThermophlUc and thermotolerant actlnomycetes 1n
sewage-sludge compost. Devel. Indus. Mlcroblol. 23;61-78.
30. Rowe, R., R. Todd, and J. Walde. 1977. Microtechnique for
most-probable-number analysis. Appl. Environ. Mlcroblol. 33_:675-680.
31. Slkora, L.J., h.A. Ramirez, and T.A. Troeschel. 1983. Laboratory
composter for simulation studies. J. Environ. Qual. ^2:219-224.
32. Faegrl, A., V.L. Torsvlk, and J. Goksoyr. 1977. Bacterial and fungal
activities 1n soil: separation of bacteria and fungi by a rapid
fractionated centrlfugatlon technique. Soil B1ol. Blochem. 2:105-H2.
33. Makawl, A.A.M. 1973. The survival of salmonellae 1n compost prepared
with straw of different plants and sewage sludge. Zentralbl.
Bakterlol. Parasltenkd. Infekt. Hyg., Abt. 2. 128:203-208.
34. Makawl, A.A.M. 1980. The effect of thermophlllc actlnomycetes
Isolated from compost and animal manure on some strains of Salmonella
and Shlgella. Zentralbl. Bakterlol. Parasltenkd. Infekt. Hyg., Abt. 2,
135:12-21.
35. Bhakru, K. and B.N. Johrl. 1980. Antibiotic activity of thermophlllc
actlnomycetes. Curr. Scl. 49:446.
36. Epstein, E., G.B. Wlllson, W.D. Burge, D.C. Mullen, and N.K. Enk1r1.
1976. A forced aeration system for composting wastewater sludge.
J. Water Poll. Control Fed. 48:688-694.
37. Gomez, R.F., A.J. Slnskey, R. Davles, T.P. Labuza. 1973. Minimal
medium recovery of heated Salmonella typh1mur1um LT2. Jour. Gen.
Mlcroblol. 74:267-274.
38. Monod, Jacques. 1949. The growth of bacterial cultures. Ann. Rev.
Mlcroblol. 3:371-394.
52
-------
-------
APPENDIX A
TABLE A-l. PARAMETERS FOR EQUATIONS (LOGin) FOR THE CURVES SUMMING THE
POPULATIONS OF SALMONELLA TYPHIMURIUM AND S. NEWPORT DURING GROWTH AND
UEATH IN TTOMPT35T
F1g.
1
1
1
1
2
2
2
3
3
3
Compost
6175
6175
6175
6175
6243
6240
6238
6243
6240
6238
Water (%)
11
11
45
45
45
45
45
10
6
7
Phase
Growth
Death
Growth
Death
Death
Death
Death
Death
Death
Death
Intercept
5.629
8.700
5.689
6.180
14.83
16.35
14.14
6.55
6.58
6.93
Slope
0.0623
-0.0259
0.1280
-0.0338
-0.2093
-0.3493
-0.1747
-0.4755
-0.5808
-0.5046
54
-------
APPENDIX B
TABLE B-1. DATA FROM THE STUDY OF THE GROWTH
OF SALHONELLAE IN A MINERAL-SALTS MEDIUM AS
INFLUENCED BY ADDED AMOUNT OF EXTRACT FROM
COMPOST 16252.
Experiment
no.
39
39
39
39
39
39
39
39
39
39
39
39
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
41
Compost
mg/ml
200
200
200
200
200
200
200
200
200
200
200
200
2
2
20
20
200
200
2
2
20
20
200
200
2
2
20
20
200
200
2
2
20
20
200
200
2
2
20
20
200
200
Time
h
0
0
2
2
4
4
6
6
8
8
24
24
0
0
0
0
0
0
2
2
2
2
2
2
4
4
4
4
4
4
6
6
6
6
6
6
8
8
8
8
8
8
Salmonellae
log cfu/ml
2.46
2.55
2.83
2.75
3.87
4.25
4.67
5.70
5.40
6.92
7.48
8.12
2.47
2.40
2.39
2.43
2.54
2.52
2.69
2.74
2.78
2.78
3.01
2.99
3.69
3.58
3.84
3.89
4.24
4.32
4.42
4.67
4.95
5.22
5.71
5.96
5.44
5.24
6.05
6.27
7.24
7.30
55
-------
Table b-l (continued)
41 2 24 6.79
41 2 24 6.80
41 20 24 7.54
41 20 24 7.69
41 200 24 8.77
41 200 24 8.74
56
-------
APPENDIX C
TABLE C-l. POPULATION AND RATE DATA FOR SALMONELLA GROWTH IN
EXTRACTS OF THREE COMPOSTS: A-6252, 8=6175. C-6266
A
A
A
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
c
c
c
w
c
w
c
Amount,
mg/ml
100.00
-rr
h-l
1.16
200.00 0.98
200.00
100.00
100.00
50.00
50.00
2.00
2.00
20.00
20.00
200.00
200.00
1.25
2.50
5.00
1.25
2.50
5.00
5.00 1
5.00 1
5.00 1
5.00 1
60.00 1
1.25 1
1.25 1
100.00 1
100.00
20.00
20.00
40.00
40.00
1.61
1.59
.08
.50
.56
.04
3.99
.26
.36
.63
.68
.34
.59
.65
.27
.41
.48
.38
.37
.49
.11
.65
.14
.14
.76
1.83
1.52
1.61
1.59
1.58
100.00 0.97
200.00
200.00
20.00
20.00
2.00
b • W v
2.00
1.77
1.78
1.28
1.41
0.56
0.65
Population,
cf u/ml XI 06
15.
29.
132.
98.
38.
95.
102.
6.
6.
35.
48.
588.
556.
15.
20.
24.
12.
20.
20.
08
89
69
14
25
32
72
20
29
03
66
21
53
31
85
38
66
62
90
12.47
14.
14
18.96
11.
280.
9.
10.
272.
287.
87.
89.
148.
157.
15.
159.
35
96
91
12
61
25
22
80
90
72
58
16
172.27
20.
25.
3.
18
97
07
3.08
Amount/K
86.
203.
124.
62.
93.
33.
32.
1.
2.
15.
14.
122.
119.
0.
1.
3.
0.
1.
3.
3.
3.
3.
4.
36.
1.
1.
56.
54.
13.
12.
25.
25.
102.
112.
112.
15.
14.
3.
3.
19
67
42
83
00
43
14
93
02
92
72
70
37
93
58
03
98
77
37
62
64
35
52
29
10
09
93
65
19
42
19
39
86
75
00
58
14
60
08
Population/I^
12.99
30.
82.
61.
35.
63.
66.
5.
6.
27.
35.
360.
332.
11.
13.
14.
9.
14.
14.
9.
10.
12.
10.
169.
8.
8.
155.
156.
57.
55.
93.
100.
16.
89.
96.
15.
18.
5.
4.
44
54
67
57
72
02
99
37
88
82
87
16
40
15
77
97
61
08
03
30
70
25
92
71
83
19
98
54
78
77
11
02
72
47
71
36
52
76
57
------- |