-------
                                 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

-------