Population Changes in Enteric Bacteria and Other Microorganisms During Aerobic Thermophilic Windrow Compostingl .JhCOB SAVAGE.: THEODORE CHASE, .JR., AND JAMES D. MACSIILLAN Department of Biochemistry and Microbiology, Rutgers University, New Brunswick, Neal Jersey 03903 Received for publication 17 May 1973 Composting of wastes from swine feeding operations was studied. The effects of the frequency of turning the wastes and addition of straw to improve the physical structure were studied to determine the most effective technique to rapidly increase the temperature and, consequently, destroy coliforms and Salmonella. Four different treatments were studied; the results showed that, with addition of 55 (wt/wt) straw and mechanical turning of the compost 20 times per week, the temperature reached 60 C within 3 days and enteric bacteria were destroyed within 14 days. According to Wadleigh (12), animal farm waste production in the United States amounts to about two billion tons annually. Traditional mkthods for utilizing these wastes as fertilizers are not being employed widely, and the vast amounts that are accumulating present serious threats from both environmental and public health standpoints. Solutions to the problems of disposal of animal solid waste require complex environmental contra!. Methods must be uti- lized that do not cause contamination of water, pollution of the atmosphere, or desecration of the land. Burning of waste in open dumps or poorly designed incinerators is a major source of air pollution. Other disposal methods, such as sanitary landfill. aerobic and anaerobic lagoons, and spreading of liquid and solid on soil, may cause odors and chemical and microbial pollu- tion of surface and ground waters. Many of these problems can be avoided by composting; with proper control of this method, wastes are incorporated into the biogeochemical cycle without serious detriment to the ecosys- tem. Composting can be defined as the decom- position of organic wastes under semi-dry conditions by aerobic, thermophilic organisms. The products of composting are carbon dioxide, water. heat, and stable humus-like organic material. Several investigators (2, 5, 10) have con- cluded that since relatively high temperatures are attained during the cornposting process any 1 Paper of the .Joumal Series. New Jersey Agricultural Experiment Station. pathogens present should be destroyed. Re- cently, Wiley and Westerberg (13) selected representatives from four groups of pathogens as indicators and related the temperature of a laboratory composter to survival of&hese orga- nisms. Few studies of a similar nature have been reported on a large-scale field operation. This paper describes the sumiva! of Salmonella, other enteric bacteria. and fecal streptococci during the composting of swine waste in win- drows that were turped mechanically with a composting machine. Changes in the levels of actinomycetes, filamentous fungi, and cel- lulolytic microorganisms occurring during com- posting are also discussed. MATERIALS Aw 3lETHODS Preparation of compost. Swine waste was ob- tained from a hog feeding operation in southern New Jersey. The swine were fed hotel and restaurant garbage, which was cooked with live steam and dumped outdoors in concrete-floored pens. The swine waste-a combination of uneaten garbage, bottles, cans, plastic, paper. bones. other inedible material from the garbage. and swine feces-was scraped up daily with a front-loading tractor and removed. The material was trucked to the compost site and stock- piled over a s-week period until a sufficient amount of waste accumulated for construction of four 40-ft (about 12.2 m) windrows, which were installed on a concrete floor. The swine waste was turned by a self-propelled composring machine (Rota-Shredder, Roto-Shredder Co., Division ui Imco, Crestline, Ohio) which straddled the row and traveled its length, shredding. grinding. and pclverizing the material (Fig. 1). As the machine moled forward. the waste fell behind the machine. reform%:: rhe windrow. Through this mixing action (turning!. oxygen (air) was incor- SAVXE. CH.A.cjE. AND !vIAC\IILI,hN APL. MICKCHIOI.. acetoin [Voges.Proakauer]: and utilizcltion of citrate, urea, malonate. and triple sugar-iron). :\hnut ;5c- were indicated to be Snlmooeila ant1 2.jL+ were Protetis types (urease positive). Serological tests were not performed. Counts in tab!?: are based on presumed Salmonella from plate county. Feca: streptococci were counted on KF agar plates (1); M-enterococcus agar w'as also used and gave similar results. For total bacterial counts. plates were poured with nutrient agar (Difco) and incubated at 37 C for 2 days. Only typical bacterial colonies were included in the counts; these nere generally mucous or small, _ - yetlow, lenticular colonies. Totai counts were also investigated on compost agar (91, but &rive; colonies were observed. Fungi were enumerated on acid-agar FIG. 1. Front uieuj of the Roto-Shredder &&xzch- or potato-glucose-novobiocin agar plates (9) incu- inq the Lcindrow. bated at 28 C for 5 days, Colonies counted as fungi were typically filamentous or large, round, mucous colonies with yetist-like appearance. Actinomycetes porated into the mass. After turning, the windrow was were enumerated on caseinate agar and Czapek-Dvs less compact, and the particle size of materials such agar (Difco). These plates were incubated for 7 days as paper, cardboard, vegetable matter, and hones uas at 37 C, and only the small. powdery, wrinkled. or greatly reduced. pasty colonies were counted as actinomycetes. When Windrow 1 contained approximately 40 tons (33 these colonies tvere isolated. their plates had the metric tons) of unsuppiemented swine waste and :vsas typical earthy odor associated with actinomycete~. turned twice a week. CVindrows 2,s. and 4 were turned 20 times per week. 1Vindrow 2 contained approG- Cellulolytic bacteria associated with coml&t were enumerated by plating the compost homogenate on a mately 40 tons of swine waste; windrow 3 conrained cellulose-mineral salts agar medium (6). The medium approximately 23 tons (21 metric tons) of. waste and contained per liter: (NH,),SO,. 0.5 g; K,HPO,, 0.5 g; 18 tons (16 metric tons) of old compost that has heen KH,PO,, 0.2 g; CaCl,. 0.5 g; MgSO,, 0.5 g: SaCI, 1.0 previously composted for 180 days with turning twice g; agar, 15 g; and 150 ml of a slurry of cotton-fiber weekly. Windrow 4 contained 40 tons of stockpiled cellulose. ball milled for 24 h in 4% HCI and washed waste and 1.5 tons (1.36 metric tons) of straw. with distilled water. The colonies which produced Sampling procedure. Samples of compost to be clearing zones on this medium were recorded as assayed were collected from the windrow s&ace cellulolytic microorganisms. The plates were incu- immediately after turning on the days indicated in bated aerobically for 7 days at :X C. Fig. 5 to 9. Samples weighing approximateI>- :3 g were Determination of windrow conditions. Tempe:a- taken randomly along the windrow and combined in a tures of the cornposting windrows were determined by sterile quart Mason jar. The total composite sample a thermocouple potentiometer and recorded on a weighed about .l kg. In the laboratory, the sample w-a.; battery-operated, 12-paint Brown recorder (Min- mixed thoroughly. and duplicate portions (10 go wer? neapolis-Honeywell Regulator Co., Industrial Divi- taken for testing. These were suspended in 90 ml of sion. Philadelphia. Pa.). The thermocouples were sterile water, and l-ml volumes of the suspension were diluted serially (wide-mouthed pipette). Samples mounted on a probe that could be inserted in the piles (one is visiblE at the lower left in Fig. I). This probe (100 ml) of run-off water were coliected from seepage had three parailel rods of -:,-inch (about 1.89 cm) at the foot of the pile; I-ml volumes of these samples diameter galvanized pipe with an aluminum point on were diluted serially. one end and a junction box on the other. Each rod had Determination of microorganism numbers. The two openings-one each at 6 and 24 inch* (about standard plate count method (1) was used forestimat- 15.24 and 60.96 cm. respectively) from the tip of the ing the numbers of microorganisms in the compos:. aluminum point. At each opening was a thermocouple Coliforms were counted on Levine rosin meth:;iene (electrically insulated from the metal rod). The rods blue agar (Dit'c~ul'plates, as justified by Poelma 15) were 24 inches apart. They were inserted vertically and used by the Food and Drug Administrarion. The into the top of the windrow and forced down until the identification methods used for Salmonella xere. in tip touched the concrete base, so the readings were general, those of Poelma (8). Plates for countinp made 6 and 2-l inches from the bottom of the piles. Salmonella were poured with brilliant green agar. The three thermocoupIes at each of the t\ acetic acid. 5% nacconol. xater. 107 !i,O,. and distilled H,O. The electrodes were standardized with solutions of known E, value: the meter was standardized by adjustment to zero H- ion and checked by connecting the electrodes- in standard soiution-to an Elec- tronic Associates Inc. 13200 digital volt meter. The moisture content Las determined by weight loss of 23-g samples which acre dried for 3 days at 103 C or unti! a constant weight \\`a~ relc,hed. RESULTS Windrow temperature, pH. and E,. Changes in temperature within the windrows during c;>mpostin,o are shown in Fig. 2. Win- drow 1 s%:wed a lag period of :I$ da?-s before a rapid rise in temperature occurred. The temper- ature in t,`le center of windrow 1 [swine waste turned 20 times per week) reached a thermo- philic range of 55 to 65 C in 2.5 days. The temljerature in windrow 3 (approximately 503 old compost) reached this range in 15 days, where&s windrow 4 (waste and straw mixture) 80 u FIG. 2. Temperature cariaticm in the windrou durmg crunpnsting. The temperature readings uere taken at 24 inches from the bottom of the uindrolc. The symbols representing the ccindro:cx are: 0, I; 0, :`: A, 3; and 0, 4. had reached 60 C within :j La--s :.rt' con~oi:~tin~. The highest temperature rv;~lrded at an!- ftc~- tion in anv win&ox SW T2 C in wincir:nv 4. After reaching thermophiiic temperature. ivind- rows 2 and 3 remained thermopbilic wii the:. were removed from the cxtrrete floor in ciay do. Windrow 1 cooled tu amhixtt temperarxre (20 to 30 C) by day 3. Changes in the pH c f the waste during cornposting are shown ir Fig. 3. LVinrirou- -t reached a pH of 8.0 in on;>- irlur days--indica- tive of rapid decomposition in this windrow- whereas windrow 3 tocl'k li days to reach this pH; windrow 2 reached pH 7.5 in 25 days but did not reach pH 3.V: 2nd windrow 1 to& approximately 80 days to reach pH 8.0. The initinl E, of the wr-indrowj was -1.3) rn\- (Fig. 4). In windrow 1, th? E, rose gradually to -200 to -250 mV over the first 1s days ot composting and then rem~ineci constant for rbe next 20 days; active theraophilic compostinq. indicated by pH and temperature Se. began onI> after this period. The E, rose to -3~ to 9 8 7 z 6 5 3 FIG. 3. Infltxnce of co-lposition on th? pro,~rew of decomposifion. cs indic;trd b) pH @tern. The symbols ore: 0. zcindrcr. ; lun.wpplrrlented .uine waste turned ttcice a cc-eec 1: 0, rcindrou 1 lun.~:ippi?- mented waste t rned 20 rimes a xeek): 0. windrocc. 3 (waste plus 18 -da\-o (maste plrts 5rr l?r,,,,, !d compost); h. windrow 4 0 5 10 15 20 25 30 33 40 TlXiE ttIayj\ FIG. 4. Redox pot~~rials in !l,ind;*jrcs of rtie com- posting process. .A11 rcrdinps were :r;ken hi.fs,re tha daily turning at n P.&:.zzi! drprh in [I indwu.. i (,I. 2 (O), 3 (CO, and 4 lilt. -+-lo0 mV immediately after the windrow was turned but decreased within 60 min to the leve! observed before the turning. In windrows 2. :i. and 4 during the active cornposting cpH .>7. temperature >6O C). the E, was -50 to -It>*) mV before the turning. Windrow 4 reached this level (from the initial -450 mV) within 3 days. These more aerobic conditions presumable re- sult from greater porosity of the pile. especiaily in windrow 4. Active composting (pH >7, temperature >50 C) began in windrow 1 only after the moisture content fell below 40%, 40 days after the composting process was begun; horvever. since windrows 3, 3. and 4 reached the active composting stage when moisture content was still 45 to 55%, dryness of the material was not solely responsible for active composting. Effects of thermophilic composting on in- testinal microorganisms. Enumeration of Salmonella in windrow 1 and its run-off water at various times indicated the number of orga- nisms that could survive and possibly pollute water supplies. Salmonelfa numbers, after an initial drop, increased in the windrow (Table 1. 40 days). Cvhen the temperature rose aSoT.-e 48 C, the population of presumptive Salmone!ic decreased sharply and continued to decrease as temperature increased. Similarly, the numbers of Salmonella, coliforms, and streptococci in the run-off water increased initially (after a drop in coliforms), then decreased as the temperature in the center of the windrow passed 52 C (Table 2). The count of fecal coliforms, shown in Table 2, appeared to remain higher than fecal strepto- cocci or Salmonella. The coliform count de- creased rapidly as composting proceeded in windrows 2, 3, and 4 (Fig. 5). The coliform test was negative in windrow 4 at day 14 of compost- ing when the temperature was `il C. When windrow 4 was removed on day 40, presumptive Salmonella colonies could not be detected. The TABLE 1. Presumptive Salmonella colonies from windrow I Days of I Temp at center 01 Bacteria ( x 1N celldgi on cornposting aindrom (Cl Bs" SS" BC' 0 36 55 83 41 19 41 9 12 ' 11 40 48 320 370 510 60 52 6 5 $8 51 4 3 i Tj 187 68 0.0s 0.07 0.x ' BS, Bislnuth sulfite agar; SS. Salmonel!n- Shi.eelln agar; BG, brilliant green agar. TARI.E 2. Presumpti-e Salm:Jnelln. fwal cnliiorm. . and fecal Streptocnccus colonie.< in run-of/ waterfrom windrou 1 180 90 6 SiO 41 12 --l---r 590 53 90 - 11; 210 160 2 0.5 ' Isolated on bismuth sulfite agar. FIG. 5. Effect of composting action on the num her offecal coliforms in windrorcs 2 (0). 3 CO), and 4 (A). Each point is the overage value from counting four replicate plates. number of coliforms was,reduced by IO'-fold in windrows 2 and 3. In 40 days of composting. the number of mesophilic bacteria decreased between lo'- and 10S-fold in Gindrows 3 and 4 (Fig. 6). The bacterial population in windrow 2 increased ' during the first 15 days and then started to decline. The decline in mesophilic bacteria occurred at the time when the temperature began to rise rapid!); in these windrows. The number of mesophlhc fungi did not drop as much as the number of bacteria (Fig. 7). In windrow 4, fungi were reduced 103-fold, but in windrows 2 and 3 the decrease was less than lOO-fold. The population of mesophilic ac- tinomycetes responded differently from either bacteria or fungi. This population increased by factors of over lo3 in windrow 4, lo3 in windrow 3, and less than lo* in windrow 2 (Fig. 8). The increase in windrow 3 followed a decrease in the first sveek from the high initial numbers present as a result of the addition of old compost. The number of cellulolytic organisms showed a re- \`UL. `26, 1973 XEROMC THERMOPHILIC kVISZ)ROLV COMPOS'I'ISC 9x3 21 I r-J-2-J 0 5 10 15 20 25 30 35 TIME (days) FIG. 6. Number of mesophilic bacteria during the composting process. The symbols represent windrows 2 (0). 3 (0). and 4 (A). - 01 I i 8 1 / 0 5 10 15 20 25 33 35 T I V\riE (days) FIG. 7. Number of mesophific fungi during the composting process. The symbols represect xindrows 2 (0),3 (Cl), and 4 (Ai. sponse similar to that of the population of actinomycetes. Xgain, the population in wind- row 4 developed more rapidly and attained a much higher number iFig. 9). DISCUSSION Although considerable use is made of com- posting in the disposal of municipal. wastes, garbage, sewage sludge, etc. (4, 51, little infor- mation is available on the numbers of enteric bacteria actually resulting from such opera- tions. Typically, thermal death points of com- mon pathogens present have been determined, and the assumption was made that attainment of these temperatures during the composting process would destroy the pathogens (2, 51. In one study (13), introduced indicator organisms (poliovirus, Candida albicans, Ascaris lumbri- coides, Salmonella newport) were shown to dis- appear within -23 h from a pilot-scale composter maintained at 60 to 70 C. The possibility of survival of pathogens at the cool surface of a rvindrow operation, however, has not been dis- Proven. Fro. 8. Number of mesophilic actinomycetes dur- ing the composting process. The symbols represent windrows 2 (0). 3 (Cl), and 1 (Ai. Our results indicate clearly a marked de- crease in coliforms, salmonellae, and en- FIG. 9. Number of celIutoI>tic organisms during the composting process. The symbols represent tcin- tcrococci during the thermophilic stage of com- draws 2 (0). 3 (Cl). and 4 (A.). posting. Before the thermophilic stage, there was presumahly anaerobic decomp&tion of carbohydrates, proteins. and fats to form or- ganic acids and other intermediate compounds that could be used by Salmonella. colitorms. and enterococci for growth under partial!y an- aerobic conditions (7). Therefore, an increase in these organisms during the first stages of com- posting was expected. From a public health standpoint, it is desirable that this early in- crease be minimized and that the windrows maintain a temperature above 48 C long enough to destroy pathogens like Salmonella. These results indicate that, for maxima1 sanitary safety, the thermophilic stage of cornposting should be reached as soon as possible. The noteworthy practical observation is that inclusion of straw (windrow f) fostered more aerobic conditions and thereby facilitated very rapid attainment of thermophilic conditions and destruction of enteric bacteria. Frequent turning of the windrows was much less effective. The original rationale for addition of straw was increase of the carbon-nitrogen ratio, since excess nitrogen is liable to be eliminated as ammonia and other malodorous amines. How- ever, the primary effect of addition of straw 10 , I I I I I 974 SAVAGE, CHASE. appeared to be imp:ovement of the physical structure of the windrow, allowing more natural aeration (compare Fig. 4) and a rapid rise in temperature consequent upon intense aerobic microbial activity. This had the beneficial ef- fect of rapid destruction of pathogens. Waste materials with similar structural properties, such as cornstalks, chipped wood, shredded municipal refuse. etc., should also have this effect. The possibility of use of vegetable wastes such as cornstalks, which are also be- coming a disposal problem as open air burn- ing is banned. to improve the composting of animal wastes is particularly attractive. X succession of microbial populations was observed during the cornposting process. The bacteria increased in number before the tem- perature of the windrows rose and then de- clined, whereas cellulolytic organisms and ac- tinomycetes in general increased in the thermo- philic stage. Presumably, the mesophilic bacte- ria rapidly attack the more readily available organic constituents, resulting in a temperature increase. The increased temperature favors the cellulolytic organisms. and the mesophilic bac- teria largely disappear. The actinomycetes ap- peared in the final stage to such an extent that the surfaces of the compost piles were white or gray. These organisms are known to play a role in the humification of organic matter, which results in a stabilized product (3). ACKSOWLEDGMENTS This inves:iga:iw was carried out by the Department of Ri.~chemktry and ~licrrblolo~y, College of Agricalture and Er;rironmental Science. R?!rgers-The State Univer>it)-. as part of a study on cornposting of swine waste being conducted by the Department of Rgicultural Engineering. The investi- gation was supported by Cnited States Department of Agri- culture Cooperative Agreement 12.14.lOWllJO78 (42) and was directed by Martin Decker. LITERATURE CITED 2. Golueke. C. G.. and H. B. Gotaaa. 19% Public he&h aspects of rvaste disposal b:: comp+:ing. .An:e:. J. Public Hrzlth 44:MB-399. 3. Colueke, C. G.. and C. P. NcGauhey. I'J?IJ. .A critical evaiuation of inoculum~ io cornposting. Appl. Yiclo- biol. 24.5. 1. Gotaas. H. B. 1956. Compostinx. Sanitay dispo>al and reiismation of organic wwtes. R'orld Health .kocia- tinn, Geneva. 5. Grind:od. J. 1961. Six years of refuse compns;rirtg in Britain. Public \Vorks !J2:1 IlClll. 6. Hungate. R. E. i9.30. The anaerobic celluiolxtic bacteria. Bacrrriol. Rev. 14:1-41. 7. McCalla. T. SI.. L. R. Frederick. sod ti. P. I,. P.llmer. 1970. %Ianure decomposition and t'are of breskdnwn products in soil, p. 2:1-2%. In T. L. \V`l!!rich axd G. E. Smith ted.). Xgrirultural practices and water quality. Iowa Stare Cniv. Press, Ames. 8. P&ma. P. L. 1967. Deteciirm and identifirsrian of SoImonella from dried vholr egg. dried egg ylt!l;. and dried eg: white: report of a co!iaborntiw study. .I. Ass. Offic. Anal. Chem. 50:753-X3. 9. Pramer. 0.. and E. L. Schmidt. 1964. Experimental soil microbiolcrg~. p. 29-37. BurgesPuh. Cv.. Slinneapolk. 10. Schillze, K. L. 1962. Continuous !hermo?hilic compost- ing. Appl. .Ilicrobioi. 10:108-122. 11. Starkey, R. L.. and K. 81. UYght. 194.5. Anaerobic cormsion of iron in soil, p. l-108. Ir: American Gas Association. report of Technical Section Disrribution Committee. 12. \Vadleiph, C. H. 1967. Agriculrural pollution of water resources. Soil Conserv. Map. 1lJ:?7-30. 13. Wiley. B. B.. and S. \Vesterberg. 1969. Human pathogens in composted sewage. Appl. \licrohiol. l&994-1UOi.