Vol. 41, No. 2APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1981, p. 422-4290099-2240/81/020422-08$02.00/0

Prevention of Surface Mold Growth on Italian Dry Sausage byNatamycin and Potassium Sorbatet

RICHARD A. HOLLEY

Food Research Institute, Research Branch, Agriculture Canada, Ottawa, Ontario KIA OC6, Canada

Inhibition of uncontrolled mold growth on three types of raw cured Italian drysalami was studied under commercial production conditions. Salami were dippedor sprayed with natamycin (pimaricin) or were given a combined organic acid-plus-potassium sorbate treatment. Acetic and citric acids potentiated the inhibi-tory effects of potassium sorbate significantly, but lactic and succinic acids showedlittle or no effect. Treatment of salami by dipping in 2.5% (wt/vol) potassiumsorbate or 2,000 ppm (mg/liter) of pimaricin did not successfully prevent thegrowth of surface molds. At 10% potassium sorbate on all types of salami and at2.5% sorbate on Casalingo salami, visual inhibition of mold growth was observed,but numbers of viable fungi on all salami types treated with 2.5% sorbate werenot significantly (95% confidence) different from numbers found in the untreatedcontrols. Pimaricin spray (2 x 1,000 ppm) was as good as or slightly better than2.5% potassium sorbate, but greater concentrations of each were required tosatisfactorily inhibit surface mold growth during the 25- to 50-day ripening period.

Dried sausages with various degrees of surfacemold growth are preferred by ethnic groups frommany countries. Unfortunately, overgrowth ofthe surface by undesirable molds is not uncom-mon and has attracted the interest of scientistsin Italy, Hungary, Germany, and the UnitedStates. Dragoni et al. (10) and Cattaneo et al.(7) concluded that molds were not a part of thenormal microflora of Italian dry sausage. Ayreset al. (2) noted that mold growth on genuineItalian salami was somewhat scanty, whereassimilar products produced in the United Stateswere heavily covered with mold.

In North America there is generally consumerreluctance to purchase moldy food items, anddried sausages do not appear to be an exception.Although sausages evenly covered with densewhite mycelium are considered desirable bysome groups, the uncontrolled mold growth seenon most dry cured meat products is usuallypigmented green, gray, or yellow and thus haslittle appeal.Mold growth on cured meat products can take

place not only during curing but also duringwholesale distribution and in the retail show-case. At each stage in the distribution chain,pigmented mold is removed by washing to im-prove product appearance. Vacuum packagingof the fully mature cured product is also used;however, washing is usually necessary beforepackaging, and the product becomes morecostly. The danger inherent in the practice ofmold removal by washing is that the procedure

t Contribution no. 431 from the Food Research Institute.

does nothing to remove toxic and allergenic moldmetabolites which may have diffused into themeat as a result of the original mold growth.Although Ayres et al. (1) observed mycotoxinformation in dry sausages under controlled con-ditions, it is felt that if storage temperaturesbelow 15°C are maintained, mycotoxin produc-tion will not occur (1, 5, 6).

Surface molding of dry sausages and curedmeats generally was a problem in three of fourmeat processing plants recently viewed. Ofthese, products manufactured in the Italianplants were most affected. This was probablybecause, unlike German-type sausages, Italiansalami are not smoked, and smoking is known toinhibit mold development on the surface of thesausages (9, 17, 20).

Italian dry salami has a water activity whichranges between 0.671 to 0.920, a moisture con-tent of 24.3 to 43.4%, and a pH of approximately4.3 to 7.1 (12, 16, 21, 24, 25). Thus, under condi-tions of relative humidity (RH) greater than70%, mold will readily form on the salami sur-face.Both potassium sorbate (3, 15, 17, 26) and

pimaricin, also known as natamycin (7, 14, 19),have been found effective as antifungal treat-ments on cured meat products under specificconditions. It was therefore considered impor-tant to evaluate the ability of these compoundsto reduce or eliminate the surface molding prob-lem on several types of raw cured pork-typeItalian sausages in a single plant under normalproduction conditions which were conducive tomold growth.

422

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

MOLD INHIBITION ON SALAMI CASING 423

MATERIALS AND METHODS

Salami. Three types of Italian dry fermented sau-sage, which contained from 21 to 25% fat, 53 to 58%moisture, and about 16% protein immediately afterformulation, were studied during and after their com-mercial manufacture. The salami contained pork, salt,nonfat dry milk powder, dextrose, glucono-delta lac-tone, spice, monosodium glutamate, sodium erythor-bate, garlic, potassium nitrate, and sodium nitrite. Nocommercial starter cultures were used.Minor differences in spice formulation were noted

between the Genoa (each weighed about 5 lb, or about2.3 kg) and Casalingo (about 3 lb, or 1.4 kg) salami.These coarse-ground sausages differed from Abruzzesesalami (about 1 lb, or 0.45 kg) in that the latter werefinely ground and contained 1-cm2 pieces of pork fatsimilar to that in mortadella formulations. Casingsused for Genoa and Casalingo salami were of a multi-layered seamless regenerated collagen type. A naturalhog casing (porcine bung) was used for the Abruzzesesalami.The complete meat, spice, and curing ingredients

formulation was stuffed into casings at 2°C and trans-ferred to the curing (green) room, where the sausageswere held at 20 to 24°C and 95% RH for 2 to 3 days.During this period, major bacterial growth and fer-mentation took place. Alterations in temperature to20°C and RH to 60% were made according to theamount of salami in the chamber and their individualsizes. On the 4th and 5th days, the temperature andRH were gradually changed to 11°C and 77%, respec-tively. Salami were then transferred to a drying roomfor an additional 20- to 45-day ripening period at 11°Cand 77% RH. Since these salami were manufacturedin the traditional Italian manner, they were neithercooked nor smoked during processing. Experimentscarried out to determine the effect of surface moldinhibitors during salami maturation were conductedunder commercial ripening conditions at Santa MariaFoods Ltd., Belleville, Ontario, Canada.

Pimaricin treatments. Immediately after stuffing,six samples each of Casalingo and Genoa salami weresprayed with 1,000 ppm (mg/liter) of pimaricin (Del-vocid; Gist-Brocades Inc., Montreal) or dipped for 30s in 2,000 ppm of pimaricin. These samples had notbeen pretreated with organic acids. The pimaricin hadbeen formulated to contain 5% (wt/vol) NaCl, andsince the Delvocid powder is 50% active pimaricin, 2g/liter was used to give a 1,000-ppm solution and 4 g/liter was used to yield a 2,000-ppm solution. Sprayingwas carried out with a plastic, pistol-type, hand-oper-ated atomizer. Spray was applied until the salami wasgenerously covered and liquid dripped from the sam-ples. These pimaricin-treated samples were placed inthe green room for 5 days as previously described. Atthe end of this period, samples sprayed with 1,000 ppmof pimaricin were sprayed again as before and returnedto the drying room until mature.

Only a single pimaricin treatment (1,000-ppm spray)was given to the Abruzzese salami, but casings weregiven the normal 3-min soak in 2% acetic acid beforestuffing. Six Abruzzese salami were removed from thegreen room on day 2 of fermentation. These weresprayed until dripping with 1,000 ppm of pimaricin as

previously described and then returned to the greenroom and treated in the normal manner.Acid pretreatments. Casalingo, Genoa, and

Abruzzese salami were treated with organic acids be-fore being treated with potassium sorbate. All Abruzz-ese casings were soaked for 3 min in 2% (vol/vol)acetic acid before stuffing since this was normal plantpractice for natural casings. Other casings were moist-ened in tap water and stuffed, and then the salami wasgiven an organic acid pretreatment. Casalingo andGenoa salami were assembled in 4 groups of 18 salamieach. Each group was dipped for 30 s in one of thefollowing acids: 2% (wt/vol) citric, 2% (wt/vol) suc-cinic, and 2% (vol/vol) lactic acid. The fourth group ofCasalingo salami was dipped in commercial vinegar(5% [vol/vol] acetic acid), whereas the fourth group ofGenoa salami was dipped in vinegar diluted to contain2% (vol/vol) acetic acid. An untreated control groupof each type of salami was also used. After treatments,Abruzzese and Genoa salami were placed in the greenroom at 20 to 24°C and 95% RH as described earlier.Casalingo samples were treated in the same mannerbut were stored at 4°C overnight before being placedin the green room.Potassium sorbate treatments. At the end of 5

days of normal incubation in the green room, sixGenoa and Casalingo salami from each acid pretreat-ment (acetic, citric, lactic, and succinic) were dippedfor 60 s in aqueous 2.5, 10, or 20% (wt/vol) solutions ofpotassium sorbate (Sorbistat-K; Pfizer Canada, Ltd.,Montreal) and were placed in the curing room at 11°Cand 77% RH. At the end of 2 days of incubation in thegreen room, six Abruzzese salami were dipped for 60s in 2.5, 10, or 20% (wt/vol) solutions of potassiumsorbate and were returned to the green room to con-tinue the normal maturation schedule.Sampling pimaricin- and sorbate-treated sal-

ami. Whole Genoa salami were sampled on days 0, 5,9, 16, 34, and 50. These samples were frozen andreturned to the laboratory for study. Whole Casalingosalami were sampled on days 0, 5, 9, 12, 15, and 26 andfrozen. Similarly, Abruzzese were sampled on days 3,7, 9, 11, 17, and 25 and frozen for later study at thelaboratory. Identical sampling patterns were followedfor untreated control groups, which were carried forall experiments. Genoa, Casalingo, and Abruzzese sal-ami were allowed to mature for 50, 26, and 25 days,respectively.Measurement of salami surface mold growth.

A cross-sectional center piece of salami measuring 5to 7.5 cm long was cut from each type of salami tested.The casing was slit longitudinally and peeled asepti-cally from the meat. The casing was placed in a sterile24-oz (ca. 720 ml) Whirl-Pak bag (Nasco EducationalMaterials Ltd., Guelph, Ontario), and the area of thesample was determined. The casing was transferred toa sterile Stomacher bag, 100 ml of 0.1% (wt/vol) pep-tone was added, and the mixture was stomached for 1min. The casing was removed and the peptone wasfurther diluted as necessary to obtain viable counts offungi, using potato dextrose agar (PDA) and the pourplate technique. Plates were incubated at 22°C for 3days. Results were expressed as colonies per squarecentimeter of casing surface area.

Visual estimates of surface growth by fungi were

VOL. 41, 1981

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

made on a scale of 0 to 5 after examination of salamisamples during maturation. A photographic record ofthe maturation process was kept.

Statistical evaluation. Viable surface mold countsper square centimeter were transformed to logarithms(base 10). These data were subjected to one-way anal-ysis of variance and were further subjected to theNeuman-Keuls test for equality of means (27).

RESULTSPimaricin treatment. The 2,000-ppm pi-

maricin dip was found to be ineffective in pre-venting surface mold growth on either Genoa or

Casalingo salami aged 50 or 26 days, respectively(Table 1). Pimaricin spray (2 x 1,000 ppm) waseffective on both salami types, but mold growthwas replaced by a dusting of yeast colonies thatcovered the surface to a degree that varied di-rectly with the length of ripening. Photographsof pimaricin-treated fully ripened Casalingo (26days) and Genoa (50 days) salami are shown inFig. 1A and 2A.Abruzzese salami sprayed with pimaricin

showed no mold growth for a week after stuffing,and then a light dusting of mold became visible.This growth continued until the end ofthe ripen-ing period, when there was only a slight improve-ment of pimaricin-treated salami over the un-treated control samples. Abruzzese salami areshown, after 26 days of ripening, in Fig. 1C.

Acid pretreatments. Genoa salami treatedwith the four organic acids at day 0 appeared tobe equally susceptible to surface mold growthwhen examined before being dipped in potas-sium sorbate at day 5 (Table 1).

Casalingo salami, pretreated with citric acid,showed no visible surface mold growth 5 daysafter treatment. Very slight mold growth wasapparent on samples treated with acetic acid,but greater growth was present on lactic acid-and succinic acid-treated salami. Succinic aciddid not significantly prevent surface moldgrowth on either the Casalingo or Genoa salami(Table 1). Casings of Abruzzese salami, whichhad been dipped in 2% (vol/vol) acetic acid,showed no mold growth at day 2, when theywere given pimaricin or sorbate treatment. Thenatural casings were more resistant to moldgrowth.Potassium sorbate treatment. From visual

examination of Genoa salami, it became appar-ent that the 2.5% potassium sorbate treatmentwas only marginally better than no treatment(Table 1). A marked improvement was notedwith 10% sorbate, and very little mold growthwas visible with 20% potassium sorbate (Fig.2B-D).Treatment with 2.5% potassium sorbate sig-

nificantly reduced the degree of visible mold

TABLE 1. Extent of salami surface mold growth assessed by visual inspectionGrowthb

Treatment Acida pretreat- Genoa Casalingoment ________________ ___________5C 16 26 50 5 16 26

Potassium sorbate2.5% (wt/vol) Acetic 2.0 1.5 3.5 4.0 1.0 1.0 1.0

Citric 2.0 1.5 3.5 4.0 0 0.5 1.0Lactic 2.0 2.0 3.5 4.0 2.0 2.0 2.0Succinic 2.0 2.0 3.5 4.0 2.5 2.5 2.5

10% (wt/vol) Acetic 2.0 1.0 1.0 1.0 1.0 0.5 0.5Citric 2.0 1.0 1.5 1.0 0 0.5 0.5Lactic 2.0 1.0 1.0 1.0 2.0 1.5 1.5Succinic 2.0 1.5 1.5 1.5 2.5 2.0 2.0

20% (wt/vol) Acetic 2.0 0 0.5 0.5 1.0 0.5 0Citric 2.0 0 0.5 1.0 0 0.5 0.5Lactic 2.0 0.5 0.5 1.0 2.0 0.5 0.5Succinic 2.0 0.5 0.5 1.5 2.5 0.5 0.5

PimaricinSprayd 2.0 1.0 2.0 2.0 2.5 1.5 1.0Dip 2.0 1.5 2.5 5.0 2.5 4.0 3.5

Untreated control 2.5 3.0 4.0 5.0 2.5 4.0 4.0a Salami were given a 30-s acid pretreatment on day 0 followed on day 5 by a 60-s dip in 2.5, 10, or 20%

potassium sorbate.b Surface growth by fungi assigned numerical values: no growth, 0; very slight, 1; slight, 2; moderate, 3;

marked, 4; confluent, 5.C Days of ripening.d Salami were either dipped for 30 s in 2,000 ppm or sprayed with 1,000 ppm of pimaricin on day 0. On day

5 the sprayed samples were sprayed again with 1,000 ppm of pimaricin.

424 HOLLEY

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

MOLD INHIBITION ON SALAMI CASING 425

growth on Casalingo salami compared with un-treated controls (Fig. 1A, B), but the best sup-pression of mold growth on Casalingo was evi-dent upon examination of salami treated with20% potassium sorbate. Under these conditions,there was slightly less mold growth on the sam-ples which received the acetic acid pretreatment(Table 1). In fact, acetic acid alone had somevisible effect on mold growth even after 26 daysof ripening (Fig. 1B). As seen with Genoa salami(Fig. 2C), individual centers of mold growthbecame enlarged where treatments had a visibleinhibitory effect (Fig. 1B).

Abruzzese salami did not appear to be signif-icantly affected by surface mold growth whenfully ripened (Fig. 10). There was very slightmold growth with 2.5% sorbate, somewhat lesswith 10%, and essentially none with 20% potas-sium sorbate. A single treatment with 1,000 ppmof pimaricin spray was as effective as 10% sor-

PIMARION IM MRION SPAY

ml% . ITE

'

FIG. 2. Fully ripened Genoa salami, aged 50 days.(A) Control, untreated; pimaricin dip, 2,000 ppm;pimaricin spray, 2 x 1,000 ppm. (B) Control, un-treated; succinic, 2% (wt/vol) succinic acid pretreat-ment; lactic, 2% (vol/vol) lactic acid pretreatment;citric, 2% (wt/vol) citric acidpretreatment; acetic, 2%(vol/vol) acetic acidpretreatment. All were dipped atday 5 for 60 s in 2.5% potassium sorbate. (C) Same as(B) except 10% potassium sorbate at day 5. (D) Sameas (B) except 20% potassium sorbate at day 5.

FIG. 1. Fully ripened dry-cured salami. (A) Cas-alingo salami, aged 26 days; untreated control; pi-maricin dip, 2,000 ppm; pimaricin spray, 2 x 1,000ppm. (B) Acetic acid-pretreated, potassium sorbate-treated Casalingo, aged 26 days; no sorbate, 0; 2.5%sorbate, 2.5; 10% sorbate, 10; 20% sorbate, 20. (C)Abruzzese, aged 25 days. Untreated control; pimari-cin spray, 1 x 1,000ppm; 2.5%potassium sorbate, 2.5;10% potassium sorbate, 10; 20% potassium sorbate,20.

bate on Abruzzese salami.Viable surface mold growth. Mold growth

on the surface of Genoa salami did not appearto be retarded by a 1-min 2.5% potassium sorbatedip (Table 2). After one-way analysis of varianceand application of the Newman-Keuls test forequality of means (27), the acetate-sorbate datawere separable into two significantly differentgroups at the 95% confidence level. The firstgroup contained Genoa salami treated with 20%

VOL. 41, 1981

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

426 HOLLEY

potassium sorbate. The second group was madeup ofthe remaining treated Genoa salami, whichwere not significantly different from the un-

treated control salami. There was no statisticaldifference between viable surface growth at day50 on salami pretreated with acetate or citrate.

Casalingo salami casings pretreated withacetic acid were examined similarly, and it wasevident that treatment means could be arrangedin several groups (Table 3). Viable numbers werelowest (95% confidence level) after treatmentwith 20% sorbate, a trend which was seenthroughout ripening. The Neuman-Keuls testalso showed that the 10 and 2.5% sorbate treat-ment means were equivalent. This test alsogrouped the 2.5% sorbate and untreated salamitogether, but the 10% sorbate results were dif-ferent from results for the control salami. Pi-maricin data were similar to those obtained for10 and 2.5% sorbate treatments.Abruzzese salami casings sprayed once with

1,000 ppm of pimaricin or dipped in the variouspotassium sorbate solutions were also examinedfor viable fungi (Table 4). There was no differ-

TABLE 2. Surfacea yeast and mold count' of Genoasalami casing during ripening

Yeast and mold count/cm2Treatment

5C 9 16 34 50 50d

Control 1.653 3.311 4.890 6.243 6.292 7.900Sorbate

2.5% 3.322 4.742 6.787 7.144 7.462 5.99610% 4.885 3.642 3.626 3.494 3.238 4.72820% 0.699 1.996 2.691 2.973 2.703 4.610

Pimaricin' 1.740 2.217 2.303 4.673 5.444

a Surface area of salami sampled (mean + standard devia-tion) = 144.4 ± 10.5 cm2.

'Results expressed as the logarithm (base 10).'Days of ripening.d Samples in this column received a 2% citric acid pretreat-

ment. With the exception of pimaricin treatments, all othersamples were given a pretreatment with 2% acetic acid.

'Pimaricin sprayed on day 0 and day 5 at 1,000 ppm, no

pretreatment.

TABLE 3. Surface' yeast and mold countb ofacetate-pretreated Casalingo salami ceasing during

ripening

Yeast and mold count/cm2Treatment

5c 9 12 15 26

Control 6.489 6.068 6.923 7.605 6.828Sorbate

2.5% 4.814 4.645 4.678 7.846 6.813

10% 5.310 4.820 5.155 4.768 4.886

20% 5.818 3.659 2.940 2.763 2.230

PimariCind 5.814 5.917 6.017

aMean + standard deviation of casing area per salamisampled = 170.8 + 13.8 cm2.

bc See Table 2.

d 2 x 1,000 ppm of spray, no pretreatment.

APPL. ENVIRON. MICROBIOL.

TABLE 4. Surface' yeast and mold countb ofAbruzzese salami casing during ripening

Yeast and mold count/cm2Treatment

9c 11 17 25

Control 3.371 4.375 3.857 3.158Sorbate

2.5% 1.987 3.508 2.672 3.522l1o0 2.303 1.556 1.944 1.74820% 2.068 1.799 1.301 1.591

Pimaricind 3.330 3.130 4.244a Mean ± standard deviation of casing area per

salami sampled = 111.9 ± 14.2 cm2.bceSee Table 2.d 1,000 ppm of spray, 2% acetic acid casing dip

before being stuffed.

ence among the untreated control, the 2.5% sor-bate, and the pimaricin treatments. The Neu-man-Keuls test grouped the 10 and 20% sorbatetreatments together as equally effective, al-though a very slightly better result was visuallyevident after treatment with 20% potassium sor-bate (Fig. 10).

DISCUSSIONPimaricin treatments. In view of the suc-

cess reported by others in using pimaricin toprevent mold growth on salami (7, 14, 19), it wasdisappointing to find that dipping the salami in2,000 ppm of pimaricin had essentially no effect(Fig. 1A and 2A). This was not too surprisingsince it had been shown (19) that immersiontreatment with pimaricin was not very satisfac-tory unless salami were stuffed in natural cas-ings. When sprayed, a 1,000-ppm pimaricintreatment protected smoked Dutch salamistuffed in protein fiber casings and stored for 5weeks at 18°C and 65 to 85% RH (19). WhenGenoa and Casalingo salami were sprayed twice(days 0 and 5) with 1,000 ppm of pimaricin, therewas an improvement which was slightly betterthan that obtained after treatment with aceticacid plus 2.5% potassium sorbate (Table 1, Fig.1A and 2A). However, there was a noticeableincrease in small yeast colonies on the surface ofpimaricin-sprayed samples of Genoa and Casal-ingo which gave them a slight grayish hue. Inthe case of a pimaricin-sprayed sample visible inFig. 2A, mold growth appeared significant, butthis patch of growth was probably due to thetwo salami touching when hung. This can causea local increase in water activity, thus pernittingeasier growth (19). In addition to the betterresult reported here for sprayed salami, Hech-elman and Leistner (14) also had better antimoldresults when pimaricin (0.25%) was sprayed onfreshly stuffed salami. It is difficult to explainwhy yeast growth was prominent on pimaricin-

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

MOLD INHIBITION ON SALAMI CASING 427

treated salami. Perhaps the yeast simply re-placed the inhibited mycelial forms. Theseyeasts either had a high tolerance for pimaricinor developed resistance to inhibition by pimari-cin. The latter possibility seems unlikely in viewof the study conducted by deBoer et al. (8).They surveyed 16 meat plants in Germany todetermine whether resistance to pimaricin wasdeveloping in populations of yeasts and moldspresent in the plants. Comparisons were madeof the minimal inhibitory concentrations ob-tained in plants where pimaricin had been in usefor as long as 5 to 9 years with those obtained inplants where pimaricin had never been used. Nodifference in the susceptibility of organisms wasfound among the plants.Mold growth on pimaricin-sprayed (1,000

ppm) Abruzzese salami was not visually exten-sive; however, neither was growth on the un-treated control sample. Comparison of viablecolony-forming units on the surface of thesesalami (Table 4) with those of the Genoa andCasalingo types (Tables 2 and 3) support thisobservation. The less extensive mold growth onthe surface of Abruzzese salami was probablydue to the barrier action of the natural casing tomold invasion. The porcine bungs used weresignificantly thicker than the artificial collagencasings and contained many fatty deposits whichdid not appear to support surface growth.Acid pretreatments. To prevent the forma-

tion of off-flavor or discoloration of the salami,potassium sorbate was applied after green roomprocessing (17). This necessitated the use of aninterim inhibitor at the time the salami werestuffed so that the surface fungi would notachieve significant growth before sorbate appli-cation. Many processors apply a short smokecycle during this period which effectively retardsmold growth; however, Italian processors of rawcured pork meat products do not traditionallyuse smoke (24). Leistner et al. (17) suggestedthat a lactic acid dip of casings before salamiripening might prevent mold growth on non-smoked products. On the basis of previous work,it was felt that other organic acids might also beeffective (11, 18, 22).Acid pretreatments used during the present

study appeared to cause only very slight inhibi-tion of mold growth on Genoa salami, and therewas no visible difference among acid treatments(Table 1). On Casalingo salami differences werenoted (Table 1), and at day 5 the order ofefficacyfrom the best to worst inhibitor was citric >acetic > lactic > succinic acid. The result is ingeneral agreement with the findings of Levineand Fellers (18), who reported that the order ofinhibitory activity against mold was acetic >hydrochloric > lactic acid; however, the result

disagrees with the findings of Erickson and Fa-bian (11), who found acetic acid best but alsofound citric acid less effective than lactic acidagainst molds in laboratory media. In contrastto the results obtained in these mold studies,Raccach and Baker (22) found succinic acid tobe better than lactic, acetic, or citric acid atequivalent pH (4.7) for the inhibition of Staph-ylococcus aureus in fermentable poultry meat.Snijders et al. (23) found that 0.75% lactic acidwas effective against bacteria on beef carcasses,but Levine and Fellers (18) suspected that As-pergillus niger used lactic acid as a carbonsource during their inhibition study. Citric acidappeared useful during the present investigation.Although its inhibitory effects on molds werenot discussed, it has been reported that at Kulm-bach, a spray of 11.7% citric acid and 21.6%sodium chloride was used successfully to reducemicrobial growth on the surface of animal car-casses (Anon, Meat Industry 26:28-29, 1980).The improved performance of acetic acid (5%)

on Casalingo salami over that obtained with 2%acetic acid on Genoa salami cannot be explainedtotally by the difference in concentration. Citricacid was also more inhibitory on Casalingo sal-ami, and the same concentration was usedthroughout these experiments. It may be thatstorage of the Casalingo salami at 4°C overnightafter acid pretreatments or the smaller diameterof the Casalingo salami enhanced the antimicro-bial action of pretreatments. The inadequacy ofacid treatment alone, applied when salami werestuffed for prevention of mold growth, wasshown by the failure of the 5% acetic acid dip(Fig. 1B).The concentration of acetic acid used on the

salami (2 to 5%, vol/vol) was greater than thatreported to be effective in laboratory media (4,11, 18) and the 1.0 to 1.2% found effective againstmold in food (13); however, the effectiveness of2% (wt/vol) citric acid was better than had beenexpected. Concentrations of lactic and succinicacids greater than those used here would benecessary to yield acceptable control of moldgrowth.Potassium sorbate treatment. Treatment

of the Genoa, Casalingo, and Abruzzese salamiwith 20% (wt/vol) potassium sorbate caused themost complete inhibition of surface moldgrowth. There were statistically (95% confidencelevel) fewer viable colony-forming units on sal-ami treated with 20% sorbate (Tables 2-4, Fig.3). For Genoa and Casalingo salami, this conclu-sion held regardless of the type of acid pretreat-ment used; however, the best combination was20% sorbate with an acetic acid pretreatment.At 10% sorbate there was still significant visualmold inhibition and fewer viable fungi. This

VOL. 41, 1981

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

APPL. ENVIRON. MICROBIOL.

ui

. 8

j7

a 4

.2

0

to 20 30

DAYS OF RIPENING

FIG. 3. Inhibition of yeast and mold aon the surface of Genoa and Casalingosalami during ripening. Casing samplesize were peeled from the salami, and v,and molds were counted on potato de:after incubation at 22°C for 3 days. CasGenoa casing data from days 5 and 9 wer

and averaged. Additional Genoa sampleson days 16, 34, and 50. The remaining doanalysis of Casalingo casings. The firsspray (1,000 ppm) was given immediatelying; the second spray (1,O0OO ppm) was gitof curing. Other samples were dipped onsium sorbate for 60 s on day 5 and haotreated immediately after stuffing withSymbols: 0, untreated control; 0, 2.5%tassium sorbate; EL, 10% potassium sorbpotassium sorbate; A, 2 x 1,000 ppm ospray.

difference was statistically significantface of Casalingo but not Genoa salaand 10% sorbate, many yeasts and npresent on the surface which grewwashed from the casing and placedmedium containing no sorbate. At thconcentrations, inhibitory but nonletof sorbate were apparent.Although 2.5% (wt/vol) potassium

used in the United States for the presurface mold growth on dry cured s;under the conditions of its use heresalami, 2.5% potassium sorbate was iIn fact, higher viable numbers of i

present on 2.5% potassium sorbate-treripened) salami than on the untreatcasings (Table 2). Genoa, Casalingo, a]ese types have diameters of about 8.5.6 to 7.5 cm, respectively. Since it is kssorbate penetrates from the surface 4

be speculated that as sorbate pene

larger diameter salami it was diluted tdegree, which caused a more rapid loiitory ability at the surface.

In contrast to the results found her(al. (15) found that a 2.5% potassium s

at the time of deboning was satisfactory forcontrol of surface mold growth on boneless dry-cured ham. But although those hams were heldlonger (84 days) and part of the time at a highertemperature (24°C for 4 weeks) than used here,the hams were smoked. Elastic netting appearedto retard mold development on inside surfacesbecause the meat pieces were held together bet-

________. ter. Again in contrast, the string on the surfaceof Casalingo salami appeared to elevate the localwater activity and permitted fungal growth un-derneath, particularly yeast growth on pimaricintreatments.

40 50 Kemp et al. (15) found that there was a ben-eficial effect of sorbate treatment on ham ac-

ievelopment ceptability. No off-flavors or off-colors weredry Italian noted on any of the experimental salami used insabfe kyneoaswtns this study, although the string on salami dippedxtrose agar in 10% or greater concentrations of sorbate wassalingo and stained yellow. Leistner et al. (17) did not findre combined off-colors, off-flavors, or odor when salami wereswere taken treated with 20% sorbate after curing in theita are from green room.;t pimaricin Baldock et al. (3) studied the mold-inhibitingy after stuff- properties of 2.5, 5.0, and 10% potassium sorbatev'en at day 5 solutions on fully ripened country cured hams.dce in potas- The sorbate was applied as a 1-mm spray andacetic acid was believed to be partly effective because of(wt/vol)ao. the removal of mold spores by the rinsing actionPate; I, 20% of the spray. This action was seen during pimar-if pimaricin icin spraying of Genoa and Casalingo salami and

may also explain the success of the pimaricinspray over the dip. Baldock et al. (3) found that

on the sur- a spray of 5% sorbate was the lowest effective.mi. At 2.5 concentration which prevented mold develop-nolds were ment over the 30-day storage period. A 10%only when solution was more effective, but inhibition wason growth lost somewhere between 30 and 60 days at 210Cese sorbate and 70% RH.;hal effects To improve statistical evaluation, viable mold

count data for acetate-pretreated Genoa andsorbate is Casalingo salami were pooled and subjected toZvention of the Neuman-Keuls test. Three statistically dif-alami (24), ferent (95% confidence level) groups then be-on Genoa came evident (Fig. 3). The most efficacious treat-ineffective. ment was 20% potassium sorbate. The secondfungi were best treatment was either a 10% sorbate dip orated (fully double 1,000-ppm pimaricin spray, although vis-;ed control ually 10% sorbate performed better on bothnd Abruzz- types of salami. The 2.5% potassium sorbate dip.8, 5.6, and and the untreated control were placed in the last:nown that group as not being significantly different. It(17), it can should be noted that there was visible inhibitiontrated the of surface mold by 2.5% sorbate but only on;o a greater Casalingo salami, and this was not supported byss of inhib- viable count data.

Abruzzese (natural casing) samples were lesse, Kemp et affected by mold growth, but viable count data;orbate dip showed that 10% sorbate treatments gave results

428 HOLLEY

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from

MOLD INHIBITION ON SALAMI CASING 429

equivalent to those obtained after 20% sorbatetreatment. This probably reflects the difficultysurface fungi face when attempting to grow onnatural casings.An examination of seven isolates from the

surface of these salami samples revealed thatthe organisms present, which were mainly Pen-icillium species plus a strain each ofAspergillusand Saccharomyces, were typical of the typesfound on the surface of dry-cured salami (2).

In cases where original mold growth was mod-erate, Leistner et al. (17) found that treatingsausage by dipping in 10% potassium sorbatewas sufficient to protect the meat from surfacemold growth for about 14 days. This concentra-tion of sorbate with an acetate pretreatmentappeared adequate for protection during thematuration of the three types of salami used inthis study.

ACKNOWLEDGMENTSI thank the Multilingual Services Division, Translation

Bureau of the Department of the Secretary of State, Canada,for translation of articles to English and Judy Bohm for hertechnical assistance.

The interest, help, and cooperation of Toni Pistilli andItalo Rosati of Santa Maria Foods Limited are gratefullyacknowledged.

Pimaricin (natamycin) was kindly supplied by P. F. Harri-son, GB Fermentation Industries Inc., Montreal.

LITERATURE CITED1. Ayres, J. C., L. Leistner, M. Sutic, P. E. Koehler, M.

T. Wu, N. A. Halls, E. Strzelecki, and F. Escher.1974. Mold growth and mycotoxin production on agedhams and sausages, p. 218-227. In Proceedings of theIV International Congress on Food Science and Tech-nology, vol. III. Selegraf, Valencia, Spain.

2. Ayres, J. C., D. A. Lillard, and L. Leistner. 1967. Moldripened meat products, p. 156-175. In Proceedings ofthe 20th Annual Reciprocal Meat Conference, Ameri-can Meat Science Association. National Livestock andMeat Board, Chicago, Ill.

3. Baldock, J. D., P. R. Frank, P. P. Graham, and F. F.Ivey. 1979. Potassium sorbate as a fungistatic agent incountry ham processing. J. Food Prot. 42:780-783.

4. Buchanan, R. L, and J. C. Ayres. 1976. Effect ofsodium acetate on growth and aflatoxin production byAspergillus parasiticus NRRL 2999. J. Food Sci. 41:128-132.

5. Bullerman, L. B., and J. C. Ayres. 1968. Aflatoxin-producing potential of fungi isolated from cured andaged meats. Appl. Microbiol. 16:1945-1946.

6. Bullerman, L. B., P. A. Hartman, and J. C. Ayres.1969. Aflatoxin production in meats. II. Aged dry sal-amis and aged country cured hams. Appl. Microbiol.18:718-722.

7. Cattaneo, P., S. D'Aubert, and A. Righetti. 1978. At-tivita antifungina della pimnaricina in Salumi crudi sta-gionati. Ind. Aliment. (Pinerolo, Italy) 17:658-664.

8. deBoer, E., H. Labots, M. Stolk-Horsthuis, and J. N.

Visser. 1979. Sensitivity to natamycin of fungi in fac-tories producting dry sausage. Fleischwirtschaft 59:1868-1869.

9. Deibel, R. H. 1974. Technology of fermented, semi-driedand dried sausages, p. 57-0. In Proceedings of theMeat Industry Research Conference. American MeatInstitute Foundation, Chicago, Ill.

10. Dragoni, I., M. Apostolo, and C. Cantoni. 1978. Evo-luzione della flora microbica superficiale di insaccatistagionati. Ind. Aliment. (Pinerolo, Italy) 17:759-760.

11. Erickson, F. J., and F. W. Fabian. 1942. Preserving andgermicidal action of various sugars and organic acids onyeasts and bacteria. Food Res. 7:68-79.

12. Genigeorgis, C. A. 1976. Quality control for fermentedmeats. J. Am. Vet. Med. Assoc. 169:1220-1228.

13. Hayashi, K., M. Terada, T. MizunuJma, and T. Yok-otauka. 1979. Retarding effect of acetic acid on growthof contaminated bacteria during shoyu-koji makingprocess. J. Food Sci. 44:359-362.

14. Hechelman, H., and L Leistner. 1969. Hemmung vonunerwunschtem Schimmelpilzwachstum auf Rohwur-sten durch Delvocid (Pimaricin). Fleischwirtschaft 49:1639.

15. Kemp, J. D., B. E. Langlois, M. B. Solomon, and J. D.Fox. 1979. Quality of boneless dry-cured ham producedwith or without nitrate, netting or potassium sorbate. J.Food Sci. 44:914-915.

16. Lee, I. C., L. G. Harmon, and J. F. Price. 1977. Growthand enterotoxin production by staphylococci in Genoasalami. J. Food Prot. 40:325-329.

17. Leistner, L., L. Y. Maing, and E. Bergmann. 1975.Verhinderung von unerwuenschtem Schimmelpilz-wachstum auf Rohwurst durch Kaliumsorbat. Fleis-chwirtschaft 55:559-561.

18. Levine, A. S., and C. R. Fellers. 1940. Action of aceticacid on food spoilage microorganisms. J. Bacteriol. 39:499-514.

19. Moerman, P. C. 1972. Schimmelwering op vleeswarendoor Pimaricine. Voedingsmiddelen Technol. 3:261-264.

20. Palumbo, S. A., J. L. Smith, and S. A. Ackerman.1973. Lebanon bologna. 1. Manufacture and processing.J. Milk Food Technol. 36:497-503.

21. Pullen, M. M., and C. A. Genigeorgis. 1977. A study ofcoagulase-positive staphylococci in salami before fer-mentation. J. Food Prot. 40:704-708.

22. Raccach, M., and R. C. Baker. 1979. Fermented me-chanically deboned poultry meat and survival of Staph-ylococcus aureus. J. Food Prot. 42:214-217.

23. Snijders, J. M. A., M. J. G. Schoenmakers, G. E.Gerats, and F. W. Pijper. 1979. DekontaminationSchlachtwarmer Rinder Korper mit organischen Sau-ren. Fleischwirtschaft 59:656-663.

24. Terrell, R. N., G. C. Smith, and Z. L. Carpenter. 1978.Practical manufacturing technology for dry and semi-dry sausage, p. 39-44. In Proceedings of the 30th AnnualReciprocal Meat Conference, American Meat ScienceAssociation. National Livestock and Meat Board, Chi-cago, Ill.

25. Vicini, E., and R. G. Racznski. 1977. Indagini sulla floramicrobica del salame tipo "Felino.' Ind. Conserve 52:235-242.

26. Wallhausser, K. H., and E. Luck. 1978. Zur Wirkungvon Sorbinsaure gegen Mykotoxinbildner. Z. Lebensm.Unters. Forsch. 167:156-157.

27. Zivin, J. A., and J. J. Bartko. 1976. Statistics for dis-interested scientists. Life Sci. 18:15-26.

VOL. 41, 1981

on May 29, 2018 by guest

http://aem.asm

.org/D

ownloaded from