Influence of Antimicrobial Agents on the Spoilage of a Meat-Based Entomophage Diet

BIOLOGICAL AND MICROBIAL CONTROL Inﬂuence of Antimicrobial Agents on the Spoilage of a Meat-Based Entomophage Diet

1 2,3 G. D. INGLIS AND A. C. COHEN

J. Econ. Entomol. 97(2): 235Ð250 (2004) ABSTRACT The microbial decomposition of a meat-based entomophage diet presented in ParaÞlm packets was investigated. Considerable bacteria but not fungi were associated with components used to prepare the diet (i.e., hensÕ eggs, liver, and ground beef). At the initial sampling time, there were no differences among diet treatments in the size of bacterial or fungal populations. Bacterial populations in diets not containing antibacterial agents rapidly increased and reached an asymptote by 24h (Ϸ1010 colony-forming units per gram). Bacterial populations also increased in diets containing antibacterial agents, but they were signiÞcantly smaller than in diets not containing antibacterial agents. The most prevalent bacteria isolated were Carnobacterium piscicola, Carnobacterium divergens, Lactobacillus curvatus, Lactobacillus sakei, Leuconostoc mesenteroides, and Enterococcus spp., regardless of the antibacterial treatment used. The proliferation of fungi was delayed relative to bacteria, but signiÞcant differences were observed among the diet treatments. Fungi were most inhibited by sorbic acid and propionic acid in the absence of antibacterial agents. The most common fungi isolated were the yeasts Candida zeylanoides, Torulaspora globosa, and Yarrowia lipolytica. The pH of diets not containing antibacterial agents decreased rapidly and was highly correlated with increases in bacteria but not fungi. The results of this study demonstrate that antimicrobial agents signiÞcantly inhibit spoilage microorganisms in a meat-based diet and that alternative management strategies to delay the decomposition of such diets presented in ParaÞlm packets should target lactic acid spoilage bacteria, particularly Carnobacterium and Lactobacillus species.

KEY WORDS artiÞcial diet, entomophagous, lacewings, Chrysoperla, spoilage

THE PRODUCTION OF ENTOMOPHAGOUS insect species for hen 1995). The presentation of solids allows the pred- biological control is a rapidly expanding industry, and ators to recapture their digestive enzymes for further success of this industry is dependent on the econom- use in their gut, a very important aspect of their normal ical production of high-quality insects. Chrysoperla feeding process (Cohen 1998a). Based on the physi- species (lacewings) are the most widely used preda- ology of food utilization in entomophagous insects, a tory insects in biological control programs, and until semisolid diet comprised of cooked eggs, and raw liver, recently, commercial production of Chrysoperla and ground beef, brewerÕs yeast, sucrose, honey, and ace- other entomophagous insects relied primarily on the tic acid was developed by Cohen and Smith (1998). use of natural hosts. Considerable effort has gone into This diet has demonstrated its effectiveness for rearing the development of artiÞcial diets in an attempt to entomophagous insects, and it is currently being used reduce the cost of production and thereby enhance commercially to produce lacewings (Chrysoperla car- the potential for success of augmentative biological nea [Stevens] s. lat.). control programs with Chrysoperla. Despite the obvi- Despite the tremendous advantages of the meat- ous advantages of an artiÞcial diet to the industry, based diet for producing entomophagous insects, the early attempts to develop diets were met with limited requirement for ground beef, beef liver, and whole success, and as recently as 1994, no artiÞcial diet-based eggs, and other ingredients such as brewersÕ yeast, is commercial rearing of entomophages existed (Gre- not only an inviting substrate for microbial catabolism nier et al. 1994). The realization that most entomopha- but also contains many microbial contaminants (Jay gous insects, including chrysopids, use extra-oral di- 1996). Microbial contaminants have been shown to gestion to process complex solid substrates has directly and/or indirectly affect insect health. For facilitated the development of solid-based diets (Co- example, they have been shown to incite mortality (Sikorowski et al. 1980, 1992), reduce the size of in- 1 Agriculture and Agri-Food Canada, Lethbridge, AB T1J 4B1, Can- sects and delay development (Sikorowski and Thomp- ada. son 1984), cause damage to the midgut (McLaughlin 2 Biological Control and Mass Rearing Research Unit, USDAÐARS, Mississippi State, MS 39762. and Sikorowski 1978, MacGown and Sikorowski 1980), 3 Current address: Insect Diet and Rearing Institute, Tucson, AZ decrease pheromone production (Gueldner et al. 85728. 1977, Wiygul and Sikorowski 1986), decrease quanti- 236 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2 ties of amino acids and fatty acids within the body with heat, 1% sodium hypochlorite and/or 70% etha- (Thompson et al. 1977, Thompson and Sikorowski nol before use. The meat-based entomophage diet 1978), increase amounts of uric acid in insect blood consisted of the following ingredients: 300 g of ground (Thompson and Sikorowski 1982, 1984), and increase beef (Ϸ30% fat), 300 g of beef liver, 300 g of hensÕ eggs, oxygen utilization (Wiygul and Sikorowski 1981). Al- 45 g of sucrose, 30 g of brewerÕs yeast (Saccharomyces though the adverse effects of microbial contaminants cerevisiae), 75 g of honey (20% solution), 15 ml of on insects may be dire, the economics of diet prepa- acetic acid (10% solution), and 195 ml of tap water. ration also is an important consideration. The cost of The antimicrobial agents incorporated into the diet the diet ingredients and of the labor involved with its were the antibacterial agents streptomycin sulfate (0.3 preparation are signiÞcant, and increasing the time g) (Sigma-Genosys, St. Louis, MO) and chlortetracy- that insects can feed on the diet without compromis- cline (0.3 g) (Sigma-Genosys) and the antifungal ing their health (i.e., due to microbial spoilage) will agents propionic acid (1.8 g) (Sigma-Genosys) and have a positive impact on the economics of produc- potassium sorbate (1.8 g) (Sigma-Genosys). Initially, tion. For example, packets containing the meat-based the ground beef, beef liver, brewerÕs yeast, honey entomophage diet are placed on the top of cages solution, and 60 ml of tap water were homogenized in containing lacewings, and the packets are replaced a Hamilton Beach blender (high setting) for Ϸ2 min. approximately every 48 h due to compromised quality. Representative samples of ground beef, liver, eggs Delaying the replacement time by even 24h would (i.e., obtained from a local grocery store), and tap substantially decrease costs of production. water (containing honey) were taken before their Antimicrobial agents, including the antibiotic placement in the blender. In a 4-liter beaker, 135 ml of agents streptomycin sulfate and chlortetracycline, are water was heated to 85Ð95ЊC, and sucrose was added currently incorporated into the diet in an attempt to and allowed to dissolve. Acetic acid and the appro- suppress spoilage microorganisms. These antibiotics priate antimicrobial agents (i.e., according to treat- have a minimal impact on C . carnea larvae (Cohen and ment) were subsequently added to the sucrose solu- Smith 1998), but their efÞcacy in preventing spoilage tion, followed by the hensÕ eggs. In a preliminary has not been investigated. In the case of the meat- experiment, the sucrose and acetic acid solutions were based entomophage diet, conventional sterilization shown to be free of microorganisms. Given the small and pasteurization methods cannot be used because amount of antimicrobial agents added (Ͻ0.2%), no the chemical and physical changes that occur are very attempt was made to compensate for volume differ- detrimental to Chrysoperla larvae (A.C.C., unpub- ences in diet treatments missing two or more agents. lished data). In the majority of diets, the microbiology The mixture was stirred constantly until the eggs had of spoilage is poorly understood, and insect diet pro- gelled but were of a sticky, stringy consistency (Ϸ2Ð3 fessionals often rely on a relatively limited number of min). The meat was then added to the hot egg slurry wide-spectrum antimicrobial agents. To illustrate this, and blended for 3Ð5 min until the entire mixture was in a survey of 58 synthetic diets used for phytophagous of a stringy, paste-like consistency. The following an- insects, 32% used chlortetracycline, 100% used meth- timicrobial treatments were included in the experi- yl-p-hydroxybenzoate, 10% used sodium benzoate and ment: 1) diet one contained both antibacterial (strep- formaldehyde, and 72% used sorbic acid (Singh and tomycin sulfate and chlortetracycline) and antifungal House 1970). A reliance on antimicrobial agents is not agents (propionic acid and potassium sorbate); 2) diet without its problems. Antibiotics are expensive, and two contained antibacterial agents only; 3) diet three they can be toxic to insects (Sikorowski et al. 1984). contained antifungal agents only; and 4) diet four Furthermore, serious concerns exist about the poten- contained neither antibacterial or antifungal agents. tial development of bacterial resistance, particularly The order in which the diet treatments were prepared with regard to the prophylactic use of antibiotics. was randomly selected in each replicate, and prepared Research to elucidate the microbial taxa responsible diets were stored aseptically at 5ЊC until they could be for spoilage is a prerequisite for the development of dispensed (Ϸ1 h). nonantibiotic methods to suppress spoilage microor- The diets were aseptically dispensed into diet pack- ganisms, and thereby reduce their adverse impact on ets in a laminar ßow hood. Each diet packet was made diet and subsequently insect quality. Using a meat- from ParaÞlm M (Fisher, Atlanta, GA) (oxygen per- based entomophage diet, the objectives of the study meability Ϸ150 cc/m2/24h and carbon dioxide per - were to 1) measure changes in bacterial and fungal meability Ϸ400 cc/m2/24hat23ЊC and 80% RH) and populations over time; 2) quantify temporal changes was Ϸ10 by 15 cm in dimension. The packet was folded in diet pH; 3) identify the bacterial and fungal taxa over on itself, and sealed on two sides with a heat responsible for decomposition; and 4) measure the sealer (Ϸ1 cm from the edge of the packet) creating efÞcacy of conventional antimicrobial agents on fun- a packet with three sealed sides. The prepared diet was gal and bacterial populations. then dispensed into the packets (20Ð25 g per packet), and the open end was subsequently sealed with the heat sealer (1 cm from the edge). The diet was evenly Materials and Methods distributed throughout the packet resulting in a thick- Preparation and Maintenance of Diet. The diet ness of Ϸ2Ð3 mm. Diet packets were immediately (Cohen 1998b) was prepared and dispensed in a lam- placed in a controlled environment chamber (CEC) inar airßow hood, and all equipment was sanitized at 27ЊC in the dark, and the various treatment-time April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 237

combinations were arbitrarily arranged within the slant cultures containing the same medium on which CEC. A temperature micrologger (HOBO H8 Pro, they were isolated, the slant cultures were incubated Onset Computer Corp., Bourne, MA) was placed in at 27ЊC, and isolates exhibiting good growth after 48Ð the CEC during the experiment, and the temperature 120 h were placed at 5ЊC. After Ϸ6 mo, isolates were within the chamber was found to be within Ϯ 1ЊCof transferred to a fresh slant tube, grown at 27ЊC, and the target temperature. restored at 5ЊC. In several instances, it was necessary Enumeration and Isolation of Microorganisms. to resurrect isolates on blood agar before identiÞca- Ground beef, liver, egg, and water samples were pro- tion or restorage. cessed for microorganisms within3hofdiet prepa- Measurement of pH. For pH determinations,1gof ration (samples were aseptically maintained at 5ЊC diet was mixed with 4ml of double distilled water, during this period). Approximately 1 g each of ground mixed thoroughly (Ϸ5 min), and the pH was mea- beef, liver, and egg (before cooking) for each treat- sured using an Accumet AB15 pH meter equipped ment was homogenized in 4ml of sterile 100 mM with an Accumet probe (13-620-531, Fisher). potassium phosphate buffer (pH 7.0) by using a tissue Characterization of Bacteria. Bacteria were tearer at high setting (Biospec Products Inc., Bartles- grouped according to colony morphology, cell shape ville, OK). The tissue homogenizer was surface-sani- and size, oxidase, catalase, gram reaction, and on their tized in 1% sodium hypochlorite followed by three ability to assimilate carbohydrates by using Biolog rinses in sterilize distilled water between samples, and microplates (Biolog System, Hayward, CA). The fatty sterility was conÞrmed in a preliminary experiment. acid proÞles of selected representatives from each The meat and egg suspensions, and water sample were group were determined using gas-liquid chromatog- diluted three to Þve times in a 10-fold dilution series, raphy (MIDI Inc., Newark, DE) according to a stan- and 100-␮l aliquots for each dilution were spread onto dard protocol (Sasser 1990). To obtain biomass for trypticase soy agar (TSA) (Difco, Detroit, MI) MIDI, bacteria were grown on TSA, and if necessary, amended with 200 ␮g/ml cycloheximide (Sigma-Ge- on MRS or Preston blood-free medium. nosys) for the isolation of bacteria, and SabouraudÕs Based on the groupings from the MIDI analyses, dextrose agar (SDA) (Difco) containing 100 ␮g/ml identiÞcations for representative isolates were ob- each of tetracycline (Sigma-Genosys) and chloram- tained from 16S rRNA genes. The 16S rRNA gene was phenicol (Sigma-Genosys) for the isolation of yeasts ampliÞed by polymerase chain reaction (PCR) by and Þlamentous fungi. All cultures were incubated at using the eubacterial primers UNI27 F (5Ј-AGA GTT 27ЊC for 48Ð120 h, and where possible, the number of TGA TCC TGG CTC AG-3Ј) and UNI1492R (5Ј-TAC colonies was counted at the dilution yielding 30Ð300 GG(C/T) TAC CTT GTT ACG ACT-3Ј). The ampli- colony-forming units (CFU) per dish. Dry weights Þcation program consisted of an initial step at 95ЊC for were determined by drying samples at 65ЊC for 48 h, 15 min, followed by 30 s at 94ЊC, 1 min 30 s at 58ЊC, and and the number of CFU/g (dry weight) of diet was 2 min at 72ЊC for 35 cycles, terminating with a 10-min calculated. extension at 72ЊC. The reaction mixture consisted of a Microorganisms within the diet also were also iso- total volume of 20 ␮l and contained 1ϫ reaction ␮ lated. An arbitrarily selected packet for each diet buffer, 0.2 mM dNTPs, 2.0 mM MgCl2, 0.5 M of each treatment was collected immediately after prepara- primer (Sigma-Genosys, Oakville, ON, Canada), 0.2 tion (Ϸ1 h), and at 6, 12, 24, 48, 72, and 96 h after ␮g of bovine serum albumin (Promega, Madison, WI), placement at 27ЊC. The surface of the packet was and 0.0025 U of HotStar Taq polymerase (QIAGEN, swabbed with 70% ethanol, and two perpendicular Mississauga, ON, Canada). The PCR products were cuts were aseptically made into the upper layer of puriÞed using the PCR puriÞcation kit (QIAGEN) ParaÞlm M dividing the packet into four quadrants. according to the manufacturerÕs instructions. To ob- From each quadrant, Ϸ0.25-g samples of diet were tain partial sequences, the UNI338 F (5Ј-ACT CCT aseptically removed, added to 4ml of sterile buffer, ACG GGA GGC AG-3Ј) primer was used with the ABI and the diet was homogenized in a Potter-Eljvehjm PRISM Big Dye Terminator cycle sequencing ready tube (Fisher). The suspension was diluted eight times reaction kit (Applied Biosystems, Foster City, CA). in a 10-fold dilution series, and 100-␮l aliquots for each Before sequencing, excess dye was removed with the dilution were spread onto TSA with cycloheximide DyeEx spin kit (QIAGEN). Sequences were obtained and SDA with tetracycline and chloramphenicol. All with an ABI PRISM 377 automated DNA sequencer cultures were incubated at 27ЊC for 48Ð120 h, and the (Applied Biosystems). Sequences were compared di- number of colonies was counted at the appropriate rectly with the Ribosomal Database Project II data- dilution. Similarly to the meat and egg ingredients, diet base by using Sequence Match for prokaryote small dry weights were determined by drying samples at subunit rDNA. If necessary, sequence data were also 65ЊC for 48 h. compared with the National Center for Biotechnology To isolate bacteria and fungi, a petri dish with rep- Information GenBank database using BLASTN. resentative colony number and diversity was chosen. IdentiÞcations of all isolates of prominent taxa Colonies were selected on the basis of colony appear- within the MIDI and 16S groups were subsequently ance and prevalence. In addition, rare occurring but conÞrmed using genera- and/or species-speciÞc prim- unique colonies were sampled. All isolates were ers according to Yost and Nattress (2000). The am- streaked for purity and their colony morphologies pliÞcation program consisted of an initial step at 95ЊC were conÞrmed. Isolates where then transferred to for 15 min, followed by 45 s at 94ЊC,45sat55ЊC, and 238 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

1 min at 72ЊC for 35 cycles, terminating with a 10-min Table 1. Total bacterial and fungal populations associated with extension at 72ЊC. The same reaction mixture and ground beef, beef liver, and hens’ eggs just before the preparation of the meat-based entomophage diet amount of template as indicated previously was used. ␮ Each PCR reaction was performed with a total of 1 l Bacteria Fungi Substrate/Treatment of a suspension of cells in exponential growth phase. Log CFU/g (dry weight) For Carnobacterium species, the primers Cb1 (5Ј-CCG Ј Ground beef TCA GGG GAT GAG CAG TTA C-3 ) and Cb2r Diet 1a 7.06 Ϯ 0.64b 2.84 Ϯ 0.24 (5Ј-ACA TTC GGA AAC GGA TGC TAA T-3Ј) spe- Diet 2 7.17 Ϯ 0.61 2.91 Ϯ 0.20 ciÞc for the 16S rRNA gene were used (Nissen et al. Diet 3 6.96 Ϯ 0.58 2.56 Ϯ 0.33 1994). The primer set of Cb1-Cb2r possess a high level Diet 47.15 Ϯ 0.71 2.72 Ϯ 0.26 Beef liver of homology with Enterococcus, and ampliÞcation Diet 1 5.15 Ϯ 0.58 2.20 Ϯ 0.35 products are also obtained for Enterococcus species Diet 2 4.78 Ϯ 0.53 2.35 Ϯ 0.32 (Scarpellini et al. 2002). To distinguish, between Car- Diet 3 4.59 Ϯ 0.58 2.05 Ϯ 0.36 nobacterium and Enterococcus, bacteria were observed Diet 44.84 Ϯ 0.86 2.30 Ϯ 0.37 HensÕ eggs microscopically to determine whether cells were rods Diet 1 3.57 Ϯ 1.07 2.47 Ϯ 0.76 or cocci. Carnobacterium isolates were identiÞed to Diet 2 3.58 Ϯ 1.88 3.42 Ϯ 1.72 species based on their ability to produce acid from Diet 3 4.05 Ϯ 0.82 2.79 Ϯ 1.09 Ϯ Ϯ inulin, mannitol, ␣-methyl-D-glucoside and ␣-methyl- Diet 44.21 1.26 3.72 0.48 D-mannoside (Collins et al. 1987). For Lac. sakei, the Ј Ј The meat-based diet was comprised of ground beef, beef liver, hensÕ primers 16 (5 -GCT GGA TCA CCT CCT TTC-3 ) and eggs, sucrose, Brewers yeast, honey, acetic acid, and tap water. Ls (5Ј-ATG AAA CTA TTA AAT TGG TAC-3Ј) coding a Treatments consisted of 1) diet containing both antibacterial for the eubacterial 16S rRNA gene and Lac. sakei streptomyoin sulfate and chloretetracycline) and antifungal agents 16S/23S rRNA gene spacer region, respectively, were propionic acid and potassium sorbate); 2) diet containing antibacterial agents only; 3) diet containing antifungal agents only; or 4) diet used (Berthier and Erhlich 1998). For Lac. curvatus, containing neither antibacterial or antifungal agents. the primers 16 and Lc (5Ј-TTG GTA CTA TTT AAT b Mean Ϯ SEM. TCT TAG-3Ј) and Lac. curvatus 16S/23S rRNA gene spacer region, respectively, were used (Berthier and Erhlich 1998). For Leuconostoc species, the primers Lu1r (5Ј-CCA CAG CGA AAG GTG CTT GCA C-3Ј) mixed procedure of SAS (SAS Institute 1999). Repli- and Lu2 (5Ј-GAT CCA TCT CTA GGT GAC GCC cates were conducted on three separate occasions, G-3Ј) coding for the Leuconostoc 16S rRNA gene were and they were treated as a random effect; and values used (Nissen et al. 1994). IdentiÞcation of Leu. mes- indicated by a “Ϯ” represent standard errors of the enteroides isolates to species was accomplished by mean. The univariate procedure of SAS was used to

sequencing of the 16S rRNA gene, and/or by using the ensure normality (using residual values). A log10 Biolog identiÞcation system. transformation to the microbial CFU data was neces- Characterization of Fungi. Yeasts were initially sary, but all other dependent variables were untrans- grouped based on colony morphology, cell shape and formed. In conjunction with a signiÞcant F test, the size, diazonium blue B reaction, urea hydrolysis, as- least-square means function of SAS was used to com- similation of L-arabinose, D-galactose, glucose, glyc- pare means. PearsonÕs correlation coefÞcients be- erol, inositol, lactose, maltose, rafÞnose, sucrose, me- tween log bacterial and fungal populations with pH of lezitose, and xylitol by using a disk method (Van der the diet were determined using the correlation pro- Walt and Yarrow 1984). In addition, the ability of cedure of SAS; before analysis, variables were plotted yeasts to ferment glucose, galactose, lactose, maltose, and the relationships seemed to be linear. The relative

sucrose, and trehalose were determined using the frequency of microbial taxa was calculated as log10 Wickerham method (Van der Walt and Yarrow 1984). [CFU/g (dry weight) times the proportion of the Tentative identiÞcations of representative isolates taxon]. The diversity index of bacteria isolated from from each group were then obtained using the auto- the meat-based diet was calculated as the number of mated Biolog System. IdentiÞcations were conÞrmed taxa isolated (across the three replicates) at each time using additional tests (e.g., nitrogen assimilation, cy- divided by the total number of taxa isolated (n ϭ 50). cloheximide resistance, growth at high temperatures) according to Barnett et al. (1990), Kurtzman and Fell Results (1998), and Van der Walt and Yarrow (1984). Fila- mentous fungi were identiÞed based on morphologi- Microorganisms Associated with Diet Ingredients. cal characteristics of conidiogenesis according to stan- Considerable numbers of bacteria were isolated from dard references. Where possible, at least one ground beef, beef liver, and hensÕ eggs at the time of representative isolate for each bacterial or fungal diet preparation (Table 1). Substantially more bacte- taxon recovered were stored in glycerol at Ϫ80ЊCasa ria were recovered from ground beef than from either reference strain. liver or eggs (F ϭ 16.1; df ϭ 2, 19; P Ͻ 0.001). There Statistical Analyses. The diet spoilage experiment were no differences in numbers of bacteria associated was arranged as randomized complete block design with the diet ingredients use to prepare the diets (i.e., and was analyzed as a factorial experiment with four among the diet treatments) (F ϭ 0.1; df ϭ 3, 19; P Ն levels of treatment and seven levels of time using the 0.99). In contrast to bacteria, relatively small numbers April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 239

Table 2. Bacteria isolated from the ingredients (hens’ eggs, beef liver, ground beef [GB]) and the meat-based encomophage diet

Taxon Eggs Liver GB Diet Taxon Eggs Liver GB Diet Acetobacter sp. Ea ❶b EELactococcus garvieae E ❷❶❷ Acinetobacter iwofﬁi E ❶ E ❷ Lactococcus lactis EE❶❶ Acinetobacter sp. EEE❶ Lactococcus plantarum EE❶ E Ancylobacter sp. EEE❶ Leuconostoc mesenteroides E ❷❶❸ Arthrobacter agilis EEE❶ Leuteococcus japonicus EEE❶ Arthrobacter atrocyaneus E ❶❶❶Microbacterium liquefaciens E ❶ EE Arthrobacter ilicis EE❶❶Microbacterium terregens E ❶ E ❶ Arthrobacter sp. E ❶ E ❶ Microbacterium sp. ❶ EE❶ Bacillus cereus EE❶❶Micrococcus luteus E ❷ E ❶ Bacillus megaterium E ❶❶❷Moraxella sp. E ❶❶❶ Bacillus sphaericus ❶❶❶❶Neisseria sp.c E ❶ EE Bacillus subtilis EEE❶ Ochrobactrum grignonense EEE❶ Brevibacterium sp. EEE❶ Paenibacillus polymyxa EEE❶ Brochothrix thermosphacta EE❶ E Paracoccus sp. ❶ EEE Carnobacterium divergens E ❸❸❸Pediococcus pentosaceus E ❶ EE Carnobacterium piscicola E ❸❷❷Planococcus okeanokoites EEE❶ Carnobacterium sp. EE❶❸Pseudomonas fragi EE❶❶ Citrobacter sp. EEE❶ Pseudomonas putida EE❷❷ Cornyebacterium callunae EEE❶ Pseudomonas sp. EE❶❷ Corynebacterium sp. EEE❶ Rhodococcus equi ❶ EE❶ Curtobacterium albidumc EE❶ E Rhodococcus erythropolis EEE❶ Enterobacter cloacae EEE❶ Rhodococcus sp. E ❶ EE Enterobacter intermedius EEE❶ Sphingobacterium multivorum ❶ EEE Enterococcus faecalis E ❶ E ❶ Staphylococcus aureus E ❶ EE Enterococcus faecium E ❶ EEStaphylococcus epidermidis ❸ E ❶❶ Enterococcus hirae EEE❶ Staphylococcus hominis ❶ EEE Enterococcus sulfureus EEE❷ Staphylococcus pasteuri ❷ EEE Enterococcus sp. E ❷❷❸Staphylococcus sciuri EE❶ E Klebsiella terrigena EEE❶ Staphylococcus warneri ❶ EEE Kocuria kristinae EEE❶ Stenotrophomonas maltophilia EEE❶ Kocuria varians ❶ E ❶❶Streptococcus pyogenes E ❶❶❷ Kurthia gibsonii EEE❶ Terrabacter sp.c E ❶ EE Lactobacillus curvatus E ❶❶❸Yersinia intermedia EEE❶ Lactobacillus sakei ❶❷❷❸Unidentifed ❷❸❷❸

The meat-based diet was comprised of ground beef, beef liver, hensÕ eggs, sucrose, brewerÕs yeast, honey, acetic acid, and tap water. a Empty circles indicate that the corresponding bacterium was not isolated. b Dark circles indicate positive isolation, and the number within the symbol indicates the number of replicates (x of 3) from which the taxon was isolated. c IdentiÞcations indeÞnite.

of fungi were recovered from all three substrates (Ta- cluded Can. zeylanoides, Cry. curvatus, Penicillium ble 1). A slight but signiÞcant difference in fungal citrinum, and Yar. lipolytica (Table 3). Six of 10 (60%), populations were observed among the different in- and 8 of 9 (89%) of the taxa isolated from liver and gredients (F ϭ 3.8; df ϭ 2, 19; P ϭ 0.041). More fungi ground beef, respectively, were subsequently isolated were recovered from eggs than beef liver (P ϭ 0.014), in the entomophage diet. Only Fusarium spp. was but similar numbers were isolated between ground isolated from eggs and subsequently recovered from beef and eggs (P ϭ 0.23), and between ground beef the diet. and liver (P ϭ 0.12). There was no difference in fungal populations among the diet treatments (F ϭ 0.82; df ϭ Bacterial Populations in the Diet. 3, 19; P ϭ 0.50). No microorganisms were recovered from water containing honey. The rate of proliferation of bacteria differed among The most frequently isolated bacteria from eggs the four diet treatments over the 96-h duration of the were Staphylococcus species, including Sta. epidermi- experiment (F ϭ 5.1; df ϭ 18, 54; P Ͻ 0.001) (Fig. 1A). dis, Sta. hominis, Sta. pasteuri, and Sta. warneri (Table At the initial sampling time, there were no differences 2). From liver, Car. divergens, Car. piscicola, Entero- (P Ͼ 0.05) in bacterial populations (5.5Ð6.0 log coccus spp., Lac. sakei, Lac. garvieae, Leu. mesen- CFU/g) among the treatments. At the 6-h sampling teroides, and Mic. luteus were commonly isolated. Car. time, populations in the diets not containing antibac- divergens, Car. piscicola, Enterococcus spp., Lac. sakei, terial agents (diets 3 and 4) were larger (P Ͻ 0.05) and Pseudomonas putida were most frequently iso- than the diets containing antibacterial agents (diets 1 lated from ground beef. Seven of 11 (64%), 17 of 25 and 2). Bacterial numbers in these diets not containing (68%), and 21 of 25 (84%) of the taxa isolated from antibacterial agents continued to increase rapidly un- eggs, liver, and ground beef, respectively, were sub- til 24h, at which point populations had reached an sequently isolated in the entomophage diet. Fre- asymptote at Ϸ9.6 log CFU/g. At the 6- and 12-h quently isolated fungi from the diet ingredients in- sampling times, a difference between the two diets not 240 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

Table 3. Filamentous fungi and yeasts isolated from the ingredients (hens’ eggs, beef liver, ground beef [GB]) and the meat-based entomophage dieta

Group/Taxon Eggs Liver GB Diet Filamentous fungi Aspergillus ustus Ea EE❶c Cunninghamella elegans E ❶ E ❶ Fusarium sp. ❶❶❶❶ Mucor hiemalis EEE❶ Mucor sp. EEE❶ Penicillium citrinum E ❷❷❸ Penicillium janthinellum E ❶ EE Penicillium oxalicum EEE❶ Penicillium simplicissimum E ❶ EE Penicillium sp. E ❶ E ❸ Trichoderma sp. EEE❶ UnidentiÞed E ❶ E ❶ Yeasts Candida albicans E ❶ EE Candida glabrata EEE❷ Candida parapsilosis EEE❷ Candida pararugosa EE❶❸ Candida zeylanoides EE❸❸ Cryptococcus curvatus EE❷❸ Cryptococcus marinus EE❶❷ Leucosporidium scottiic EEE❶ Pichia pastorisd EEE❶ Rhodotorula acheniorumc EE❶❷ Rhodotorula minuta EE❶ E Saccharomyces dairensis E ❶ E ❶ Schizoblastosporion starkeyi-henriciic E ❶ EE Torulaspora globosa EEE❶ Trichosporon brassicaec EEE❶ Yarrowia lipolytica E ❷❷❸ UnidentiÞed ❶❶❷❸

Fig. 1. Changes in pH and microbial populations (log10 containing antibacterial agents was observed; more CFU/g dry weight) recovered from the meat-based ento- (P Յ 0.05) bacteria were isolated from the control diet mophage diet over a 96-h period. (A) Bacteria isolated on relative to diet three (i.e., no antibacterial with anti- trypticase soy agar; (B) fungi isolated on SabouraudÕs dextrose agar; and (C) pH. Diet treatments consisted of 1) diet fungal agents). Bacteria in diets containing antibac- Յ containing antibacterial agents (chloramphenicol and tetra- terial agents were signiÞcantly inhibited (P 0.02) cycline) and antifungal agents (sorbic acid and propionic relative to those not containing antibacterial agents acid); 2) diet containing antibacterial agents but no antifun- after the initial sampling time. There was no difference gal agents; 3) diet containing antifungal agents but no anti- (P Ͼ 0.05) in populations of bacteria in diets 1 and 2 bacterial agents; and 4) diet containing neither antibacterial until the 96-h sample time, at which point, more (P ϭ or antifungal agents. The minimum detection level for both 0.008) bacteria were observed in the diet one (i.e., fungi and bacteria was 50 CFU/g, and bars represent SEM ϭ containing both antibacterial and antifungal agents). (n 3). SEM bars not observed are within markers. In diets containing antibacterial agents, substantially greater variability was observed among the replicates. and there was no obvious selection of particular taxa In numerous instances, conspicuous buildup of gases resulting from the presence of the antimicrobial treat- (i.e., swelling) of the ParaÞlm M packets containing ments. The diversity index of bacteria in diets con- diets 3 and 4were observed at the 24Ð96-h sampling taining antibacterial agents was 0.24and 0.26 com- times. pared with 0.16 and 0.18 in diets not containing In excess of 700 bacterial isolates were character- antibacterial agents (Table 4). In general the diversity ized from the various diet treatments. The primary of bacteria recovered from all diet treatment de- taxa recovered from all four diet treatments included creased with incubation time. Populations of Lac. sakei Car. divergens, Car. piscicola, Lac. curvatus, Lac. sakei, and to a lesser extent, Car. divergens, were substan- and Leu. mesenteroides (Table 4). Substantial popula- tially larger in diets not containing antibacterial agents tions of these taxa were observed in all diet treatments, (Fig. 2). Considerably fewer Carnobacterium spp. but April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 241

Table 4. Bacteria isolated from the meat-based entomophage diet at various times (h) after placement at 27°C

Treatment/Taxona 0 6 12 244872 96 Diet 1b Acinetobacter iwofﬁi Ec EEEEEE Ancyclobacter sp. EEE❶d EEE Arthrobacter ilicis EEE❶ EE❶ Arthrobacter sp. E ❶ EEEEE Bacillus cereus ❶❶❶❶❶❶E Bacillus megaterium ❶ EEEEEE Bacillus sphaericus EE❶ E ❶ EE Bacillus subtilis ❶ EEEEEE Carnobacterium divergens ❷❸❸❷❷❶❷ Carnobacterium piscicola ❶❷❷❶❶❶E Carnobacterium sp. EEEEEEE Corynebacterium sp. ❶ E ❶ EEEE Enterococcus sp. E ❶❶❶E ❶❶ Kocuria varians EE❶ EEEE Lactobacillus curvatus E ❶ EE❶❶❶ Lactobacillus sakei ❶❶EE❸❸❸ Lactococcus garvieae ❶ EEEEEE Leuconostoc mesenteroides E ❶ E ❶ E ❶ E Micrococcus luteus EEE❶ EEE Microbacterium terregens EEE❶ EEE Microbacterium sp. EE❶ EEEE Moraxella sp. EEE❶ EEE Planococcus okeanokoites E ❶ EEEEE Pseudomonas fragi ❶ EEEEEE Pseudomonas putida ❷ EE❶ E ❶ E Pseudomonas sp. EEE❶ EEE Rhodococcus equi EE❶ EEEE Rhodococcus erythropolis EEEE❶ EE Staphylococcus epidermidis EEEE❶ EE Streptococcus pyogenes ❶ EEEEEE Yersinia intermedia ❶ EEEEEE UnidentiÞed ❷❷❶❶E ❶❶ Diversity indexe 0.240.18 0.18 0.240.16 0.18 0.10 Diet 2 Acinetobacter iwofﬁi EEE❶ EEE Acinetobacter sp. EEE❶ EEE Arthrobacter atrocyaneus EE❶ EEEE Bacillus cereus ❶ E ❶ EEEE Bacillus megaterium ❶ EEEEEE Bacillus sphaericus ❶ EEEEEE Brevibacterium sp. ❶ EEEEEE Carnobacterium divergens ❷❶❶❶❷❶E Carnobacterium piscicola ❷❷❷❶E ❶❶ Carnobacterium sp. EEEE❶ EE Enterobacter intermedius EE❶ EEEE Enterococcus cloacae EEEEE❶❶ Enterococcus hirae ❶ EEEEEE Enterococcus sulfureus ❶ EEEEEE Enterococcus sp. EEE❶ EEE Lactococcus garvieae ❶ EEEEEE Lactococcus japonicus EE❶ EEEE Lactobacillus curvatus EE❶ EEE❶ Lactobacillus sakei ❶❶❶❶❶❶❷ Leuconostoc mesenteroides EEE❶ E ❶❶ Micrococcus luteus ❶ EEEEEE Ochrobactrum grignonense EEE❶ EEE Staphylococcus epidermidis ❶ E ❶❶EEE Stenotrophomonas maltophilia EEE❶ EE❶ Streptococcus pyogenes EEE❶ EEE Yersinia intermedia ❶ EE❶ EEE UnidentiÞed ❶❶❶❶❶❷❷ Diversity Index 0.26 0.06 0.18 0.240.06 0.10 0.12 Diet 3 Carnobacterium divergens ❷❷❷❷❷❷❶ Carnobacterium piscicola ❷❶E ❶❶EE Corynebacterium callunae ❶ EEEEEE Enterococcus sp. EEE❶❶EE Kocuria varians EEEEEE❶ Lactococcus garvieae ❶ EEEEEE Lactobacillus curvatus E ❶❶E ❷❶E Lactobacillus sakei E ❶❸❸❸❷❶ 242 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

Table 4. Continued

Treatment/Taxona 0 6 12 244872 96 Leuconostoc mesenteroides ❶ EEEEE❶ Microbacterium sp. ❶ EEEEEE Paenibacillus polymyxa ❶ EEEEEE Staphylococcus epidermidis ❶ EEEEEE Streptococcus pyogenes E ❶ EEEEE UnidentiÞed ❶❶❶❶❷❶❶ Diversity Index 0.16 0.10 0.06 0.08 0.10 0.06 0.08 Diet 4 Bacillus megaterium ❶ E ❶ EEEE Carnobacterium divergens ❷❷❷❸❷❷❷ Carnobacterium piscicola ❸❶E ❷ EEE Carnobacterium sp. EEEE❶ EE Enterococcus sp. ❶ EE❶ EEE Enterococcus faecalis ❶ EE❶❶EE Enterococcus sulfureus ❶ EEEEEE Kocuria kristinae ❶ EEEEEE Lactococcus garvieae E ❷ EEEEE Lactococcus lactis ❶ EEEEEE Lactobacillus curvatus E ❶❶❶❶E ❶ Lactobacillus sakei E ❸❸❸❸❸❸ Leuconostoc mesenteroides EEEEE❸❶ Ochrobactrum grignonense EEEE❶ EE Pseudomonas sp. ❶ EEEEEE UnidentiÞed ❶❶❷❷❶❶E Diversity index 0.18 0.10 0.08 0.12 0.12 0.06 0.08

a The meat-based diet was comprised of ground beef, beef liver, hensÕ eggs, sucrose, brewerÕs yeast, honey, acetic acid, and tap water. b Treatments consisted of 1) diet containing both antibacterial (streptomycin sulfate and chlortetracycline) and antifungal agents (propionic acid and potassium sorbate); 2) diet containing antibacterial agents only; 3) diet containing antifungal agents only; or 4) diet containing neither antibacterial or antifungal agents. c Empty circles indicate that the corresponding bacterium was not isolated. d Dark circles indicate positive isolation, and the number within the symbol indicates the number of replicates from which the taxon was isolated. e The diversity index was calculated as the number of taxa isolated (across the three replicates) at each time divided by the total number of taxa isolated (n ϭ 50).

not Lactobacillus spp., were isolated from the diet in Can. pararugosa, Can. zeylanoides, Cry. curvatus, Rho. the third replicate (Fig. 3). acheniorum, Sac. dairensis, Tor. globosa, and Yar. lipo- Fungal Populations in the Diet. Differences were lytica. Of these taxa, Tor. globosa, and Yar. lipolytica observed in the rate of fungal proliferation among the were particularly prevalent at latter stages of decom- four diet treatments (F ϭ 5.05; df ϭ 18, 54; P Ͻ 0.001) position (Fig. 4). No taxa of Þlamentous fungi were (Fig. 1B). There were no differences (P Ͼ 0.05) isolated from more than one replicate. among any of the diets at the 0Ð12-h sample times. pH and Moisture. There were no differences among Thereafter, substantiative (P Յ 0.05) differences were diet treatments (F ϭ 1.5; df ϭ 3, 54; P ϭ 0.22) or observed among the diet treatments. After the two changes in moisture content over time (F ϭ 2.0; df ϭ initial sampling times, no fungi were isolated from the 6, 54; P ϭ 0.084). The dry weights of diets ranged from diet containing antifungal but no antibacterial agents 49.7 Ϯ 1.6 to 51.8 Ϯ 1.3% at time 0 to 47.3 Ϯ 0.9 to 48.4 Ϯ (i.e., diet 3). More (P Ͻ 0.001) fungi were isolated 1.5 after 96 h. from the diet containing antifungal and antibacterial The pH of diets differed among treatments over agents (i.e., diet 1) than from diet three at the 72- and time (F ϭ 34.0; df ϭ 18, 54; P Ͻ 0.001) (Fig. 1C). At 96-h sampling times. Fungi were less inhibited in diets time 0, the pH of diets containing propionic acid (i.e., not containing sorbic acid and propionic acid (i.e., diets 1 and 3) were lower (P Յ 0.001) than those not diets 2 and 4), and more (P Յ 0.001) fungi were amended with propionic acid (i.e., diets 2 and 4). With recovered from these diets relative to diets 1 and 3 time, the pH of diets containing antibacterial agents from 48 to 96 h. Of the diets not containing antifungal (i.e., diets 1 and 2) did not signiÞcantly decrease (P Ն agents, more (P Ͻ 0.05) fungi were isolated from diet 0.072), with the exception of diet one at 96 h where the two (i.e., with antibacterial agents) than the control pH decreased (P ϭ 0.003) from 5.5 Ϯ 0.07 to 5.3 Ϯ 0.23. diet (i.e., diet 4) at the 24-and 48-h sample periods. In contrast, the pH of diets not containing antibacte- More than 270 fungal isolates were characterized, rial agents (i.e., diets 3 and 4) decreased (P Ͻ 0.001) and yeasts were more prevalent than Þlamentous rapidly within a 24-h period, falling Ϸ1 pH unit during fungi. However, there was considerable variation in this period. The rapid decrease in the pH of diets 3 and the composition of the mycoßora isolated from the 4was subsequently maintained relative to the other diets among the three replicates (Table 5). Yeast taxa two diet treatments throughout the experiment. A isolated from diets in two or more replicates included more rapid (P Յ 0.046) drop in pH was observed for April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 243

Fig. 2. Popu- lations of prominent lactic acid bacteria isolated from a meat-based entomophage diet over a 96-h period. (AÐD) Car. divergens. (EÐH) Car. piscicola. (IÐL) Lac. curvatus. (MÐP) Lac. sakei. (QÐT) Enterococ- cus species. (UÐX) Leu. mesenteroides. Graphs, A, E, I, M, Q, and U represent bacteria isolated from diet containing antibacterial agents (chloramphenicol and tetracycline) and antifungal agents (sorbic acid and propionic acid) (diet 1). Graphs B, F, J, N, R, and V represent bacteria isolated from diet containing antibacterial agents but no antifungal agents (diet 2). Graphs C, G, K, O, S, and W represent bacteria isolated from diet containing antifungal agents only (diet 3). Graphs D, H, L, P, T, and X represent bacteria isolated from diet containing no antibacterial or antifungal agents (diet 4). Bars represent SEM (n ϭ 3). 244 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

limited information is available on the microbial decomposition of meats and meat products at higher temperatures. In the current study, we observed that bacteria and fungi rapidly multiplied in the meat- based entomophage diet incubated at 27ЊC and that the increases in bacteria and to a lesser extent, fungi, were relatively consistent in the three replicates (i.e., conducted on separate occasions). Using a general purpose, nonselective medium, we observed that bacterial populations within the meat-based entomophage diet rapidly increased in size (Ϸ4orders of magnitude within 24h), reaching an asymptote at Ϸ1010 CFU/g (dry weight). The maximum level that bacteria normally reach during refrigerated storage of meat and meat products is 107 to 109 CFU/g (Borch et al. 1996). The proliferation of fungi was delayed relative to bacteria, but fungal populations also increased Ϸ4 orders of magnitude within the 96-h duration of the experiment. Although the placement of the diet at 27ЊC likely facilitated the rapid increases in numbers of microorganisms that we observed, incubation at such temperatures is necessary for entomophagous insects, such as larvae of Chr. carnea to complete their development. Spoilage Bacteria. A number of lactic acid bacteria were isolated from the meat-based entomophage diet, including bacteria in the genera Carnobacterium, En- terococcus, Lactobacillus, Lactococcus, Leuconostoc, Fig. 3. Populations of prominent lactic acid bacteria iso- and Streptococcus. In addition, Pediococcus pentosaceus lated from a meat-based entomophage diet over a 96-h period on three separate occasions (i.e., replicates). (A) Carnobac- was isolated from liver but not from the meat-based terium species. (B) Lactobacillus species. Means were cal- diet. The most prevalent bacterial taxa isolated from culated across treatments, and bars represent the SEM (n ϭ the diet were Car. divergens and Lac. sakei with pop- 9 4). ulations reaching 10 CFU/g in some instances. We observed that the species composition was relatively consistent across the three replicates, with the excep- diet 4than diet 3 (e.g., 12 and 24h) despite the higher tion that Lactobacillus species were common in all initial pH of diet 4. A signiÞcant correlation was ob- three replicates, whereas Carnobacterium species served between bacterial populations and pH (P Ͻ were commonly isolated in only two replicates. Lactic 0.001), but not between fungal populations and pH acid bacteria are commonly associated with modiÞed- (P ϭ 0.14); the Pearson correlation coefÞcient for the atmosphere-packaged (MAP) and vacuum-packaged relationship between bacterial numbers and pH was (VP) meats and meat products (Jay 1996). In the 0.93. current study, the diet was placed in packets con- structed from ParaÞlm M. The diet is presented in such packets because most entomophagous insects use ex- Discussion traoral digestion to process complex solid substrates Microbial Spoilage of the Meat-Based Diet. The (Cohen 1998a). The ParaÞlm mimics the cuticle of development of artiÞcial diets is facilitating the eco- soft-bodied insects, which allows the predator to in- nomical production of entomophagous insects and is sert its mouthparts through the Þlm, stabilizing them, having a tremendous positive impact on the manage- and then to secrete its digestive enzymes into the food ment of pest insects by using nonchemical strategies. slurry. The enzymes can then be recaptured for fur- The majority of artiÞcial diets for insects are amended ther use in the gut of the predator, an important aspect with antimicrobial agents to prevent spoilage, but in of their digestive physiology (Cohen 1998a). Although most if not all cases, the microorganisms causing spoil- ParaÞlm is an ideal material for mimicking the cuticle age and their impact on insect diets have not been of an insect and protecting the diet from loss of mois- investigated. In the current study, we monitored the ture, it also possesses low gas permeability properties microbial spoilage of a meat-based entomophage diet. (i.e., oxygen permeability is Ϸ150 cc/m2/24h and Four companies currently hold a license to produce carbon dioxide permeability is Ϸ400 cc/m2/24h). We predatory insects by using this diet. did not measure the oxidation-reduction potential The microbiology of meat spoilage has been exten- within the ParaÞlm M packets, but the predominance sively documented, but essentially all of the research of microaerophilic bacteria such as Carnobacterium conducted in this area in the past 40 yr has been and Lactobacillus would suggest increased levels of conducted at low temperatures (Jay 1996). Relatively CO2 and a lowered oxidation-reduction potential typ- April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 245

Table 5. Fungi isolated from the meat-based entomophage diet at various times (hours) after placement at 27°C

Treatment/Taxona 0 6 12 244872 96 Diet 1b Aspergillus ustus Ec E ❶d ❶ EEE Candida glabrata EEEE❶ EE Candida parapsilosis EEE EEE❶ Candida zeylanoides E ❶ E EEEE Cryptococcus curvatus EE❶ EEEE Fusarium sp. ❶ EE EEEE Penicillium citrinum ❶ EE EEEE Penicillium oxalicum ❶ EE EEEE Penicillium sp. EE❶❶EEE Rhodotorula acheniorume E ❶ E EEEE Torulaspora globosa EEE ❶❶❷❷ Trichosporon brassicaee ❶ EE EEEE Yarrowia lipolytica EE❶ EE❶ E UnidentiÞed ❶❶❶❶❶EE Diet 2 Candida parapsilosis EEE ❶❶E ❶ Candida pararugosa EEE ❷❶EE Candida zeylanoides ❶ EE ❶❶E ❶ Cryptococcus curvatus EEE ❷❶E ❶ Cryptococcus marinus EEE EEE❶ Leucosporidium scottiie EEE EEE❶ Pichia pastorise EEE ❶ EEE Rhodotorula acheniorume EEEE❷❷❷ Torulaspora globosa EEEEE❷❷ Yarrowla lipolytica ❶ E ❶ ❶❷❷❸ UnidentiÞed EE❶❶E ❶❶ Diet 4 Candida glabrata EEEE❶❶❶ Candida zeylanoides E ❶❷ ❶❶EE Mucor hiemalis EEE ❶ EEE Penicillium sp. E ❶ E EEEE Saccharomyces dairensis EEE EEE❶ Torulaspora globosa EEEE❶❶❶ Yarrowla lipolytica ❶❶E ❶❶❶E UnidentiÞed ❶ E ❷ EEEE

a The meat-based diet was comprised of ground beef, beef liver, hensÕ eggs, sucrose, brewerÕs yeast, honey, acetic acid, and tap water. b Treatments consisted of 1) diet containing both antibacterial (streptomycin sulfate and chlortetracycline) and antifungal agents (propionic acid and potassium sorbate); 2) diet containing antibacterial agents only; and 4) diet containing neither antibacterial or antifungal agents. No fungi were isolated from diet treatment 3 (diet containing antifungal agents only) at sample times Ն12 h. c Empty circles indicate that the corresponding fungus was not isolated. d Dark circles indicate positive isolation, and the number within the symbol indicates the number of replicates from which the taxon was isolated. e IdentiÞcation indeÞnite.

Ն ical of meats packaged in O2-impermeable Þlms (Sei- packets at latter stages of decomposition (e.g., 72 h) deman et al. 1976, Vanderzant et al. 1982, Seideman consistent with the evolution of carbon dioxide. Fur- and Durand 1983). Lac. sakei and Lac. curvatus are thermore, the substantial decreases in pH of the diet often the predominant species on VP meat maintained (Ϸ2 U) that we observed are indicative of the prolif- at low temperatures (Morishita and Shiromizu 1986, eration of lactic acid bacteria. Brochothrix thermo- Hastings and Holzapfel 1987, Grant and Patterson sphacta is often observed in VP meats and can be an 1991). Car. divergens and Car. piscicola also are com- important spoilage taxon (Jay 1996). Interestingly, we monly encountered on/in unprocessed, refrigerated only observed B. thermosphacta in ground beef before red meat, but vacuum packaging can create a favorable it was incorporated into the diet. There is evidence ecosystem for these bacteria (Hammes et al. 1992). In that this bacterium is inhibited by lactobacilli (Roth contrast to packages possessing low oxygen tensions, and Clark 1975), which may explain its absence or low Pseudomonas species often predominate in meats frequency in the meat-based entomophage diet. Leu. packaged in Þlms that are highly permeable to oxygen mesenteroides is a lactic acid bacterium that also is (i.e., Ϸ900 cc/m2/24h), and we only observed commonly isolated from VP beef (Hitchener et al. pseudomonads in the diet at low frequencies. In the 1982). We observed that Leu. mesenteroides numbers Ϸ 8 current study, we did not remove O2 from the packets were substantial ( 10 CFU/g) in some instances, but before sealing them. However, residual oxygen can be populations were generally smaller than for Lactoba- rapidly consumed by the aerobic microßora and the cillus and Carnobacterium. meat itself, resulting in microaerophilic conditions Spoilage Fungi. The primary fungi that we isolated (Jay 1996). We observed inßation of the ParaÞlm M from the diet were the yeast taxa Can. zeylanoides, Tor. 246 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

Fig. 4. Populations of the prominent yeasts isolated from a meat-based entomophage diet over a 96-h period. (AÐC) Can. zeylanoides. (DÐF) Tor. globosa. (G and H) Yar. lipolytica. Graphs A, D, and G represent yeasts isolated from the diet containing antibacterial agents (chloramphenicol and tetracycline) and antifungal agents (sorbic acid and propionic acid) (diet 1). Graphs B, E, and H represent yeasts isolated from diet containing antibacterial agents but no antifungal agents (diet 2). Graphs C, F, and I represent yeasts isolated from diet containing no antibacterial or antifungal agents (diet 4). No fungi were isolated from diet 3 (i.e., diet containing antifungal agents only) at sample times Ն12 h. Bars represent SEM (n ϭ 3). globosa, and Yar. lipolytica. Both Can. zeylanoides and the diet packets to permit aerobic development of Yar. lipolytica are commonly isolated from various these yeasts. types of meats at low temperatures, including beef that Impact of Microorganisms. Most of the research on is vacuum packaged (Deak and Beuchat 1996). Tor. spoilage of meat and meat products has focused on globosa has been isolated from feces and soil (Barnett sensorial factors in relation to human consumption of et al. 1990), but it is not considered to be a meat- these products. Bacterial numbers, in particular those spoilage yeast to our knowledge. Nonetheless, we ob- of lactic acid bacteria, are rarely correlated with sen- served substantial populations of this yeast (Յ106 sorial spoilage (Korkeala et al. 1987, De Pablo et al. CFU/g) at latter sampling times (Ն48 h). Interest- 1989, Yano et al. 1995). Bacterial spoilage of refriger- ingly, of the prominent yeasts that we isolated from ated meats is most often associated with sour off- the meat-based entomophage diet, only Tor. globosa is ßavors, off-odors, discoloration, gas production, slime capable of vigorous growth under anaerobic condi- production, and decreased pH (Borch et al. 1996). tions (Barnett et al. 1990). However, the prominence Off-ßavors in VP or MAP meats are typically described of obligate aerobes, such as Yar. lipolytica, at latter as sour and acid (Gardner 1983), due to the accumu- sampling times suggests that sufÞcient oxygen exists in lation of acetic and/or lactic acids in the meat. Car. April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 247 divergens, Car. piscicola, and Leu. mesenteroides are ciated with eggs, liver, and fresh ground beef (Jay heterofermenative bacteria, producing equimolar 1996), we observed both small numbers (primarily Յ 3 amounts of lactic acid, CO2, ethanol, and/or acetic 10 CFU/g) and a limited diversity of fungi associ- acid (Hammes et al. 1992). Lac. sakei and Lac. curvatus ated with these substrates. However, a number of the are homofermentative bacteria, primarily producing yeast taxa (e.g., Can. zeylanoides, Cryptococcus spp., lactic acid as the end product of glucose fermentation. and Yaw. lipolytica) that we isolated from ground beef In addition to acidiÞcation, heterofermentative Lac- are commonly associated with this substrate (Deak tobacillus spp., Leuconostoc spp. and Car. divergens are and Beuchat 1996). Of the microbial taxa associated able to produce H2O2 (Borch and Molin 1989), which with the entomophage diet, ground beef and liver may cause green discoloration through the oxidation seem to be the primary sources of bacteria and fungi. of nitrosohaemochrome to choleomyoglobin (Borch For example, lactic acid bacteria (e.g., Carnobacte- et al. 1996). Furthermore, Lac. sakei has been impli- rium, Enterococcus, Lactobacillus, and Leuconostoc) cated with slime formation on meat products (Ma¨kela¨ were commonly associated with both liver and ground et al. 1992) and has been demonstrated to produce H2S beef and subsequently with the meat-based ento- (Borch et al. 1996). Carnobacterium spp. do not seem mophage diet. Beef liver and ground beef also seem to to produce H2S or other foul-odor compounds, and be the primary source of fungi. Pen. citrinum and Yar. their impact on meat spoilage is not well understood lipolytica were commonly isolated from liver and (Jay 1996). At reduced temperatures, yeasts are not ground beef and subsequently from the diet. Very few considered to be of great importance in the spoilage of the bacteria or fungi that we isolated from hensÕ of meats (Jay 1987, Dillon and Board 1991). Although eggs were subsequently recovered from the ento- yeasts are rarely the primary cause of chemical alter- mophage diet, and most of the taxa we recovered are ations to refrigerated meats during storage, increases not commonly associated with VP or MAP meats or in yeast populations will result in changes to the com- meat products (Jay 1996). However, eggs were heated position of the substrate. Although bacteria associated to Ϸ95ЊC for 3Ð5 min before dispensing the diet, and with meat products at low oxygen tensions can cause this may have reduced bacterial populations; cooking spoilage in relation to human consumption, physical, of the eggs is essential to render the eggs into a semi- and/or chemical alterations that impact on the nutri- solid, to denature avidin to avert losses of biotin, and tion of insects will no doubt differ from those impact- to capture meat nutrients in a solid, stable form, that ing on humans. simulates the texture and concentration of nutrients of Source of Microorganisms. The relative importance insect prey (Cohen and Smith 1998). An additional of the primary components of the entomophage diet reason for heating the egg/sugar/acetic acid mixture (i.e., sucrose, brewersÕ yeast, honey, tap water, hensÕ is to blanch the other diet components, especially the eggs, liver, and ground beef) as a source of spoilage meat paste (Cohen 1998b, Cohen and Smith 1998). microorganisms also was investigated. The sucrose, Impact of Antimicrobial Agents. The primary honey, acetic acid, and tap water were found to be means by which spoilage microorganisms are man- sterile. Despite the incorporation of large numbers of aged in artiÞcial diets for insects is through the incor- viable Sac. cerevisiae, cells into the diet and the suit- poration of antimicrobial agents. We observed that the ability of SDA as a growth medium for this yeast, we presence of streptomycin sulfate, chlortetracycline, did not isolate this species. Mesophilic bacterial counts propionic acid, and potassium sorbate had profound on meat are normally Ϸ102 to 103 CFU/g (Mol et al. effects on bacterial populations but not on the com- 1971, Blickstad et al. 1981, Blickstad and Molin 1983, position of the bacterial ßora. The two antibacterial Egan 1983, Jackson et al. 1992), but it is not uncommon agents used (alone and in combination with antifungal to observe substantially higher numbers. For example, agents) signiÞcantly inhibited bacteria. Even after 12% of fresh ground beef samples possessed bacterial 96 h, bacterial populations in diets containing anti- counts higher than 107 CFU/g (Pivnick et al. 1976). In bacterial agents were signiÞcant smaller than in un- the current study, we consistently observed substan- amended diets. Streptomycin sulfate is primarily ac- tial numbers of bacteria associated with ground beef tivity against gram-negative bacteria, and many lactic (Ϸ107 CFU/g) and also with eggs (Ϸ104 CFU/g) and acid bacteria, including Lactobacillus and Carnobac- liver (Ϸ105 CFU/g). Beef livers and ground beef terium, possess resistance to this antibiotic (Vidal and stored at low temperatures are subject to spoilage by Collins-Thompson 1987, Baya et al. 1991). In contrast a number of microorganisms, including lactic acid to streptomycin sulfate, chortetracycline is a wide- bacteria, pseudomonads, streptococci, moraxellas, spectrum antibiotic (i.e., active against both gram- acinetobacters, Bro. thermosphacta, coryneforms, and positive and gram-negative bacteria). However, Lac. yeasts (Shelef 1975, Hanna et al. 1982a,b, Egan and curvatus and Lac. sakei in meat products have dem- Roberts 1987, Jay 1996). HensÕ eggs are normally ster- onstrated resistance to tetracycline (Gevers et al. ile when laid, but within a relatively short period of 2000). Furthermore, streptomycin sulfate is very heat time, numerous microorganisms may be found on the labile, whereas chlortetracycine is relatively stable at outside, and under the proper conditions, within the high temperatures (i.e., it would only be partially egg itself (Jay 1996). Staphylococcus is commonly as- inactivated after brief cooking of the diet). Given the sociated with hensÕ eggs (Jay 1996), and we frequently natural resistance of many lactic acid bacteria to strep- isolated Staphylococcus species from eggs in the cur- tomycin sulfate and possible deactivation of this an- rent study. Although a number of fungi may be asso- tibiotic during the brief cooking phase of the diet, it is 248 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2 likely that most of the antibacterial action imparted by ponents, and most of the taxa responsible for spoilage the antibacterial agents was due to chlortetracycline. of the diet originated from liver and ground beef. The Interestingly, the presence of antifungal agents (i.e., effect of reducing initial bacterial numbers on/in liver propionic acid and potassium sorbate) without anti- and ground beef (e.g., by using irradiation or electron bacterial agents also delayed bacterial proliferation. beams) alone and in combination with other manage- The sorbates possess bacteriostatic properties and in- ment strategies to reduce spoilage of meat-based diet hibit a wide range of bacteria (Block 1991), but the for entomophagous insects warrants investigation. We lactic acid bacteria are relatively resistant (Jay 1996). observed that the incorporation of antibacterial agents The antimicrobial activity of sorbates is related to pH substantially suppressed populations of bacteria but (i.e., Ͻ6.0) and potassium salts exhibit a lower pre- that it did not affect the species composition. Regard- servative effect than sorbic acid because such salts are less of the antibiotic treatment, the most prevalent neutral. Propionic acid possesses minimal antibacte- species were lactic acid bacteria, primarily Car. diver- rial properties (Block 1991), but its use in conjunction gens and Lac. sakei. Substantial populations of fungi with sorbic acid likely enhanced the activity of the were observed in diets not containing antifungal latter by decreasing pH Ϸ0.5 U. The addition of acetic agents, and Can. zeylanoides, Tor. globosa, and Yar. acid intended to lower the pH of the diet (Cohen lipolytica were commonly isolated. The results of this 1998b) may also have conferred some degree of an- study clearly demonstrate that management strategies timicrobial activity, considering the fact that this com- to delay the decomposition of the meat-based ento- pound is bacteriostatic (Jay 1996). mophage diet presented in ParaÞlm M packets at 27ЊC The antimicrobial agents tested also demonstrated should target lactic acid spoilage bacteria, particularly signiÞcant effects against fungi. After the two initial Carnobacterium and Lactobacillus species. Future re- sampling times, no fungi were isolated from the diet search will focus on the integration of combinations of containing potassium sorbate and propionic acid intrinsic and extrinsic strategies in a “hurdle” approach alone. Interestingly, the presence of antibacterial to managing spoilage microorganisms. agents decreased the fungistatic activity of the antifungal agents (i.e., diet 1 versus diet 3). Reasons for the decreased efÞcacy of the antifungal agents applied in Acknowledgments combination with streptomycin sulfate and chlortetracycline are speculative. However, a substantially We thank the following people at Agriculture and Agri- lower pH, resulting from fermentation by lactic acid Food Canada Research Centre at Lethbridge: Jenny Gusse for assistance with the isolation and preliminary character- bacteria, was observed in the diet not containing an- ization of microorganisms; Lisa Kalischuk for help with the tibacterial agents (i.e., diet 3). The lower pH would be maintenance of bacteria and the identiÞcation of bacteria expected to enhance the activity of the sorbic acid; the based on rDNA sequences; Kathaleen House for assistance pKa of sorbic acid is 4.8, and at a pH of 4.0, 86% of the with the preparation of bacteria for MIDI analysis and se- compound is undissociated compared with only 6% of quencing of rDNA; Jay Yanke for providing MIDI identiÞ- the molecule being undissociated at pH 6.0 (Block cations of bacteria and for conÞrming identiÞcations of bac- 1991). Sorbic acid is both a mold and yeast inhibitor, teria by using physiological methods. We also thank Gay whereas propionic acid is primarily active against McCain (BCMRRU, USDAÐARS, Mississippi State) for pre- molds and possesses limited activity against yeasts paring the diets, and Brent Selinger (University of Leth- bridge) for allowing access to the automated Biolog system. (Block 1991). Interestingly, we also observed larger This is LRC contribution #03079. populations of fungi in diets containing antibacterial agents than in diets containing no antibacterial agents in the absence of antifungal agents (i.e., diet 2 versus References Cited diet 4). The antibacterial properties of lactic acid bacteria, including Lac. sakei, in meat and meat products Barnett, J. A., R. W. Payne, and D. Yarrow. 1990. Yeasts, have been documented (Bredholt et al. 2001), and a characteristics and identiÞcation. Cambridge University variety of mechanisms such as production of antibi- Press, Cambridge, United Kingdom. otics, hydrogen peroxide, depressed pH, nutrient de- Baya, A. M., A. E. Toranzo, B. Lupiani, T. Li, B. S. Roberson, and F. M. Hetrick. 1991. Biochemical and serological pletion, and bacteriocins or bacteriocin-like factors characterization of Carnobacterium spp. isolated from have been implicated in “lactic antagonism” (Jay farmed and natural populations of striped bass and catÞsh. 1996). It is possible that lactic acid bacteria also pro- Appl. Environ. Microbiol. 57: 3114Ð3120. duce antifungal factors. It is uncertain whether inoc- Berthier, F. F., and S. D. Erhlich. 1998. Rapid species iden- ulation of insect diets with speciÞc taxa or stains of tiÞcation with two groups of closely related lactobacilli lactic acid bacteria or other microorganisms will prove using PCR primers that target the 16S/23S rRNA spacer useful in suppressing spoilage of meat-based ento- region. FEMS Microbiol. Let. 161: 97Ð106. mophage diets but this warrants investigation. Blickstad, E., and G. Molin. 1983. The microbial ßora of In conclusion, to achieve efÞcacious management smoked pork loin and frankfurter sausage stored in different gas atmospheres at 4ЊC. J. Appl. Bacteriol. 54: 45Ð of spoilage microorganisms, it is necessary to under- 56. stand the microbial taxa responsible for spoilage, and Blickstad, E., S.-O. Enfors, and G. Molin. 1981. Effect of how differing management strategies affect the hyberbaric carbon dioxide pressure on the microbial ßora growth of speciÞc taxa. We observed that considerable of pork stored at 4ЊCor14ЊC. J. Appl. Bacteriol. 50: numbers of bacteria were associated with diet com- 493Ð504. April 2004I NGLIS AND COHEN:MICROBIAL SPOILAGE OF LACEWING DIET 249

Block, S. S. 1991. Disinfection, sterilization, and preserva- Hammes, W. P., N. Weiss, and W. H. Holzapfel. 1992. Lac- tion. Lea & Febiger, Philadelphia, PA. tobacillus and Carnobacterium, pp. 1535Ð1594. In A. Borch, E., and G. Molin. 1989. The aerobic growth and Balows, H. G. Truper, M. Dworkin, W. Harder, and K. H. product formation of Lactobacillus, Leuconostoc, Brocho- Schleifer [eds.], The prokaryotes, 2nd ed. Springer, New thrix and Carnobacterium in batch cultures. Appl. Micro- York biol. Biotechnol. 30: 81Ð88. Hanna, M. O., G. C. Smith, J. W. Savell, F. K. McKeith, and Borch, E., M. L. Kant-Muermans, and Y. Blixt. 1996. Bac- C. Vanderzant. 1982a. Microbial ßora of livers, kidneys terial spoilage of meat and cured meat products. Int. J. and hearts from beef, pork, and lamb: effects of refrig- Food Microbiol. 33: 103Ð120. eration, freezing and thawing. J. Food Protect. 45: 63Ð73. Bredholt, S., T. Nesbakken, and A. Holck. 2001. Industrial Hanna, M. O., G. C. Smith, J. W. Savell, F. K. McKeith, and application of an antilisterial strain of Lactobacillus sakei C. Vanderzant. 1982b. Effects of packaging methods on as a protective culture and its effect on the sensory ac- the microbial ßora of livers and kidneys from beef or pork. ceptability of cooked, sliced, vacuum-packaged meats. J. Food Protect. 45: 74Ð81. Int. J. Food Microbiol. 66: 191Ð196. Hastings, J. W., and W. H. Holzapfel. 1987. Conventional Cohen, A. C. 1995. Extra-oral digestion in predaceous ter- taxonomy of lactobacilli surviving radurization of meat. restrial Arthropoda. Annu. Rev. Entomol. 40: 85Ð103. J. Appl. Bacteriol. 23: 130Ð135. 1998a. Solid-to-liquid feeding: the inside(s) story of extra- Hitchener, B. J., A. F. Egan, and P. J. Rogers. 1982. Char- oral digestion in predaceous Arthropoda. Am. Entomol. acteristics of lactic acid bacteria isolated from vacuum- 44: 103Ð117. packaged beef. J. Appl. Bacteriol. 52: 31Ð37. 1998b. ArtiÞcial media for rearing entomophages compris- Jackson, T. C., G. R. Acuff, C. Vanderzant, T. R. Sharp, and ing cooked, whole egg. U.S. Patent #5,834,177. November J. W. Savell. 1992. IdentiÞcation and evaluation of vol- 10: 1998. atile compounds of vacuum and modiÞed atmosphere Cohen, A. C., and L. K. Smith. 1998. A new concept in packaged beef strip loins. Meat Sci. 31: 175Ð190. artiÞcial diets for Chrysoperla carnea: the efÞcacy of solid Jay, J. M. 1987. Meats, poultry, and seafoods, pp. 155Ð173. In diets. Biol. Contr. 13: 49Ð54. L. R. Beuchat [ed.], Food and beverage mycology. Von Collins, M. D., J.A.E. Farrow, B. A. Phillips, S. Ferusu, and Nostrand Reinhold, New York D. Jones. 1987. ClassiÞcation of Lactobacillus divergens, Jay, J. M. 1996. Modern food microbiology. Chapman & Lactobacillus piscicola, and some catalase-negative, as- Hall, New York porogenous, rod-shaped bacteria from poultry in a new Korkeala, H., S. Lindroth, R. Ahvenainen, and T. Alanko. genus Carnobacterium. Int. J. Syst. Bacteriol. 37: 310Ð316. 1987. Interrelationship between microbial numbers and Deak, T., and L. R. Beuchat. 1996. Handbook of food spoil- other parameters in the spoilage of vacuum-packed age yeasts. CRC, Boca Raton, FL. cooked ring sausage. Int. J. Food Microbiol. 5: 311Ð321. De Pablo, B., M. A. Asensio, B. Sanz, and J. A. Ordo´n˜ ez. 1989. Kurtzman, C. P., and J. W. Fell. 1998. The yeasts, a taxo- The D(Ϫ)lactic acid and acetoin/diacetyl as potential nomic study, 4th ed., Elsevier, Amsterdam. indicators of the microbial quality of vacuum-packed MacGown, M. W., and P. P. Sikorowski. 1980. Histopathol- pork and meat products. J. Appl. Bacteriol. 66: 185Ð190. ogy of midgut of mass reared, irradiated boll weevils Dillon, V. M., and R. G. Board. 1991. Yeasts associated with contaminated with Staphylococcus aureus and Streptococ- red meats. J. Appl. Bacteriol. 71: 93Ð108. cus sp. J. Econ. Entomol. 73: 81Ð87. Egan, A. F. 1983. Lactic acid bacteria of meat and meat Ma¨kela¨, P., U. Schillinger, H. Korkeala, and W. H. Holzapfel. products. Antonie van Leeuwenhoek 49: 327Ð336. 1992. ClassiÞcation of ropy slime-producing lactic acid Egan, A. F., and T. A. Roberts. 1987. Microbiology of meat bacteria based on DNA-DNA homology, and identiÞca- and meat products, pp. 167Ð197. In J. R. Norris and G. L. tion of Lactobacillus sake and Leuconostoc amelibiosum as Pettipher [eds.], Essays in agricultural and food micro- dominant spoilage organisms in meat products. Int. J. biology. John Wiley, Chichester, United Kingdom. Food Microbiol. 16: 167Ð172. Gardner, G. A. 1983. Microbial spoilage of cured meats, pp. McLaughlin, R. E., and P. P. Sikorowski. 1978. Observations 179Ð202. In T. A. Roberts and F. A. Skinner [eds.], Food of boll weevil midgut when fed natural food or on bac- microbiology: advances and prospects. Academic, Lon- terially contaminated artiÞcial diet. J. Invertebr. Pathol. don, United Kingdom. 32: 64Ð70. Gevers, D., G. Huys, F. Devlieghere, M. Uyttendaele, and J. Mol, J.H.H., E. A. Hietbrink, H.W.M. Moller, and J. Van Debevere. 2000. Isolation and identiÞcation of tetracy- Tinteren. 1971. Observations on the microßora of vac- cline resistant lactic acid bacteria from pre-packed sliced uum packed sliced cooked meat products. J. Appl. Bac- meat products. Syst. Appl. Microbiol. 23: 279Ð284. teriol. 34: 377Ð397. Grant, I. R., and M. F. Patterson. 1991. A numerical taxo- Morishita, Y., and K. Shiromizu. 1986. Characterization of nomic study of lactic acid bacteria isolated from irradi- lactobacilli isolated from meats and meat products. Int. J. ated pork and chicken packaged under various gas atmo- Food Microbiol. 3: 19Ð29. spheres. J. Appl. Bacteriol. 70: 302Ð307. Nissen, H., A. Holck, and R. H. Dainty. 1994. IdentiÞcation Grenier, S., P. D. Greany, and A. C. Cohen. 1994. Potential of Carnobacterium spp. and Leuconostoc spp. in meat by for mass release of insect parasitoids and predators genus-speciÞc 16S rRNA probes. Lett. Appl. Microbiol. through development of artiÞcial culture techniques, pp. 19: 165Ð168. 181Ð206. In D. Rosen, F. D. Bennett, and J. L. Capinera Pivnick, H., I. E. Erdman, D. Collins-Thompson, G. Roberts, [eds.], Pest management in the subtropics: biological M. A. Johnston, D. R. Conley, G. Lachapell, U. T. Purvis, control - a Florida perspective. Intercept, Andover, R. Foster, and M. Milling. 1976. Proposed microbiolog- Hants, United Kingdom. ical standards for ground beef based on a Canadian sur- Gueldner, R. C., P. P. Sikorowski, and J. M. Wyatt. 1977. vey. J. Milk Food Technol. 39: 408Ð412. Bacterial load and pheromone production in the boll Roth, L. A., and D. S. Clark. 1975. Effect of lactobacilli and weevil Anthonomus grandis. J. Invertebr. Pathol. 29: 397Ð carbon dioxide on the growth of Microbacterium thermo- 398. sphactum of fresh beef. Can. J. Microbiol. 21: 629Ð632. 250 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 97, no. 2

SAS Institute 1999. UserÕs guide, version 8.0. SAS Institute, 1982. Haemolymph analysis of irradiated and Dimilin- Cary, NC. treated boll weevils, Anthonomus grandis. J. Invertebr. Sasser, M. 1990. IdentiÞcation of bacteria by gas chroma- Pathol. 39: 158Ð163. tography of cellular fatty acids. Technical Note 101. Mi- Thompson, A. C., P. P. Sikorowski, and J. M. Wyatt. 1977. crobial ID, Inc., Newark, DE. The effect of bacterial load on fatty acids in the boll Scarpellini, M., D. Mora, S. Colombo, and L. Franzetti. 2002. weevil Anthonomus grandis (Coleoptera: Curculionidae). Development of genus/species-speciÞc PCR analysis for J. Invertebr. Pathol. 30: 274Ð275. identiÞcation of Carnobacterium strains. Curr. Microbiol. Van der Walt, J. P., and D. Yarrow. 1984. Methods for the 45: 24Ð29. isolation, maintenance, classiÞcation, and identiÞcation Shelef, L. A. 1975. Microbial spoilage of fresh refrigerated of yeasts, pp. 45Ð104. In N.J.W. Kreger-van Rij [ed.], The beef liver. J. Appl. Bacteriol. 39: 273Ð280. yeasts, a taxanomic study. Elsevier, Amsterdam. Seideman, S. C., and P. R. Durand. 1983. Vacuum packaging Vanderzant, C., M. O. Hanna, J. G. Ehlers, J. W. Savell, G. C. of fresh beef: a review. J. Food Qual. 6: 29Ð47. Smith, D. B. Grifﬁn, K. D. Terrell, and D. E. Galloway. Seideman, S. C., C. Vanderzant, G. C. Smith, M. O. Hanna, 1982. Centralized packaging of beef loin steaks with dif- and Z. L. Carpenter. 1976. Effect of degree of vacuum ferent oxygen-barrier Þlms: microbiological characteris- and length of storage on the microßora of vacuum pack- tics. J. Food Sci. 47: 1070Ð1079. aged beef wholesale cuts. J. Food Sci. 41: 738Ð742. Vidal, C. A., and D. L. Collins-Thompson. 1987. Resistance Sikorowski, P. P., A. D. Kent, O. H. Lindig, G. Wiygul, and and sensitivity of meat lactic acid bacteria to antibiotics. J. Roberson. 1980. Laboratory and insectary studies on J. Food Prot. 50: 737Ð740. the use of antibiotics and antimicrobial agents in mass- Wiygul, G., and P. P. Sikorowski. 1981. The effect of various rearing of boll weevils. J. Econ. Entomol. 73: 106Ð110. levels of contamination by two species of bacteria (Staph- Sikorowski, P. P., and A. C. Thompson. 1984. Effects of ylococcus aureus and Streptococcus sp.) on oxygen uptake bacterial contamination on development and blood of the boll weevil. Comp. Biochem. Physiol. 68(A): 527Ð chemistry of Heliothis virescens. Comp. Biochem. Physiol. 530. 77(A): 283Ð285. Wiygul, G., and P. P. Sikorowski. 1986. The effect of staph- Sikorowski, P. P., G. R. Burkett, A. C. Thompson, and A. D. ylococcal enterotoxin B on pheromone production in fat Kent. 1984. Effects of antimicrobials on development of bodies isolated from male boll weevils. J. Invertebr. Heliothis virescens (F.) (Lepidoptera: Noctuidae). J. Ga. Pathol. 47: 116Ð119. Entomol. Soc. 19: 327Ð332. Yano, Y., N. Kataho, M. Watanabe, and T. Nakamura. 1995. Sikorowski, P. P., J. E. Powell, and A. M. Lawrence. 1992. Changes in the concentration of biogenic amines and Effects of bacterial contamination on development of application of tyramine sensor during storage of beef. Microplitis croceipes [Hym: Braconidae]. Entomophaga Food Chem. 54: 155Ð159. 37: 475Ð481. Yost, C. K., and F. M. Nattress. 2000. The use of multiplex Singh, P., and H. L. House. 1970. Antimicrobials: “safe” lev- PCR reactions to characterize populations of lactic acid els in a synthetic diet of an insect, Agria afﬁnis. J. Insect bacteria associated with meat spoilage. Lett. Appl. Mi- Physiol. 16: 1769Ð1782. crobiol. 31: 129Ð133. Thompson, A. C., and P. P. Sikorowski. 1978. The effect of bacterial load on amino acids in the boll weevil Anthono- Received for publication 5 May 2003; accepted 19 October mus grandis. J. Invertebr. Pathol. 32: 388Ð389. 2003.