Screening of Salmonella Isolates from a Turkey Production Facility for Antibiotic Resistance 1

Screening of Salmonella Isolates from a Turkey Production Facility for Antibiotic Resistance 1 R. Nayak* and P. B. Kenney,2 *US Food and Drug Administration, National Center for Toxicological Research, Division of Microbiology, Jefferson, Arkansas 72079; and Division of Animal and Veterinary Sciences, West Virginia University, Morgantown, West Virginia 26506-6108 ABSTRACT An ecological survey was conducted from followed by environmental swabs (43%), feeder content April 1997 to June 1999 on four turkey flocks (F1 to F4). Turkey cecal contents, litter, waterers, feed, feeders, and environmental swabs were analyzed. Presence of Salmonella was determined using conventional microbiological screening techniques and confirmed by serology. Positive isolates were serotyped and screened for antibiotic resistance. From a total of 69 Salmonella isolates 25% were resistant to one or more antibiotics including gentamicin (G), spectinomycin (SP), streptomycin (S), tetracycline (T), tobramycin (TO), and trimethoprim/sulfamethoxazole. Isolates included 45 S. heidelberg, 13S. senftenberg, 7S. muenster, 2S. anatum, and 2 S. worthington. Resistance to antibiotic(s) was highest among waterer isolates (55%) samples (33%), turkey cecal contents (26%), and litter samples (5%). Frequencies of antibiotic-resistant Salmonella in F1, F2, and F4 were 27, 13, and 40%, respectively. Salmonella was undetected in F3. In F1, S. heidelberg from cecal content and waterer samples was resistant to G, SP, S, and T, whereas S. anatum from waterer samples was resistant to T and S. In F2, S. worthington from litter and feeder content samples was resistant to T, and in F4, S. muenster from environmental swabs was resistant to TO, S, SP, and G. Identifying preharvest sources and characterizing serotype and antibiotic-resistance profile can assist poultry producers and integrators in tracking movement of Salmonella on turkey farms. (Key words: Salmonella, turkey production, antibiotic resistance, preharvest sources) 2002 Poultry Science 81:1496 1500 INTRODUCTION Antibiotics are added to poultry feed and water at low levels (15 to 25 ppm) to improve bird performance and at high levels (100 to 200 ppm) to treat specific bacterial diseases (Stavric and D Aoust, 1993). Widespread use of antibiotics in food animals has resulted in emergence of antibiotic-resistant strains of food borne pathogens, such as Salmonella, Escherichia coli, and Campylobacter, as well as bacteria endogenous to the microflora of animals (Davis et al., 1999; van den Bogaard and Stobberingh, 1999). These resistant strains may be transmitted to humans through food. Multiple drug-resistant isolates account for 20 to 25% of human Salmonella infections in the US (Holmberg et al., 1984). Several strains of multiple-drug-resistant Salmonella 2002 Poultry Science Association, Inc. Received for publication February 6, 2002. Accepted for publication June 6, 2002. 1 This research was supported partially by funds appropriated under the Hatch Act and in part from a grant provided by British United Turkeys of America (BUTA). It is published with the approval of the Director of the Agricultural and Forestry Experiment Station as scientific paper no. 2814. 2 To whom all correspondence should be addressed: bkenney @wvu.edu. strains have been isolated from farm animals and foods of animal origin (DuPont and Steele, 1987; Threlfall et al., 1993; Ansari and Khatoon, 1994; Nair et al., 1995; Aarestrup et al., 1998; Tollefson, et al., 1998; Klien, 1999; van der Wolf et al., 1999; Wagner and Hahn, 1999). Every tenth nontyphoidal Salmonella isolate received by the state public health laboratories in the US from July 1994 to June 1995 was resistant to 12 antimicrobial agents (Herikstad et al., 1997). Salmonella and other food borne pathogens acquire antibiotic resistance by random chromosomal mutations, mutation of existing genes, and through specific mechanisms such as transduction, tranformation, and conjugation (Okolo, 1986). These mechanisms involve transfer of drugresistant genes by means of circular DNA plasmids such as R-factor, conjugative plasmid, or chromosomal elements (Vidon et al., 1978; O Brien et al., 1982; Holmberg et al., 1984; Poppe and Gyles, 1987; Schuman et al., 1989; Salyers and Whitt, 1994; Poppe et al., 1996; Wagner and Hahn, 1999). One such strain, S. typhimurium DT 104, is resistant Abbreviation Key: BHI = brain heart infusion; F1 = flock 1; F2 = flock 2, F3 = flock 3; F4 = flock 4; G = gentamicin; LIA = lysine iron agar; SP = spectinomycin; S = streptomycin; T = tetracycline; TO = tobramycin; TSI = triple sugar iron; UP = universal pre-enrichment. 1496

ANTIBIOTIC RESISTANT-SALMONELLA 1497 TABLE 1. Frequency of antibiotic resistant Salmonella serotypes isolated Sample Flock 1 Flock 2 Flock 4 Total Turkey ceca 6/22 0/1 0/0 6/23 (26%) 1 Litter 0/10 1/10 0/1 1/21 (5%) Waterers 6/11 0/0 0/0 6/11 (55%) Environmental swabs 1/5 0/0 2/2 3/7 (43%) Feed 0/0 0/4 0/0 0/4 Feeder contents 0/0 1/1 0/2 1/3 (33%) Total 13/48 (27%) 2/16 (13%) 2/5 (40%) 17/69 (25%) 1 Figures in parentheses indicate percentage positive. to antibiotics such as ampicillin, chloramphenicol, streptomycin (S), sulfonamides, tetracycline (T), and trimethoprim (Herikstad et al., 1997). Approval of drugs such as sarafloxacin in poultry has contributed to the emergence and spread of Salmonella strains resistant to fluoroquinolone (nalidixic acid and ciprofloxacin) (Herikstad et al., 1997). Human infections with DT 104 are associated with consumption of contaminated meat and meat products, and the pathogen has been traced back to infected farm animals. Manie et al. (1998) found several strains of multiple antibiotic-resistant [S, methicillin, T, and gentamicin (G)] Salmonella strains in chicken. The objective of this study was to evaluate antibiotic resistance patterns of Salmonella collected from turkeys and environmental sources in the turkey production facility. MATERIALS AND METHODS A survey was conducted at Wardensville, WV. This facility contained 12 pens on either side of a centrally located service area, for a total of 24 pens. Four consecutive turkey flocks (F1 to F4) were sampled in this facility from April 1997 to June 1999. Sampling Procedures Litter, waterer, air, and feed samples, and environmental swabs were collected from each flock throughout the growout period (Table 1). Litter, waterer, and turkey sampling was synchronized so that positive samples could be identified by the pen source. Poults entering the facility and box liners in which poults were shipped were sampled for an indication of Salmonella contamination from the breeder flocks or the hatchery. Liners were sampled only for F3. Before placement of the poults in F3 (Day 0), litter, waterers, air, feed, and environmental swab samples were collected from various locations in the facility to determine the initial level of contamination. In F3, 12 poults from each of three hatcheries, for a total of 36 poults, were euthanized by cervical disarticulation (American Veterinary Medical Association, 1993) at Day 0, and the entire intestine and yolk sac were transferred to 100 ml of univer- 3 Difco Laboratories, Detroit, MI. 4 Fisher Scientific, Pittsburgh, PA. 5 Spiral Biotech, Inc., Bethesda, MD. 6 Becton Dickinson and Company, Cockeysville, MD. sal preenrichment (UP) broth. 3 At Week 2, intestinal samples were transferred to 100 ml of UP broth. From Week 10 until the end of the grow-out period, the entire ceca were removed, the blind end was snipped with sterile scissors, and cecal contents were emptied into sterile stomacher bags and sealed. Crop contents were sampled in F3 and F4 by removing the crop, making an incision with a sterile scalpel, and transferring the contents into a sterile stomacher bag. Litter samples were collected from the top 5.08-cm layer of litter and placed in sterile bags. These bags were held on ice during transport to the laboratory. Waterers were swabbed with sterile cloth gauze (25 cm 2 ) held by a pair of sterilized forceps. Environmental swabs were collected from various surfaces in the facility with a 16-cm 2 template using a sterile swab moistened with sterile UP broth. Locations included walls, ventilation fans, feathers of dead birds, employee shoes, the feed truck, fans inside the pens, feed storage bins, and door handles. Swabs were transferred to 10 ml sterile UP broth. Air samples were collected on Rodac plates 4 (65 15 mm) containing brain heart infusion (BHI) agar 3 using a SAS portable high flow air-sampler. 5 The air sampler was set to collect 60 L of air in 20 s. The agar was aseptically transferred to sterile stomacher bags. Feed and feeder samples were collected randomly from each feed shipment by placing a sterile bag in the flow of feed from the auger or from the feed cart. Samples were stored at 4 C at the production facility prior to transport to the laboratory for analyses. All samples were held on ice during transport to the laboratory. Transport time did not exceed 4 h. Laboratory Procedures Samples were mixed with UP broth (1:10 wt/vol) and sealed. Gauze and agar from waterer and air samples, respectively, were mixed with 100 ml UP broth. Approximately 5 g of litter and 10 g of feed or feeder content samples, respectively, were transferred to bottles containing 100 ml UP broth. Samples were thoroughly mixed with UP broth and incubated for 24 h at 37 C. One milliliter of this pre-enriched sample was transferred to 9 ml of tetrathionate 6 broth and incubated for 24 h at 37 C. Selectively enriched samples from tetrathionate broth were streaked to isolation on xylose lactose tergitol (XLT4) 3 plates. These plates were incubated at 37 C for 24 h. A single presumed positive Salmonella colony was stabbed and streaked on triple sugar iron (TSI) 6 and lysine iron

1498 NAYAK AND KENNEY TABLE 2. Antibiotic resistance 1 among Salmonella serotypes isolated from Flock 1 Week 2 Week 10 Week 18 Sample Serotype Frequency 2 Serotype Frequency Serotype Frequency Total Turkey ceca S. heidelberg 0/6 S. heidelberg 0/1 S. heidelberg 6/15 6/22 (27%) 4 (5-S, SP, G) (1-T, S, SP, G) Litter S. heidelberg 0/1 S. heidelberg 0/2 S. heidelberg 0/6 0/10 S. muenster 0/1 Waterers S. anatum 2/2 S. heidelberg 0/1 S. heidelberg 4/7 6/11 (55%) (2-T, S) 3 (2-S, SP, G) (2-T, S, SP, G) S. muenster 0/1 Environmental swabs S. heidelberg 0/1 S. heidelberg 1/4 1/5 (20%) (1-T, TS) Total 2/11 0/5 11/32 13/48 (27%) (18%) (34%) 1 T = tetracycline; S = streptomycin; SP = spectinomycin; G = gentamicin; TS = trimethoprim/sulfmethoxazole. 3 Number of isolates and resistant antibiotics. agar (LIA) 3 slants, respectively, using the same inoculating loop. The TSI and LIA slants were incubated at 37 C for 24 and 48 h, respectively. At least two positive isolates, identified by TSI or LIA, were maintained on BHI agar slants and incubated for 24 h at 37 C. Serology was performed on fresh (18 to 24 h) BHI cultures using Salmonella O (Group A E; Vi) polyvalent antisera 6 to confirm the presence of Salmonella. Isolates were shipped to the Salmonella Reference Center at the University of Pennsylvania for serotyping and antibiotic resistance screening. RESULTS AND DISCUSSION Sixty-nine isolates were serotyped from 58 Salmonellapositive samples in F1, F2, and F4. Salmonella was not detected in F3. Salmonella heidelberg was the most prevalent serotype, accounting for 65% of the isolates from these flocks. Other serotypes were identified as S. senftenberg (19%), S. muenster (10%), S. anatum (3%), and S. worthington (3%). Of the 48 isolates serotyped in F1, 92% were S. heidelberg, 4% S. muenster, and 4% S. anatum (Nayak, 2000). Of the 69 Salmonella serotypes screened, 25% were resistant to one or more of the following antibiotics: G, spectinomycin (SP), S, T, tobramycin (TO), and trimethoprim / sulfamethoxazole. Among the isolates screened, 100% of S. anatum and S. worthington, 29% of S. muenster, and 17% of S. heidelberg were resistant to the above antibiotics. Based on source, in F1, F2, and F4, the percentage of isolates that were antibiotic resistant were 55% of waterers, 43% of swabs, 33% of feeder contents, 26% of turkey ceca, and 5% of litter samples (Table 1). Of the 48 isolates in F1, 27% of serotypes were resistant to two or more antibiotics (Table 2). Two isolates of S. anatum from waterers at Week 2 were resistant to T and S. Of the six S. heidelberg isolates collected from turkey ceca at Week 18, five isolates were resistant to S, SP, and G, whereas the remaining isolate was resistant to T, S, SP, and G. Two isolates of S. heidelberg collected from waterers at Week 18 were resistant to S, SP, and G and the remaining two were resistant to T, S, SP, and G. One isolate of S. heidelberg from an environmental swab taken at Week 18 was resistant to T and trimethorprim/sulfamethoxazole. Antibiotic-resistant strains of S. heidelberg isolated from waterer samples at Week 18 might have been horizontally transmitted by turkeys. Of the 16 isolates in F2, 13% were resistant to T (Table 3). At Week 21 of the grow-out period S. worthington, detected in litter and feeder content samples, was resistant to T. Of the five isolates in F4, two were resistant to TO, S, SP, and G (Table 4). These isolates were S. muenster, taken from environmental swabs. Antimicrobial agents are fed at subtherapeutic levels to poultry to enhance growth rate and feed efficiency, and at therapeutic levels to prevent bacterial infections. Polymyxin B and combinations of 1) trimethoprim and sulfadiazine, 2) neomycin and polymyxin, and 3) trimethoprim and polymyxin B have been administered in feed or drinking water to reduce the levels of Salmonella in chickens (Craven, 1995). Williams (1985) reported that feeding oxytetracycline or neomycin, or both reduced S. typhimurium levels in broiler intestines. On the other hand, feeding antibiotics such as avoparicin and lincomycin favored colonization of S. typhimurium, whereas nitrofurazone enhanced colonization of S. infantis (Glisson, 1998). Spread of Salmonella to contact chickens was subsequently observed. The increase in Salmonella colonization could have resulted from decreased natural gut microflora, therefore, less competition. Antibiotics, by a process of selection, can facilitate the proliferation of resistant population of microorganisms. Gast et al. (1988) reported 40% isolation of drug-resistant S. typhimurium from kanamycin-treated rats fed a diet containing liver from kanamycin-treated poults. In another

ANTIBIOTIC RESISTANT-SALMONELLA 1499 TABLE 3. Antibiotic 1 resistance among Salmonella serotypes isolated from Flock 2 Week 13 Week 21 Sample Serotype Frequency 2 Serotype Frequency Total Turkey ceca...... S. senftenberg 0/1 0/1 Litter S. senftenberg 0/8 S. worthington 1/1 (1-T) 3 1/10 (10%) 4 S. heidelberg 0/1 Feed S. senftenberg 0/2 S. senftenberg 0/2 0/4 Feeder contents...... S. worthington 1/1 1/1 (100%) (1-T) Total 0/11 2/5 (40%) 2/16 (13%) 1 T = tetracycline. 3 Number of isolates and resistant antibiotic. study, Gast and Stephens (1988) found that administration of kanamycin in drinking water increased the frequency of isolation of kanamycin-resistant transconjugant S. typhimurium from the intestines and livers of poults inoculated with drug-sensitive S. typhimurium. Of the 1,824 Salmonella serotypes isolated from broiler carcasses, 57% were resistant to T, S, sulfisoxazole, G, or a combination of trimethoprim and sulfamethoxazole (Lee et al., 1993). Salmonella typhimurium, S. heidelberg, S. agona, and S. enteritidis were the major serotypes isolated from the carcasses. Bokanyi et al. (1990) found that S. hadar, S. heidelberg, and S. johannesburg, the major strains isolated from readyto-cook broilers or parts, were resistant to triple sulfa and T. Poppe et al. (1996) reported that 97.8% of Salmonella strains isolated from 295 layer flocks in Canada were resistant to SP and other antibiotics such as amikacin, ciprofloxacin, cephalothin, chloramphenicol, cotrimoxazole, G, kanamycin, neomycin, and polymyxin B. Salmonella anatum var. O15+ and S. typhimurium var. copenhagen strains were resistant to multiple antimicrobial agents. Salmonella binza, S. anatum var. O15+, S. schwarzengrund, and S. heidelberg isolated from 294 broiler flocks were resistant to ampicillin, carbenicillin, chloramphenicol, G, kanamycin, neomycin, nitrofurantoin, T, sulfisoxazole, and SP (Poppe et al., 1996). A survey component of this study found that low frequency of Salmonella detection in F2, F3, and F4 was attributed to the use of Termin-8, an antimicrobial feed additive used to inhibit Salmonella (Nayak, 2000). Therefore, based on the limitation of low levels of Salmonella detection, the current study does not infer antibiotic resistance associated with a specific source. Nonetheless, on the average approximately 25% of Salmonella serotypes isolated were antibiotic resistant, and throughout the study S. heidelberg was identified most often for antibiotic resistance. Identifying preharvest sources, characterizing serotypes, and determining antibiotic-resistance profiles can assist poultry producers and integrators in tracking Salmonella transmission pathways and in planning antibiotic use strategies. ACKNOWLEDGMENTS The authors gratefully acknowledge help offered by Chestina Barr and Mark Satterfield at the Reymann Memorial Farm, Wardensville, WV. REFERENCES Aarestrup, F. M., F. Bager, N. E. Jensen, M. Madsen, A. Meyling, and H. C. Wegener. 1998. Survelliance of antimicrobial resis- TABLE 4. Antibiotic 1 resistance among Salmonella serotypes isolated from Flock 4 Week 2 Week 20 Sample Serotype Frequency 2 Serotype Frequency Total Litter S. muenster 0/1...... 0/1 Environmental swabs...... S. muenster 2/2 2/2 (100%) 4 (2-TO, S, SP, G) 3 Feeder contents...... S. muenster 0/2 0/2 Total 0/1 2/4 (50%) 2/5 (40%) 1 TO = tobramycin; S = streptomycin; SP = spectinomycin; G = gentamicin. 3 Number of isolates and resistant antibiotics.

1500 NAYAK AND KENNEY tance in bacteria isolated from food animals to antimicrobial growth promoters and related therapeutic agents in Denmark. APMIS 106:606 622. American Veterinary Medical Association. 1993. Report of the AVMA panel on euthanasia. J. Am. Vet. Med. Assoc. 202:229 249. Ansari, F. A., and H. Khatoon. 1994. Multiple antibiotic resistance among gram negative bacteria isolated from poultry. Indian J. Exp. Biol. 32:211 212. Bokanyi, R. P. Jr., J. F. Stephens, and D. N. Foster. 1990. Isolation and characterization of Salmonella from broiler carcasses or parts. Poult. Sci. 69:592 598. Craven, S. E. 1995. Salmonella typhimurium, hydrophobic antibiotics, and the intestinal colonization of broiler chicks. J. Appl. Poult. Res. 4:333 340. Davis, M. A., D. D. Hancock, T. E. Besser, D. H. Rice, J. M. Gay, C. Gay, L. Gearhart, and R. DiGiacomo. 1999. Changes in antimicrobial resistance among Salmonella enterica serovar typhimurium isolates from humans and cattle in the Northwestern United States. Emerg. Infect. Dis. 5:802 806. DuPont, H. L., and J. H. Steele. 1987. The human health implication of the use of antimicrobial agents in animal feeds. Vet. Q. 9:309 320. Gast, R. K., and J. F. Stephens. 1988. Effects of kanamycin administration to poultry on the proliferation of drug-resistant Salmonella. Poult. Sci. 67:689 698. Gast, R. K., J. F. Stephens, and D. N Foster. 1988. Effects of kanamycin administration to poultry on the interspecies transmission of drug-resistant Salmonella. Poult. Sci. 67:699 706. Glisson, J. R. 1998. Use of antibiotics to control Salmonella in poultry. Pages 173 175 in Proceedings of the International Symposium on Food-Borne Salmonella in Poultry. American Association of Avian Pathologists, Baltimore, MD. Herikstad, H., P. Hayes, M. Mokhtar, M. L. Fracaro, E. J. Threlfall, and F. J. Angulo. 1997. Emerging quinoline-resistant Salmonella in the United States. Emerg. Infect. Dis. 3:371 372. Holmberg, S. D., M. T. Osterholm, K. A. Senger, and M. L. Cohen. 1984. Drug-resistant Salmonella from animals fed antimicrobials. New Engl. J. Med. 311:617 622. Klien, G. 1999. Food as a potential vector for antibiotic resistance. 1. Relevance of residues and selected foodborne infections and intoxicants. Berl. Munch. Tierarztl. Wochenschr. 112:365 369. Lee, L. A., V. L. Threatt, N. D. Puhr, P. Levine, K. Ferris, and R. V. Tauxe. 1993. Antimicrobial-resistant Salmonella isolated from healthy broiler chickens during slaughter. J. Am. Vet. Med. Assoc. 202:752 755. Manie, T., S. Khan, V. S. Brozel, W. J. Veith, and P. A. Gouws. 1998. Antimicrobial resistance of bacterial isolated from slaughtered and retail chickens in South Africa. Lett. Appl. Microbiol. 26:253 258. Nair, U. S., A. M. Saeed, P. M. Muriana, R. A. Kreisle, B. Barrett, C. L. Sinclair, and M. L. Fleissner. 1995. Plasmid profiles and resistance to antimicrobial agents among Salmonella enteritidis isolates from human beings and poultry in the midwestern United States. J. Am. Vet. Med. Assoc. 206:1339 1344. Nayak, R. 2000. Foodborne pathogens in poulty production and post-harvest control. Ph.D. Dissertation. West Virginia University, Morgantown, WV. O Brien, T. F., J. D. Hopkins, E. S. Gilleece, A. A. Mederios, R. L. Kent, B. O. Blackburn, M. B. Holmes, J. P. Reardon, J. M. Vergeront, W. L. Schell, E. Christenson, M. L. Bissett, and E. V. Morse. 1982. Molecular epidemiology of antibiotic resistance in Salmonella from animals and human beings in the United States. N. Engl. J. Med. 307:1 6. Okolo, M. I. 1986. Bacterial drug resistance in meat animals: a review. Int. J. Zoonoses 13:143 152. Poppe, C., and C. L. Gyles. 1987. Relation of plasmids to virulence and other properties of salmonellae from avian sources. Avian Dis. 31:844 854. Poppe, C., K. A. McFadden, and W. H. Demczuk. 1996. Drug resistance, plasmids, biotypes and susceptibility to bacteriophages of Salmonella isolated from poultry in Canada. Int. J. Food Microbiol. 30:325 344. Salyers, A. A., and D. D. Whitt, eds. 1994. Antibiotics: Mechanisms of action and mechanisms of bacterial resistance. Pages 97 110 in Bacterial Pathogenesis: A Molecular Approach. ASM Press, Washington, DC. Schuman, J. D., E. A. Zottola, and S. K. Harlander. 1989. Preliminary characterization of a food-borne multiple antibioticresistant Salmonella typhimurium strain. Appl. Environ. Microbiol. 55:2344 2348. Stavric, S., and J.-Y. D Aoust. 1993. Undefined and defined bacterial preparations for the competitive exclusion of Salmonella in poultry A review. J. Food Prot. 56:173 180. Threlfall, E. J., B. Rowe, and L. R. Ward. 1993. A comparison of multiple drug resistance in salmonellas from humans and food animals in England and Wales, 1981 and 1990. Epidemiol. Infect. 111:189 197. Tollefson, L., F. J. Angulo, and P. J. Fedorka-Cray. 1998. National surveillance for antibiotic resistance in zoonotic enteric pathogens. Vet. Clin. North Am. Food Anim. Pract. 14:141 150. van den Bogaard, A. E., and E. E. Stobberingh. 1999. Antibiotic usage in animals. Impact on bacterial resistance and public health. Drugs 58:589 607. van der Wolf, P. J., J. H. Bongers, A. R. Elbers, F. M. Franssen, W. A. Hunneman, A. C. van Exsel, and M. J. Tielen. 1999. Salmonella infections in finishing pigs in the Netherlands: bacteriological herd prevalence, serogroup and antibiotic resistance of isolates and risk factors for infection. Vet. Microbiol. 67:263 275. Vidon, D. J., S. Jacob, and M. Ganzenmuller. 1978. Incidences of simple and transferable drug resistance in Escherichia coli and Salmonella isolated from various foods: Identification of a R plasmid in S. saint-paul. Ann. Microbiol. (Paris) 129:155 159. Wagner, J., and H. Hahn. 1999. Increase of bacterial resistance in human medicine by resistance genes of bacteria from meat supplying animals. Berl. Munch. Tierarztl. Wochenschr. 112:380 384. Williams, B. J. 1985. The effect of neomycin and oxytetracycline alone or combined upon the incidence of salmonellosis in broiler chickens. Poult. Sci. 64:1455 1457.