HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Santos, F. B. O. Articles by Ferket, P. R. Search for Related Content PubMed PubMed Citation Articles by Santos, F. B. O. Articles by Ferket, P. R.

Influence of Housing System, Grain Type, and Particle Size on Salmonella Colonization and Shedding of Broilers Fed Triticale or Corn-Soybean Meal Diets¹

Department of Poultry Science, North Carolina State University, Raleigh 27695-7608

² Corresponding author: brian_sheldon{at}ncsu.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Salmonella colonization in poultry may be influenced by grain type and particle size. Broilers reared either in nonlitter cage-based housing or in a conventionally floored litter house from 0 to 42 d were assigned to 1 of 4 dietary treatments: 1) ground corn-soybean meal (C, 560 µm), 2) coarsely ground corn-soybean meal (CC, >1,700 µm), 3) ground triticale-soybean meal (T, 560 µm), or 4) whole triticale-soybean meal (WT). A 4-strain cocktail of Salmonella enterica was orally gavaged into each chick at placement. Growth performance, cecal and fecal Salmonella populations, gizzard and proventriculus pH, intestinal size, jejunum histomorphometry, and carcass yields were measured. Broilers responded differently to the dietary treatments according to the housing system used. At 42 d, birds reared on litter and fed ground grain had greater BW than those fed coarse grain (2.87 vs. 2.71 kg), whereas cage-reared broilers fed ground triticale were heavier than those fed corn (2.75 vs. 2.64 kg). Broilers raised on litter had a better feed conversion ratio than those raised in cages (1.71 vs. 1.81 g/g). Independent of the housing system, relative eviscerated carcass weights of birds fed T and C were heavier than those of CC- and WT-fed broilers (762 vs. 752 g/kg). Generally, the jejunum villus area and mucosal depth were larger, whereas the small intestine was lighter and shorter in broilers raised on litter. Relative gizzard weights of broilers raised on litter and fed the coarser diets were heavier than those of broilers reared in cages and fed finely ground diets. Feeding whole or coarsely ground grains decreased cecal Salmonella populations in 42-d-old broilers (3.8, 3.9, 4.4, and 4.4 log most probable number/g for CC, WT, C, and T, respectively). Additionally, 42-d-old broilers reared on litter had lower cecal Salmonella populations than those in cages (3.8 vs. 4.4 log most probable number/g). In conclusion, as a feed ingredient, triticale is a good alternative to corn, resulting in improved BW and reduced Salmonella colonization. Broilers raised on litter may have achieved lower cecal Salmonella populations than caged birds because access to litter may have modulated the intestinal microflora by increasing competitive exclusion microorganisms, which discouraged Salmonella colonization.

Key Words: Salmonella • housing system • grain particle size • triticale • broiler

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Food-producing animals are the main reservoir for several human bacterial pathogens. Salmonella species are among some of the pathogens causing thousands of foodborne infections annually by the consumption of contaminated meat and other products of animal origin (Oosterom, 1991). Although salmonellosis cases are occasionally associated with exposure to pets, reptiles, and contaminated water, it has been estimated that approximately 95% of cases are primarily of foodborne origin (Mead et al., 1999). Salmonella has been closely associated with the consumption of poultry and poultry products (Hoszowski et al., 1996; Byrd et al., 1997), and several human infections likely originate at the poultry production level. The prevention of poultry-related salmonellosis has been a major goal of governmental agencies, processing plants, and producers. An important means of preventing human salmonellosis is by preventing the infection of live poultry at the farm level (Porter, 1998).

Competitive exclusion products, probiotics, prebiotics, organic acids, and enzymes are often used to reduce or eliminate Salmonella and other pathogens associated with poultry production. The competitive exclusion approach was first proposed by Nurmi and Rantala (1973), and this method has been shown to have a protective effect on Salmonella colonization of broilers (De Oliveira et al., 2000). More recently, the use of probiotics, which are defined microbial species, to protect poultry from pathogen colonization has been widely accepted. Improvement of productive performance and feed conversion ratio (FCR), along with decreased intestinal Salmonella enterica serovar Typhimurium colonization and bird mortality, have been reported after the administration of probiotics to broiler chickens (Corrier et al., 1995). Prebiotics are dietary components that are not digested by the host but that stimulate the growth, activity, or both of one or a limited number of beneficial commensal bacteria in the gastrointestinal tract (GIT), predominantly those that produce short-chain fatty acids (Lan, 2004). Because bacterial species differ from each other in relation to their substrate preferences and growth requirements, the bacterial community structure is very much dependent on the diet as the ultimate source of substrates for metabolism (Wagner and Thomas, 1987; Savory, 1992). Therefore, both probiotics and prebiotics can encourage the colonization of commensal bacteria in the lower intestinal tract.

Dietary inclusion of cereals has been shown to influence the ecology of the intestinal tract of poultry, including Salmonella colonization. Dietary inclusion of whole wheat can decrease intestinal Salmonella colonization in broilers (Bjerrum et al., 2005). Similar effects were observed with turkeys fed wheat- or triticale-based diets (Santos, 2006). Altering the feed structure by changing grain particle size is another approach used to control salmonellae in the poultry industry (Bjerrum et al., 2005; Huang et al., 2006). Improvements in feed efficiency (Plavnik et al., 2002; Lentle et al., 2006) and nutritive value of the feed (Svihus et al., 2004) have been reported when broilers were fed whole wheat-based diets.

In addition to affecting growth performance, changing the feed structure or form can influence Salmonella infection, as observed in some studies with broiler chickens. Bjerrum et al. (2005) reported that broilers fed pelleted feed had a greater Salmonella population in the gizzard than those consuming pelleted feed supplemented with whole wheat. Similarly, Huang and coworkers (2006) demonstrated an increased incidence of Salmonella Typhimurium in the gizzard and cecal contents of broilers fed finely ground, as opposed to coarsely ground, mash feed. Moreover, the physical characteristics of the feed, such as particle size, can influence intestinal pH of the digestive tract of broilers (Svihus et al., 2004; Huang et al., 2006), which may influence Salmonella colonization. Therefore, feed structure may influence Salmonella colonization by changing the gastrointestinal ecology of broilers.

Disposal of poultry litter and preventing the recycling of infectious pathogens is also a great concern for the poultry industry. Poultry litter consists of manure, a bulking agent to absorb moisture, and other components (i.e., feathers and soil; Kelley et al., 1994). To address the concerns about litter-rearing systems, nonlitter systems are being considered for the broiler and turkey industry. A commercially available nonlitter system allows broilers to be raised in large cages with plastic-covered nylon floors. These trampoline-like floors provide a soft, non-abrasive surface with ample rigidity, allowing the birds to be raised on a surface nearly free of feces, yet avoiding many of the animal welfare problems associated with standard cage systems such as breast blisters, folliculitis, and wing and leg problems. A belt beneath the cages collects and dries the manure before it is automatically conveyed from the facility. Havenstein et al. (1998) reported that broilers raised in this system have better weight gains and lower mortality rates than those raised on litter floors and that nitrogen volatilization from the manure is greatly reduced, thus creating a nutrient-dense by-product that is more easily handled.

The study reported herein was designed to study the effects of grain, particle size, and housing system. It was hypothesized that dietary inclusion of triticale, especially the whole grain, could help maintain intestinal health and discourage the enteric pathogen colonization of broilers challenged with Salmonella. It was also hypothesized that Salmonella colonization in broilers could be influenced by the type of flooring system (litter vs. nonlitter), which differed in the degree of bird contact with manure. To test these hypotheses, experimental diets were formulated either with corn or triticale, and broilers were raised either in a conventionally floored litter house or in a nonlitter cage system. We also examined how the shift in Salmonella population by the dietary treatments was related to changes in growth performance, intestinal histomorphometry, and gizzard pH.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bird Husbandry

A total of 2,561 one-day-old Ross 508 (Aviagen, Huntsville, AL) broiler chickens obtained from the North Carolina State University hatchery (Raleigh, NC) were weighed, neck-tagged, and then orally gavaged with 1 mL (8 x 10⁵ cfu) of a cocktail of S. enterica as described below. These birds were then randomly assigned to 2 experimental housing systems: an 11 x 30 m curtain-sided house with conventionally floored litter, and an 11 x 24 m power-ventilated house equipped with a nonlitter cage-based system (Farmer Automatic of America Inc., Register, GA). Forty broilers were randomly assigned to each of the 32 pens used in each house. In the cage-equipped house, the birds were randomly placed into 2 battery units consisting of 2 tiers spaced 24 cm apart, with 8 pens per tier. Each pen was 120 cm wide, 194 cm long, and 42 cm high, providing 582 cm² of floor space per bird. Each battery tier was individually equipped with Farmer automatic chain feeders and Lubing nipple drinkers (Cleveland, TN). In the conventional house, the birds were randomly placed into 2 blocks of 16 pens, with each block separated by a center work area. Each floor pen was 122 cm wide x 366 cm long, providing 1,116 cm² of floor space per bird. Each floor pen was provided with 10 cm of soft pine wood shavings. House temperature and ventilation rate were maintained similarly across the 2 houses during the experimental period, which ran from March to April 2005. House temperatures throughout the trial period included highs of 19 to 39°C (average of 27.5°C) and lows of 14 to 37°C (average of 24.8°C), whereas ambient temperatures included daily highs of 4 to 34°C (average of 18.9°C) and daily lows of –11 to 22°C (average of 6.7°C). Birds were kept on a 24-d light schedule in both houses. Feed and water were given ad libitum. Broilers were inspected daily and birds with visual health problems or poor body condition were removed, weighed, and euthanized by cervical dislocation. All mortality was weighed so that an appropriate adjustment of FCR could be made.

Experimental Design and Diets

The experimental design consisted of 4 dietary treatments each with 8 replicate pens of 40 broiler chickens per house. The dietary treatments were randomized within 4 blocks in each house to account for position effects. Over the entire experimental period (1 to 42 d), all broilers were fed either a corn- or a triticale-soybean meal (SBM) diet. From 1 to 14 d of age (starter diet), feed was offered in crumble form, and from 15 to 42 d (grower and finisher diets), feed was offered in pellet form (Table 1). The experimental diets were formulated without antimicrobials or coccidiostats by using the least-cost linear programming to meet or exceed the NRC (1994) nutrient requirements. The corn included in the feed was either finely ground or coarsely ground for treatments 1 (C) and 2 (CC), respectively; and the triticale was either finely ground or whole for treatments 3 (T) and 4 (WT), respectively. The corn used to prepare the fine and coarse diets was supplied and ground by Southern States Feed Mill (Farmville, NC). The finely ground corn had an average particle size of 560 µm, whereas the coarsely ground corn had a final average particle size greater than 3,000 µm. The triticale (Tritical-498, Lot No. TC-1101-B, Virginia, USA) was supplied by Resource Seeds Inc. (Golroy, CA) and ground at the North Carolina State University feed mill (Raleigh, NC) by using a hammermill (Bliss Industries Inc., Ponca City, OK) equipped with a 3-mm screen, resulting in a final average particle size of 560 µm. The forth experimental diet was prepared by using the whole triticale grain. A vertically oscillating sieve shaker (W. S. Tyler Inc., Mentor, OH) was used to determine particle size distributions of the pelleted diets (grower and finisher; Grant and Colenbrander, 1990). Approximately 500 g of feed was weighed and placed on the top screen of a stack of 8 sieves (US Standard Test Sieve, Fisher Scientific, Bohemia, NY) and then shaken for 20 min. The time was predetermined as the point at which the weight of feed particles recovered in the pan did not change. The screen sizes used were 4.75, 3.35, 2.00, 1.70, 1.40, 1.18, 1.00, and 0.425 mm. The arithmetic mean feed particle size and percentage of feed particles recovered on the sieves were calculated.

The triticale-based diets were supplemented with Avizyme 1502 (Danisco Animal Nutrition, Wiltshire, UK), which is a commercial fine granular enzyme preparation obtained from fermentation of Bacillus subtilis and genetically modified Trichoderma longibrachiatum. The genetically modified T. longibrachiatum produces a heat-stable endoxylanase. The enzyme preparation contained standardized activities of at least 600 endo-1,4-β-xylanase units, 8,000 units of subtilisin, and 800 units of -amylase per gram of product.

Bacterial Strains and Inoculum

A cocktail of S. enterica subsp. enterica serotypes Typhimurium (ATCC 700408), Newport (ATCC 6962), Heidelberg (ATCC 8326), and Kentucky (field isolate) was used as the inoculum. The serovar Kentucky had been isolated previously from turkey feces and was serotyped by the National Veterinary Service Laboratories (Animal and Plant Health Inspection Services, USDA, Ames, IA). For preparation of the inoculum, the 4 serovars were grown separately overnight at 37°C in brain-heart infusion broth (Oxoid Ltd., Ogdensburg, NY). The cultures were then mixed together and serially diluted in buffered peptone water (BPW, Oxoid Ltd.) to a final concentration of 8 x 10⁵ cfu/mL. The cell count was determined by direct plating on brain-heart infusion agar plates and incubation overnight at 37°C. Negative controls were used for all plating procedures to ensure that the media had been properly sterilized.

Serotyping

Salmonella isolates recovered from meconium samples were restreaked on modified Lys iron agar (Oxoid Ltd.) plates, and 1 isolated colony from each sample was randomly picked, streaked onto tryptic soy agar (Oxoid Ltd.) slants, and shipped to the National Veterinary Service Laboratories for serotype determination. Serotyping was done based on the Kauffmann-White scheme. For the purposes of this document, serovars of S. enterica subsp. enterica are referred to as Salmonella accompanied by the serotype name. For example, S. enterica subsp. enterica serovar Typhimurium is referred to as Salmonella Typhimurium.

Data Collection and Analytical Methods

At each sampling time, 1 bird was randomly chosen from each pen, weighed, and euthanized by cervical dislocation. The abdomen was opened and the gizzard, small intestine (duodenum, jejunum, and ileum), and ceca were collected. These tissue samples were collected at 7, 14, 21, 28, 35, and 42 d of age, with the exception for the intestinal samples used for histology analysis, which were collected on d 3. Feed consumption by pen and individual bird BW were recorded at 1, 14, 28, and 42 d of age.

Gizzard Weight and pH. After being removed from the abdominal cavity, the gizzard was placed in a Whirl-Pak bag (Fisher Scientific) and stored on ice during transport to the laboratory. Upon arrival at the laboratory, the contents of the gizzards were flushed with sterile deionized (DI) water for pH determination and then the organ was weighed. After being flushed with DI water (approximately 1:10 dilution, content:water ratio) into sterile Whirl-Pak bags, suspensions were manually homogenized by shaking for 1 min. A pH probe (Fisher Scientific) was then inserted directly into each bag and the pH was recorded (Fisher Scientific). The probe was washed between readings by using sterile DI water.

Salmonella Isolation. Ceca were aseptically removed, weighed, and stored on ice before being quantitatively cultivated for Salmonella isolation. Cultivation for Salmonella was performed immediately after transport to the laboratory by using the most probable number (MPN) procedure described by Santos et al. (2005) with minor modifications. Briefly, cecal samples were placed in separate 17.78 x 30.48 cm (7 x 12 in.) sterile filtered stomacher bags (Spiral Biotech Inc., Norwood, MA) followed by the addition of BPW at a 1:10 dilution in each bag. The 3-tube MPN technique was used, with BPW as a preenrichment broth, Rappaport-Vassiliadis broth for selective enrichment, and modified Lys iron agar for selective-differential plating.

On the day of placement, meconium samples from all poults were collected and examined for salmonellae by using the MPN procedure as described above. Meconium samples were collected prior to Salmonella challenge. Fecal samples were collected on d 14 and 28 and cultivated for Salmonella by using the MPN procedure to ensure that broilers were colonized by Salmonella and were actively shedding the pathogen. A composite sample of 10 g of fresh feces was collected beneath each cage and on the litter surface. Twelve hours before fecal collection, the belts of the nonlitter cage system were cleaned and the old fecal material removed.

Intestinal Histomorphometry. The small intestine segments (duodenum, jejunum, and ileum) were collected and intestinal weights and lengths were measured. The length of the small intestine was measured for each segment as defined by the following: duodenum, from the gizzard to the pancreatic and bile duct; jejunum, from the bile duct to Meckel’s diverticulum; and ileum, from Meckel’s diverticulum to the ileo-cecal-colonic junction (Samanya and Yamauchi, 2002). Intestinal segment weights (g) were recorded after the digesta content had been manually removed. Intestinal weights and lengths were calculated relative to live bird BW (kg; Bjerrum et al., 2005).

For histology analysis, approximately 3 cm of tissue was sampled from each bird at 3 d of age (Salgado et al., 2002) and prepared following the procedure described by Santos (2006). Ten individual villi were assessed per section to calculate villus height, crypt depth, muscularis depth, villus surface area, and the villus height-to-crypt depth ratio (Iji et al., 2001). An average of the 10 measurements assessed per section was expressed as a mean for the corresponding jejunum segment. Each bird was the experimental unit for statistical analysis.

Carcass Yield. On d 42 of the trial, 2 birds per pen and cage (16 birds per treatment per house) were selected at random for estimation of carcass characteristics. Feed was removed for 12 h before processing. Birds were allowed access to water during the first 4 h of the feed withdrawal period. At the time of slaughter, the fasted BW (live weight) of each bird was measured immediately prior to stunning and exsanguination. All birds were electrically stunned, killed by hand with a conventional unilateral neck cut to sever the carotid artery and jugular vein, bled for 180 s, scalded at 63°C for approximately 120 s, and then placed into a rotary drum mechanical picker for 30 s.

After the head, shanks, and feet were removed, carcasses were eviscerated by cutting around the vent to remove the abdominal fat pad and all of the viscera except the kidneys. After evisceration, the hot carcasses were submerged overnight in an ice-water bath without agitation. After overnight chilling, the carcasses were drained and cut into their component parts: wings, drumsticks, thighs, pectoralis major, pectoralis minor, and rack (i.e., the thoracic vertebrae and ribs with overlying skin and muscle, the clavicle, the sternum, and neck) along with the back posterior. Wings, drumsticks, thighs, pectoralis major, and pectoralis minor weights were recorded to the nearest 0.01 g. Further processing was completed for all birds within approximately 1 h from the time the birds were removed from the ice bath.

Statistical Analysis

Statistical analysis of results was performed by using the GLM procedure of SAS (SAS Institute, 1996) according to the following general model: Y_ijklm = M + A_i + B_j(l) + D_k + H_l + (AD)_ik + (AH)_il + (DH)_kl + e_ijklm, where Y_ijklm is the observed dependent variable; M is the overall mean; A_i is the age effect; B_j is the block effect, which is nested in the house effect; D_k is the dietary treatment effect; H_l is the house effect; (AD)_ik is the interaction between age and dietary treatment; (AH)_il is the interaction between age and house; (DH)_kl is the interaction between dietary treatment and house; and e_ijklm is the random error. Replicate pens or cages of 40 birds each served as experimental units. When treatment effects were found to be significant with the F-test (P < 0.05), the treatment means were separated by the least squares means function of SAS (SAS Institute, 1996), with a confidence level of P < 0.05. Before statistical analysis, all MPN data were transformed to the base-10 logarithm and the intestinal measurement data were transformed relative to BW.

Animal Ethics

The experiments reported herein were conducted according to the guidelines of the Institutional Animal Care and Use Committee at North Carolina State University. All husbandry practices and euthanasia were performed with full consideration of animal welfare.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Diet Particle Size and Performance

The arithmetic mean particle sizes of the 4 dietary treatments were 3.39, 3.69, 4.25, and 2.99 mm for treatments 1 (C), 2 (CC), 3 (T), and 4 (WT), respectively (Table 2). The overall average recovery of the 4 dietary treatment samples from the screen sieves was 99.86%. The most noticeable differences between the particle size distributions were in the percentages of particles retained on the first 3 screens (4.75, 3.35, and 2.00 mm). Finely ground triticale produced the best quality pellet of the 4 treatments, whereas WT was the worst.

Gizzard

Gizzard weight was influenced by housing design and dietary treatment (Figure 1). No housing x diet interaction effects were observed for the gizzard relative weights. On average, the relative gizzard weight of broilers raised on litter was significantly greater than those in cages (22.2 vs. 20.8 g/kg). Regardless of grain type, broilers fed the coarser diets had heavier gizzards than those fed the finely ground grain diets (23.4 vs. 19.6 g/kg).

View larger version (14K): [in this window] [in a new window]   Figure 1. Effect of housing design, grain type, and particle size on relative gizzard weight (RGW) of broilers fed finely ground corn (C), coarsely ground corn (CC), finely ground triticale (T), or whole triticale (WT) from 1 to 42 d of age. Relative weights (g of organ/kg of BW) are the least squares means of organ weights of 8 birds per dietary treatment per house collected on d 7, 14, 21, and 28. Means with different superscript letters within dietary treatments or within housing designs differ (A,BP < 0.01).

Relative small intestine weights and lengths are summarized in Tables 8 and 9. The relative weight and length of the total small intestine were significantly less in broilers raised on litter than in those raised in cages. At 42 d, the litter-raised broilers had a significantly lower relative jejunum length (20.6 vs. 22.8 cm/kg) and ileum weight and length (7.4 vs. 8.7 g/kg; 20.4 vs. 23.1 cm/kg).

Histological measurements of the jejunum were minimally affected by dietary treatment but were influenced by housing type (Table 10). Contrast analysis revealed that only the villus apical width increased when broilers were fed triticale (58.96 vs. 54.36). Villus height, villus area, the villus height-to-crypt depth ratio, and mucosal depth were significantly larger for broilers reared on litter, regardless of the dietary treatment. There was no significant difference in villus apical and basal width, crypt depth, and muscularis depth between the birds raised on litter and those raised in cages.

Dietary treatments and housing design significantly influenced carcass yields (relative breast weight and total meat yield) of broilers at 42 d of age (Table 11). Contrast analysis showed that raising broilers on litter and feeding finely ground grain diets, regardless of grain type, increased the relative breast weight (214.53 vs. 202.69 g/ kg) and meat yield (525.50 vs. 511.77 g/kg) when compared with broilers fed whole or coarsely ground grain diets. In contrast, raising broilers in cages and feeding triticale-based diets, regardless of grain particle size, increased the relative breast weight (210.80 vs. 193.76 g/ kg) and meat yield (532.58 vs. 510.03 g/kg) compared with broilers fed corn-based diets.

Most probable number and serotyping analyses of meconium samples showed that chicks were naturally infected with Salmonella Infantis. However, there were no significant differences in Salmonella populations recovered from the meconium samples, which averaged 2.38 log MPN/g (Table 12). Additionally, cecal Salmonella populations at 7 d were not affected by dietary treatment (Figure 2) or housing type (Figure 2), confirming that all birds started with an equivalent dose of Salmonella.

View larger version (27K): [in this window] [in a new window]   Figure 2. Effect of dietary treatments on Salmonella cecal populations of broilers fed finely ground corn (C), coarsely ground corn (CC), finely ground triticale (T), or whole triticale (WT) in crumble form (1 to 14 d) or pellet form (15 to 42 d). Means with different letters within an age group differ (a,bP < 0.05, A–CP < 0.01). MPN = most probable number.

View larger version (10K): [in this window] [in a new window]   Figure 3. Effect of housing design on Salmonella cecal populations of broilers raised in a conventional litter-floored house or in the nonlitter cage-system house (Broilermatic). Means with different letters within an age group differ (A,BP < 0.01). MPN = most probable number.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Several differences in growth performance and pathogen colonization were observed between broilers reared in the litter-free cage system and those raised on litter. Moreover, these differences were in some cases significantly influenced by the type of grain in the diet and the particle size of that grain. Broilers raised on litter had better performance when fed finely ground grain, regardless of the grain type. This improvement in growth rate by feeding diets containing finely ground grain was most likely associated with the improved pellet quality, which enhanced feed consumption. Engberg et al. (2002) and Bennett et al. (2002) demonstrated increased feed intake and BW gains by feeding pellets. Supplementing pelleted diets with whole wheat and whole barley decreased the growth rate of chickens (Bennett et al., 2002), similar to what was found in the present study. Conversely, the influence of grain type (corn vs. triticale) on the growth performance of broilers raised in cages was much greater than the influence of grain particle size. Regardless of particle size, triticale-based diets increased BW gains yet had no effect on feed consumption or FCR. Other published studies have shown that increased addition of high-fiber cereals in broiler (Plavnik et al., 2002) and turkey (Santos, 2006) diets improved weight gain and FCR.

The major difference between the nonlitter cage system and the conventionally floored litter house was access to litter. In the conventional house, broilers were grown on litter and had free access to the pine shavings, whereas the only source of fiber for broilers raised in the cage system was from the nonstarch polysaccharides (NSP) present in the diets. The difference in total NSP between the corn-based diets and triticale-based diets was, on average, 40 g/kg. In addition, triticale-based diets had approximately 33% more soluble NSP than did corn-based diets. These data suggest that there may be a minimum fiber intake requirement for birds. Studies with humans (Marlett et al., 2002), broilers (Jimenez-Moreno et al., 2004, 2006), and many other animals (NRC, 1995) have shown the importance of a minimum daily intake of fiber for normal gastrointestinal function. Thus, the corn-based diets may not have provided sufficient fiber to sustain normal intestinal development and function. However, access to the litter likely fulfilled the fiber requirements of the birds, although the actual amount of litter consumed was not measured in the present study. This conclusion is supported by the observation that the litter-reared broilers had significantly greater jejunum villus height, villus area, villus height-to-crypt depth ratio, and mucosal depth than cage-reared birds, regardless of the dietary treatment. Furthermore, the FCR was significantly lower, by approximately 5%, when broilers were reared on litter, confirming that access to litter material improved the efficiency of nutrient uptake.

The experimental diets had no significant effect on the total size of the small intestine. However, litter-raised broilers had lighter and shorter relative small intestines than cage-reared broilers. The difference was attributed mostly to reduced ileum and jejunum weights and lengths. A possible reason for this finding is that broilers raised on litter had reduced small intestinal mass, which may have improved the microbial stability in the intestinal tract and lowered the pathogen load. Humphrey et al. (2002) reported that germ-free chickens had a lower intestinal mass than conventional chickens. Sonnenburg et al. (2004) reported that a healthy intestine is composed of protective dense communities of commensal microorganisms that are permanent residents of the intestinal tract. They concluded that an increased microbial diversity would result in the development of stable microbial communities and decreased intestinal mass while maintaining intestinal efficiency. Intestinal microbial community diversity is recognized as an important host defense mechanism against infection (Abrams and Bishop, 1966). Accordingly, we observed that broilers raised on litter had lower Salmonella populations than birds raised in cages.

Gizzard pH was also influenced by housing system and dietary treatment. In general, gizzard pH was more acidic when broilers were reared on litter, probably related to the extra fiber intake because of litter consumption. Feeding whole grains to broilers resulted in a more acidic gizzard environment. Previous studies have also shown a significant reduction in the gizzard pH in response to feeding whole wheat (Engberg et al., 2004; Bjerrum et al., 2005). As the authors reported, the low pH was caused either by stimulation of bacterial fermentation or an increased secretion of hydrochloric acid by the proventriculus, or both. Results from the present study support some of these findings.

Feeding broilers coarser diets significantly increased the relative gizzard weight of broilers throughout the trial (7 to 42 d), similar to the findings of Engberg et al. (2004) and Santos et al. (2006). The increased gizzard weight was associated with increased mechanical stimulation in response to the ingestion of coarse feed particles. Engberg et al. (2002) observed an increased luminal content in the upper small intestine (duodenum and jejunum) after birds were fed pelleted feed. Svihus and Hetland (2001) also reported that the stimulation of gizzard activity by coarsely ground grain or whole grain may modulate pancreatic enzyme activity in the duodenum, thereby regulating the transit of starch by avoiding starch overload in the small intestine and improving starch digestibility. Access to litter material, and thus greater insoluble fiber, also increased relative gizzard weights by stimulating the organ. Santos et al. (2006) observed increased relative gizzard weights after supplementing corn-based turkey diets with 4% pine shavings. Similar results were reported in broilers fed diets supplemented with oat hulls (Hetland and Svihus, 2001).

The observed carcass yield data were in accordance with the performance results. Relative gizzard weights of the broilers processed at 42 d were similar to the relative gizzard weights observed from birds sampled throughout the study (7 to 42 d). Broilers raised on litter also had heavier livers, lighter thighs, and more abdominal fat than those raised in the litter-free system. In general, birds fed finely ground grain produced heavier eviscerated carcasses than those fed coarser diets. However, the yields of white meat and total meat followed a response trend similar to BW. Feeding finely ground grain in the litter house and feeding triticale-based diets in the cage system resulted in greater white and total meat yields. In comparison, Bennett et al. (2002) observed that carcass yields of broilers were not adversely affected by feeding whole wheat or barley. The authors also reported no significant difference in eviscerated carcass weights between diets containing ground and whole wheat and barley.

In addition to growth performance analysis, the experimental design of the present study allowed us to test the hypothesis that Salmonella colonization would decrease as the grain particle size and dietary NSP level increased. The soluble NSP are primary substrates for the fermentation of important commensal bacteria found in the GIT (i.e., Lactobacillus and Bifidobacterium), which can competitively prevent intestinal colonization by pathogens (Lan, 2004). To enhance the effect of dietary NSP, the triticale-based diets were supplemented with 600 endoxylanase units of endoxylanase/kg of feed. Previous studies have shown that enzyme supplementation of wheat- or triticale-based diets decreased Salmonella cecal colonization in turkeys (Santos, 2006). Appropriate enzyme supplementation of cereal-based diets was shown to decrease the intestinal viscosity and increase the performance and nutrient digestibility of broilers (Flores et al., 1994; Brum et al., 2000; Silva and Smithard, 2002) and laying hens (Coon et al., 1988; Lázaro et al., 2003). Additionally, enzyme supplementation of diets high in NSP content can promote the development of a healthier enteric microflora ecosystem (Steenfeldt et al., 1998).

As hypothesized, Salmonella colonization of the broiler intestine was discouraged by a diet high in NSP content. Although all treatments had similar Salmonella populations at 7 d, at 28 d the Salmonella populations were greater in broilers fed the corn-based diets compared with those fed the triticale-based diets, but by 42 d, no statistical difference was attributable to grain type among the dietary treatments. These findings agree with Santos (2006), who reported a faster rate of reduction in Salmonella populations among turkey toms fed wheat- and triticale-based diets than among those fed corn-based diets. The author demonstrated that toms were more capable of recovering from a Salmonella infection when fed wheat- or triticale-based diets. Conversely, the influence of grain particle size on Salmonella colonization took longer to develop than the dietary fiber content. Broilers fed the coarse diets had lower Salmonella cecal populations by approximately 1 and 0.5 log MPN/g at 28 and 42 d, respectively. Similar responses were observed by other investigators. Bjerrum et al. (2005) showed that whole wheat feeding decreased Salmonella Typhimurium populations in the ileum of broilers. Engberg et al. (2004) observed a reduction in the population of lactose-negative enterobacteria throughout the GIT, with the exception of the ceca. Salmonella was more effectively reduced in our study because of the inclusion of greater concentrations of coarser particles in the CC and WT experimental diets.

Dietary NSP, including the high NSP content present in litter, is believed to have beneficial effects on the GIT of poultry. These benefits include the shift of microbial populations toward one that competitively excludes harmful microorganisms that disrupt the host-microflora ecosystem. Additionally, dietary NSP indirectly stimulate the immune system by encouraging the growth of enteric lactic acid bacteria, which, through their cell wall components (Takahashi et al., 1993; Haller et al., 1999), evoke an immune response (Stewart-Tull, 1980), especially on the intestinal mucosal surface (Link-Amster et al., 1994). Therefore, feed ingredients containing high levels of NSP and access to litter would serve as potential prebiotic sources that could both promote intestinal health by encouraging the proliferation of commensal microflora and discourage the colonization of Salmonella species in the intestinal tract of broilers.

Cecal Salmonella populations and fecal shedding of Salmonella were significantly influenced by housing design. The difference in Salmonella populations between the 2 housing systems reached a maximum of 1.3 log MPN/g at 35 d for the cage-reared birds compared with those reared on litter. By 42 d, the difference between these 2 systems declined to 0.6 log MPN/g. Considering that the broilers raised in both housing facilities were fed the same diets, the difference between the 2 is presumably related to their access to litter. Contrary to the general assumption that litter consumption increases Salmonella colonization through reinfection, the present study showed a consistent reduction in Salmonella populations for birds reared on litter. The coarse wood components in litter may play a major role in reducing pathogen colonization in the avian ceca, perhaps by the mechanical stimulation of the gizzard and proventriculus or by serving as a seeding agent for competitive exclusion microorganisms.

In conclusion, the results of this study show that dietary NSP content, housing design, grain type, and particle size influenced intestinal size and histomorphometry as the bird adapted to changes in nutrient digestibility. This study also demonstrated that replacing corn with triticale in broiler diets and using coarsely ground or whole grain can have a significant effect on Salmonella colonization and the intestinal health of broilers. Furthermore, dietary inclusion of coarse grain particles or a grain high in dietary fiber content, such as triticale, can discourage Salmonella colonization in broiler ceca. Finally, rearing broilers in a conventional litter-based house was superior to a nonlitter cage-based housing system in terms of growth performance and resistance to colonization after a Salmonella challenge.

	ACKNOWLEDGMENTS

 
This study was supported by the USDA Initiative for Future Agriculture and Food Systems grant. The authors wish to thank Annette Israel, Jamie Warner, Jean de Oliveira, Ondulla Foye, Renee Plunske, Mike Mann, Robert Neely, and the North Carolina State University Poultry Educational Unit farm employees (Raleigh, NC) for their technical assistance during this study. Appreciation is also extended to Southern States Feed Mill (Farmville, NC) for providing and grinding the corn used in the experimental diets, and to Resource Seeds Inc. (Golroy, CA) for providing the triticale used in the experimental diets.

	FOOTNOTES

 
¹ Use of trade names in this publication does not imply endorsement by the North Carolina Agriculture Research Service or the North Carolina Cooperative Extension Service of the products mentioned, nor criticism of similar products not mentioned.

Received for publication December 7, 2006. Accepted for publication November 21, 2007.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Abrams, G. D., and J. E. Bishop. 1966. Effect of the normal microbial flora on the resistance of the small intestine to infection. J. Bacteriol. 92:1604–1608.[Abstract/Free Full Text]

Bennett, C. D., H. L. Classen, and C. Riddell. 2002. Feeding broiler chickens wheat and barley diets containing whole, ground and pelleted grain. Poult. Sci. 81:995–1003.[Abstract/Free Full Text]

Bjerrum, L., K. Pedersen, and R. M. Engberg. 2005. The influence of whole wheat feeding on salmonella infection and gut flora composition in broilers. Avian Dis. 49:9–15.[CrossRef][Web of Science][Medline]

Brum, P. A. R., D. L. Zanotto, A. L. Guidon, P. S. Rosa, G. J. M. M. Lima, and E. S. Viola. 2000. Triticale in diets for broilers. Pesqi. Agropecu. Bras. 35:229–239.

Byrd, J. A., D. E. Corrier, J. R. DeLoach, and D. J. Nisbet. 1997. Comparison of drag-swab environmental protocols for the isolation of Salmonella in poultry houses. Avian Dis. 41:709–713.[CrossRef][Web of Science][Medline]

Coon, C. N., I. Obi, and M. L. Hamre. 1988. Use of barley in laying hen diets. Poult. Sci. 67:1306–1313.[Web of Science]

Corrier, D. E., D. J. Nisbet, C. M. Scanlan, A. G. Hollister, and J. R. DeLoach. 1995. Control of Salmonella typhimurium colonization in broiler chicks with a continuous-flow characterized mixed culture of cecal bacteria. Poult. Sci. 74:916–924.[Web of Science][Medline]

De Oliveira, G. H., A. Berchieri Jr., and P. A. Barrow. 2000. Prevention of Salmonella infection by contact using intestinal flora of adult birds and/or a mixture of organic acids. Braz. J. Microbiol. 31:116–120.

Engberg, R. M., M. S. Hedemann, and B. B. Jensen. 2002. The influence of grinding and pelleting of feed on the microbial composition and activity in the digestive tract of broiler chickens. Br. Poult. Sci. 44:569–579.

Engberg, R. M., M. S. Hedemann, S. Steenfeldt, and B. B. Jensen. 2004. Influence of whole wheat and xylanase on broiler performance and microbial composition and activity in the digestive tract. Poult. Sci. 83:925–938.[Abstract/Free Full Text]

Flores, M. P., J. I. R. Castanon, and J. M. McNab. 1994. Effects of enzyme supplementation of wheat and triticale based diets for broilers. Anim. Feed Sci. Technol. 49:237–243.[CrossRef]

Grant, R. J., and V. F. Colenbrander. 1990. Milk fat depression in dairy cows: Role of silage particle size. J. Dairy Sci. 73:1834–1842.[Abstract]

Haller, D., C. Bode, and W. P. Hammes. 1999. Cytokine secretion by stimulated monocytes depends on the growth phase and heat treatment of bacteria: A comparative study between lactic acid bacteria and invasive pathogens. Microbiol. Immunol. 43:925–935.[Web of Science][Medline]

Havenstein, G. B., J. L. Grimes, P. R. Ferket, C. R. Parkhurst, F. W. Edens, J. Brake, and J. H. van Middelkoop. 1998. Recent experiences with reduced or non-litter systems for growing broilers and turkeys. Pages 225–240 in Proc. Natl. Poult. Waste Manage. Symp., Springdale, AR. Auburn Univ. Print. Serv., Auburn Univ., AL.

Hetland, H., and B. Svihus. 2001. Effect of oat hulls on performance, gut capacity and feed passage time in broiler chickens. Br. Poult. Sci. 42:354–361.[CrossRef][Web of Science][Medline]

Hoszowski, A., A. D. E. Fraser, B. W. Brooks, and E. M. Riche. 1996. Rapid detection and enumeration of Salmonella in chicken carcass rinses using filtration, enrichment and colony blot immunoassay. Int. J. Food Microbiol. 28:341–350.[CrossRef][Web of Science][Medline]

Huang, D. S., D. F. Li, J. J. Xing, Y. X. Ma, Z. J. Li, and S. Q. Lv. 2006. Effects of feed particle size and feed form on survival of Salmonella typhimurium in the alimentary tract and cecal S. typhimurium reduction in growing broilers. Poult. Sci. 85:831–836.[Abstract/Free Full Text]

Humphrey, B. D., N. Huang, and K. C. Klasing. 2002. Rice expressing lactoferrin and lysozyme has antibiotic-like properties when fed to chicks. J. Nutr. 132:1214–1218.[Abstract/Free Full Text]

Iji, P. A., A. A. Saki, and D. R. Tivey. 2001. Intestinal development and body growth of broiler chicks on diets supplemented with non-starch polysaccharides. Anim. Feed Sci. Technol. 89:175–188.[CrossRef]

Jimenez-Moreno, E., J. M. Gonzales-Alvarado, A. Coca, R. Lazaro, and G. G. Mateos. 2006. Influence of cereal type, heat processing of the cereal, and inclusion of fiber in the diet on organ weights and ileal digestibility of nutrients in broilers. Poult. Sci. 85(Suppl. 1):132. (Abstr.)

Jimenez-Moreno, E., J. M. Gonzales-Alvarado, D. G. Valencia, R. Lazaro, and G. G. Mateos. 2004. Influence of cereal, heat processing of the cereal, and inclusion of fiber in the diet on performance of broilers. Poult. Sci. 83(Suppl. 1):323. (Abstr.)

Kelley, T. R., O. C. Pancorbo, W. C. Merka, S. A. Thompson, M. L. Cabrera, and H. M. Barnhart. 1994. Fate of selected bacterial pathogens and indicators in fractionated poultry litter during storage. J. Appl. Poult. Res. 3:279–288.[Abstract/Free Full Text]

Lan, Y. 2004. Gastrointestinal health benefits of soy water-soluble carbohydrates in young broiler chickens. PhD Thesis. Wageningen Univ., Wageningen, the Netherlands.

Lázaro, R., M. García, M. J. Araníbar, and G. G. Mateos. 2003. Effect of enzyme addition to wheat-, barley-, and rye-based diets on nutrient digestibility and performance of laying hens. Br. Poult. Sci. 44:256–265.[CrossRef][Web of Science][Medline]

Lentle, R. G., V. Ravindran, G. Ravindran, and D. Thomas. 2006. Influence of feed particle size on the efficiency of broiler chickens fed wheat-based diets. J. Poult. Sci. 43:135–142.[CrossRef]

Link-Amster, H., F. Rochat, K. Y. Saudan, O. Mignot, and J. M. Aeschlimann. 1994. Modulation of a specific humoral immune response and changes in intestinal flora mediated through fermented milk intakes. FEMS Immunol. Med. Microbiol. 10:55–64.[CrossRef][Web of Science][Medline]

Marlett, J. A., M. I. McBurney, and J. L. Slavin. 2002. Position of the American Dietetic Association: Health implications of dietary fiber. J. Am. Diet. Assoc. 102:993–1000.[CrossRef][Web of Science][Medline]

Mead, P. S., L. Slutsker, V. Dietz, L. F. McCaig, J. S. Bresee, C. Shapiro, P. M. Griffin, and R. V. Tauxe. 1999. Food-related illness and death in the United States. Emerg. Infec. Dis. 5:607–625.[Web of Science][Medline]

NRC. 1994. Nutrient Requirements of Poultry. 9th rev. ed. Natl. Acad. Press, Washington, DC.

NRC. 1995. Nutrient Requirements of Laboratory Animals. 4th rev. ed. Natl. Acad. Press, Washington, DC.

Nurmi, E., and M. Rantala. 1973. New aspects of Salmonella infection in broiler production. Nature 241:210–211.[CrossRef][Medline]

Oosterom, J. 1991. Epidemiological studies and proposed preventive measures in the fight against human salmonellosis. Int. J. Food Microbiol. 12:41–52.[CrossRef][Web of Science][Medline]

Plavnik, I., B. Macovsky, and D. Sklan. 2002. Effect of feeding whole wheat on performance of broiler chickens. Anim. Feed Sci. Technol. 96:229–236.[CrossRef]

Porter, R. E., Jr. 1998. Bacterial enteritides of poultry. Poult. Sci. 77:1159–1165.[Abstract/Free Full Text]

Salgado, P., J. P. B. Freire, M. Mourato, F. Cabral, R. Toullec, and J. P. Lalles. 2002. Comparative effects of different legume protein sources in weaned piglets: Nutrient digestibility, intestinal morphology and digestive enzymes. Livest. Prod. Sci. 74:191–202.[CrossRef]

Samanya, M., and K. Yamauchi. 2002. Histological alterations of intestinal villi in chickens fed dried Bacillus subtilis var. natto. Comp. Biochem. Physiol. A 133:95–104.[CrossRef][Medline]

Santos, F. B. O., A. A. Santos Jr., P. R. Ferket, and B. W. Sheldon. 2006. Influence of grain particle size and insoluble fiber content on Salmonella colonization and shedding of turkeys fed corn-soybean meal diets. Int. J. Poult. Sci. 5:731–739.

Santos, F. B. O., X. Li, J. B. Payne, and B. W. Sheldon. 2005. Estimation of most probable number Salmonella populations on commercial North Carolina turkey farms. J. Appl. Poult. Res. 14:700–708.[Abstract/Free Full Text]

Santos, A. A., Jr. 2006. Poultry intestinal health through diet formulation and exogenous enzyme supplementation. PhD Thesis. North Carolina State Univ., Raleigh.

SAS Institute. 1996. SAS/STAT User’s Guide, Version 6.12. 4th ed. SAS Inst. Inc., Cary, NC.

Savory, C. J. 1992. Enzyme supplementation, degradation and metabolism of three U-¹⁴C-labelled cell-wall substrates in the fowl. Br. J. Nutr. 67:91–102.[CrossRef][Web of Science][Medline]

Silva, S. S. P., and R. R. Smithard. 2002. Effect of enzyme supplementation of a rye-based diet on xylanase activity in the small intestine of broilers, on intestinal crypt cell proliferation and on nutrient digestibility and growth performance of the birds. Br. Poult. Sci. 43:274–282.[CrossRef][Web of Science][Medline]

Sonnenburg, J. L., L. T. Angenent, and J. Gordon. 2004. Getting a grip on things: How do communities of bacterial symbionts become established in our intestine? Nat. Immunol. 5:569–573.[CrossRef][Web of Science][Medline]

Steenfeldt, S., M. Hammershoj, A. Mullertz, and F. Jensen. 1998. Enzyme supplementation of wheat-based diets for broilers. 2. Effect on apparent metabolisable energy content and nutrient digestibility. Anim. Feed Sci. Technol. 75:45–64.[CrossRef]

Stewart-Tull, D. E. S. 1980. The immunological activities of bacterial peptidoglycans. Annu. Rev. Microbiol. 34:311–340.[CrossRef][Web of Science][Medline]

Svihus, B., and H. Hetland. 2001. Ileal starch digestibility in growing broiler chickens on a wheat-based diet improved by mash feeding, dilution with cellulose or whole wheat inclusion. Br. Poult. Sci. 42:633–637.[CrossRef][Web of Science][Medline]

Svihus, B., E. Juvik, H. Hetland, and Å. Krogdahl. 2004. Causes for improvement in nutritive value of broiler chicken diets with whole wheat instead of ground wheat. Br. Poult. Sci. 45:55–60.[Web of Science][Medline]

Takahashi, T., T. Oka, H. Iwana, T. Kuwata, and Y. Yamamoto. 1993. Immune response of mice to orally administered lactic acid bacteria. Biosci. Biotechnol. Biochem. 57:1557–1560.

Wagner, D. D., and O. P. Thomas. 1987. Influence of diets containing rye or pectin on the intestinal flora of chicks. Poult. Sci. 57:971–975.

This article has been cited by other articles:

				E. Teirlynck, F. Haesebrouck, F. Pasmans, J. Dewulf, R. Ducatelle, and F. Van Immerseel The cereal type in feed influences Salmonella Enteritidis colonization in broilers Poult. Sci., October 1, 2009; 88(10): 2108 - 2112. [Abstract] [Full Text] [PDF]

				M. Frikha, H. M. Safaa, M. P. Serrano, X. Arbe, and G. G. Mateos Influence of the main cereal and feed form of the diet on performance and digestive tract traits of brown-egg laying pullets Poult. Sci., May 1, 2009; 88(5): 994 - 1002. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Santos, F. B. O. Articles by Ferket, P. R. Search for Related Content PubMed PubMed Citation Articles by Santos, F. B. O. Articles by Ferket, P. R.