Effects of wild blueberry (Vaccinium angustifolium) pomace feeding on gut microbiota and blood metabolites in free-range pastured broiler chickens Md. Rashedul Islam,∗ Dion Lepp,∗ David V. Godfrey,† Steve Orban,† Kelly Ross,† Pascal Delaquis,† and Moussa S. Diarra∗,1 ∗Guelph Research and Development Centre, Agriculture and Agri-Food Canada (AAFC), Guelph, ON N1G 5C9, Canada; and †Summerland Research and Development Centre, AAFC, Summerland, BC V0H 1Z0, Canada ABSTRACT There is a need to develop cost- effective approaches to modulate gut microbiota, pro- mote bird health, and prevent infections in pasture- raised broiler chickens. The present study evaluated the efficacy of organic wild blueberry (Vaccinium angusti- folium) also called low-bush blueberry pomace (LBBP)- supplemented feed to modulate the chicken gut micro- biota, and blood metabolites in order to improve bird health and productivity. Slow-growing broiler chickens were reared on pasture up to 64 D for sampling af- ter 2 wk of treatment during brooding with 0, 1, and 2% LBBP in feed. Intestinal samples were collected at different time-points throughout the trial for bac- terial culture and microbial community analysis by 16S rRNA gene sequencing using Illumina MiSeq. Blood sera were also analyzed for metabolites at each sam- pling time. Of the 14 bacterial phyla, the predominant taxa across all sampling time-points were Firmicutes, Proteobacteria, Bacteroidetes, and Tenericutes, repre- senting >97% of all sequences. Bacteroidetes seemed to be replacing Firmicutes by LBBP supplementation, with the most noticeable effect at day 64 with 1% LBBP. LBBP inclusion enriched Lactobacillus, Bac- teroides, and Bifidobacterium, while Escherichia coli, Clostridium Clostridiaceae, Helicobacter, and Entero- coccus showed higher abundances in control birds at the end of trial. Principal co-ordinate analysis showed a clear clustering of the intestinal samples from con- trol and LBBP-treated groups at day 29. Application of LBBP resulted in a decrease (P < 0.05) in cholesterol at day 21, and an increase (P < 0.05) in high-density lipoprotein cholesterol in 14-day-old broilers. Higher (P < 0.05) levels of phosphorus, magnesium, and glob- ulin at day 21 as well as iron and albumin at day 36 were also observed in 1% LBBP-fed birds. Despite lim- itations consisting essentially of low sampled birds for measurements, this study indicated that dietary sup- plementation of LBBP could positively influence gut microbiota and blood metabolites that may contribute to the overall health of pasture-raised broiler chickens. Key words: blueberry pomace, pastured broiler, gut microbiota, 16S rRNA gene, blood metabolites 2019 Poultry Science 98:3739–3755 http://dx.doi.org/10.3382/ps/pez062 INTRODUCTION Organic poultry production standards require free- range access to improve the overall welfare of birds. Pastures can provide 5 to 20% of feed intake of birds and thus become a good source of nutrients and other bioac- tive compounds for broilers raised in free-range systems (https://attra.ncat.org/attra-pub/download.php?id= 452). Although there is an increasing interest in free-range pastured poultry meat production, control of pathogens and the food safety associated with this type of production remain challenging. In conventional broiler production, antimicrobial agents such as baci- tracin and the ionophore salinomycin are used in feed to control necrotic enteritis and coccidiosis caused by C© Her Majesty the Queen in the Right of Canada, as represented by the minister of Agriculture and Agri-Food Canada. Received April 24, 2018. Accepted March 18, 2019. 1Corresponding author: Moussa.Diarra@canada.ca Clostridium perfringens and Eimeria sp., respectively, resulting in performance enhancement by improving feed conversion and body weight gain (Knarreborg et al., 2002; Collier et al., 2003; Singer and Hofacre, 2006). Efficient and cost-effective methods for main- taining gut health, reducing public health risks and lessening negative environmental impacts (for example bioaerosols, litters containing antibiotic resistant bacteria, and antibiotic residues) are urgently needed for both conventional and organic broiler productions (Jacob and Pescatore, 2012; Merchant et al., 2012; Vela et al., 2012). In Canada, approximately 27,649 hectares of land are devoted to the cultivation of wild blueberry (Vaccinium angustifolium) also called low-bush blueberries (LBB), primarily in Quebec and the Atlantic Provinces (AAFC, 2012). The majority of the blueberry harvest is frozen or processed into jams, syrups, purees, or juices (AAFC, 2011). We recently showed that organic low-bush blueberry pomace (LBBP), a major 3739 https://attra.ncat.org/attra-pub/download.php?id=452 https://attra.ncat.org/attra-pub/download.php?id=452 mailto:Moussa.Diarra@canada.ca 3740 ISLAM ET AL. by-product of the juice industry, contains carbohy- drates, proteins, lipids, minerals, and several phenolic compounds including flavonoids such as anthocyanins and flavonols (Ross et al., 2017). Therefore, pomace could be an important resource for value-added applications in chicken production. Many of the phenolic components in blueberry have been reported to exhibit biological effects ranging from antioxidant, anti-inflammatory, vasodilatory, and antimicrobial activities (Herken and Guzel, 2010; Biswas et al., 2012; Shen et al., 2014). Despite all these potential benefits, there have been no in-depth studies documenting the effects of blueberry pomace on poultry health and performance. Gut microbiota plays an important role in maintain- ing overall animal health as it is implicated in the de- velopment of the immune system, energy homeostasis, and protection against pathogens (Pan and Yu, 2014; Kers et al., 2018). The broiler ileum microbial compo- sition has also been shown to influence intestinal func- tion, digestion, and nutrient absorption (Gong et al., 2002). The chicken gut microbiota is thought to contain more than 1,000 bacterial species, with approximately 90% of the species being unknown, which illustrates the scarcity of knowledge about the chicken intestinal mi- crobiota (Bjerrum et al., 2006; Chambers and Gong, 2011; Oakley et al., 2014). Moreover, little is known about the influence of pastures and other feeding prac- tices on the gut microbiota of birds during free-range rearing. The potential of berry products to modulate gut microbiota, health, disease, and immunity in food ani- mals has been the subject of recent investigations (Das et al., 2017; Islam et al., 2017). In the present study, 16S rRNA gene sequencing was used to examine the ileum bacterial community composition and diversity in free-range pastured broiler chickens fed with or without LBBP-supplemented diets. The impacts of LBBP on blood metabolites were also evaluated at various ages. Additionally, associations between gut microbiota and blood parameters were estimated to investigate the biological responses of the chickens upon feeding with LBBP. MATERIALS AND METHODS Product Use: Pomace Preparation and Composition Frozen organic LBB (V. angustifolium Ait.) were purchased from Fruit d’Or (Villeroy, QC, Canada). Pomace was prepared according to Ross et al. (2017). Briefly, fruits were thawed and juice was removed by using a hydraulic rack and cloth. The pomace was then freeze dried and ground through 2 mm mesh screen using a Cutting Mill (SM 2000 Retsch; Haan, Germany). The dried pomace was stored at -20◦C until used. As described by Ross et al. (2017), the LBBP used in this study contained 5.43% lipids, 8.41% proteins, 84.91% carbohydrates, 31.13 mg gallic acid eq/g of total phenolics, 5.35 mg caffeic acid eq/g of tartaric esters, 6.17 mg quercetin eq/g of flavonols, and 17.41 mg cyaniding-3-glucoside eq/g of anthocyanins. Birds, Diets and Management A total of 500 day-old Cobb 308 broiler chicks were purchased from a local commercial hatchery (Okana- gan Valley, BC, Canada) and placed in a commercial farm (Rosebank Farms, Armstrong, BC, Canada) from April 21 to June 23 (slaughter day). The birds were randomly divided into 3 groups: a control group fed an all-vegetarian feed (basal diet); a group receiving the basal diet supplemented with 1% LBBP; and another group receiving the basal diet supplemented with 2% LBBP. Each bird group was assigned to a separate pen (100 birds/pen) in the same rearing conditions during brooding and on pasture. Water and feed were offered ad libitum throughout the experiment. Feeds formulated in accordance with requirements for organic broiler diet were mixed at the commercial farm. The composition of the basal and supplemented diets during the starter and grower (Table 1) included wheat and barley as the principal cereals, and pea and soybean as protein concentrates to meet the National Research Council nutrient requirements for broiler chickens (NRC, 1994). Analysis of dry matter (DM), total proteins, soluble carbohydrates, fatty acids, and some of the most common minerals were performed by the Cumberland Valley Analytical Services, Inc. (Laboratory Services for Agriculture, BC, Canada). Treatments (LBBP-supplemented feed) were applied at day 7 (April 27) until 21 D of age (May 11) in the brooder pens (Figure 1), because a decrease of early morality rate in birds fed cranberry product (Leusink et al., 2010) was previously reported. Birds were reared slowly to accommodate their physiology for a final target dressed carcass weight of 5.40 to 5.70 lbs (2.45 to 2.59 kg) for the fall rearing period without undue leg stress or heart attacks. From day 1 to 21, birds from all treatment groups were held in separate pens in the same brooder house (same rearing conditions). The birds were then moved to free-range outdoor pasture and the treatment groups were contained in their own separate rectangular areas of 160 sq. ft. each (1.6 sq. ft./bird) on the same farm until they reached 64 D of age (2015 April 21 to 2015 June 23). On the pasture, each group had its own shelters and range area pro- tected by electrified poultry netting. Each group was managed the same way and on same type and quality pasture that was a dryland forage mix under seeded with Dutch white clover at 10%. All experimental procedures were conducted at the Rosebank Farms in conformity with the general practices for commercial free-range broiler chicken production described in the guidelines of the Canadian Council on Animal Care (http://www.ccac.ca/en/standards/guidelines). The protocol was approved by Rosebank Farms, http://www.ccac.ca/en/standards/guidelines WILD BLUEBERRY POMACE IN BROILER CHICKENS 3741 Table 1. Composition of the feeds used in the study during the starter and grower period. Percentage (%) of inclusion in basal and LBBP1 supplemented diets Ingredient and nutrient profile Starter (days 0 to 14) Grower (days 14 to 36) Control 1% LBBP 2% LBBP Control 1% LBBP 2% LBBP Ingredient Wheat 39.15 39.15 39.15 56.11 56.11 56.11 Organic peas 32.81 32.81 32.81 21.64 21.64 21.64 Barley and peas 10.67 10.67 10.67 14.07 14.07 14.07 Vitamin premix blend (Macro)2 17.37 17.37 17.37 8.18 8.18 8.18 Analysed nutrient3 DM 88.40 88.70 88.60 87.90 88.20 88.20 Moisture 11.60 11.30 11.40 12.10 11.80 11.80 Protein Crude protein 24.70 25.40 23.90 18.90 19.20 19.00 Adjusted protein 24.70 25.40 23.90 18.90 19.20 19.00 Soluble protein 7.60 8.80 8.30 9.90 10.50 10.10 Ammonia 0.37 0.85 0.62 0.14 0.10 0.09 Fiber Acid detergent fiber (ADF) 6.30 7.30 6.40 4.70 5.50 4.90 Neutral detergent fiber (NDF) 14.70 16.20 13.30 13.50 13.50 14.70 Carbohydrate Silage acids 0.30 0.30 0.30 0.10 0.10 0.10 Ethanol soluble aldehyde (CHO) 12.60 3.40 3.30 2.70 3.40 4.10 Starch 39.20 41.60 43.60 52.40 48.30 47.00 Crude fat 2.70 2.76 3.05 1.95 2.09 2.23 Mineral Ash 5.94 6.22 6.14 3.63 4.80 5.48 Calcium 0.88 0.93 0.76 0.46 0.79 0.79 Phosphorus 0.83 0.83 0.71 0.58 0.68 0.69 Magnesium 0.27 0.26 0.24 0.21 0.23 0.22 Potassium 1.03 0.99 0.89 0.61 0.67 0.64 Sodium 0.28 0.32 0.25 0.11 0.21 0.20 Iron (ppm) 320.00 316.00 270.00 195.00 281.00 279.00 Manganese (ppm) 134.00 151.00 124.00 69.00 90.00 84.00 Zinc (ppm) 166.00 159.00 119.00 62.00 101.00 85.00 Copper (ppm) 24.00 30.00 19.00 10.00 16.00 15.00 Fermentation pH 5.72 5.68 5.79 6.05 5.93 5.86 Total volatile fatty acid (TVFA) 0.34 0.33 0.27 0.14 0.14 0.13 Lactic acid 0.30 0.30 0.20 0.10 0.10 0.10 Lactic acid as % of total VFA 100.00 103.00 77.00 111.00 100.00 111.00 Acetic acid 0.04 0.03 0.07 0.04 0.04 0.03 Titratable acidity (meq/100 g) 0.67 0.79 0.73 0.25 0.37 0.34 Calculated energy and index Total digestible nutrient (TDN) 77.30 76.70 78.30 79.60 78.30 77.10 Net energy lactation (mcal/lb) 0.81 0.80 0.82 0.83 0.82 0.81 Net energy maintenance (mcal/lb) 0.84 0.83 0.86 0.88 0.86 0.84 Net energy gain (mcal/lb) 0.56 0.55 0.57 0.58 0.57 0.55 Non-fiber carbohydrates 51.90 49.40 53.60 62.10 60.30 58.20 Non-structural carbohydrates 51.80 45.00 46.90 55.10 51.70 51.1 1Organic LBBP was added to the basal diet at a rate of 1% and 2%. 2Vitamin premix blend for starter and grower diets contain 11.47% Soy, 4.0% Wetaskiwin-chick-macro, 0.9% Biolys, 0.85% Limestone, 0.15% Methionine, and 4.0% Wetaskiwin-chick-macro, 3.84% Distillers wheat, 0.34% Lysine, respectively. 3The nutrient contents were analyzed on a dry matter (DM) basis. associate members of the North Okanagan Organic As- sociation, which use exclusively a “Customer Certified” system (inspected 2 to 3 times through the rearing period and the night before slaughter by customers). Performance and Mortality Grazing chickens consume 5 to 20% of their in- take from the pasture (https://attra.ncat.org/attra- pub/download.php?id=452). Given the type of the management and pasture used in this study, the con- sumption rate was estimated at 10 to 12%. Feed weights were recorded throughout the trial to estimate feed uptake; carcass weights of all remaining birds were measured at processing to estimate the apparent feed conversion ratio (FCR, indication of benefit to the farmer). The FCR was calculated from the total con- sumed (served) feed (kg) divided by the dressed carcass weight (kg) at slaughter (over 60 birds/replicate) at day 64. The entire gut and its contents removed by eviscer- ation from individual birds at slaughter were weighed and the gut weight index (GWI) was determined by di- viding gut weight (g) by the corresponding bird’s body weight (kg). The birds were monitored 3 times a day (morning, afternoon, and evening) for general health, and all mortalities were recorded. https://attra.ncat.org/attra-pub/download.php?id=452 https://attra.ncat.org/attra-pub/download.php?id=452 3742 ISLAM ET AL. Figure 1. Study time-line of broiler chicken trial. I = ileal samples collection for bacteriological analysis by culture-dependent and culture- independent approaches; B = blood collection for blood serum metabolites analysis. Culture-dependent Bacteriological Analysis The digesta samples from the ilea were aseptically collected (1 bird/pen, 2 pens/treatment) for bacteri- ological analysis after 2 (April 22, before application of the treatment), 14, 21, 29, 36, and 42 D. At the end of the trial at day 64 (June 23), 2 birds/pen (four birds/treatment) were sampled for analysis. Briefly, the section of the ilea extending from the vitelline diver- ticulum to a point 10 mm proximal to the ileo–cecal junction was removed from the gut in a Biosafety cab- inet. The ileal contents were expelled into a sterile test tube by squeezing and were mixed prior to analysis. Total aerobic Gram-positive bacteria were quantified on Tryptic Soy Agar (BD Bioscience) supplemented with benzyl carbinol (2.50 g/L) at 35◦C for 24 h. Total aero- bic Gram-negative bacterial populations were enumer- ated using MacConkey agar plates incubated at 35◦C for 24 h. Lactobacillus populations were counted on Lactobacilli MRS Agar (Oxoid, Nepean, ON, Canada) incubated at 37◦C for 48 h under anaerobic con- ditions. Enterococcus and C. perfringens populations were quantified using KF Streptococcal agar CM0701 (Oxoid) incubated at 37◦C for 48 h (Hayes et al., 2003), and tryptose sulfite agar (Oxoid) supplemented with cycloserine (SR088E, Oxoid) incubated anaerobically for 24 h at 35◦C (Knarreborg et al., 2002), respectively. The number of Escherichia coli was estimated using E. coli and coliform Petrifilm (3 M, St. Paul, MN, USA) as described elsewhere (Diarrassouba et al., 2007). Analysis of Bacterial Community Structure Genomic DNA was extracted from ileal digesta col- lected as described above at days 2, 14, 21, 29, 36, 42, and 64, using the PowerSoil DNA Isola- tion Kit (MO BIO Laboratory Inc.) according to manufacturer instructions. Eluted DNA was stored at -20◦C until further analysis. DNA yield and purity were determined using PicoGreen (Thermo Fisher Scientific, USA). Sequencing libraries of the 16S V3–4 region were prepared according to the Illumina 16S Metagenomic Sequencing Library Prepa- ration Guide Rev. B. Briefly, primers Bakt˙341F (5′-CCTACGGGNGGCWGCAG-3′) and Bakt 805R (5′-GACTACHVGGGTATCTAATCC-3′) (Herlemann et al., 2011) containing 5′ Illumina overhang adapter se- quences (5′ CGTCGGCAGCGTCAGATGTGTATAA- GAGACAG and 5′ GTCTCGTGGGCTCGGAGAT- GTGTATAAGAGACAG, respectively) were used to amplify an ∼550-base-pair (bp) fragment (including linkers and barcode sequences) of the 16S rRNA V3–4 region. Each reaction contained 12.5 ng of template DNA, 200 nM each primer, and 1 × KAPA HiFi HotStart ReadyMix (VWR, Mississauga, ON, Canada) in a 25-μl volume. Polymerase chain reaction (PCR) products were purified with AmpureXP beads (Beck- man Coulter, Mississauga, ON, Canada) and sequenc- ing adapters containing 8-bp indices were added to the 3′ and 5′ ends by PCR using the Nextera XT Index kit (Illumina, San Diego, CA, USA) and purified am- plicons were pooled in equimolar ratios, for sequencing on an MiSeq instrument using the MiSeq 600-cycle v3 kit (Illumina). The fastq files containing 300-bp dual-indexed paired-end reads were analyzed using QIIME version 1.8.0 (Caporaso et al., 2010a). Paired-end reads were joined with fastq-join (Aronesty, 2011), and quality filtered and demultiplexed in QIIME using default settings. The reads were clustered at 97% sequencing identity (similar to the species level) using uclust (Edgar, 2010) and operational taxonomy units (OTUs) were picked against the Greengenes database (gg otus 13 8) using an open-reference approach WILD BLUEBERRY POMACE IN BROILER CHICKENS 3743 (DeSantis et al., 2006). Taxonomic assignment of the sequences was performed using the uclust consensus taxonomy assigner. Taxa that could not be assigned a genus were presented as “unclassified” using the highest taxonomic level that could be assigned to them. The sequences were aligned against the Greengenes core set with PyNast (Caporaso et al 2010b) and a phylogenetic tree was constructed with FastTree (Price et al., 2009). Alpha-diversity (within group) metrics were then calculated by QIIME and a β-diversity (between group) distance matrix based on unweighted UniFrac metric (Lozupone and Knight, 2005) was calcu- lated, which was used for principal co-ordinate analysis (PCoA). All 16S rRNA gene sequencing data derived from this project were deposited in National Center for Biotechnology Information (NCBI’s) Sequence Read Archive (http://www.ncbi.nlm.nih.gov/sra/) under accession number SRP100528. Analysis of Blood Metabolites Blood samples were collected from the same birds sampled for bacteriology and 16S rRNA sequencing. The samples were allowed to clot at room temperature before centrifugation at 2000 × g for 10 min for serum collection. Collected sera were transferred to sterile Eppendorf tubes and stored at -80◦C until further analysis. Blood serum samples were assessed for biochemistry parameters including cholesterol (CHO), high-density lipoprotein cholesterol (HDLC), triglyceride (TG), minerals (calcium, Ca; iron, Fe; magnesium, Mg; phosphorus, P), and proteins (total proteins, TP; albumin, ALB; globulin, GLO; ALB- GLO ratio, AGR) at the Animal Health Laboratory Services (University of Guelph, Guelph, ON, Canada). Statistical Analysis Statistical analyses of data were conducted according to a randomized block design using the General Linear Model procedure of the Statistical Analysis System (SAS, 2000) with treatment groups as sources of variation and the individual pens as experimental units. Data from bacterial counts were log-transformed before analysis along with the blood parameters by repeated measurement option using treatment groups and sampling times (bird ages = source of variations) as factors. Means were estimated, and the least significance difference was used to separate treatments means whenever the F-value was significant. Correla- tions between microbiota and blood metabolites were estimated using QIIME version 1.9.1 (Caporaso et al., 2010a). Spearman correlation method was used to test the degree of association. The P-values were calculated by Fisher’s Z transform and these values were corrected by the Benjamini–Hochberg FDR method (Benjamini and Hochberg, 1995). The P-value of 0.05 was used to declare significance. RESULTS Feed Composition The composition of the both starter and grower diets is presented in Table 1. The DM contents were slightly higher in the LBBP-supplemented feed (both 1 and 2%) than in the control feed. The concentration of soluble protein, acid detergent fiber (ADF), crude fat, ash, and titratable acidity was generally higher, while the con- tent of non-structural carbohydrates was lower in the 2 LBBP-supplemented feed compared to the control one. The contents of minerals such as potassium (K), zinc (Zn), Mg, and Fe were lower in the starter, but higher in the grower diet for the LBBP-supplemented feeds com- pared to the control. Similarly, the amount of ammonia and starch was higher in the starter diet; however, both were lower in the grower diet. In the starter diet, the concentration of ethanol-soluble aldehyde (12.60% DM) was substantially higher in the control than 1% (3.40% DM) and 2% (3.30% DM) LBBP-supplemented feeds. Comparison of feed supplemented with 1 and 2% LBBP showed that the levels of proteins (crude, ad- justed, or soluble), ADF, Mg, K, Fe, Zn, sodium (Na), manganese (Mn), copper (Cu), and titratable acidity were higher in 1% LBBP, whereas the concentration of crude fat was higher in 2% LBBP-supplemented feed, in both the starter and grower diets. Performance and Mortality The effects of diets with or without LBBP sup- plementation on consumed feed, carcass weight, and FCR (a standard estimate by Rosebank Farms to have an indication of benefit) at slaughter, as well as GWI and mortality were measured. The highest (P > 0.05) GWI was obtained with the 2% LBBP treatment (181.41 g/kg) compared to 1% LBBP (145.53 g/kg) and control (150.44 g/kg) diets. At pro- cessing, the average carcass weight (kg) per bird was 2.38, 2.32, and 2.29 in the birds fed 1% LBBP, 2% LBBP, and control diets, respectively (P = 0.11). The FCR at slaughter (day 64) was lower (P < 0.01) in 2% LBBP-fed birds (4.19) than in the control (5.58) and 1% LBBP-treated birds (5.15). It should be noted that the apparent FCR used in study is not the traditional feed efficiency that is calculated by dividing feed in- take by the bodyweight gain. However, it is well known that feed efficiency (or FCR) is better in the younger broilers, than at target market weight. In general all birds were healthy, but mortality rates were lower in 1 and 2% LBBP treatment groups than the control group, from the beginning of the trial un- til 28 D of age. The mortality rate in control birds was higher (P < 0.05) from day 21 to 28 than that in LBBP- treated birds (Figure 2). Similarly, the lowest (P > 0.05) cumulative mortalities were observed in LBBP- treated birds (7.39 and 10.0% for 1 and 2% LBBP feed supplementation, respectively) when compared to http://www.ncbi.nlm.nih.gov/sra/ 3744 ISLAM ET AL. Sampling day D1-13 D14-20 D21-28 D29-35 D36-41 D42-64 D1-64 M or ta lit y (% ) 0 2 4 6 8 10 12 14 16 Control 1% LBBP 2% LBBP * * Figure 2. Effects of organic low-bush blueberry pomace (LBBP) on mortality rates of broiler chickens. Results are expressed as Mean ± SEM (Standard error of the mean). Asterisks indicate the time/rearing period at which the mortality rates were statistically different (P < 0.05). the untreated control birds (11.70%). Nevertheless, our results suggest that LBBP-supplemented diets could decrease mortality rates in younger free-range broiler chickens. Culture-dependent Bacterial Enumeration Populations of total aerobic Gram-positive and Gram-negative, as well as Lactobacillus, Enterococcus, C. perfringens, and E. coli were determined in broiler chicken ileal samples at 2, 14, 21, 29, 36, 42, and 64 D of age (Figure 3). Diet supplementation with LBBP resulted in an increase (P < 0.05) in aero- bic Gram-positive bacterial populations in the LBBP- treated groups at day 29 (1 and 2% LBBP) and day 42 (2% LBBP). However, their population size decreased (P < 0.05) in LBBP-treated birds compared to con- trol birds at the end of trial (day 64) (Figure 3A). Lactobacillus numbers were higher in LBBP treat- ments than the control diet from days 21 to 42 (Fig- ure 3C). In parallel with Gram-positive bacteria, Lac- tobacillus populations in LBBP treatment groups were lower (P < 0.05) than in the control at day 64. At days 21 and 64, Enterococcus populations were higher (P < 0.05) in control birds than those fed LBBP (Fig- ure 3D). No significant differences were noticed between the LBBP-treated and control groups for population sizes of total aerobic Gram-negative, and 2 specific bac- terial species (C. perfringens and E. coli) during the entire trial (Figures 3B, E, and F). As expected, total aerobic Gram-positive and Gram-negative bacteria as well as Lactobacillus populations were consistently more abundant (5.27 to 9.13 Log CFU/g) than Enterococcus, C. perfringens and E. coli (1.0 to 7.46 Log CFU/g) in birds from all ages irrespective of the treatment group. Culture-independent Gut Microbiota Structure The composition of the ileum bacterial community in chickens fed with or without LBBP was analyzed by 16S rRNA gene sequencing. A total of 4292,296 paired-end reads were generated. After joining the paired-end reads and quality control, 2578,958 sequences with a median length of 465 bp remained. The number of sequences per sample ranged from 20,969 to 163,756, with an av- erage of 66,127 sequences per sample. The sequences were clustered into 297 OTUs. Results revealed that the bacterial community was highly diverse, consisting of 14 known phyla detected in proportions that var- ied according to treatment group and sampling interval (Figure 4A). Treatment or sampling intervals notwith- standing, Firmicutes was always the highest abundance phylum, followed by Proteobacteria, Bacteroidetes, and Tenericutes that collectively accounted for up to 98% of all sequences. The remaining phyla included Defer- ribacteres, Actinobacteria, Cyanobacteria, Fusobacte- ria, Verrucomicrobia, Synergistetes, TM7, Elusimicro- bia, WPS-2, and Lentisphaerae. At day 14 (a week after initiation of LBBP feeding), the ileum bacterial commu- nity was mainly composed of Firmicutes, whereas this phylum together with Bacteroidetes was predominant at day 21 in control birds and those fed 1% LBBP. At day 29 (a week after cessation of LBBP feeding), birds fed LBBP showed an abundance of Proteobacte- ria population, whereas Tenericutes was recovered at a proportion of 21.05% (control), 8.98% (1% LBBP), WILD BLUEBERRY POMACE IN BROILER CHICKENS 3745 Figure 3. Effects of organic LBBP on bacterial population dynamics in intestinal samples of free-range pastured broiler chickens: (A) Gram- positive, (B) Gram-negative, (C) Lactobacillus, (D) Enterococcus, (E) Clostridium perfringens, and (F) Escherichia coli. Results are expressed as Mean ± SEM (Standard error of the mean). Asterisks indicate the sampling day at which the treatment effects were statistically different (∗P < 0.05; ∗∗P < 0.01). �, Clostridium perfringens for control and 1% LBBP at day 14 was not determined. and 12.85% (2% LBBP). In contrast, the relative abun- dance of Bacteroidetes decreased in the control birds from days 36 to 42. However, an increase in the abun- dance of Tenericutes and Proteobacteria was noticed in all bird groups during the same time interval, with the exception of Proteobacteria in birds fed 1% LBBP. At the end of the trial, Bacteroidetes (44.19%) appeared to replace other bacterial phyla in the gut of birds fed 1% LBBP, particularly Firmicutes (42.76%) that was the dominant phylum at all previous sampling intervals. 3746 ISLAM ET AL. Figure 4a. Taxonomic profiles of bacterial communities in the intestines of free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) Bacterial composition at the phylum level, (B) bacterial composition at the genus level (only taxa representing >1% of all sequences, together with the genera Clostridium and Bifidobacterium, are shown), and (C) important bacterial genera at the end of trial (day 64). Sampling time-points are represented along the horizontal axis, and relative abundance is shown by the vertical axis. Taxa that could not be assigned a genus were displayed “unclassified” using the highest taxonomic level that could be assigned to them. §Assigned as “unclassified Enterobacteriaceae” by Greengenes taxonomy database included in QIIME, however, the OTUs were later identified as Escherichia/Shigella by NCBI BLAST. At this time-point, Firmicutes:Bacteroidetes (F/B) ra- tio was 1.85, 0.97, and 1.76 in the control, 1 and 2% LBBP-fed broilers, respectively. Lactobacillus was the most prevalent (36.40%) bacterial genus detected by 16S rRNA gene se- quencing, irrespective of treatments and sampling intervals. Lactobacillus together with unclassified Helicobacteraceae, Candidatus Arthromitus (recently proposed as Candidatus Savagella, Thompson et al., 2012), unclassified Mycoplasmataceae, unclassified Clostridiales, Bacteroides, unclassified Clostridiaceae, Parabacteroides, Streptococcus, Ruminococcus, and unclassified Enterobacteriaceae accounted for >83% of all sequences (Figure 4B). Other prevalent genera included Faecalibacterium, Helicobacter, unclassified Ruminococcaceae, Enterococcus, Clostridium, and Bifidobacterium. The sequences (1.90%, 31 OTUs) belonging to unclassified Enterobacteriaceae were found WILD BLUEBERRY POMACE IN BROILER CHICKENS 3747 Figure 4b. Taxonomic profiles of bacterial communities in the intestines of free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) Bacterial composition at the phylum level, (B) bacterial composition at the genus level (only taxa representing >1% of all sequences, together with the genera Clostridium and Bifidobacterium, are shown), and (C) important bacterial genera at the end of trial (day 64). Sampling time-points are represented along the horizontal axis, and relative abundance is shown by the vertical axis. Taxa that could not be assigned a genus were displayed “unclassified” using the highest taxonomic level that could be assigned to them. §Assigned as “unclassified Enterobacteriaceae” by Greengenes taxonomy database included in QIIME, however, the OTUs were later identified as Escherichia/Shigella by NCBI BLAST. to be aligned with Escherichia/Shigella 16S rRNA se- quences with 99 to 100% identity by NCBI Basic Local Alignment Search Tool (BLAST) against the GenBank nr database. After the end of LBBP feeding at day 21, the highest proportion (86.79%) of Lactobacillus was recovered from birds fed 2% LBBP. Parabacteroides (28.82%) and Faecalibacterium (5.12%) were abundant at the same time-point, and the former appeared to replace Lactobacillus in birds fed 1% LBBP. After 29 D, the proportion of Lactobacillus in the gut of control birds was drastically decreased and Candidatus Savagella (58.14%) was dominant. Unlike control and 2% LBBP treatment, a proportion of unclassified Helicobacteraceae decreased in birds fed 1% LBBP between 36 and 42 D. At day 64, the populations of Lactobacillus were replaced by other bacterial genera, particularly Bacteroides in all bird groups, leaving 2.56, 6.19, and 3.41% of Lactobacillus in the control, 1 and 2% LBBP-fed birds, respectively (Figure 4C). A com- paratively higher proportion of Bacteroides (25.07 to 34.73%), unclassified Ruminococcaceae (2.50 to 2.55%), Faecalibacterium (1.05 to 1.51%), and Bifidobacterium (0.08 to 2.0%) was observed in LBBP-treated groups than in the control, where the proportions of these 4 genera were 21.86, 2.17, 0.59, and 0.04%, respectively. On the other hand, the relative proportions of Candidatus Savagella (1.86 to 5.80%), Es- cherichia/Shigella (1.45 to 2.01%), Enterococcus (0.46 to 0.81%), Streptococcus (0.07 to 0.08%), un- classified Helicobacteraceae (0.07 to 0.09%), and Clostridium˙Clostridiaceae (0.05 to 0.07%) were lower in LBBP-fed birds, whereas the percentages of these genera in the control group were 6.07, 6.83, 4.81, 2.07, 0.61, and 0.31%, respectively. These results suggest that feeding LBBP promotes the growth of beneficial bacteria and reduces bacterial groups potentially associated with diseases. The phylogenetic composition of gut microbiota was characterized by comparison of the α-diversity param- eters in the 3 bird groups. The average index values of the phylogenetic diversity (PD) whole tree and Shan- non diversity were higher in 29-day-old chickens fed LBBP (Figure 5), although these differences (P > 0.05) were not significant. Together with Chao1, Simpson, and observed species, the above-mentioned 2 param- eters were different (P < 0.01) at slaughter (day 64) when compared to all other sampling intervals. An un- weighted UniFrac PCoA based on phylogenetic rela- tionships was used to assess the β-diversity of chicken gut microbiota. Samples from days 14 to 21, 29 to 42, and 64 showed a clear separation, indicating the devel- opment and succession of microbiota composition and diversity with ageing (Figure 6A). While a progres- sive change in the bacterial communities was observed 3748 ISLAM ET AL. Figure 5. Alpha-diversity comparison of the microflora in the in- testines of free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) phylogenetic diversity (PD) whole tree and (B) Shannon diversity index. Sampling days are repre- sented along the horizontal axis, and α-diversity parameters are shown by the vertical axis. ∗Asterisk indicates the sampling time-point (day 64) at which the parameter was statistically different (P < 0.01) as compared with all other sampling time-points. The treatment effects were not statistically different (P > 0.05) at any sampling day. with broiler age independently to treatments applied, a distinct clustering between samples from control and LBBP groups was noted at day 29 (Figure 6B), with samples from 1% LBBP-fed broilers exhibiting variabil- ity at day 36 (data not shown). This result indicated that LBBP may have an impact on the composition of the bacterial community in the gut of broiler chickens. Blood Biochemistry Parameters The effects of LBBP feeding on blood serum choles- terols and other lipids of broiler chickens are shown in Figure 7. A decrease (P = 0.002) of serum CHO was observed in both 1 and 2% LBBP treatments at day 21 (Figure 7A). Feed supplementation with 1% LBBP increased (P < 0.05) the “good” cholesterol HDLC con- centration in chickens at days 14, 42, and 64 (Fig- ure 7B). Comparison of all the treatments revealed that the concentration of HDLC increased with 2% LBBP until day 29 but that this was the lowest concentration measured until the end of trial. The lowest TG level was observed in birds fed 1% LBBP a week after the initiation of pomace feeding until the end of treatment at days 14 and 21 of age (Figure 7C); however, from day 29, the TG level tended to increase in pomace-fed birds until the end of the trial. No significant differ- Figure 6. Unweighted UniFrac principal co-ordinate analysis by microbiota in the intestines of free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) Samples from days 14–21, days 29–42, and day 64 show a clear separation, suggesting the changes in bacterial diversity and composition with the birds’ age, and (B) Samples from day 29, where the treatment (1 and 2% LBBP) and control trend to cluster separately. Percent of data set variability explained by each principal component (PC1, PC2, PC3) is shown in brackets in the axis titles. Each group is represented by a different symbol and/or color (control, red; 1% LBBP, blue; 2% LBBP, green). ence was observed between the bird groups for serum Ca, although the concentration tended to be higher in LBBP groups than in the control (Figure 8A). Higher (P < 0.05) concentrations of Fe (d 36), Mg, and P (d 21) were observed in birds fed 1% LBBP diet (Fig- ures 8B, C, and D). Although LBBP-fed broilers gen- erally showed higher serum TP, ALB, and GLO levels than chickens fed the control diet at days 14, 21, 42, and 64, these effects were not significant (P > 0.05), except for GLO at day 21 with 1% LBBP (P < 0.05) (Figures 9A, B, and C). A higher (P < 0.01) concen- tration of ALB and AGR was noted in 1% LBBP-fed birds at day 36 (Figures 9B and D). In contrast, TP, ALB, and GLO values were higher (P < 0.05) in the control birds at day 29. In general, LBBP at 1% in feed seemed to be the most effective dose on blood bio- chemistry parameters than 2% LBBP, which suggested a possible concentration-dependent response threshold. WILD BLUEBERRY POMACE IN BROILER CHICKENS 3749 Figure 7. Serum cholesterol and other lipids in free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) CHO, (B) HDLC, and (C) TG. Results are expressed as Mean ± SEM (standard error of the mean). Asterisks indicate the sampling day at which the treatment effects were statistically different (P < 0.05). CHO = cholesterol, HDLC = high-density lipoprotein cholesterol, TG = triglycerides. As expected, broiler age clearly influenced all of the blood metabolites measured in the course of the study. Association Between Gut Microbiota and Blood Metabolites Correlation between gut microbiota composition and blood serum metabolic parameters was estimated to determine their possible relationships (Table 2). Results showed that the 2 most prevalent bacterial genera i.e., Lactobacillus and Bacteroides in LBBP- fed birds were associated with blood metabolites dif- ferently. The genus Lactobacillus correlated negatively with the parameters CHO, HDLC, ALB, and TP, whereas Bacteroides was positively correlated with these parameters in addition to Ca, but showed a nega- tive association with TG (P < 0.05). Likewise Parabac- teroides, which was predominant in 1% LBBP-fed birds with 3-fold higher than control birds at day 21, was found to be correlated (P < 0.05) with blood parame- ters Ca, ALB, and TP. On the other hand, bacterial populations such as Escherichia/Shigella, Clostrid- ium Clostridiaceae, Enterococcus, and Candidatus Sav- agella that were dominant in control birds, particu- larly compared to 1% LBBP, were positively associated (P < 0.05) with the parameters CHO, HDLC, or TG. DISCUSSION Understanding the impact of feeding practices on bird health, performance, and gut microbiota is impor- tant for the development of strategies employing alter- natives, such as berry by-products, to antibiotics for poultry production. The current work investigated for the first time the effects of organic LBBP in feed on gut microbiota composition and blood biochemistry in free-range pastured broiler chickens at various ages. Dietary supplementation with grape pomace and xylo-oligosaccharide has been reported to improve per- formance in broiler chickens (Viveros et al., 2011; Pourabedin et al., 2015). In general, young birds seem to be more susceptible to several clinical and sub- clinical infectious diseases than mature birds. In this study, no differences were noticed between LBBP- treated and control birds for dressed carcass weight at slaughter. Despite the lack of significant differences be- tween treatments for the overall mortality rate, a signif- icantly lower mortality rate was observed in LBBP-fed 3750 ISLAM ET AL. Figure 8. Serum minerals in free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) Ca, (B) Fe (C) Mg, and (D) P. Results are expressed as Mean ± SEM (standard error of the mean). Asterisks indicate the sampling day at which the treatment effects were statistically different (P < 0.05). Ca = calcium, Fe = iron, Mg = magnesium, P = phosphorus. birds between 21 and 28 D. As discussed below, well controlled studies are warranted to establish the effects of the blueberry pomace on the performance, mortality, and the overall health of broiler chickens. The effects of berry pomace on bacterial popula- tion dynamics in the gut of free-range broiler chick- ens are not well investigated. In the current study, dietary LBBP-feeding considerably altered cultivable ileal bacterial populations. In accordance with our re- sults on Lactobacillus, Viveros et al. (2011) noticed a higher count of this bacterium in ileal and cecal sam- ples of 21-day-old broilers fed grape pomace. Feeding LBBP also significantly reduced Enterococcus popula- tions in 21- and 64-day-old birds, a result that was in- consistent with a previous study (Viveros et al., 2011), where the authors reported an increase of Enterococcus in grape pomace-fed birds. Although E. coli, Entero- coccus, Clostridium, and Lactobacillus are considered normal residents of the intestinal tract of chickens, this study indicated that C. perfringens population size was comparatively low. Regardless of treatment group, a gradual change in the composition of all bacterial pop- ulations was observed throughout the trial, which con- firmed that bird’s age is an important determinant in bacterial dynamics in the gut (Barnes et al., 1972). In the present study, the observed increase in cultivable Lactobacillus in LBBP-fed birds compared to control from days 21 to 42 and the concomitant reduction in Enterococcus, C. perfringens, and E. coli at the end of trial was most likely due to the change in nutritional composition of the broiler diets brought by the inclu- sion of LBBP in diet during the late starter and the beginning of the grower phase. Diet can play a central role in shaping the gut mi- crobiota in broiler chickens since components that es- cape host digestion and absorption serve as the sub- strates for the growth of gut bacteria (Rodriguez et al., 2012; Pan and Yu, 2014; Corrigan et al., 2015). In agree- ment with most previous studies, our data showed that the broiler ileum microbiota was mainly composed of Firmicutes, Bacteroidetes, and Proteobacteria (>90%), with Firmicutes being the most predominant phylum (67.30%) (Wei et al., 2013; Pan and Yu, 2014; Waite and Taylor, 2014). It should be noted, however, that Singh et al. (2012) reported Proteobacteria to be the most prevalent phylum (>48%) in fecal samples from broil- ers. A gradual shift in the proportion of Bacteroidetes to Firmicutes was evident in control and 1% LBBP- fed birds over the duration of the trial. Temporal shifts in the proportion of Bacteroidetes at the expense of WILD BLUEBERRY POMACE IN BROILER CHICKENS 3751 Figure 9. Serum proteins in free-range pastured broiler chickens fed with or without organic LBBP supplemented diets: (A) TP, (B) ALB, (C) GLO, and (D) AGR. Results are expressed as Mean ± SEM (standard error of the mean). Asterisks indicate the sampling day at which the treatment effects were statistically different (P < 0.05). TP = total proteins, ALB = albumin, GLO = globulin, AGR = ALB-GLO ratio. Firmicutes in the chicken gut microbiota were previ- ously reported by Wei et al. (2013). Moreover, the abun- dance of Bacteroidetes in broilers as a result of feed sup- plementation with mannan-oligosaccharides has also been described (Corrigan et al., 2015). Bacteroidetes and Firmicutes contribute to the fermentation of indi- gestible carbohydrates to short-chain fatty acids (SC- FAs), and while the former phylum mainly produces butyrate, the latter produces acetate and propionate as its primary metabolic end products (Macfarlane and Macfarlane, 2003). SCFAs are known to have antimi- crobial activities against some pathogenic bacteria and anti-inflammatory properties, as well as immunomodu- latory roles. On the other hand, an increased F/B ratio has been shown to be associated with obesity in humans and mice, owing to the improved energy harvesting ca- pacity of Firmicutes species (Ley et al., 2006). At the end of the trial, the F/B ratio in the birds fed the con- trol diet was almost double than in 1% LBBP-fed birds, suggesting that the gut microbiota composition could be very different in LBBP-treated and LBBP-untreated broilers of common genetics, flock, and reared under identical conditions. Both culture-dependent and 16S rRNA gene sequence analysis confirmed the dominance of Lactobacillus in the ileum of chickens, which has been reported previously (Wei et al., 2013; Pan and Yu, 2014). Lactobacillus has been shown to produce bacteri- ocins, and moreover it produces lactic acid that lowers the pH in the surrounding environment and inhibits the growth of pathogenic bacteria such as E. coli, C. per- fringens, and Campylobacter jejuni (Murry et al., 2004; Neal-McKinney et al., 2012). In this study, the relative proportions of Lactobacillus were considerably higher (1.33 to 2.42-fold), and those for Escherichia/Shigella and Clostridium Clostridiaceae were drastically lower (3.40 to 6.20-fold) in LBBP-fed chickens at the end of production cycle. These results confirm the notion that Lactobacillus is antagonistic to enterobacterial pathogens (Servin, 2004; Klose et al., 2010). Contrary to our findings, lower Lactobacillus and higher Clostrid- ium population sizes were reported by Mohd Shaufi et al. (2015) in broilers fed a commercial diet, and the authors concluded that it is crucial to modulate the broiler gut microbiota in order to improve gut health. In the present study, Bacteroides, Bifidobac- terium, and Faecalibacterium were more prevalent in LBBP-fed birds while Enterococcus was 6 to 10-fold more abundant in the control birds at the end of the trial. The greater abundance of Enterococcus in 3752 ISLAM ET AL. Table 2. Relationship between free-range pastured chicken gut bacterial community at genus level and various blood serum metabolites. Genus1 Correlations with blood metabolic parameters (P < 0.05)2 Bacteroides, Desulfovibrio, Mucispirillum, Eubacterium, Sutterella, Odoribacter, RFN20, Phascolarctobacterium Ca, TG, HDLC, CHO, ALB, TP Peptococcus, Megamonas Ca, HDLC, CHO, ALB, TP Megasphaera Ca, TG, HDLC, CHO, ALB Lactobacillus TG, HDLC, CHO, ALB, TP Fusobacterium TG, HDLC, CHO, ALB Dorea Ca, HDLC, CHO Paludibacter, Anaerobiospirillum, Kurthia, Elusimicrobium, Lysinibacillus, 5–7N15 TG, HDLC, CHO Veillonella, Turicibacter HDLC, CHO, ALB Parabacteroides, Helicobacter Ca, ALB, TP Escherichia/Shigella3, Clostridium Clostridiaceae, Enterococcus, Ruminococcus, Butyricimonas, Gallibacterium, Akkermansia, Aerococcus, HDLC, CHO Campylobacter ALB, TP Oscillospira Ca Blautia CHO Epulopiscium HDLC Candidatus Savagella TG 1Genera assigned (at 97% identity) by Greengenes taxonomic scheme are mentioned here. Taxa failed to classify down to genus level are not listed. 2Results from Spearman’s correlations with P-values assigned by Fisher’s Z transform as corrected by Benjamini–Hochberg FDR procedure. 3Identified by NCBI BLAST. Ca = calcium, TG = triglycerides, HDLC = high-density lipoprotein cholesterol, CHO = cholesterol, ALB = albumin, TP = total protein. the control birds suggested possible inhibition of this bacterium by LBBP. Members of the genus Bacteroides are among the most effective degraders of indigestible carbohydrates (Al-Sheikhly and Al-Saieg, 1980), and are characterized as SCFA producers (Kaakoush et al., 2014). The predominance of Bacteroides in LBBP- fed birds could provide an advantage in digesting indigestible carbohydrates that exist in the host. Bac- teroides and Bifidobacterium can contribute to mucin degradation (Hooper et al., 2002; Ruas-Madiedo et al., 2008), while the genus Faecalibacterium produces the SCFA butyrate that functions as an anti-inflammatory in host cells (Louis et al., 2010). On the other hand, the genus Enterococcus is reported to be associated with colorectal cancer (Balamurugan et al., 2008), and Enterococcus faecalis can damage eukaryotic cellular DNA in colonic epithelial cells by producing extracellular superoxides and hydroperoxides (Huycke et al., 2002; Jones et al., 2008). Observations from the present work indicate that the dietary inclusion of LBBP may beneficially alter the intestinal microbiota and their activities in broiler chickens. Diversity parameters of chicken gut bacterial commu- nities in the treatment groups (particularly 1% LBBP) were relatively stable from days 14 to 42. In paral- lel with α-diversity, the unweighted UniFrac PCoA also showed that the intestinal samples tended to clus- ter more tightly by time-point than by treatment, in- dicating that bird age was a major factor in shap- ing the bacterial community. Similar results have been reported previously upon feeding broilers with mannan- oligosaccharides and xylo-oligosaccharide (Corrigan et al., 2015; Pourabedin et al., 2015). However, a clear clustering was evident at day 29, providing evidence that diet supplementation with LBBP altered the com- position and structure of bacterial communities in the broiler chicken gut. Examination of blood biochemistry in birds provides valuable information on their health and performance. We showed that feeding organic broilers with LBBP influenced the levels of several blood metabolites, in- cluding CHO and TG. The decrease of these 2 metabo- lites levels in the present study may be related to the increase in Lactobacillus population size. Accordingly, Kalavathy et al. (2003) reported 8 to 11% and 16 to 25% lesser levels of CHO and TG, respectively, at days 21 to 42 in broilers fed Lactobacillus. However, these au- thors did not observe any change in HDLC level, which is contrary to the present findings showing a significant increase in HDLC level in 1% LBBP-fed birds at days 14, 42, and 64. Minerals are essential nutrients involved in many di- gestive, physiological, and biosynthetic processes re- quired for broiler growth and development. Organic LBBP has been reported to be a good source of minerals such as Ca, Fe, Mn, Zn, and Cu (Ross et al., 2017). In this work, the concentration of Ca and Mg was higher in blood serum of LBBP-fed birds from days 14 to 29. The increase of these 2 minerals might be explained by their availability in the LBBP-supplemented grower diet compared to the control diet. Likewise, Fe and P content was much higher in the LBBP-supplemented WILD BLUEBERRY POMACE IN BROILER CHICKENS 3753 grower diet than in the control, which has been re- flected in blood serum of 1% LBBP-fed birds, showing significantly elevated amount of Fe at day 36 and P at day 21. These results underline the importance of mineral content in broiler diets for the maintenance of good health and performance. In this work, considering all bird groups throughout the trial, the maximum of 43.50 (P = 0.07) and 27.50 g/L (P < 0.05) of TP and GLO was ob- served, respectively, in 1% LBBP-fed birds at day 21. These results could be explained by the composition of the LBBP described earlier (Ross et al., 2017), which was different to the product used by Kim et al. (2015) who did not find significant differences in serum TP, ALB, GLO, and AGR at day 35 after feeding chickens with α-lipoic acid-supplemented diet. Protein content (crude, adjusted and soluble) was generally higher in LBBP-supplemented broiler feed that resulted in a higher concentration of TP in all sampling time-points (except for days 29 to 36) in 1% LBBP-fed birds. It is essential to carefully investigate the influence of dietary factors on the physiology and metabolism of the gut bacterial community, and possible effects on the absorption of fermentation end products. Nevertheless, the present work demonstrates that supplementation of broiler diets with LBBP can positively influence metabolic responses in broiler chickens. The data from the correlation study show that Lactobacillus and Bacteroides, dominant in LBBP- fed birds, were associated with most of the blood metabolites. The genus Lactobacillus negatively corre- lated with CHO, which supports the notion that this bacterium could reduce the serum CHO level in broilers (Jin et al., 1998). On the other hand, positive corre- lations between some potentially pathogenic bacteria (Escherichia/Shigella, Clostridium Clostridiaceae, and Enterococcus) and blood CHO, as well as between Can- didatus Savagella and TG indicate that these bacte- rial taxa may play a role in modulating these serum metabolites in broilers. The apparent correlation be- tween the abundance of specific bacterial taxa and blood metabolites suggests the possibility of biological response control through manipulation of the bacterial flora in broiler chickens. Further studies are required to characterize these relationships in more detail with a view to develop feeding strategies using LBBP to improve overall health and performance in organic poultry production. The present study was subject to limitations im- posed by research in an open pasture environment. Understanding of chicken-ranging behavior is presently insufficient to ensure optimal range design and moni- toring free-range bird performance in this environment is challenging (Taylor et al., 2017). Like most stud- ies associated with 16S rRNA gene-sequencing data (reviewed in Kers et al., 2018), the small number of birds (low sample size) investigated in this study was restrictive and a larger number of samples would undoubtedly have provided stronger data concerning overall microbiota composition. However, the present study generated data that implied the relevance of an integrated approach for the assessment of feed supple- mentation on overall chicken health and productivity. Future research in our laboratory will seek to more accurately define the relationship between microbiota composition and blood metabolites. Collectively, results of this study indicated that in- clusion of LBBP in diets for free-range pastured broiler chickens could modulate gut microbiota and blood metabolic responses. The present study confirmed that bacterial community composition and clustering was primarily influenced by the age of the birds. However, LBBP supplementation augmented the proportion of beneficial bacteria such as Lactobacillus, Bacteroides, Parabacteroides, and Bifidobacterium, and reduced the bacteria among which some species are known to be as- sociated with infectious diseases in broilers. These re- sults provide an interesting insight into the intricacy of chicken gut bacterial communities and help understand the impact of feed supplementation with berry pomace in free-range pastured broilers. More study is needed to better understand the underlying mechanisms of ac- tion of LBBP interacting with gut microbiota and the host to further enhance the overall free-range pastured broiler health. ACKNOWLEDGMENTS This work was supported by Agriculture and Agri- Food Canada (AAFC) under the Organic Science Clus- ter II Program (AIP CL-02 AGR-10383). We thank the Certified Organic Association of British Columbia and the participating farmers (A. and S. Gunner, Arm- strong, BC) for their supports. We acknowledge X. Yin (AAFC, Guelph) for help and assistance. Conflict of interest statement The authors declare that there are no conflicts of interest. REFERENCES AAFC. 2011. Canadian blueberries: A true taste of nature. AAFC No. 11509E. Agriculture and Agri-Food Canada. AAFC. 2012. Statistical overview of the Canadian blueberry indus- try, 2010. AAFC No. 11743E. Agriculture and Agri-Food Canada. Al-Sheikhly, F., and A. Al-Saieg. 1980. Role of Coccidia in the oc- currence of necrotic enteritis of chickens. Avian Dis. 24:324–333. Aronesty, E. 2011. ea-utils: “Command-line tools for processing bi- ological sequencing data”. Balamurugan, R., E. Rajendiran, S. George, G. V. Samuel, and B. S. Ramakrishna. 2008. Real-time polymerase chain reaction quan- tification of specific butyrate-producing bacteria, Desulfovibrio and Enterococcus faecalis in the feces of patients with colorectal cancer. J. Gastroenterol. Hepatol. 23:1298–1303. Barnes, E. M., G. C. Mead, D. A. Barnum, and E. G. Harry. 1972. The intestinal flora of the chicken in the period 2 to 6 weeks of age, with particular reference to the anaerobic bacteria. Br. Poult. Sci. 13:311–326. 3754 ISLAM ET AL. Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B. 57:289–300. Biswas, D., N. E. Wideman, C. A. O’Bryan, A. Muthaiyan, J. M. Lingbeck, P. G. Crandall, and S. C. Ricke. 2012. Pasteurized blueberry (Vaccinium corymbosum) juice inhibits growth of bacterial pathogens in milk but allows survival of probiotic bacteria. J. Food Safety. 32:204–209. Bjerrum, L., R. M. Engberg, T. D. Leser, B. B. Jensen, K. Finster, and K. Pedersen. 2006. Microbial community composition of the ileum and cecum of broiler chickens as revealed by molecular and culture-based techniques. Poult. Sci. 85:1151–1164. Caporaso, J. G., K. Bittinger, F. D. Bushman, T. Z. DeSantis, G. L. Andersen, and R. Knight. 2010. PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinformatics. 26:266–267. Caporaso, J. G., J. Kuczynski, J. Stombaugh, K. Bittinger, F. D. Bushman, E. K. Costello, N. Fierer, A. G. Peña, J. K. Goodrich, J. I. Gordon, G. A. Huttley, S. T. Kelley, D. Knights, J. E. Koenig, R. E. Ley, C. A. Lozupone, D. McDonald, B. D. Muegge, M. Pirrung, J. Reeder, J. R. Sevinsky, P. J. Turnbaugh, W. A. Walters, J. Widmann, T. Yatsunenko, J. Zaneveld, and R. Knight. 2010. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods. 7:335–336. Chambers, J. R., and J. Gong. 2011. The intestinal microbiota and its modulation for Salmonella control in chickens. Food Res. Int. 44:3149–3159. Collier, C. T., J. D. van der Klis, B. Deplancke, D. B. Anderson, and H. R. Gaskins. 2003. Effects of tylosin on bacterial mucol- ysis, Clostridium perfringens colonization, and intestinal barrier function in a chick model of necrotic enteritis. Antimicrob. Agents Chemother. 47:3311–3317. Corrigan, A., M. de Leeuw, S. Penaud-Frézet, D. Dimova, and R. A. Murphy. 2015. Phylogenetic and functional alterations in bacte- rial community compositions in broiler ceca as a result of man- nan oligosaccharide supplementation. Appl. Environ. Microbiol. 81:3460–3470. Das, Q., M. R. Islam, M. F. Marcone, K. Warriner, and M. S. Diarra. 2017. Potential of berry extracts to control foodborne pathogens. Food Control. 73:650–662. DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu, and G. L. Andersen. 2006. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72:5069–5072. Diarrassouba, F., M. S. Diarra, S. Bach, P. Delaquis, J. Pritchard, E. Topp, and B. J. Skura. 2007. Antibiotic resistance and viru- lence genes in commensal Escherichia coli and Salmonella isolates from commercial broiler chicken farms. J. Food Prot. 70:1316– 1327. Edgar, R. C. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 26:2460–2461. Gong, J., R. J. Forster, H. Yu, J. R. Chambers, R. Wheatcroft, P. M. Sabour, and S. Chen. 2002. Molecular analysis of bacterial populations in the ileum of broiler chickens and comparison with bacteria in the cecum. FEMS Microbiol. Ecol. 41:171–179. Hayes, J. R., L. L. English, P. J. Carter, T. Proescholdt, K. Y. Lee, D. D. Wagner, and D. G. White. 2003. Prevalence and antimicro- bial resistance of Enterococcus species isolated from retail meats. Appl. Environ. Microbiol. 69:7153–7160. Herken, E. N., and S. Guzel. 2010. Total antioxidant capacity and to- tal phenol contents of selected commercial fruit juices in Turkey. Int. J. Food Prop. 13:1373–1379. Herlemann, D. P., M. Labrenz, K. Jurgens, S. Bertilsson, J. J. Waniek, and A. F. Andersson. 2011. Transitions in bacterial com- munities along the 2000km salinity gradient of the Baltic Sea. ISME J. 5:1571–1579. Hooper, L. V., T. Midtvedt, and J. I. Gordon. 2002. How host- microbial interactions shape the nutrient environment of the mammalian intestine. Annu. Rev. Nutr. 22:283–307. Huycke, M. M., V. Abrams, and D. R. Moore. 2002. Enterococcus faecalis produces extracellular superoxide and hydrogen peroxide that damages colonic epithelial cell DNA. Carcinogenesis. 23:529– 536. Islam, M. R., D. V. Oomah, and M. S. Diarra. 2017. Potential im- munomodulatory effects of non-dialyzable materials of cranberry extract in poultry production. Poult. Sci. 96:341–350. Jacob, J., and A. Pescatore. 2012. Prebiotics, Pages 379—405 in Organic Meat Production And Processing. S. C. Ricke, E. J. Van Loo, M. G. Johnson, and C. A. O’Bryan, ed. John Wiley & Sons Inc., Hoboken, NJ. Jin, L. Z., Y. W. Ho, N. Abdullah, and S. Jalaludin. 1998. Growth performance, intestinal microbial populations, and serum choles- terol of broilers fed diets containing Lactobacillus cultures. Poult. Sci. 77:1259–1265. Jones, B. V., M. Begley, C. Hill, C. G. M. Gahan, and J. R. Marchesi. 2008. Functional and comparative metagenomic analysis of bile salt hydrolase activity in the human gut microbiome. Proc. Natl. Acad. Sci. 105:13580–13585. Kaakoush, N., N. Sodhi, J. Chenu, J. Cox, S. Riordan, and H. Mitchell. 2014. The interplay between Campylobacter and Helicobacter species and other gastrointestinal microbiota of commercial broiler chickens. Gut Pathog. 6:18 (Pages 1-10). doi:10.1186/1757-4749-6-18. Kalavathy, R., N. Abdullah, S. Jalaludin, and Y. W. Ho. 2003. Ef- fects of Lactobacillus cultures on growth performance, abdominal fat deposition, serum lipids and weight of organs of broiler chick- ens. Br. Poult. Sci. 44:139–144. Kers, J. G., F. C. Velkers, E. A. J. Fischer, G. D. A. Hermes, J. A. Stegeman, and H. Smidt. 2018. Host and environmental factors affecting the intestinal microbiota in chickens. Front. Microbiol. 9:235 doi: 10.3389/fmicb.2018.00235. Kim, D. W., M. M. Mushtaq, R. Parvin, H. K. Kang, J. H. Kim, J. C. Na, J. Hwangbo, J. D. Kim, C. B. Yang, B. J. Park, and H. C. Choi. 2015. Various levels and forms of dietary α-lipoic acid in broiler chickens: impact on blood biochemistry, stress response, liver enzymes, and antibody titers. Poult. Sci. 94:226–231. Klose, V., K. Bayer, R. Bruckbeck, G. Schatzmayr, and A. P. Loibner. 2010. In vitro antagonistic activities of animal intesti- nal strains against swine-associated pathogens. Vet. Microbiol. 144:515–521. Knarreborg, A., M. A. Simon, R. M. Engberg, B. B. Jensen, and G. W. Tannock. 2002. Effects of Dietary Fat Source and Sub- therapeutic Levels of Antibiotic on the Bacterial Community in the Ileum of Broiler Chickens at Various Ages. Appl. Environ. Microbiol. 68:5918–5924. Leusink, G., H. Rempel, B. Skura, M. Berkyto, W. White, Y. Yang, J. Y. Rhee, S. Y. Xuan, S. Chiu, F. Silversides, S. Fitzpatrick, and M. S. Diarra. 2010. Growth performance, meat quality, and gut microflora of broiler chickens fed with cranberry extract1. Poult. Sci. 89:1514–1523. Ley, R. E., P. J. Turnbaugh, S. Klein, and J. I. Gordon. 2006. Human gut microbes associated with obesity. Nature. 444:1022– 1023. Louis, P., P. Young, G. Holtrop, and H. J. Flint. 2010. Diversity of human colonic butyrate-producing bacteria revealed by anal- ysis of the butyryl-CoA:acetate CoA-transferase gene. Environ. Microbiol. 12:304–314. Lozupone, C., and R. Knight. 2005. UniFrac: a new phylogenetic method for comparing microbial communities. Appl. Environ. Mi- crobiol. 71:8228–8235. Macfarlane, S., and G. T. Macfarlane. 2003. Regulation of short- chain fatty acid production. Proc. Nutr. Soc. 62:67–72. Merchant, L. E., H. Rempel, T. Forge, T. Kannangara, S. Bittman, P. Delaquis, E. Topp, K. A. Ziebell, and M. S. Diarra. 2012. Char- acterization of antibiotic resistant and potentially pathogenic Es- cherichia coli from soil fertilized with litter of broiler chickens fed antimicrobial-supplemented diets. Can. J. Microbiol. 50:1084– 1098. Mohd Shaufi, M. A., C. C. Sieo, C. W. Chong, H. M. Gan, and Y. W. Ho. 2015. Deciphering chicken gut microbial dynamics based on high-throughput 16S rRNA metagenomics analyses. Gut Pathog. 7:4 (Pages 1-12). doi: 10.1186/s13099-015-0051-7. Murry, A. C., Jr., A. Hinton, Jr., and H. Morrison. 2004. Inhi- bition of growth of Escherichia coli, Salmonella Typhimurium, and Clostridia perfringens on chicken feed media by Lactobacillus salivarius and Lactobacillus plantarum. Int. J. Poult. Sci. 3:603– 607. WILD BLUEBERRY POMACE IN BROILER CHICKENS 3755 National Research Council (NRC). 1994. Nutrient Requirements of Poultry. Washington, DC. National Academy Press. Neal-McKinney, J. M., X. Lu, T. Duong, C. L. Larson, D. R. Call, D. H. Shah, and M. E. Konkel. 2012. Production of organic acids by probiotic lactobacilli can be used to reduce pathogen load in poultry. PLoS One. 7:e43928. Oakley, B. B., H. S. Lillehoj, M. H. Kogut, W. K. Kim, J. J. Mau- rer, A. Pedroso, M. D. Lee, S. R. Collett, T. J. Johnson, and N. A. Cox. 2014. The chicken gastrointestinal microbiome. FEMS Microbiol Lett. 360:100–112. Pan, D., and Z. Yu. 2014. Intestinal microbiome of poultry and its interaction with host and diet. Gut Microbes. 5:108–119. Pourabedin, M., L. Guan, and X. Zhao. 2015. Xylo-oligosaccharides and virginiamycin differentially modulate gut microbial composition in chickens. Microbiome. 3:15 (Pages 1-12). doi: 10.1186/s40168-015-0079-4. Price, M. N., P. S. Dehal, and A. P. Arkin. 2009. FastTree: computing large minimum evolution trees with profiles instead of a distance matrix. Mol. Biol. Evol. 26:1641–1650. Rodriguez, M. L., A. Rebole, S. Velasco, L. T. Ortiz, J. Trevino, and C. Alzueta. 2012. Wheat- and barley-based diets with or without additives influence broiler chicken performance, nutrient digestibility and intestinal microflora. J. Sci. Food Agric. 92:184– 190. Ross, K. A., D. Ehret, D. Godfrey, L. Fukumoto, and M. S. Diarra. 2017. Characterization of pilot scale processed Canadian organic cranberry (Vaccinium macrocarpon) and blueberry (Vaccinium angustifolium) juice pressing residues and phenolic enriched ex- tractives. Int. J. Fruit Sci. 17:202–232. Ruas-Madiedo, P., M. Gueimonde, M. Fernández-Garćıa, C. G. de los Reyes-Gavilán, and A. Margolles. 2008. Mucin degradation by Bifidobacterium strains isolated from the human intestinal microbiota. Appl. Environ. Microbiol. 74:1936–1940. Servin, A. L. 2004. Antagonistic activities of lactobacilli and bifi- dobacteria against microbial pathogens. FEMS Microbiol. Rev. 28:405–440. Shen, X., X. Sun, Q. Xie, H. Liu, Y. Zhao, Y. Pan, C. Hwang, and V. C. H. Wu. 2014. Antimicrobial effect of blueberry (Vaccinium corymbosum L.) extracts against the growth of Listeria mono- cytogenes and Salmonella Enteritidis. ARS USDA Submissions. 35:159–165. Singer, R. S., and C. L. Hofacre. 2006. Potential impacts of antibiotic use in poultry production. Avian Dis. 50:161–172. Singh, K., T. Shah, S. Deshpande, S. J. Jakhesara, P. G. Koringa, D. N. Rank, and C. G. Joshi. 2012. High through put 16S rRNA gene-based pyrosequencing analysis of the fecal microbiota of high FCR and low FCR broiler growers. Mol. Biol. Rep. 39:10595– 10602. Statistical Analysis System (SAS). 2000. Release 8.02. SAS Institute Inc., Cary, NC. Taylor, P. S., P. H. Hemsworth, P. J. Groves, S. G. Gebhardt- Henrich, and J. L. Rault. 2017. Ranging behaviour of commercial free-range broiler chickens 1: Factors related to flock variability. Animals 7:54 (Pages 1-14). doi: 10.3390/ani7070054. Thompson, C. L., R. Vier, A. Mikaelyan, T. Wienemann, and A. Brune. 2012. ‘Candidatus Arthromitus’ revised: segmented filamentous bacteria in arthropod guts are members of Lach- nospiraceae. Environ. Microbiol. 14:1454–1465. Vela, J., K. Hildebrandt, A. Metcalfe, H. Rempel, S. Bittman, E. Topp, and M. S. Diarra. 2012. Characterization of Staphylococcus xylosus isolated from broiler chicken barn bioaerosol. Poult. Sci. 91:3003–3012. Viveros, A., S. Chamorro, M. Pizarro, I. Arija, C. Centeno, and A. Brenes. 2011. Effects of dietary polyphenol-rich grape products on intestinal microflora and gut morphology in broiler chicks. Poult. Sci. 90:566–578. Waite, D., and M. Taylor. 2014. Characterising the avian gut microbiota: membership, driving influences, and po- tential function. Front. Microbiol. 5:223 (Pages 1-12). doi: 10.3389/fmicb.2014.00223. Wei, S., M. Morrison, and Z. Yu. 2013. Bacterial census of poultry intestinal microbiome. Poult. Sci. 92:671–683. Effects of wild blueberry (Vaccinium angustifolium) pomace feeding on gut microbiota and blood metabolites in free-range pastured broiler chickens INTRODUCTION MATERIALS AND METHODS Product Use: Pomace Preparation and Composition Birds, Diets and Management Performance and Mortality Culture-dependent Bacteriological Analysis Analysis of Bacterial Community Structure Analysis of Blood Metabolites Statistical Analysis RESULTS Feed Composition Performance and Mortality Culture-dependent Bacterial Enumeration Culture-independent Gut MicrobiotaStructure Blood Biochemistry Parameters Association Between Gut Microbiota and Blood Metabolites DISCUSSION ACKNOWLEDGMENTS Conflict of interest statement REFERENCES