Growth performance and total tract digestibility in broiler chickens fed different corn hybrids an,* J. F. Perez,* G. Gonzalez-Ortiz,y S. Villagomez-Estrada,* M. R. Bedford,y D. Melo-Dur H. Graham,y and D. Sola-Oriol*,1 * Servei de Nutricio i Benestar Animal (SNiBA), Department de Ciencia Animal i dels Aliments, Universitat Autonoma de Barcelona (UAB), Barcelona 08193, Spain; and yAB Vista, Marlborough, Wiltshire, SN8 4AN, United Kingdom ABSTRACT The aim of the present study was to 8 diet (P < 0.05), together with varieties 3 and 5. Energy investigate the variability in nutrient digestibility asso- digestibility was higher in varieties 2, 4 and 7 (P < 0.05). ciated with corn genetic background and its influence on Multivariate analysis revealed that corn protein concen- the feeding value for broiler chickens. A total of 960 1- tration was positively correlated with vitreousness day-old male broiler chicks (Ross 308) were distributed (r = 0.60, P = 0.054) and the arabinose:xylose ratio in eight treatments, with 12 pens per treatment and 10 (r = 0.67, P < 0.05) and negatively correlated with birds per pen in a 42-day study. Eight corn samples starch (r = -0.62, P < 0.05). Soluble non-starch polysac- (Variety 1 to Variety 8) were selected based on their charide content was negatively correlated with the pro- nutrient composition. A fixed amount of each corn tein solubility index (r = -0.88, P < 0.05). In addition, (577 g/kg in the starter diets and 662 g/kg in the finisher corn protein concentration was negatively correlated (P diets) was used to formulate feeds. Diets were offered ad < 0.05) with 21-d BW (r = -0.71) and weight gain (r = - libitum in pellet form. Performance parameters were 0.62). In conclusion, the corn genetic background influ- determined at d 21 and d 42, and excreta samples col- enced the nutrient digestibility and growth performance lected at d 21 to determine energy, organic matter and of broiler chickens. The content and nature of the non- dry matter (DM) whole-tract digestibility. The results starch polysaccharides were found to be two of the main revealed a decrease (P < 0.05) in body weight (BW) factors affecting the solubility and availability of and feed intake in birds fed variety 8 compared to other nutrients in corn, and could be the reason for the nega- varieties at d 21. The lowest whole tract DM and energy tive effects on the performance of broiler chickens as apparent digestibility were also observed for the variety shown in the present study. Keywords: corn, near-infrared spectroscopy, total tract digestibility, chickens 2021 Poultry Science 100:101218 https://doi.org/10.1016/j.psj.2021.101218 INTRODUCTION can be very variable (Summers, 2001; Cowieson, 2005). Genetics, agronomic conditions, proximate composition, Corn is the most common energy source used in com- and pre- and post-harvest processing are considered major mercial animal diets, especially in American, Southern factors affecting the nutrient variability. Of these, genetics European and Asian countries, where corn grain is the pri- has been demonstrated to be an important source of bio- mary cereal for all poultry feed (Larbier et al., 1994; chemical and nutrient variability (Uribelarrea et al., 2004; Dei, 2017). Due to its high dietary inclusion rate, corn can Reynolds et al., 2005). Phenotypic characteristics such as contribute up to 65% of metabolizable energy and 20% of the grain-filling duration, related to physiological matu- protein in poultry diets (Gehring et al., 2013; rity, and composition, growth rate and moisture of the ker- Naderinejad et al., 2016). Although corn has a high and nel are specific for each genotype and could affect the consistent nutritional value for livestock, its feeding value nutrient value (Seebauer et al., 2010; Prado et al., 2014). The main differences in corn composition include protein Ó 2021 The Authors. Published by Elsevier Inc. on behalf of Poultry solubility, zein content, amylose to amylopectin ratio, and Science Association Inc. This is an open access article under the CC vitreousness (Gehring et al., 2013). Moreover, the apparent BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ metabolizable energy (AME) value of corn can fluctuate 4.0/). Received April 17, 2020. by more than 470 kcal/kg from batch to batch (Cowie- Accepted April 14, 2021. son, 2005). 1 Corresponding author: David.Sola@uab.cat 1 2  ET AL. MELO-DURAN Near-infrared spectroscopy (NIRS) is an efficient tool land area and the harvest weights of each variety were for assessing the nutritive value of raw ingredients before used to calculate maize yield per hectare. The plot for manufacturing the feed, and thus it enhances precision each variety consisted of 8 rows with 17 cm of separation feed formulation (Pujol et al., 2007). In addition to pre- between plants within the row, and 70 cm of separation dicting the conventional composition, NIRS can also be between rows of the same variety. The cobs used in the used to predict the non-conventional composition, includ- present study were collected at random from the experi- ing anti-nutritional factors (Rahman et al., 2015). The mental field. Briefly, for each variety a total of 50 cobs anti-nutritional factors of corn include non-starch poly- were obtained from groups of five cobs in 10 sampling saccharides (NSP), phytate, lectins, and resistant points that followed a diagonal pattern along the 4 central starches (Englyst, 1989; Eeckhout and De Paepe, 1994; rows from the plot of each variety. Eight corn samples Cowieson, 2005). In this context, it was hypothesized that (variety 1 to variety 8) were selected considering their the corn genetic background could influence energy and nutrient composition, in order to cover a wide range of nutrient digestibility, and thus broiler performance. nutrient variation, mainly focused on protein, starch and Therefore, the aim of the present study was to research NSP content. The corn proximate and physiochemical the differences in growth performance and energy and analyses are the averages from 10 cobs of each hybrid nutrient digestibility in broilers fed diets based on eight selected randomly from the experimental field, analyzed different corn hybrids harvested in the same field/area by NIRS and expressed on a fresh weight basis. and under similar environmental growing and fertilizing conditions. Experimental Diets MATERIALS AND METHODS The experimental diets based on corn and soybean- meal were formulated to meet all nutrient requirements Birds and Housing recommended by FEDNA (2018) (Table 1). Broilers A total of 960 Ross 308-day-old male broilers were were fed in 2 different phases: starter, from 0 to 21 d of purchased from a local hatchery. Upon arrival, birds age; and finisher, from 22 to 42 d of age. A fixed amount were individually weighed and assigned to 96 floored of each corn (577 g/kg in the starter diets and 662 g/kg pens in an environmentally controlled room, with ten in the finisher diets) was used in the formula regardless birds per pen. Each pen (150 cm x 75 cm) was equipped of their chemical composition. The nutrient contents of with a bell feeder and drinker. Test diets and water were each corn hybrid, determined by NIRS, are shown in provided ad libitum throughout the trial. During the first Table 2. Diets were formulated to have the same content 2 d, the temperature was set at 32°C and was gradually of protein (starter: 200 g/kg and finisher: 180 g/kg), reduced to 20°C until the end of the trial (d 42). For the ether extract (starter: 7.7 g/kg and finisher: 5.5 g/kg) first 10 d, 23 h of light were provided and this was and corrected nitrogen AME (AMEn) (starter: 2,900 reduced to 18 h from d 11 onwards. −2,940 kcal/kg and finisher: 3,045−3,087 kcal/kg). In order to include the same quantity of corn and be isoni- trogenous, an ingestible ingredient (silicon dioxide; Corn Samples IBERSIL) was added to the diets as required. Diets were pelleted at 70°C using a pellet press (Pellet Mill 3020-4) Sixteen corn varieties were sowed in the same field with with a capacity of 4,500−5,500 kg/h and using a 2.2 mm the same environmental growing conditions (Gimenells, die for starter diets and 3.3 mm die for finisher diets. Catalunya, Spain), including fertilizer and harvesting. Diets contained 5 g/kg of titanium dioxide as an indi- The rainfall of the area was 207 mm in the cultivation gestible marker for estimating nutrient digestibility. period and the average minimum/maximum temperatures varied between 15.2 and 29.3°C (AEMET, 2020). The planting density was 92,000 seeds/ha. Fixed cane sprin- Experimental Procedures kling was used for irrigation from sowing until an average plant height of 50 cm, at a rate of 1 h of watering per day, The experimental procedures were approved by the applied at night. From 50 cm until the plants had 12 to 14 Animal Experiment Committee of the Universitat leaves, they were watered according to the need for soil Autonoma de Barcelona, and were in compliance with moisture, based on rainfall and visual assessment of the the European Union guidelines for the care and use of plants. From 12 to 14 leaves until the flowering spike was animals in research (European Parliament, 2010). lost, water was applied for 30 min daily; and from flower- Birds were weighed individually at d 0, 21 and 42, and ing to harvest, plants were watered for 1 h daily at a rate feed consumption was determined by pen per diet. Body of 50 min at night and 10 min at the time of maximum weight (BW) uniformity was evaluated as the BW coeffi- heat during the day. No pesticide was applied during culti- cient of variation. Mortality was monitored twice daily, vation, the fertilizer scheme was 170 kg of N/ha before and the weights of dead birds were used to adjust the feed sowing, 50 kg of liquid N/ha at 8 leaves, and 50 kg of liq- conversion ratio. Excreta samples were collected at d 21 to uid N/ha plus 5 L of organic fertilizer, through the irriga- determine energy, organic matter (OM) and dry matter tion system, at 12 leaves. The maize hybrids were sown in (DM) digestibility. The excretas were collected in plastic April 2018 and harvested in October 2018. The size of the containers and immediately frozen at -20°C. Excretas CORN NUTRIENT VARIABILITY IN BROILER DIETS 3 Table 1. Ingredients and calculated composition of the experi- NSP components using high performance liquid chroma- mental diets1. tography, following the method of Englyst et al. (1994). Diets d 0−21 d 22−42 Diet and Excreta Nutrient Analyses The contents of DM, ash, and gross energy (GE) were analyzed in feed Ingredients, (g/kg) Mean Range Mean Range and excreta, as well as crude protein (CP) in feed. Proxi- Corn 577 - 662 - mate analyses were performed according to Soybean meal 47 317 311−331 263 256−269 Soy oil 53 50−55 33 30−36 AOAC (2005) Official Methods: Method 968.06 (CP), DL-Methionine 3.3 3.2−3.4 2.8 2.7−2.9 Method 934.01 (DM), Method 942.05 (ash). GE was Lysine HCl 3.6 3.3−3.8 2.8 2.6−3.1 determined using an isoperibolic calorimeter (Parr Threonine 1.3 1.2−1.4 1 0.9−2.0 L-Valine 1.2 1.1−1.3 0.3 0.2−0.5 Instrument Company, Moline, Illinois, USA). Salt 3 - 3 - Whole-tract apparent digestibility of crude protein, Limestone 11 - 11 - energy, DM, and OM were calculated with the index Monodicalcium 10 - 9 - phosphate method using the following equation: Vitamin premix2 4 - 4 - Sepiolite 14 0−24 8 0−17 Whole Tract Apparent Digestibility Phytase3 0.1 0.1 Calculated composition ¼ 1  ðð½TiD =½TiE Þ  ð½NE =½ND ÞÞ (g/kg) AMEn kcal/kg 2,920 2,900−2,950 3,064 3,032−3,087 Crude protein 200 - 180 - Where [TiD] is the concentration of Ti in the diet, Calcium 9.3 - 9.0 - Available P 4.5 - 4.3 - [TiE] is the concentration of Ti in excreta, [NE] is the Fat 77 - 55 - nutrient content in the excreta, and [N] is the nutrient D Met+Cys 8.6 - 7.6 - content in the diet. D Lys 11.8 - 10.0 - Analyzed composition, DM Solubility and Water Retention Capacity The (g/kg) hydrolysis capacity of the nutrients and the ability to Gross energy, Kcal/kg 4,150 4,075−4,224 3,938 3,914−3,958 retain water by the fiber components (NSP) were deter- Dry matter 910 906−922 870 864−874 Crude protein 196 190−199 170 168−179 mined using the solubility of the DM (solDM) and water Ether extract 74 71−76 50 48−52 retention capacity (WRC), respectively. All corns and Ash 110 104−138 90 61−103 diets (n = 5) were analyzed using a modification of the Crude fiber 20 14−24 20 15−22 NFD 60 50−63 60 55−65 protocol described by Anguita et al. (2006). Samples 1 were treated following an in vitro procedure that simu- Eight diets with different corn hybrid varieties were obtained; using NIRS predictions, all diets were formulated in order to contain the same lates gastric pH. In short, 0.5 g of each sample was calculated nutrient composition. weighed into a 10 ml screw cap tube and incubated with 2 Provided per kg of feed: vitamin A (retinol acetate) 10,000 UI; vitamin 5 ml of 0.1M sodium phosphate buffer and 2 ml of 0.2M D (vitamin D3) (cholecalciferol) 4,800 UI; vitamin E/tocopherol) 45 mg; vitamin K3 (MNB, menadione nicotinamide bisulfate) 3 mg; vitamin B1 hydrochloric acid (pH = 2.5). Tubes were kept at 41°C (tiamin mononitrate) 3 mg; vitamin B2 (riboflavin) 9 mg; vitamin B6 (pir- for 2 h in a horizontal shaking water bath. The amount idoxin chlorohydrate) 4.5 mg; vitamin B12 (cyanocobalamine) 0.04 mg; of sample submitted to analysis was recorded (W0) as nicotinamide 51 mg; pantothenic acid (calcium D-pantothenate) 16.5 mg; biotin (D-(+)-biotin) 0.15 mg; folic acid 1.8 mg; choline chloride 350 mg; well as the weight of the screw cap tube plus the sample iron (iron sulfate monohydrate) 54 mg; zinc (Zn, zinc oxide) 66 mg; man- (W1). After incubation, the tubes were centrifuged for ganese (Mn, manganese oxide) 90 mg; iodine (I, calcium iodine anhydrate) 20 min at 2000 £ g. The supernatant was carefully 1.2 mg; selenium (Se, sodium selenate) 0.18 mg; copper (Cu, copper sulfate pentahydrate) 12 mg; ethoxiquin 4 mg; D,L-malic acid 60 mg; fumaric removed and tubes were kept upside down for 10 min to acid 75 mg; sepiolite 907 mg; vermiculite 2001 mg; colloidal silica 45 mg. ensure that the non-retained water was drained. Tubes 3 Quantum Blue 5G, AB Vista, Marlborough, UK; 5,000 FTU/g. with sample were weighed (W2) then dried in the oven at 100°C for 16 h to ensure the complete drying of the were then oven-dried and ground to pass through a insoluble residue, and then weighed again (W3). The sol- 0.5 mm screen in a grinder before the analyses. ubility of the DM was calculated as follows:   W1  W3 SolDM ¼ W0 Sample Analyses WRC determined after centrifugation is expressed as Corn Proximate and Physiochemical Analyses The gram of water retained by the total amount of sample fresh weight proximate and physiochemical characteris- incubated: tics of corn samples (10 per variety) included AMEn,   W2  W3 crude protein, starch, crude fat, crude fiber, neutral deter- WRCDM ¼ gent fiber, acid detergent fiber, protein solubility index W0 (PSI), vitreousness, total NSP, total arabinoxylan (AX), and soluble AX. These were determined with NIRS (Foss DS2500, Hilleroed, Denmark) using calibrations provided Statistical Analysis by AB Vista (Feed Quality Service). To validate the NIRS predicted data, 16 corn samples were also analyzed Corn sample and pen were considered the experimental for moisture, protein, starch and fat bywet chemistry, and unit for all variables. The nutrient composition of all 4  ET AL. MELO-DURAN Soluble-AX8 corns, broiler performance, solDM, WRC and coefficients <0.0001 (g/kg) 5.1ab 4.8ab 0.11 4.6b 3.2d 5.2a of digestibility were analyzed by one-way ANOVA to 2.7e 4.0c 2.1f identify treatment effects, using PROC GLM (SAS 9.4). Significantly different means were separated using Tukey adjust. Significance was declared at a probability P ≤ 0.05 Total-AX8 and tendencies were considered when P values were <0.0001 (g/kg) 41.3bcd 45.8ab 38.7cd 39.4cd 42.9bc 42.9bc 1.09 36.6d between >0.05 and <0.10. Spearman's non-parametric 49.6a correlation analysis was used to explore the associations between NIRS nutrient predictions, solDM, WRC and performance. A regression analysis between NIRS and Total-NSP7 <0.0001 (g/kg) wet chemistry of NSP and AX values, and correlation plot 60.1bcd 69.7ab 66.5bc 57.6cd 2.23 54.8d 67.5b 50.7d 77.8a between NIRS nutrient content predictions and perfor- mance, nutrient digestibility and in vitro corn and feed results were created using a linear model function and the corrplot package of R 3.6.1., respectively. Vitreou.6 Table 2. Proximate and physiochemical analysis (g/kg fresh weight) of corn hybrid samples measured by near-infrared spectroscopy (NIRS)1. <0.0001 62.1ab 58.6cd (%) 0.35 57.1d 57.2d 60.7b 63.2a 63.1a 58.9c RESULTS Corn and Diets Analyses <0.0001 PSI5 40.2bc 37.3cd 40.5bc (%) 0.78 33.9d 33.9d 42.7b 47.4a 29.4e The corn hybrid samples differed in nutritional com- position (Table 2). The average contents of the main components were: AMEn 3,501 kcal/kg; starch 677 g/ <0.0001 (g/kg) 35.9bcd ADF4 kg; crude protein 83.7 g/kg; crude fiber 21.4 g/kg; crude 37.5bc 37.9bc 33.7cd 38.1bc 1.31 31.0d 39.6b 46.1a fat 38.7 g/kg; total NSP 63.1 g/kg; total AX 42.2 g/kg; soluble AX 4.0 g/kg; ratio A:X 0.73; PSI 38.2%; and vit- reousness 60.2%. The non-starch polysaccharide content <0.0001 Standar error mean9a-fValues in the same column not sharing a common letter are significantly different (P < 0.05). (g/kg) NDF3 94.9ab 98.5ab 87.4bc 88.2bc 2.89 102.9a 81.1c 76.4c 81.6c in the corn samples is shown in Figure 1. Overall, the sum of components (crude protein + starch + crude fat + total NSP) came to approximately 863 g/kg of fresh weight, with the remaining presumably being Crude Fiber <0.0001 (g/kg) mainly moisture and sugars (mono-, di- and oligo- 21.5ab 21.6ab 21.7ab 21.4ab 0.78 20.6b 24.7a 24.6a 14.8c saccharides + fructans). The regression analyses between NIRS and wet chemistry values for moisture, protein, starch, fat, NSP and AX are shown in Figure 2. The proximate analyses of starter diets were in accor- <0.0001 (g/kg) Fat 0.83 43.1b 32.1d 49.7a 36.3c 36.7c 37.8c 37.2c 36.9c dance with the expected values for protein and ether extract. Minor differences in GE were observed, in a range from 4,075 to 4,224 kcal/kg. In the finisher diets, <0.0001 the CP showed higher variability (range 168−179 g/kg), Starch (g/kg) 687.7ab 676.5bc 672.5cd 679.4bc 676.1bc 668.0cd 3.21 660.3d 698.0a as the analyzed contents in all diets were lower than the calculated value. Diet GE showed less variability, with a range from 3,925 to 3,958 kcal/kg. <0.0001 Protein (g/kg) 85.6abc Data are means of 10 replicates from each corn hybrid. 90.3ab 80.4cd 80.9cd 82.6cd 82.7bc 1.94 73.7d 93.2a DM Solubility and WRC Results from the SolDM and WRC of corn and feed Apparent metabolizable energy for poultry. Vitreousness.6Non-starch polysaccharides. <0.0001 (kcal/kg) after in vitro incubation are shown in Table 3. The AMEn2 SolDM and WRC were influenced by corn variety (P < 3,599b 3,439d 3,432de 3,526c 3,390e 3,471d 3,457d 3,694a 10 0.05 and P < 0.001, respectively). Variety 2 had a higher WRC compared to varieties 1 and 5. The other varieties were intermediate. Variety 8 had the highest solDM fol- Trial Code Protein solubility index. Neutral detergent fiber. Variety 1 Variety 2 Variety 3 Variety 4 Variety 5 Variety 6 Variety 7 Variety 8 lowed by variety 1, with no statistically significant dif- Acid detergent fiber. SEM8 ferences between other varieties. In the starter diets, no differences were observed in WRC (P = 0.11). Differen- Arabinoxylans. ces in solDM of the starter diets were observed (P = 0.0001), as those formulated with varieties 1 and 8 Kws-2679yg Kws-16772 Kefrancos Kws-7569 Kws-7554 Kws-7661 Kws-7562 Kerubino had higher levels of solDM than the other diets. In the Varieties Variety P value finisher diets, significant differences were observed for 1 2 3 4 5 6 7 8 WRC (P < 0.0001) and solDM (P < 0.0001). Diets CORN NUTRIENT VARIABILITY IN BROILER DIETS 5 Figure 1. Soluble (A) and insoluble (B) non-starch polysaccharide and arabinoxylan (C) content in the corn samples. containing varieties 3, 4, 5, and 6 had the highest WRC, is described in Table 4. Overall broiler mortality was varieties 1 and 2 had intermediate results, and varieties 8.1% (data not shown), and no differences were observed 7 and 8 had the lowest values. The solubility of the DM between the experimental treatments (P > 0.05). in variety 8 finisher feed was higher compared to the Although no significant effects of feeding broiler chick- others except for the variety 1 diet. ens different corn varieties were observed for the whole period or the finisher phase (P > 0.10), a significant trend was observed in weight gain (P = 0.096). Birds fed Growth Performance varieties 2 and 6 had the highest BW gain, and those fed with varieties 1, 3, and 8 had the lowest BW gain. Over- The effect of diets based on the individual corn all feed intake was not influenced by corn variety (P > hybrids on broiler chicken BW and growth performance 0.10); however, it is interesting to highlight that birds 6  ET AL. MELO-DURAN Figure 2. Regression analysis between wet chemistry and NIRS predicted values for moisture, protein, starch, fat, NSP and AX using the 8 corn varieties. Table 3. Effect of corn hybrid samples and feeds on water retention capacity (WRC, g:g) and dry matter solubility (SolDM, g:g) at pH 2.51. Variety 1 Variety 2 Variety 3 Variety 4 Variety 5 Variety 6 Variety 7 Variety 8 SEM2 P value Corn WRC 1.23b 1.39a 1.32ab 1.26ab 1.25b 1.27ab 1.32ab 1.30ab 0.031 0.020 SolDM 0.116b 0.099c 0.100c 0.101c 0.103c 0.102c 0.101c 0.123a 0.0012 <0.0001 Feed d 0−21 WRC 2.17 2.16 2.18 2.23 2.24 2.19 2.20 2.11 0.029 0.110 SolDM 0.139a 0.130b 0.123b 0.129b 0.129b 0.126b 0.127b 0.139a 0.0021 0.0001 Feed d 21−42 WRC 1.88ab 1.90ab 1.92a 1.94a 2.00a 1.96a 1.77b 1.77b 0.032 <0.0001 SolDM 0.170ab 0.159c 0.164bc 0.165bc 0.156c 0.166bc 0.163bc 0.175a 0.0023 <0.0001 1 Data are means of 5 replicates per sample. 2 Standard error of the mean.3abcValues in the same row not sharing a common letter are significantly different (P < 0.05). CORN NUTRIENT VARIABILITY IN BROILER DIETS 7 Table 4. Effect of corn hybrid sample on body weight (BW), weight gain, feed intake and feed conversion ratio (FCR).1 Variety 1 Variety 2 Variety 3 Variety 4 Variety 5 Variety 6 Variety 7 Variety 8 SEM2 P value3 d 0−21 BW (g) 1,036ab 1,039ab 998ab 1,043a 1,018ab 1,011ab 993ab 949b 21 0.033 Weight gain (g/d) 994 997 956 1,001 976 966 960 907 21 0.056 Feed Intake (g/d) 1,253ab 1,233ab 1,216ab 1,286a 1,210ab 1,203ab 1,256ab 1,168b 23 0.021 FCR (g/g) 1.263 1.238 1.273 1.285 1.241 1.246 1.309 1.291 0.0183 0.065 d 21−42 BW (g) 3,028 3,140 3,016 3,070 3,042 3,095 3,068 2,975 38 0.100 Weight gain (g/d) 1,984 2,084 2,018 2,027 2,024 2,113 2,067 2,019 33 0.130 Feed Intake (g/d) 3,695 3,778 3,677 3,663 3,603 3,814 3,755 3,668 53 0.113 FCR (g/g) 1.866 1.816 1.825 1.807 1.782 1.807 1.818 1.819 0.0253 0.544 d 0−42 Weight gain (g/d) 2,978 3,081 2,974 3,028 3,001 3,080 3,028 2,926 40 0.096 Feed Intake (g/d) 4,949 5,011 4,892 4,949 4,813 5,016 5,011 4,836 64 0.147 FCR (g/g) 1.662 1.626 1.644 1.634 1.606 1.630 1.655 1.654 0.015 0.205 1 Data are means of 8 pens each with 10 birds in the starter and finisher periods. 2 Standard error of the mean. 3ab Values in the same row not sharing a common letter are significantly different (P < 0.05) Table 5. Effects of corn variety on total tract digestibility and intake of energy, dry matter (DM) and organic matter (OM) at d 21.1 Total tract apparent digestibility Intake of digestible nutrients DM OM Energy DM(g) OM(g) Energy (kcal) ab abc ab a a Variety 1 0.69 0.72 0.73 868 901 3,824a Variety 2 0.68ab 0.72abc 0.74a 844ab 899a 3,747ab Variety 3 0.67abc 0.70cd 0.71bc 820ab 858ab 3,678ab Variety 4 0.69ab 0.73ab 0.74a 896a 940a 3,979a Variety 5 0.67bc 0.70bcd 0.72ab 817ab 860ab 3,693ab Variety 6 0.68ab 0.72abc 0.73ab 821ab 870ab 3,679ab Variety 7 0.70a 0.73a 0.74a 862a 908a 3,777ab Variety 8 0.65c 0.69d 0.70c 764b 812b 3,404b SEM2 0.087 0.081 0.091 20.9 21.4 91.9 P value3 Variety 0.008 0.048 0.003 0.0004 0.001 0.001 1 Data were obtained by pooling 10 birds per pen, with 8 pens per treatment. 2 Standard error of the mean. 3abcd Values in the same column not sharing a common letter are significantly different (P < 0.05). fed varieties 5 and 8 tended to have low feed consump- correlations (P < 0.05) were observed among physio- tion for the entire period compared to the other diets. chemical corn values. For example, AMEn was posi- BW and FI were influenced by treatments in the starter tively correlated with PSI (r = 0.88) and fat (r = 0.98), period (P < 0.05). Birds fed the variety 8 diet had a and negatively correlated with soluble NSP (r = -0.83) lower BW and FI than those fed with the variety 4 diet, and soluble AX (r = -0.88). Protein was positively corre- while the other corns gave intermediate performances. lated with the total arabinose:xylose ratio (A:X) (r = 0.67) and vitreousness (r = 0.60, P = 0.054) and negatively correlated with starch (r = -0.62). Soluble Whole-Tract Apparent Digestibility and AX was negatively correlated with PSI (r = -0.88), fat Intake of Digestible Nutrients (r = -0.83) and AMEn (r = -0.88). In addition, significant correlations (P < 0.05) The effects of corn source on energy, DM and OM between physiochemical corn values and in vitro analy- whole-tract apparent digestibility and intake of digest- sis, 21-d performance and digestibility were observed. ible nutrients at d 21 are shown in Table 5. The variety DM solubilities of corn and starter feeds were negatively 8 diet had lower apparent digestibility of DM, OM and correlated with total NSP (r = -0.79 & -0.76, respec- energy compared to diets with varieties 1, 2, 4, 6 and 7 tively). DM solubility of the starter feed was negatively (P < 0.05). Intakes of digestible DM, OM and energy correlated with crude fiber content (r = -0.69), total AX were lower with variety 8 compared to varieties 1 and 4 (r = -0.88), phytic phosphorus (r = -0.90) and corn yield (764 g vs 868 and 896 g; 812g vs 901 and 940g; 3404 kcal (r = -0.81). DM solubility of finisher feeds was nega- vs 3824 and 3979 kcal, P < 0.05). Variety 7 gave a higher tively correlated with soluble NSP (r = -0.80). BW and intake of digestible DM and OM, and variety 2 gave a weight gain at d 21 were negatively correlated with corn higher intake of digestible OM than variety 8. protein content (r = -0.71 & -0.62, respectively), while BW at d 21 was positively correlated with corn starch Correlations content (r = 0.74). DM solubility of corn hybrids was negatively correlated with BW at d 42 (r = -0.57) and Figure 3 shows the correlation matrix and significance overall weight gain (r = -0.57). Corn harvest yield was for the relationships between all parameters. Significant positively correlated with phytic phosphorus (r = 0.83), 8  ET AL. MELO-DURAN Figure 3. Correlations between NIRS predicted nutrient content and broiler performance, nutrient digestibility and in vitro corn and feed results (WRC and solDM). The * (P < 0.05) symbol indicate a statistically significant correlation. The scale colors (Spearman’s r from -1 to +1) indicate whether the correlation is positive (blue) or negative (red) between factors. and DM (r = 0.79), OM (r = 0.79), and energy (r = 0.74) reported nutritional variation within one hybrid har- digestibility. vested in different regions of China with similar ranges of CP, starch and crude fiber (76−85 g/kg, 742 −757 g/kg and 25.9−26.5 g/kg, respectively). Although DISCUSSION the range in nutrients reported here was low, it revealed that the genetic background is a consistent source of Corn nutrient composition can be influenced by many nutrient variation, which is probably as important as factors, such as genetics and environmental conditions. the geographical factors. The results of the present study Genetic selection can influence several kernel character- also showed differences in total NSP (50.7−77.8 g/kg), istics and thus the high heritable parameters, including PSI (29.4−47.4%) and total AX (36.6−49.6 g/kg). Some kernel weight, volume, endosperm type, degree of dam- expected correlations were observed, such as a negative age, density and kernel breakage (Johnson and Rus- correlation between protein and starch, as previously sell, 1982; LeFord and Russell, 1985; Chen et al., 2016). described in other studies (Uribelarrea et al., 2004; In addition, the genetic background can also produce Masey O’Neill et al., 2012), while the positive correlation variation in the nutritional and anti-nutritional compo- between crude protein and A:X ratio might suggest a nents in corn (Reynolds et al., 2005). Potentially anti- relationship between AX structure and protein. A higher nutritional factors of corn, such as NSP, could interact A:X ratio indicates more arabinose single-sugar side with major nutrients, reducing their availability and chains on the xylan backbone, and has been related to consequently impairing bird digestibility and perfor- the solubility of AX (Rosicka-Kaczmarek et al., 2016). mance. Based on these considerations, the aim of the AX physicochemical characteristics, such as gelling present study was to test the effects of anti-nutritional capability, depends on many characteristics, including factors and nutrient availability of corn hybrids on molecular weight, side-chain ferulic acid concentration broiler performance and nutrient digestibility. and A:X ratio (Izydorczyk and Biliaderis, 1995). AX- In the current study, the proximate analysis by NIRS protein associations can also play an important role, of 8 corn varieties showed a range in nutrient content of especially the nature and quantity of the protein fraction CP (73−93 g/kg), starch (660−698 g/kg) and crude (Mendez-Encinas et al., 2019). Furthermore, the nega- fiber (14.8−24.7 g/kg). Masey O’Neill et al. (2012) tive correlation between PSI and soluble NSP and AX CORN NUTRIENT VARIABILITY IN BROILER DIETS 9 suggests that protein solubility is affected by these solu- performances when they were fed eight different diets ble fiber components. Thus, the availability for digestion formulated with NIRS data to be iso-nitrogenous and of corn protein could depend on the interaction with iso-energetic. NIRS could also be used to make decisions other corn kernel components, including fiber AX. In about adding carbohydrase, such as xylanase, in order addition, the results from this study showed a positive to prevent the negative effects of AX, the major compo- correlation between protein content and vitreousness, nent of NSP. with varieties 8 and 1 having the highest and lowest vit- In conclusion, the results from the present study show reousness and protein contents, respectively. Genotype, that corn genetic background influenced nutrient digest- maturity at harvest and growing environment are fac- ibility and growth performance of broiler chickens fed tors that affect corn vitreousness (Corona et al., 2006; diets based on these corns. The content and nature of Masoero et al., 2011). Therefore, cereal endosperm vitre- the NSP are shown as two of the main factors affecting ousness or hardness can influence the particle size distri- the solubility and availability of nutrients in corn, with bution of subsequent diets (Lentle et al., 2006; possible negative effects on the performance of broiler Amerah et al., 2008) and thus the nutritional value of chickens. The NIRS prediction for NSP and AX could corn for chickens (Kaczmarek et al., 2013). In addition, be useful for improving feed formulation or the efficiency corn diet particle size distributions have been related to of any fiber utilization enhancer in corn-based diets. the development of the gastrointestinal tract, perfor- mance and digestibility in broiler chickens (Jacobs et al., 2010). However, a negative correlation between corn ACKNOWLEDGMENTS solDM and d 42 BW and weight gain was observed in The authors acknowledge the financial support of the this study. This suggests that soluble corn compounds SENESCYT (Secretary for Higher Education, Science, such as soluble NSP could decrease nutrient availability Technology and Innovation of Ecuador), AB Vista Lab- and consequently performance. This would explain why oratory Services (UK) for their assistance with the NIRS birds fed the variety 8 diet, which has the highest analysis and KWS Semillas Iberica, S.L.U. (Spain) for solDM, had a lower BW and FI at d 21 and a decrease in the provision of corn varieties. whole tract DM, OM and energy digestibility compared to those fed the other corns. Corn variety 8 also showed the highest CP content. In this regard, zeins, or corn DISCLOSURES prolamins, and starch have been related to endosperm texture (Lee et al., 2006). Thus, the amount and packing On behalf of the authors, we wish to confirm that all of this protein within the kernel could affect physical authors have read and approved and that there are no properties such as vitreousness (Paulsen et al., 2019) other persons who satisfied the criteria for authorship and access to digestive enzymes (Liu et al., 2016; but are not listed. Moreover, we confirm that all the Pan et al., 2017). Overall, these factors could explain authors listed in the manuscript have been approved part of the effects observed in the current study on one to each other. broiler performance and digestibility when the animals We also wish to confirm that there are known conflicts were fed different corn hybrids. of interest associated with this publication and that Corn solubility could be affected by the nature of the there has not been financial support for this work that zeins, which are divided into several subclasses with dif- could have influenced its outcome. ferent molecular weights and solubilities (Esen, 1986; We also assume that although the corresponding Lee et al., 2006). It is thought that the matrix in the author is the sole contact with the Editor in chief, he is endosperm is associated with the nature of the protein/ responsible for communicating with the other co-author starch interactions, and this may influence the nutri- about the submission and the progress of the manuscript tional value of the cereal (Kaczmarek et al., 2013), and revision and final approval of proofs. thus affect the protein and amino acid digestibility (Kaczmarek et al., 2011). It is plausible that the nega- tive correlation observed between protein content and d REFERENCES 21 BW and weight gain could be related to the profile of AEMET. 2020. Agencia Estatal de Metereología (AEMET). Accessed corn zeins, given that the high-protein variety 8 had a Jan. 2020. http://www.aemet.es/ca/lineas_de_interes/datos_y_ negative effect on nutrient digestibility and bird perfor- estadistica. Amerah, A. M., V. Ravindran, R. G. Lentle, and D. G. Thomas. 2008. mance. It is clear that an interaction between nutrients Influence of feed particle size on the performance, energy utiliza- and other compounds influences the physicochemical tion, digestive tract development, and digesta parameters of characteristics and nutrient value of corn and, conse- broiler starters fed wheat- and corn-based diets. Poult. Sci. quently, their behavior in feed processing and digestion 87:2320–2328. Anguita, M., J. Gasa, S. M. Martín-Or ue, and J. F. Perez. 2006. processes. Study of the effect of technological processes on starch hydrolysis, Despite the small number of samples in this study, non-starch polysaccharides solubilization and physicochemical regression results between NSP and AX, predicted by properties of different ingredients using a two-step in vitro system. Anim. Feed Sci. Technol. 129:99–115. NIRS, and the wet chemistry values showed that NIRS AOAC. 2005. Official Methods of Analysis of AOAC International. is an effective nutrient prediction tool. This is supported W. Horwitz and G. W. Latimer, eds. AOAC International, Rock- by there being no differences in the 42-d bird ville, MD. 10  ET AL. MELO-DURAN Chen, J., L. Zhang, S. Liu, Z. Li, R. Huang, Y. Li, H. Cheng, X. Li, Lentle, R. G., V. Ravindran, G. Ravindran, and D. V. Thomas. 2006. B. Zhou, S. Wu, W. Chen, J. Wu, and J. Ding. 2016. The genetic Influence of feed particle size on the efficiency of broiler chickens fed basis of natural variation in kernel size and related traits using a wheat-based diets. J. Poult. Sci. 43:135–142. four-way cross population in maize (X Zhang, Ed.). PLoS One 11: Liu, S. Y., H. H. Truong, A. Khoddami, A. F. Moss, P. C. Thomson, e0153428. T. H. Roberts, and P. Selle. 2016. Comparative performance of Corona, L., F. N. Owens, and R. A. Zinn. 2006. Impact of corn vitre- broiler chickens offered ten equivalent diets based on three grain ousness and processing on site and extent of digestion by feedlot sorghum varieties as determined by response surface mixture cattle. J. Anim. Sci. 84:3020–3031. design. Anim. Feed Sci. Technol. 218:70–83. Cowieson, A. J. 2005. Factors that affect the nutritional value of Masey O’Neill, H. V., N. Liu, J. P. Wang, A. Diallo, and S. Hill. 2012. maize for broilers. Anim. Feed Sci. Technol. 119:293–305. Effect of Xylanase on performance and apparent metabolisable Dei, H. K. 2017. Assessment of maize (Zea mays) as feed resource for energy in starter broilers fed diets containing one maize variety poultry. In Poultry Science. Manafi, M., ed. Intech Open, London, harvested in different regions of China. Asian-Australasian J. United Kingdom. Anim. Sci. 25:515–523. Eeckhout, W., and M. De Paepe. 1994. Total phosphorus, phytate- Masoero, F., A. Gallo, C. Zanfi, G. Giuberti, and M. Spanghero. 2011. phosphorus and phytase activity in plant feedstuffs. Anim. Feed Effect of nitrogen fertilization on chemical composition and rumen Sci. Technol. 47:19–29. fermentation of different parts of plants of three corn hybrids. Englyst, H. 1989. Classification and measurement of plant polysac- Anim. Feed Sci. Technol. 164:207–216. charides. Anim. Feed Sci. Technol. 23:27–42. Mendez-Encinas, M. A., E. Carvajal-Millan, A. Rasc on-Chu, Englyst, H. N., M. E. Quigley, and G. J. Hudson. 1994. Determination Y. L. Lopez-Franco, and J. Lizardi-Mendoza. 2019. Arabinoxilanos y of dietary fibre as non-starch polysaccharides with gas−liquid la Relaci on de la Fracci on Proteica Remanente con la Capacidad Gel- chromatographic, high-performance liquid chromatographic or ificante del Polisacarido. Acta Univ. 29:1–19. spectrophotometric measurement of constituent sugars. Analyst Naderinejad, S., F. Zaefarian, M. R. Abdollahi, A. Hassanabadi, 119:1497–1509. H. Kermanshahi, and V. Ravindran. 2016. Influence of feed form Esen, A. 1986. Separation of alcohol-soluble proteins (Zeins) from and particle size on performance, nutrient utilisation, and gastro- Maize into three fractions by differential solubility. Plant Physiol intestinal tract development and morphometry in broiler starters 80:623–627. fed maize-based diets. Anim. Feed Sci. Technol. 215:92–104. European Parliament. 2010. Directive 2010/63/EU of the European Pan, L., Q. H. Shang, Y. Wu, X. K. Ma, S. F. Long, L. Liu, D. F. Li, and Parliament and of the Council of 22 of September 2010 on the pro- X. S. Piao. 2017. Concentration of digestible and metabolizable tection of animals used for scientific purposes. 33−79. energy, standardized ileal digestibility, and growth performance of Gehring, C. K., A. J. Cowieson, M. R. Bedford, and pigs fed diets containing sorghum produced in the United States or W. A. Dozier. 2013. Identifying variation in the nutritional value corn produced in China1. J. Anim. Sci. 95:4880–4892. of corn based on chemical kernel characteristics. Worlds. Poult. Paulsen M. R., M. Singh, and V. Singh 2019. Measurement and main- Sci. J. 69:299–312. tenance of corn quality. Pages 165−211 in Corn: Chemistry and Izydorczyk, M. S., and C. G. Biliaderis. 1995. Cereal arabinoxylans: Technology. Woodhead Publishing and AACC International advances in structure and physicochemical properties. Carbohydr. Press, Duxford, United Kingdom. Polym. 28:33–48. Prado, S. A., C. G. L opez, M. L. Senior, and L. Borr as. 2014. The Jacobs, C. M., P. L. Utterback, and C. M. Parsons. 2010. Effects of genetic architecture of maize (Zea mays L.) kernel weight determi- corn particle size on growth performance and nutrient utilization nation. G3 Genes, Genomes, Genet 4:1611–1621. in young chicks. Poult. Sci. 89:539–544. Pujol, S., A. M. Perez-Vendrell, and D. Torrallardona. 2007. Evalua- Johnson, D. Q., and W. A. Russell. 1982. Genetic variability and rela- tion of prediction of barley digestible nutrient content with near- tionships of physical grain-quality traits in the BSSS population of infrared reflectance spectroscopy (NIRS). Livest. Sci. 109:189–192. maize 1. Crop Sci 22:805–809. Rahman, A., I. Bayram, S. Khanum, and S. Ullah. 2015. Pakistan Kaczmarek, S. A., A. J. Cowieson, D. J ozefiak, and Journal of Life and Social Sciences use and calibration of near A. Rutkowski. 2013. Effect of maize endosperm hardness, drying infrared reflectance spectroscopy in feed. J. Life Soc. Sci. 13:1–7. temperature and microbial enzyme supplementation on the perfor- Reynolds, T. L., M. A. Nemeth, K. C. Glenn, W. P. Ridley, and mance of broiler chickens. Anim. Prod. Sci. 54:956–965. J. D. Astwood. 2005. Natural variability of metabolites in maize Kaczmarek, S., A. Cowieson, A. Rutkowski, F. Management, B. grain: differences due to genetic background. J. Agric. Food Chem. Road, and B. Park. 2011. Effect of maize endosperm hardness, dry- 53:10061–10067. ing temperature and microbial phytase and xylanese supplementa- Rosicka-Kaczmarek, J., A. Komisarczyk, E. Nebesny, and tion on amino acid availability and phytate phosphorus hydrolysis B. Makowski. 2016. The influence of arabinoxylans on the quality of in broilers. Proceedings of the 18th European Symposium on Poul- grain industry products. Eur. Food Res. Technol 242:295–303. try Nutrition: 50-52. (Short Communication). Seebauer, J. R., G. W. Singletary, P. M. Krumpelman, M. L. Ruffo, Larbier, M., B. Leclercq, and J. Wiseman. 1994. Nutrition and Feed- and F. E. Below. 2010. Relationship of source and sink in determin- ing of Poultry. M. Larbier and B. Leclercq, eds. Nottingham Uni- ing kernel composition of maize. J. Exp. Bot. 61:511–519. versity Press, Nottingham, UK. Summers, J. D. 2001. Maize: factors affecting its digestibility and var- Lee, K.-M., S. R. Bean, S. Alavi, T. J. Herrman, and iability in its feeding value. Pages 109−124 in Enzymes in Farm R. D. Waniska. 2006. Physical and biochemical properties of maize Animal Nutrition. M. R. Bedford, and G. G. Partridge, eds CABI hardness and extrudates of selected hybrids. J. Agric. Food Chem. Publishing, Wallingford, United Kingdom. 54:4260–4269. Uribelarrea, M., F. E. Below, and S. P. Moose. 2004. Grain composi- LeFord, D. R., and W. A. Russell. 1985. Evaluation of physical grain tion and productivity of maize hybrids derived from the Illinois quality in the BS17 and BSI (HS)CI synthetics of maize 1. Crop protein strains in response to variable nitrogen supply. Crop Sci. Sci. 25:471–476. 44:1593–1600.