Preparation of chelated TM
Bonzachicken is a supplement manufactured on the basis of self-assembly method and according to the patent US8288587B2. The structure of this supplement is highly stable because of the shear-like enclosure of central ions by polymerized chelating agents. The stability of this chelated structure at various pH levels is presented in Table 1. As can be seen, there is little variation in the ppm of different elements at alkaline and acidic pH levels as compared to total, which reveals the stability of the chelated structure of Bonzachicken and its resistance to environmental pH alterations.
Broiler chickens and dietary treatments
A total of 625 male broiler chickens (Ross 308) at 0 day old (44 ± 1.15 g) were allocated to 5 dietary treatments, in a completely randomized design. Each of the 5 treatments had 5 replicates and 25 birds per replicate. The composition of starter (d 0 to 10), grower (d 10 to 24), and ﬁnisher (d 24 to 42) diets is given in Table 2. Experimental treatments were: (1) no TM (NTM), (2) commercially recommended levels of inorganic TM (ITM; 80 mg Fe as ferrous sulfate, 92 mg Zn as zinc sulphate, 100 mg Mn as manganese sulfate, 16 mg Cu as copper sulfate, 0.3 mg Se as sodium selenite, 1.2 mg I as potassium iodide, 0.1 mg Cr as potassium dichromate, per kilogram of diet;); (3) very low CTM (CTM25; match to 25% of the ITM); (4) Low CTM (CTM50; match to 50% of the ITM); (5) high CTM (CTM100; equivalent to ITM). For the ITM100 treatment, the basal diet supplemented with 2.5 g/kg inorganic TM premix. For the CTM25, CTM50, and CTM100 treatments, the basal diet supplemented with 0.5, 1, and 2 g/kg of Bonzachicken supplement, respectively, to meet the defined TM levels for each treatment. The TM premixes were added in place of the building sand that is included in the NTM diet as inert filler. Dietary treatments varied in TM composition are shown in Table 3. Finisher diets were contained 0.5% titanium dioxide (TiO2) as an indigestible marker to estimate the apparent mineral digestibility in the ileum.
All birds had free access to water (1 hanging drinker per pen) and feed (mash form). Drinking water was analyzed for minerals using an inductively coupled plasma– atomic emission spectroscopy (ICP-AES; Optima 7000 DV, Perkin Elmer, Waltham, MA) prior to the experiment. The concentrations of Ca, Mn, and Fe in the drinking water were 5.3, 0.06, and 0.03 mg/L, respectively, whereas the concentrations of P and other TM were undetectable in the water.
All experimental groups were housed in ﬂoor pens (length 175 cm× width 170 cm) using litter top dressed with 5 cm of clean wood shavings in an environmentally controlled house. The room temperature was set at 34°C on the day of arrival, and then reduced by 0.40°C per day until 24°C, where it remained for the rest of the trial. The environmental relative humidity was maintained at 50–65% by periodical spraying the walkways with water and adjusting the humidifiers. The lighting program used was 24L∶0D from d 0 to 3 and 23L∶1D for the remainder of the experiment.
Pen body weight (BW) and feed intake were recorded at placement, 10, 24, and 42 d of age for calculation of average daily gain (ADG) and average daily feed intake (ADFI) per bird for each replicate pen. Incidences of mortality were recorded daily in order to determine the mortality rate. With the body weight of any deceased or culled chickens included, total ADG and total ADFI for each pen were used for calculating the mortality-adjusted feed conversion ratio (FCR) during each feeding period. European performance index (EPI) per treatment was also estimated using the formula described by Ghasemi and Nari .
On d 35, 2 chickens per replicate pen, each with a BW close to the average BW of each pen, were euthanized by cervical dislocation, and the digesta samples from the ileum (from Meckel's diverticulum up to 5 cm proximal to the ileocecal junction) were gently flushed into plastic tubes. The ileal digesta of two birds in a replicate were pooled, after which a representative sample was freeze-dried and ground to pass a 1 mm screen prior to chemical analysis.
At the end of the experiment (42 d), after overnight fasting, 2 birds per replicate pen were selected according to the average BW of the pen. Blood samples were collected from the brachial vein of each bird and then centrifuged (2500 × g, 15 min) at 4 °C to obtain serum samples for blood chemical analysis. The above 2 birds from each cage were then euthanized by cervical dislocation followed blood sampling; both right and left tibia of each bird was excised and cleaned of adhering tissues. Right tibia was measured for tibia weight, length, and diameter, whereas left tibia kept frozen at −20 °C until analysis for tibia ash, calcium, and phosphorous content.
Chemical analysis and mineral digestibility
Duplicate samples of basal diets were analyzed for crude protein (N × 6.25; method 994.13), crude fat (method 920.39), and crude fiber (method 978.10) according to AOAC . For mineral analysis, all diets and ileal digesta samples were milled to pass through 0.5 mm mesh sieve before analysis. Samples were subsequently analyzed for calcium (Ca), phosphorus (P), magnesium (Mg), potassium (K), sodium (Na), Fe, Zn, Mn, Cu, and TiO2. Mineral content in the diet and ileal digesta samples was determined using an ICP-AES  following digestion in concentrated HNO3. The total Se, I and Cr concentrations in the diets were also determined using an inductively coupled plasma mass spectrometry (ICP-MS) technique. Titanium dioxide analysis was done using the spectrophotometric method described by Short et al. . The apparent ileal digestibility (day 35) of minerals was calculated using the following formula:
Apparent ileal mineral digestibility = 1 − [(mineral/ TiO2) digesta/ (mineral/ TiO2) diet]
Tibia bone parameters
The weight and length of the right tibia of each bird were measured. Tibia diameters were also measured at the widest and narrowest points using a digital caliper in mm to 2 decimals, and then averaged. The left tibia samples were crushed and defatted with petroleum ether for 24 h using Soxhlet apparatus, and dried in the oven at 100 °C for 24 h. Dried bone samples were then burned (24 h) into a muffle furnace preheated to 600 °C for determination of ash percentage (dry, fat-free basis). The ash from tibia samples was solubilized with a mixture of nitric and perchloric acid, and the contents of minerals were determined by the same methods as those applied for the samples from diet and excreta.
Serum samples were used for the measurement of some metabolites, total antioxidant capacity (TAC), antioxidant enzyme activity, as well as MDA content. The serum concentrations of glucose, triglyceride, cholesterol, total protein, albumin, and uric acid were determined using a corresponding assay kit (Pars Azmoon Company, Tehran, Iran) and an automatic biochemical analyzer (Clima, Ral. Co, Spain). The activities of GSH-Px, SOD, and CAT were determined colorimetrically (enzymatically) and serum TAC and MDA contents were measured by assay kits for TAC (Randox Laboratories Ltd, Crumlin, UK), GSH-Px, SOD, CAT, MDA (Cayman Chemical Co., Ann Arbor, MI, USA).
All the data were statistically analyzed as a completely randomized design using GLM procedures of SAS 9.4 (SAS Institute Inc., Cary NC) with a pen as an experimental unit. All percentage data were tested for normality by employing the UNIVARIATE procedure of SAS, then transformed to arcsine values before analysis if normality was not met. Mean separation was conducted by the Tukey's post-hoc analysis and Bonferroni correction with differences deemed significant at P < 0.05.