The effect of calcium source and particle size on the production performance and bone quality of laying hens

Transcription

1 The effect of calcium source and particle size on the production performance and bone quality of laying hens J. L. Saunders-Blades,* J. L. MacIsaac, D. R. Korver,* and D. M. Anderson 1 * Department of Agricultural, Food, and Nutritional Science, University of Alberta, Edmonton, Canada T6G 2P5; Atlantic Poultry Research Institute, Truro, Nova Scotia, Canada B2N 5E3; and Department of Plant and Animal Sciences, Nova Scotia Agricultural College, Truro, Nova Scotia, Canada B2N 5E3 ABSTRACT The efficacy of 3 local limestone sources as potential Ca sources for laying hens was studied. Limestone sources were assessed for in vitro solubility. Four Ca sources (control, A, B, or C) and 2 particle size combinations (ground, 100% ground; or mixed, 67% ground + 33% large particle) was used. The control consisted of a commercial ground limestone for the ground Ca source and oyster shell of the large particle Ca source. DeKalb laying hens were randomly placed in 32 battery cage units (n = 12/cage). At 19 wk of age, hens received 1 of 8 Ca source and particle size treatments (4 replicates/treatment) until 74 wk. Egg production, feed consumption, BW, and egg quality were measured throughout. Bone mineral density (by quantitative computed tomography), breaking strength, ash, and Ca were assessed at the end of lay. In vitro solubility was dependent upon Ca source and particle size (P < ) with oyster shell generally having a greater in vitro solubility than test limestone sources A, B, and C of similar particle sizes. Feed consumption (100 to 117 g/bird per day), BW (1,500 to 1,800 g), egg production (92% peak egg production), egg weight (55 to 67 g), and egg specific gravity (1.090 to 1.078) did not differ among hens fed the different Ca sources (P > 0.05). Tibia bone mineral density indicated the oyster shell treatment had a lower trabecular density than Ca source A; however, all other Ca sources had similar bone mineral density for all measures (P > 0.05). Hens fed the mixed Ca particle treatments consistently had greater feed consumption from 27 to 70 wk than those fed the 100% ground Ca source (P < 0.05). Bone mineralization was enhanced in hens that received the mixed Ca particle treatments (P < 0.05). Overall, the results of this study indicate that the local limestone sources A, B, and C would be suitable alternatives to current commercial sources of Ca for laying hens. In addition, large particle Ca did improve bone quality. Key words: laying hen, calcium source, particle size, production performance, bone quality 2009 Poultry Science 88 : doi: /ps INTRODUCTION Laying hens have a high demand for Ca, especially during peak egg production when Ca output is at its greatest. Calcium metabolism is also under strain in the later stages of egg production, when hens have a decrease in Ca absorption efficiency (al-batshan et al., 1994). Egg producers primarily use 2 supplemental sources of dietary Ca, oyster shell or limestone. Producers have been using oyster shell for more than 100 yr (Roland and Bryant, 1999), in which time it has become a proven source of Ca to laying hens for maintaining good quality eggshells. Oyster shell and limestone 2009 Poultry Science Association Inc. Received July 7, Accepted October 21, Corresponding author: danderson@nsac.ca both provide Ca in the form of Ca carbonate, and each contain about 38% Ca; however, limestone costs considerably less than oyster shell. The value of both Ca sources to the hen has been the focus of many studies. In a review Roland (1986) concluded that about half of the studies comparing ground limestone to large particle oyster shell found that oyster shell resulted in better eggshell quality, whereas the rest found that eggshell quality was not affected. When comparing the 2 sources at similar particle sizes, the majority of researchers concluded that oyster shell and limestone were of equal value for eggshell quality (Roland, 1986). However, Guinotte and Nys (1991) reported superior egg and shell weights of hens fed particulate limestone than those fed oyster shell. Roland and Bryant (1999) indicated that producers are still not convinced of equality between limestone and oyster shell of similar particle size and continue to use oyster shell. 338

2 The improved eggshell quality from feeding oyster shell may be more a factor of particle size of the Ca source, rather than the Ca source per se. Scott et al. (1971) speculated that the larger particles remain in the upper digestive tract (crop and gizzard) for a longer period of time than the ground Ca sources, resulting in Ca being available to the hen for a longer period of time. A large particle Ca may therefore be beneficial to the hen during the 8 to 9 h dark period when feed is not consumed, but Ca requirements are high due to eggshell formation (Etches, 1987). Roland (1986) concluded that large particle size has no effect on shell quality when Ca levels in the diet are adequate. Poor eggshell quality still results in a significant number of eggs being unsaleable, even with adequate dietary Ca. Calcium nutrition also plays a significant role in bone quality of the laying hen. Poor bone quality in laying hens can lead to many problems, which include broken or weak bones, osteoporosis, economic losses, and difficulties at processing plants that can result in bone fragments in meat products (Whitehead and Fleming, 2000; Webster, 2004; Julian, 2005). Large particle Ca has been shown to have a positive effect on layer bone quality (Guinotte and Nys, 1991; Rennie et al., 1997; Fleming et al., 1998b). The composition of different Ca sources can vary (Reid and Weber, 1976) and may be a result of the location in which it is mined. Sources may vary in the amount of Ca and the presence of other nutrients, which can affect the utilization of the Ca source by the laying hen (Reid and Weber, 1976). Therefore the objective of this study was to examine the effects of 3 potential new locally available Ca sources and the effect of particle size on laying hen production performance, eggshell quality, and bone quality. CALCIUM SOURCE AND PARTICLE SIZE 339 MATERIALS AND METHODS Ca Sources: Mineral Composition, Particle Size Distribution, and In Vitro Solubility A commercially available ground limestone (CGL) source was used in the 100% ground control treatment, whereas a mixture of the CGL and oyster shell was used in the same proportion as the test limestone sources for the mixed particle size treatment. The test limestone sources were identified as A, a white limestone; B, a brown limestone; and C, a gray limestone. The experimental limestone sources were all locally mined (within Nova Scotia, Canada). Mineral analysis of each Ca source was performed with samples prepared using the AOAC dry ash method (Association of Official Analytical Chemists, 1990) and analyzed using inductively coupled plasma/atomic emission spectrometry (Jarrell Ash Model 9000, Thermo Elemental, Franklin, MA; Table 1). Purity of each Ca source was based on a 40% Ca in CaCO 3 (Levenson and Bockman, 1994). Individual particle sizes of each large particle calcium source (oyster shell, A, B, and C) were obtained through sifting the source using standard sieve sizes. Sieve sizes used were as follows, from small to large screen sizes: no.40 (0.425 mm), no. 35 (0.50 mm), no. 25 (0.71 mm), no. 18 (1.00 mm), no. 14 (1.41 mm), no. 10 (2.00 mm), no. 7(2.83 mm), and no. 5(4.00 mm). A sample of the calcium source was placed on the top screen and allowed to shake on a Model RX-24 Portable Sieve Shaker (W. S. Tyler Company of Canada Ltd., St. Catharines, Ontario, Canada) for 5 min, after which particle sizes were collected from each screen. The percentage of particles within each particle size range were calculated by dividing the weight of the particles col- Table 1. Mineral composition of Ca sources Source of Ca Item Oyster shell Commercial ground limestone A limestone 1 B limestone 2 C limestone 3 Purity Mineral 5 Ca (%) Total P (%) Mg (%) K (%) Na (%) Mn (mg/kg) 643 2, Zn (mg/kg) Fe (mg/kg) 3,710 1, ,284 2,635 B (mg/kg) Cu (mg/kg) Test limestone source A, a white limestone. 2 Test limestone source B, a brown limestone. 3 Test limestone source C, a gray limestone. 4 Purity of Ca source based on Ca content of Ca carbonate (40.04% Ca). 5 Mineral content reported on a DM basis.

3 340 SAUNDERS-BLADES ET AL. Table 2. Composition of experimental diets fed to laying hens from 19 to 74 wk of age Timing of diet changes Ingredient, % of diet (as fed) Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Corn Wheat Soybean meal Tallow Poultry fat Ca source Dicalcium phosphate Iodized salt dl-methionine Vitamin-mineral premix Total Ca sources consisted of oyster shell, commercial ground limestone, ground and large particle limestone A, ground and large particle limestone B, and ground and large particle limestone C. 2 Vitamin and mineral premix supplied the following per kilogram of feed: vitamin A (8,000 IU); vitamin D 3 (2,500 IU); vitamin E (20 IU); vitamin K (2.97 mg); riboflavin (7.6 mg); dl-ca-pantothenate (7.2 mg); vitamin B 12, (0.012 mg); niacin (30.69 mg); folic acid (0.66 mg); choline (640 mg); biotin (0.16 mg); pyridoxine, (3.96 mg); thiamine (1.94 mg); dl-methionine (500 mg); manganese (70 mg); zinc (80 mg); copper (25 mg); selenium (0.15 g); ethoxyquin (0.05 g); wheat middlings (2.1 g). lected at the respective screen size by the total weight of the sample of the calcium source. The ground version (<0.42 mm) of each source (CGL, A, B, and C) and each particle size of the large particle calcium sources was studied for in vitro solubility. The percentage of weight loss method was used to measure in vitro solubility of the Ca sources as described by Zhang and Coon (1997a). Five replications of each Ca source and particle size combination were performed. Briefly, a 2-g sample was allowed to react with 200 ml of 0.2 N HCl for 10 min in a 42 C oscillating (80 Hz) water bath. After 10 min, the calcium source acid mixture was removed from the water bath and filtered through a Whatman no. 41 ashless filter paper (Whatman International Ltd., Maidstone, UK). The residues on the filter paper were dried in an oven at 70 C until a consistent weight was obtained. Experimental Conditions This research was reviewed and approved by the Nova Scotia Agricultural College (NSAC) Animal Care Committee in accordance with the guidelines set by the Canadian Council on Animal Care (1993). Pullets obtained for this research had been reared under similar conditions. Three hundred eighty-four 19-wkold DeKalb laying hens were randomly assigned to 1 of 32 battery cage units in the lower 2 tiers of a 3-tier stacked battery cage system, such that each unit consisted of 12 hens. One unit consisted of 2 adjacent battery cages (6 birds/cage; 355 cm 2 /hen), with birds in both cages consuming feed from the same feeder. Hens were provided with 15 h of light from 0400 to 1900 h, and feed (mash form) and water (from a nipple drinker) were provided to the hens for ad libitum consumption. Hens were fed to 74 wk of age. Hens that died during the course of the trial were sent to a veterinary pathologist for postmortem examination. Experimental Design and Dietary Treatments A 4 2 factorial arrangement of treatments in a randomized block design, using 4 Ca sources (a control and sources A, B, and C), fed at 2 particle size levels [100% ground (ground) or a mixture of 67% ground and 33% large particle Ca (mixed), ranging in particle size from 0.5 to >4 mm] was established. Diets were balanced to meet or exceed the nutrient requirements for laying hens (National Research Council, 1994) and formulated to be isocaloric and isonitrogenous with only the Ca source in the diets differing (Tables 2 and 3). The feeding period included 5 phases with diet changes at 26, 46, 50, and 70 wk of age, each phase representing an increase in dietary Ca to reflect changing Ca requirements as hens aged (Table 3). Diets were formulated based on a feed consumption of 90 g/hen/d feed intake in phase 1; 100 g/hen/d feed intake in phase 2; and 110 g/hen/d feed intake in phases 3, 4, and 5. Feed samples were analyzed for Ca and phosphorus (Table 3). Calcium was determined by atomic absorption spectrometry on a Varian Spectra AA-20 (Varian Canada Inc., Mississauga, Ontario, Canada). A 1-g sample of feed (as fed) was ashed overnight in a muffle furnace (600ºC) and digested with concentrated hydrochloric acid as described by Clunies et al. (1992). Solutions were then diluted with deionized distilled water to volume in a 100-mL volumetric flask. A 1-mL aliquot was further diluted to 10 ml in a volumetric flask with a 1% solution of lanthanum oxide (Fisher Scientific, Nepean, Ontario, Canada) prepared by method of the Association of Official Analytical Chemists (1995).

4 CALCIUM SOURCE AND PARTICLE SIZE 341 Standard solutions were prepared from a 1,000 mg/ kg Ca reference solution (Fisher Scientific). Phosphorus was analyzed using inductively coupled plasma/atomic emission spectrometry on samples prepared using the same method as described for Ca. Feed Consumption, BW, and Production Measurements A measured amount of feed was added on a daily basis with feed consumption determined every 28 d. Individual BW was obtained at 19 wk of age and at the end of each subsequent 28 d interval, coinciding with feed intake measurements. The number of marketable and unmarketable (cracked or soft-shelled) eggs was recorded on a daily basis. Eight eggs from each unit were collected during the last 2 d of each 28 d interval, weighed, and assessed for quality by specific gravity (SG) shortly after collection that same day. Egg SG was measured using the floatation method as described by Hamilton (1982). End-of-Lay Bone Quality At 74 wk of age, 16 reproductively active hens from each treatment group (n = 128) were killed by cervi- Table 3. Calculated and analyzed nutrient content of laying hen diets used from 19 to 74 wk of age Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Timing of diet changes Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Calculated composition 1 ME (kcal/kg) 2,880 2,880 2,880 2,880 2,880 CP (%) Lys (%) Met (%) Met + Cys (%) Ca (%) Total P (%) Available P (%) Analyzed composition (as fed) n Ca (%) Control 3 Ground Mix A 6 Ground Mix B 7 Ground Mix C 8 Ground Mix Total P (%) Control Ground Mix A Ground Mix B Ground Mix C Ground Mix Calculated nutrient intake was based on 90 g/hen per day of feed intake in phase 1, 100 g/hen per day of feed intake in phase 2, and 110 g/hen per day of feed intake in phases 3, 4, and 5. 2 n = number of feed batches made during each phase for each diet. 3 A commercially used ground limestone (CGL) source was used in the 100% ground control treatment, whereas a mixture of the CGL and oyster shell were used. 4 Ground Ca source diet. 5 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground). 6 Test limestone source A, a white limestone. 7 Test limestone source B, a brown limestone. 8 Test limestone source C, a gray limestone.

5 342 cal dislocation and right tibias dissected out for subsequent analysis. Tibias were frozen with flesh intact at 20ºC until further analysis. Before analysis, bones were cleaned of all tissue and dried in a forced air oven at 100ºC for 24 h. Bone mineral density (BMD) and cross-sectional area analysis were performed using quantitative computed tomography using a Stratec Norland XCT (XCT Reseach SA, Norland Corp., Fort Atkinson, WI) scanner with a 50 kv x-ray tube as described by Korver et al. (2004). Calibration of the scanner was performed daily before bone analysis, with a multi-slice standard phantom (XCT Research SA, Norland Corp., Fort Atkinson, WI). A 1-mm cross-sectional x-ray slice with a voxel size of 0.1 mm was set for the mid-shaft of each bone. Norland XMENU software version 5.40C was used to analyze the resulting cross-sectional total, cortical, and trabecular BMD and cross-sectional areas. Total BMD and total area was the weighted average of both the cortical and trabecular bone, and reflected the amount and the density or area of each bone section. Cortical BMD was the outer shell of the bone that was determined to have a density of >500 mg/cm 3. Due to the inability of the current quantitative computed tomography equipment to distinguish between trabecular and medullary bone using this technology, bone in the trabecular space was assumed to include, and reflect changes in, medullary bone. Bone mineral content (BMC) was calculated as BMD multiplied by the cross-sectional area and is the amount of bone mineral contained in a 1-mm linear section of the scanned region of the bone. Bone breaking strength was measured as described by Riczu et al. (2004) with an Instron Materials Tester (model 4411, Instron Corp., Canton, MA) using software version 8.09, a standard 50-kg load cell and a modified sheer plate (8 cm in length and 1 mm in width). A 6-cm distance between the 2 fixed points supporting the tibia and a crosshead speed of 100 mm/min were held constant throughout each measurement. Bone ash was determined on fat extracted, dried tibias. Bones were first oven-dried at 100ºC for 48 h, and then fat was extracted for 8 h using petroleum ether (boiling range 36 to 60ºC) with a Soxhlet extraction apparatus. Tibias were again oven-dried at 100ºC for 24 h, then ashed at 600ºC in a muffle furnace for 24 h for bone ash determination as outlined by Zhang and Coon (1997b). Calcium analysis of ashed samples was determined by atomic absorption spectrometry on a Varian Spectra AA-20 (Varian Canada Inc.). Bone ash was dissolved in 15 ml of 1:1 HCl:double distilled water and 10 drops of concentrated nitric acid. The solutions were then diluted with double distilled water and further diluted (0.5:25 ml) with 0.5% lanthanum oxide (Fisher Scientific). Standard solutions were prepared by serially diluting a 1,000 mg/kg of Ca reference solution (Fisher Scientific). SAUNDERS-BLADES ET AL. Statistical Analysis The response variables were analyzed using the PROC MIXED procedure of the SAS (SAS Institute, 1999). The main effects were Ca source and particle size. For repeated measures analysis, the repeated statement in PROC MIXED was used adding the factor time (with production phases as the time factor). If significant effects (P 0.05 for analysis at each time point and P 0.01 for repeated measures analysis) of main effects or their interactions were detected, then means were compared using least squares means comparison (SAS Institute, 1999). A 1% level of significance was used for least square means comparisons when the interaction of age with both of the main effects was found to be significant. This was done to compensate for the increase in type 1 error due to the large number of comparisons being made (Moore and McCabe, 1993). RESULTS Mineral Content of Ca Sources The Ca content and purity of all Ca sources used in the current study were found to be very similar, ranging from 37.5 to 38.7% Ca and 93.7 to 96.5% purity (Table 1). However, other mineral components differed somewhat with regards to their concentration within the Ca sources that may affect its potential use by the laying hen. There was a large variation in the amount of some micro-minerals among the Ca sources. The oyster shell had 10 times the Na content of any of the limestone sources. The CGL had much greater levels of manganese and zinc than the other Ca sources (Table 1). The iron and boron content of oyster shell was much greater than in the other Ca sources (Table 1). Particle Size Distribution of Ca Sources The particle size distribution of the A, B, and C experimental limestone sources were found to be similar, all having the majority of particles ranging from 1.41 to 4.00 mm in size (Table 4). The B and C limestone sources had a very similar particle size distribution, whereas the A limestone source had fewer smaller particle sizes (<1.41 mm) and more in the to mm particle size range (Table 4). However, oyster shell had the greatest percentage (50.2%) of the largest size particles (>4.00 mm), as compared with the test limestone sources A, B, and C, which had only 3.1, 3.9, and 4.8%, respectively. This difference in particle size distribution between the limestone sources and the oyster shell can mostly be attributed to the difference in the shapes of the different Ca sources. Limestone had a hexagonal crystalline structure, whereas oyster shell has a flat and long appearance of flakes.

6 CALCIUM SOURCE AND PARTICLE SIZE 343 Table 4. Percentage of particle size distribution of large particle Ca sources 1 Ca source Particle size range (mm) Oyster shell A limestone 2 B limestone 3 C limestone 4 >0.50 to >0.71 to >1.00 to >1.41 to >2.00 to >2.83 to > Individual particle sizes of each large particle Ca source were obtained through sifting the source using standard sieve size screens. 2 Test limestone source A, a white limestone. 3 Test limestone source B, a brown limestone. 4 Test limestone source C, a gray limestone. In Vitro Solubility of Ca Sources and Particle Sizes In vitro solubility was dependent upon Ca source and particle size with a significant interaction effect (P < ; Table 5). Oyster shell, for the most part, had a faster solubility than test limestone sources A, B, and C of similar particle sizes, especially when comparing the larger particle sizes (>2 mm; Table 5). Therefore, the A, B, and C limestones may have the potential to remain in the gizzard longer than the oyster shell and supply Ca to the hen for a longer period of time. The larger particles (>2.00 mm) were less solubilized after 10 min than the ground or very small particle sizes among all Ca sources (Table 5). However, a significant decline in the amount of Ca source remaining after just 10 min for all particles indicates that the large particles would be capable of providing a rapidly released source of Ca, like the ground Ca sources. This may reduce the need of finding the right percentage of mixture of ground and large particle Ca because the large particles may be able to provide sufficient solubilized Ca in a short period of time in addition to providing Ca to the hen over a longer period of time. Ca and P Content of Experimental Diets Analyzed Ca values of experimental diets varied somewhat from the calculated values, ±0.54% on average (Table 3), except for the Ca value of the mixed particle control treatment containing oyster shell during phase 3, which was analyzed as 2.4%. This low Ca content had no apparent detrimental effects on laying hens receiving this diet. Although it is possible that the Ca content of this treatment was somewhat lower than the other treatments during this phase, it is believed that the majority of this difference was due to sampling error. The large particle size of oyster shell makes sampling for Ca analysis difficult. With this exception, Ca analysis of all other treatments indicated that Ca content of all treatments within each phase was similar (Table 3). Analyzed total P levels were all lower than the calculated total P levels (Table 3). However, all analyzed P levels among the different treatment groups were very similar. Effect of Calcium Source and Particle Size on Body Condition and Production Performance BW. Neither Ca source nor particle size affected BW during any single phase of the laying hens production cycle (P > 0.05; Table 6). All dietary treatment groups had a lower BW in phase 1 of the production study than all other phases of the laying cycle (P < 0.01; Table 6). Hens fed the A or B Ca source treatments continued to increase in BW until phase 3 after which the BW of birds appeared to stabilize between 1,860 and 1,890 g throughout the remainder of the trial. Hens fed the C or control Ca source treatments continued to increase in BW from phase 3 to phase 4 of the production cycle (P < 0.01; Table 6). However, during the final phase, BW was not different from those during phase 3 or 4 (P > 0.01; Table 6) and ranged from about 1,800 to 1,880 g over the final 3 phases. There was no difference in the pattern of weight gain for either of the particle size treatments (P > 0.05; Table 6). Birds on either of the particle size treatments gained weight until phase 3 of the laying cycle, from an average of 1,492 g during phase 1 to 1,838 g in phase 3. There was an increase in BW from phase 3 to 4, but the final BW was not different from either phase 3 or 4. Feed Consumption. Dietary Ca source did not affect feed consumption of hens during any phase of the production cycle (P > 0.05; Table 7). Hens consumed an average of 101 g/bird per day during phase 1 and from 113 to 117 g/bird per day for the remainder of the production trial (Table 7). Hens consuming the mixed particle Ca source treatments consumed 115, 117, and 114 g/bird per day, whereas hens on the ground particle size treatment consumed 113, 112, and 112 g/bird per day during phases 2, 3, and 4, respectively (P < 0.05; Table 7).

7 344 Feed consumption of all Ca source groups followed similar patterns of intakes throughout the entire production trial (Table 7). Feed consumption during phase Table 5. In vitro solubility of Ca source and different particle sizes of each Ca source 1 Item Particle size Solubility (%) Ca source Control A 2 B 3 C 4 Particle size 53.4 d 41.4 a 45.2 c 43.5 b Ground 66.2 g > f > e > e > d > c > b > a Source particle size Control Ground 66.6 l Control > l Control > k Control > k Control > i Control > j Control > c Control > c A Ground 67.3 l A > def A > fgh A > k A > de A > j A > bc A > a B Ground 66.6 l B > def B > hi B > fgh B > efg B > c B > c B > a C Ground 64.4 kl C > j C > fgh C > fgh C > ef C > d C > c C > ab ANOVA P-value Ca source < Particle size < Ca source particle size < a l Means within the same column and main and interaction effects (Ca source, particle size, Ca source particle size) with no common superscripts differ significantly (P 0.05). 1 n = 5 replications of each Ca source and particle size combination. Individual particle sizes of each large particle Ca source were obtained through sifting the source using standard sieve size screens. A 2-g sample was allowed to react with 200 ml of 0.2 N HCl for 10 min in a 42ºC oscillating water bath. After 10 min, the calcium source acid mixture was removed from the water bath and filtered. The residues on the filter paper were dried in an oven at 70 C until a consistent weight was obtained. Percentage of solubility is the amount of sample dissolved after 10 min. 2 Test limestone source A, a white limestone. 3 Test limestone source B, a brown limestone. 4 Test limestone source C, a gray limestone. SAUNDERS-BLADES ET AL. 1 was lower than all other phases of the production trial for all treatment groups (P < 0.01; Table 7). Feed intake was on average 10 g/bird per day greater during the later phases of the production trial than during phase 1. Egg Production and Quality. Hen mortality was similar among the experimental treatments and relatively low for the entire experimental group of hens (total mortality = 3.4%). Therefore, there was very little difference between percentage of hen housed and percentage of hen-day egg production. However, to account for losses in egg production due to hen mortality, hen-day egg production was used to compare egg production among the treatment groups. Egg production peaked at approximately 31 wk of age, at which time the average hen-day egg production was 92%. Percentage of hen-day egg production was not different among Ca source and particle size treatment groups during phases 1, 2, 3, or 4 (P > 0.05; Table 8). However, henday egg production (%) during phase 5 (71 to 74 wk of age) was greater for those hens fed Ca source B than those hens fed the A and C Ca sources (P < 0.05; Table 8). Total number of eggs produced for the entire production trial was not different among the Ca source or particle size treatment groups (P > 0.05; Table 8). This was expected because for the majority of the production period (phases 1 through 4) there were no differences in percentage of hen-day egg production. Repeated measures analysis indicated a significant age effect on egg production (P < 0.01; Table 8). However, no one particular Ca source or particle size treatment consistently had higher egg production. Except for the hens fed the B Ca source and the mixed particle treatments, all other Ca sources and ground particle treatment groups had lower egg production rates during phase 1 and 5 of the production trial than all other phases (P < 0.01; Table 8). For the hens fed Ca source B and the mixed particle size treatments, phase 1 had lower production rates than any of the other phases (P < 0.01; Table 8). During phase 2 of the production trial, hens yielded the highest egg production rates from all treatment groups (Table 8). Throughout the production trial there were no differences among Ca sources or particle sizes with regard to the number and percentage of unmarketable eggs (at the farm level) produced by the hens (P > 0.05; Table 8). A total of 0.6 to 0.7% of all eggs laid were found to be unmarketable (at the farm level) throughout the entire production trial. This percentage of unmarketable eggs only accounts for the number of eggs lost on the farm and does not account for eggs that would be unmarketable after grading processes. Egg weights did not differ among any of the Ca source and particle size treatments during the first 4 phases of the laying cycle (P < 0.05; Table 9). During phase 5 (71 to 74 wk of age) the hens fed the control mixed particle Ca source group had greater egg weights than all other treatment groups except for the Ca source A ground and B mixed particle size groups (P < 0.05; Table 9).

8 CALCIUM SOURCE AND PARTICLE SIZE 345 Table 6. The effect of Ca source and particle size on BW (g) of hens from 19 to 74 wk of age Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Ca source A 1 1,497 Y 1,784 X 1,861 W 1,874 W 1,862 W B 2 1,506 Y 1,792 X 1,873 W 1,890 W 1,859 W C 3 1,474 Z 1,714 Y 1,797 X 1,881 W 1,819 WX Control 4 1,492 Z 1,755 Y 1,820 X 1,830 W 1,861 WX SEM Particle size Ground 5 1,489 Z 1,752 Y 1,814 X 1,852 W 1,828 X Mixed 6 1,495 Z 1,770 Y 1,861 X 1,885 W 1,873 WX SEM ANOVA P-value Ca source (S) Particle size (P) S P Age < < < < < W Z Means within main effects (Ca source and particle size) with no common superscripts are significantly different in repeated measures analysis (age P < 0.01). 1 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 2 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 3 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 4 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age. 5 Ground Ca source diet. 6 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground). Egg weight gradually increased throughout the course of the production trial. Egg weights were lowest during the phase 1 of the production cycle compared with all other phases in all treatment groups (P < 0.01; Table 9). Egg weights increased from about 55 g during phase 1 to about 66 g by the end of the trial (Table 9). After phase 1, the increase in egg weight throughout the remainder of the trial was not consistent among treatment groups. The hens fed the control and Ca source C treatments had similar egg weights during phases 2 and 3 (P > 0.01; Table 9), whereas most of the other treatments exhibited an increase during this time (P < 0.01; Table 9). The heaviest egg weights for all treatment groups were obtained during the final phase, but these were not greater (P > 0.01) than egg weights during phase 4, except for the hens fed Ca source A ground Table 7. The effect of Ca source and particle size on feed consumption (g) of laying hens from 19 to 74 wk of age Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Ca source A Z X X Y X B Y X X X X C Y X X X X Control Y X X X X SEM Particle size Ground Z b,xy b,xy b,y X Mixed Z a,y a,y a,y XY SEM ANOVA P-value Ca source (S) Particle size (P) S P Age < < < < < a,b Means within the same column and main effect (Ca source and particle size) with no common superscripts differ significantly (P 0.05). X Z Means within main effects (Ca source and particle size) within rows with no common superscripts are significantly different in repeated measures analysis (age P 0.01). 1 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 2 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 3 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 4 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age. 5 Ground Ca source diet. 6 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground).

9 346 SAUNDERS-BLADES ET AL. Table 8. The effect of Ca source and particle size on egg production of laying hens from 19 to 74 wk of age Hen-day egg production (%) Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Total egg production per hen % unmarketable per hen Ca source A Y 92.1 V 86.2 W 77.9 X 68.3 b,y B Y 87.0 V 85.8 W 79.7 X 77.7 a,x C Y 88.0 V 82.2 W 76.3 X 68.2 b,y Control X 92.1 V 85.6 V 79.8 W 73.2 ab,x SEM Particle Ground Y 88.4 V 83.4 W 77.4 X 70.1 Y Mixed Z 90.4 V 86.5 W 79.5 X 73.6 Y SEM ANOVA P-value Ca source (S) Particle size (P) S P Age < < < < < NA7 NA a,b Means within the same column and main effect (Ca source and particle size) with no common superscripts differ significantly (P 0.05). V Z Means within main effects (Ca source and particle size) within rows with no common superscripts are significantly different in repeated measures analysis for egg production only (age P 0.01). 1 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 2 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 3 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 4 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age. 5 Ground Ca source diet. 6 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground). 7 NA = not applicable. Table 9. Effect of Ca source particle size interaction effect on egg weight (g) of eggs from laying hens 19 to 74 wk of age Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (46 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Ca source A 1 Ground Z 62.5 Y 65.9 X 65.6 X 68.5 ab,w Mixed Y 61.6 X 64.7 W 65.2 W 65.2 c,w B 4 Ground 55.7 Y 61.3 X 63.1 WX 65.3 W 65.9 bc,w Mixed 55.8 Z 60.6 Y 65.9 X 65.0 X 67.4 abc,w C 5 Ground 54.7 Y 60.6 X 62.8 WX 65.0 W 65.1 c,w Mixed 55.4 Y 62.3 X 64.3 WX 65.2 W 65.4 c,w Control 6 Ground 54.8 Y 61.3 X 62.2 X 65.2 W 66.1 bc,w Mixed 55.7 Z 61.8 Y 63.7 Y 66.4 X 69.1 a,w SEM ANOVA P-value Ca source (S) Particle size (P) S P Age < < < < < a c Means within the same column and main effect (Ca source and particle size) with no common superscripts differ significantly (P 0.05). W Z Means within main effects (Ca source and particle size) within rows with no common superscripts are significantly different in repeated measures analysis (age P 0.01). 1 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 2 Ground Ca source diet. 3 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground). 4 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 5 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 6 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age.

10 CALCIUM SOURCE AND PARTICLE SIZE 347 and B and control mixed particle treatment groups (P < 0.01; Table 9). Egg SG did not differ among Ca source and particle size groups during any phase of the laying cycle (P > 0.05; Table 10). Across all treatments, average egg SG was 1.090, 1.082, 1.078, 1.081, and for phases 1 to 5, respectively. There was a significant effect of time on SG (P < 0.01; Table 10). Egg SG followed the same pattern for all treatment groups throughout time, being greatest at the beginning of lay and continually decreasing until phase 3 (Table 10). This period corresponded to a time when the environmental temperature was high (25 to 30ºC), which affect shell quality. After phase 3, SG increased slightly during phase 4 as the outside environmental temperature began to decrease. However, SG decreased again during the phase 5 of the laying cycle, most likely due to the aging hen (Table 10). Bone Quality. There were no differences due to Ca source on total and cortical BMD as well as total and cortical BMC (P > 0.05; Table 11). However, trabecular BMD of tibias from hens that received Ca source A were greater than trabecular BMD of hens fed Ca source B or the control treatment. Because the trabecular measurements were assumed to also include medullary bone, this may indicate a larger pool of readily available skeletal Ca for those hens fed Ca source A. This can further be explained by the differences seen in the tibia cross-sectional area analysis in which there were no differences in total or trabecular areas; however, cortical area was nearly significantly greater for the Ca source A treatment group (P = 0.051; Table 11). Although the Ca source A treatment groups had a greater trabecular density, a smaller (although not significant) cross-sectional area resulted in a similar BMC as the other Ca sources (P > 0.05; Table 11). The control treatment group, which had the lowest trabecular BMD, also had the smallest cortical area, potentially indicating a greater loss of cortical bone due to a lack of labile Ca from medullary bone as the hens aged (Table 11). Tibia weight (dried), breaking strength, percentage of Ca, and percentage of ash were not different among the Ca source treatment groups (P > 0.05; Table 11). However, tibias from hens fed Ca sources A and B were longer than those tibias from hens fed Ca source C (P < 0.05; Table 11). There was a pattern of increased bone quality of tibias from hens fed the mixed particle size treatments. Total and trabecular BMD, cortical area, total and cortical BMC, tibia weight, and breaking strength were greater in the tibias from hens fed the mixed particle Ca source treatments (P < 0.05; Table 11). These improvements in laying hen bone quality demonstrate the benefits of a slower dissolving Ca source on bone quality of laying hens. The tibia trabecular area was greater from the hens fed the ground Ca source treatments, although having a lower trabecular BMD, indicating a greater portion of this bone type may have been mobilized. DISCUSSION Composition and In Vitro Solubility of Ca sources The Ca and P content of all Ca sources used in this study were similar (Table 1). Other minerals analyzed Table 10. The effect of Ca source and particle size on mean egg specific gravity of eggs from laying hens 19 to 74 wk of age Item Phase 1 (19 to 26 wk of age) Phase 2 (27 to 46 wk of age) Phase 3 (47 to 50 wk of age) Phase 4 (51 to 70 wk of age) Phase 5 (71 to 74 wk of age) Source A X Y Z Y Z B X Y Z Y Z C X Y Z Y Z Control X Y Z Y Z SEM Particle size Ground X Y Z Y Z Mixed X Y Z Y Z SEM ANOVA P-value Source (S) Particle size (P) S P Age < < < < < X Z Means within main effects (Ca source and particle size) within rows with no common superscripts are significantly different in repeated measures analysis (age P 0.01). 1 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 2 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 3 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 4 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age. 5 Ground Ca source diet. 6 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground).

11 348 SAUNDERS-BLADES ET AL. Table 11. The effect of Ca source and particle size on laying hen tibia quality at 74 wk of age Item Density 1 (mg/cm 3 ) Cross-sectional area 1 (mm 2 ) Bone mineral content 1 (mg/mm) Total Cortical Trabecular Total Cortical Trabecular Total Cortical Trabecular Bone weight, 2 g Length, cm Breaking strength, kg Ca, % Ash, % Source A 3 (n = 31) a a B 4 (n = 29) , b a C 5 (n = 32) , ab b Control 6 (n = 30) , b ab SEM Particle size Ground 7 (n = 62) b 1, b b a b b b b Mixed 8 (n = 60) a 1, a a b a a a a SEM ANOVA P-value Source (S) Particle size (P) S P a,b Means within the same column and main effect (Ca source and particle size) with no common superscripts differ significantly (P 0.05). 1 The right tibia was scanned at the midpoint after bone were cleaned of all adhering flesh and dried. Bone mineral density was measured for cortical = measurements taken on the area define as >500 mg/cm 3 and the outer part of the bone; trabecular = measurements taken in the inner part of the bone in the trabecular space; and total = the total for the entire bone; bone mineral content was calculated as BMD multiplied by the cross-sectional area and is the amount of bone mineral contained in a 1-mm linear section of the scanned region of the bone. 2 Fat-free, moisture-free bone weight. 3 Hens fed a diet consisting of test limestone source A, a white limestone, from 19 to 74 wk of age. 4 Hens fed a diet consisting of test limestone source B, a brown limestone, from 19 to 74 wk of age. 5 Hens fed a diet consisting of test limestone source C, a gray limestone, from 19 to 74 wk of age. 6 Hens fed a diet consisting of a commercially used ground limestone (CGL) source in the 100% ground control treatment, and a mixture of the CGL and oyster shell in the mixed treatment, from 19 to 74 wk of age. 7 Ground Ca source diet. 8 Mixed particle size Ca source diet (1/3 particulate + 2/3 ground).

12 CALCIUM SOURCE AND PARTICLE SIZE 349 showed variation between Ca sources (Table 1); however, the amounts of these minerals (except Mg) within the Ca sources would only provide minimal amounts (0.02 to 0.5%) of the daily National Research Council (1994) recommended levels and would likely not affect the total amount of these minerals in the diet. The Mg content of Ca sources would supply a greater proportion of the daily recommended value, with the oyster shell and CGL supplying about 2%, the A and B supplying 9 to 10%, and the C supplying 6.4%. However, Mg would only affect eggshell quality when in excess of 5,000 mg/kg (100 the NRC recommended level; Nys, 1999), which is much greater than contributed from any of these sources. Therefore, differences observed due to treatments are not likely to be confounded by levels and availability of minerals other than Ca and P. For all large particle sizes the control (oyster shell) had a greater amount solubilized than all the test limestone sources of similar particle sizes (Table 5). This may be related to the physical or chemical makeup of the oyster shell in comparison with the test limestones A, B, and C. The flat, long surface of the oyster shell may allow for a greater surface area to be exposed to the acid, which may lead to the increased solubility. The solubility of the calcium source affects the hen s ability to utilize it and thus affects eggshell and bone quality. Similar to the results of the current study, Guinotte and Nys (1991) found that particulate oyster shell had a higher in vitro solubility than particulate limestone. However, other researchers have reported oyster shell to have a slower solubility rate than limestone (Kuhl and Sullivan, 1977; Guinotte, et al., 1991). The difference in conclusions between the studies may be because comparisons between Ca sources were being made on different particle sizes because they were examined as a whole and not individual particle sizes. Calcium sources have been shown to vary by as much as 15 to 27% in solubility rates, even when comparing similar particle sizes (Rabon and Roland, 1985). Oyster shell and limestone are the most popular Ca sources studied for the laying hen. Although these Ca sources are very similar in Ca content, they have been shown to differ among studies in their effects on eggshell quality (Roland, 1986). This may be a result of differing solubility among the Ca sources used in each study (Roland, 1986). The solubility of the Ca source is affected by the composition and particle size of the calcium source. Roland (1986) reported that because comparisons made between oyster shell and limestone were comparing sources that were of different particle sizes, solubility, or both, research groups made different conclusions. In the current study, oyster shell did have a very different particle size make-up than the test limestone sources (Table 4); however, even within similar particle sizes oyster shell still had a great solubility. In contrast, all the test limestone sources had relatively similar in vitro solubility (Table 5), along with similar particle size compositions (Table 4). Within each Ca source, the particle sizes >2.00 mm had a slower solubility than the ground and small particle sizes (Table 5). Several previous studies have also found in vitro solubility rates of large particle sizes to be lower than those of smaller particles and ground forms of the same Ca source (Kuhl and Sullivan, 1977; Rao and Roland, 1989; Guinotte and Nys, 1991; Guinotte et al., 1991; Zhang and Coon, 1997a). In vitro solubility rates are inversely related to in vivo solubility (Rao and Roland, 1989; Zhang and Coon, 1997a) and a better predictor of eggshell quality and bone status than particle size (Cheng and Coon, 1990). The slower the in vitro the longer the Ca source it will remain within the gizzard of the hen, increasing the Ca retention of these slower dissolving Ca particles (Zhang and Coon, 1997a). Therefore, the results of the current study would indicate that the larger particles, in particular those of the test limestone A, B, and C, should remain in the gizzard longer, thereby having a greater Ca retention than the small particle and perhaps even similar particle sizes of oyster shell. Effect of Ca Source on Productive Performance and Bone Quality of the Laying Hen Body weight was not affected by Ca source (Table 6), a finding supported by earlier studies that also found no effect of different Ca sources on BW (Miller and Sunde, 1975; Guinotte and Nys, 1991). Previous research has shown that larger birds tend to consume more feed and lay larger eggs (Leeson et al., 1997; Riczu et al., 2004). Often, the way in which pullets are reared in the weeks before first egg will determine mature BW. However, BW during the laying phases may be affected by Ca source if, for example, it affects feed intake or health status of the hen. In the current study, feed consumption (Table 7) and bone weight (Table 11) were similar among Ca sources; therefore, BW would not be expected to be affected by Ca source. Although energy is the largest determinant of feed intake, hens have a specific appetite for Ca and may vary feed intake to accommodate Ca needs (Mongin and Sauveur, 1974; Sauveur and Mongin, 1974). Hens on low Ca diets (1%) have been shown to consume more feed than hens on adequate Ca diets (3.15%; Mongin and Sauveur, 1974). The results of the current study indicate that all Ca sources used in this study were able to supply the hens with sufficient Ca, so as to not to have to alter feed consumption to compensate for low Ca bioavailability. The Ca sources compared in the current study resulted in no differences with regards to percentage of hen-day egg production from 19 wk of age until 70 wk of age (P > 0.05; Table 8). After 70 wk of age, although egg production of hens fed test Ca sources A, B, and C did not differ from the control treatment group, hens fed Ca source B had greater egg production than hens