Effect of dietary formates on growth performance

Effect of dietary formates on growth performance, carcass traits, sensory quality, intestinal microflora, and stomach alterations in growing-finishing pigs1 M. Øverland*,2, T. Granli†, N. P. Kjos*, O. Fjetland‡, S. H. Steien†, and M. Stokstad§ ˚ s, Norway; *Department of Animal Science, Agricultural University of Norway, P.O. Box 5025, N-1432 A †Norsk Hydro ASA, Bygdøy alle` 2, N-0240, Oslo, Norway; †Norwegian Pig Health Service, Gilde Agro Fellesslakteri, BA, Norway; and §Veterinary College of Norway, Ullevaˆlsveien 72, P.O. Box 8146 Dep., N-0033 Oslo, Norway

ABSTRACT: Three experiments were conducted to evaluate the effect of adding salts of formic acid to diets for growing-finishing pigs. In Exp. 1, 72 pigs (23.1 kg and 104.5 kg initial and final BW) were used to evaluate the effect of Ca/Na-formate and K-diformate on performance and carcass traits. Treatments were organized in a 2 × 3 factorial arrangement with two feeding regimens (limit and semi-ad libitum feeding) and three diets (control, .85% Ca/Na-formate, and .8% K-diformate). No significant feeding regime × diet interaction was found. The K-diformate diet increased overall ADG of pigs compared with the control and Ca/Na-formate diets, but had no effect on ADFI or gain/feed (G/F) ratio. Neither K-diformate nor Ca/Na-formate had any effect on carcass lean or fat content. In Exp. 2, 10 limit-fed pigs (24.3 kg and 85.1 kg initial and final BW) were used to study the effect of K-diformate on performance and sensory quality of pork. Adding .8% K-diformate to diets increased ADG (P < .13) and G/F (P < .04), but had no effect on sensory quality of the pork or content of formate in liver, kidney, or muscle tissue of pigs. In

Exp. 3, 96 limit-fed pigs (27.1 kg and 105 kg initial and final BW) were used to determine the effect of adding K-diformate to diets on performance, carcass traits, and stomach keratinization and(or) lesions. Adding K-diformate (0, .6, or 1.2%) to diets increased ADG and ADFI (linear P < .01). The K-diformate reduced the percentage of carcass fat (linear P < .03) and fat area in the cutlet (linear P < .09) and increased percentage lean in the ham (linear P < .01), flank (linear P < .02), loin (linear P < .09), and neck and shoulder (linear P < .09). The K-diformate had no negative effect on stomach alterations. In Exp. 3, the concentration of coliform bacteria in the gastrointestinal tract was evaluated in eight control and eight 1.2% K-diformate-fed pigs. The Kdiformate reduced the number of coliforms in the duodenum (P < .03), jejunum (P < .02), and rectum (P < .10) of pigs. In conclusion, K-diformate improved growth performance and carcass quality of growing-finishing pigs, whereas Ca/Na-formate had no effect. K-diformate had no adverse effect on sensory quality of pork or on stomach alteration scores.

Key Words: Carcass Traits, Formate, Growth Performance, Pigs, Sensory Quality 2000 American Society of Animal Science. All rights reserved.

Introduction Cross-resistance of bacteria to antibiotic growth promoters and cross-resistance of animal and human pathogens demonstrate the need for safer, alternative growth enhancers for rearing pigs (SOU, 1997). Organic 1 The authors would like to thank Arnljot Mehl, Anne Nystad, Kjell Skoglund, and Jorunn Sylju for taking care of pigs. Mona Gjestvang for investigating stomachs, Rune Christiansen for preparing and analyzing digesta samples, Henning Sørum for his consulting on microbiology, and Hydro Nutrition for financial support. 2 Correspondence: (phone: +47 22 43 21 00; fax: + 47 22 33 75; E-mail: [email protected]). Received July 16, 1999. Accepted January 14, 2000.

J. Anim. Sci. 2000. 78:1875–1884

acids have received much attention as alternatives. Formic acid has shown to be effective against pathogenic bacteria (Gedek et al., 1992; Kirchgessner et al., 1992b), but its use has been limited by problems of handling, strong odor, and corrosion during feed processing and during its use on the farm. The product, K-diformate (trade name Formi LHS) has been shown to be easier to handle than pure formic acid and is an effective growth promoter in diets of both weanling pigs (Paulicks et al., 1996) and growing-finishing pigs (Kirchgessner et al., 1997; Øverland and Lysø, 1995; Roth et al., 1996). Adding K-diformate to diets increased apparent digestibility and retention of N in pigs (Roth et al., 1998), suggesting that it may improve carcass lean percentage by making amino acids more available for pro-

1875

1876

Øverland et al.

tein deposition within the body. Limited information exists on the effect of supplementing diets with K-diformate on carcass quality, as well as the gastrointestinal tract microflora of pigs. Also, the effect of K-diformate on sensory quality of pork has not been elucidated. Stomach alterations are becoming of more concern in modern pig production. Limited information exists on the effect of supplementing diets with K-diformate on stomach lesions and ulcers of pigs. The objective of the present experiment was tripartite: 1) to determine the effect of Na/Ca-formate and Kdiformate on growth performance and carcass traits of growing-finishing pigs; 2) to establish the effect of Kdiformate on sensory quality of pork; and 3) to assess the effect of increasing levels of K-diformate (0%, .6%, and 1.2%) in diets for growing-finishing pigs on growth performance, carcass quality, gastrointestinal microflora, and safety with respect to stomach alterations.

Materials and Methods Three growth experiments utilizing growing-finishing pigs were carried out: Exp. 1 and 2 at the experimental pig facilities of the Agricultural University of Norˆ s, Norway, and Exp. 3 at the Staur Experimental way, A Station, Ottestad, Norway. Animals, Housing, and Allocation. In Exp. 1, 72 (Norwegian Landrace × Yorkshire) growing-finishing pigs were used. Average initial and final weights were 23.0 kg and 104 kg, respectively. The experimental period lasted on average 88.2 d. The experiment was conducted as a randomized complete block design. Pigs were allotted on the basis of initial weight, litter, and sex to each of six dietary treatments with 12 replicates per treatment. There were 12 pens with six pigs per pen. At feeding time, each pig was restrained in an individual feeding stall until the feed was consumed in order to obtain individual feed intake. Thus, each pig was one experimental unit. There were equal numbers of gilts and barrows in each treatment group. In Exp. 2, 10 (Norwegian Landrace × Yorkshire) growing-finishing pigs from two separate litters were used. Average initial and final weights were 24.3 kg and 85.1 kg. There were two pens with five pigs per pen. The experimental period lasted 65 d. The experiment was conducted as a randomized complete block design. Pigs were allotted by weight, litter, and sex to two dietary treatments with five replicates per treatment. At feeding time, pigs were restrained in an individual feeding stall until the feed was consumed in order to obtain individual feed registration. Each pig was one experimental unit. In Exp. 1 and 2, pens measuring 8.2 m2 and designed for individual feeding were used. Pigs were housed in an environmentally controlled barn with partially slotted concrete floors. Experiment 3 was carried out with 96 Noroc ([Norwegian Landrace × Yorkshire] × [Norwegian Landrace × Duroc]) pigs from 16 litters. Average initial weight was 24.1 kg and average final weight was 104.1 kg. Pigs

were housed at an average ambient daily temperature between 16 and 20°C in a barn in pens with partially slotted concrete floors. Pens measuring 15.0 m2 equipped with individual feeding stalls were used with eight pigs per pen. Pigs were allotted on the basis of initial weight, sex, and litter to each of three dietary treatments with 32 replicates per treatment. As in Exp. 1 and 2, pigs were individually fed and each pig was one experimental unit. There were equal numbers of gilts and barrows in each treatment group. The experimental period lasted on average 87.0 d. Diets, Feeding, and Weighing. In Exp. 1, the dietary treatments were organized according to a 2 × 3 factorial arrangement with two feeding regimes (limit feeding and semi-ad libitum feeding) and three experimental diets (control, Ca/Na-formate and K-diformate). The Ca/Na-formate contained 48.75% Ca-formate [Ca(HCOO)2], 48.75% Na-formate (NaOOCH), 1.0% formic acid (HCOOH), and 1.5% silicate and water. This provided a total formate content of 67.0% in the product. The K-diformate contained 98.5% K-diformate, of which 30.0% was potassium, 35.4% formic acid, 34.6% formates, and 1.5% silicate and water. The formates were added to the diet at a level that corresponded to .6% formic acid, which amounted to .85% Ca/Na-formate and .8% K-diformate. The Ca/Na-formate and K-diformate were added to the diets at the expense of barley. In the Ca/Na-formate diets, adjustments were made for the additional supply of Ca and Na. This was done by omitting the limestone and reducing the salt content. In Exp. 2, the dietary treatments were a basal diet and a test diet containing .8% K-diformate. The diets were the same as in Exp. 1. In Exp. 3, the treatments consisted of a basal diet and two test diets containing .6% or 1.2% K-diformate. The K-diformate was added on top of the basal diet. The diets were conventional Norwegian diets mainly based on barley and oats and the main protein sources were soybean meal, canola meal, meat and bone meal, and fish meal. The oats were ground through a 3-mm sieve, whereas the remaining ingredients were ground through a 4-mm sieve. The feed was produced at a commercial feed mill and the formates were added to the diets during manufacture and before pelleting. The feed was pelleted in a 3-mm die. Diets were formulated to meet or exceed NRC (1988) requirements for all indispensable amino acids and other nutrients. Composition and analyses of diets are shown in Table 1 for Exp. 1 and 2 and in Table 2 for Exp. 3. Feed samples at the feed mill were taken for formate analysis. Results from the analysis in Exp. 1 and Exp. 2 showed that the control, Ca/Na-formate, and K-diformate diets contained .008, .60, and .58% formate, respectively. In Exp. 3, analyzed content of formate in the basal diet and the .6% and 1.2% Kdiformate diets was .12%, .42%, and .82%, respectively. The high level of formate in the basal diet can be explained by the presence 2.5% fish silage, which con-

1877

Formates in swine diets

Table 1. Percentage composition and chemical content of diets used in Exp. 1 and 2 (as-fed basis) Item Barley Oats Wheat bran Fish meal Meat and bone meal Soybean meal, 45% Canola meal Limestone Salt Premixb L-Lysine HCl. 98 DL-Methionine L-Threonine Choline chloride Organic acids Calculated content Metabolizable energy, MJ per kgc Analyzed content, g per kg Dry matter CP Crude fiber HCL-ether extract Ash Calcium Total P Lysine Threonine Methionine Cystined Buffering capacity, mEq

Control

Ca/Na-formatea

K-diformate

62.182 20.0 3.00 1.00 6.00 2.40 4.00 .33 .48 .098 .27 .03 .11 .10 .00

62.002 20.00 3.00 1.00 6.00 2.40 4.00 .00 .14 .098 .27 .03 .11 .10 .85

61.492 20.00 3.00 1.00 6.00 2.30 4.00 .32 .48 .098 .27 .03 .11 .10 .80

12.2

12.2

12.1

892 151 63 34.3 50 9.9 7.6 9.97 6.93 3.04 2.99 602

887 151 60 34.3 53 10.8 8.0 9.27 6.43 2.87 2.87 623

891 149 59 35.6 56 9.2 7.3 9.54 6.35 2.52 2.95 647

a

Ca/Na-formate was only used in Exp. 1. Provided the following amounts per kilogram of feed in Exp. 1 and 2: Zn (ZnO) 105 mg; Fe (FeSO2ⴢH2O) 75 mg; Mn (MnO) 60 mg; Cu (CuO) 15 mg; I (Ca(IO3)2) 7.44 mg; Se (Na2SeO3), .3 mg; vitamin A, 8,400 IU; cholecalciferol, 700 IU; DL-α-tocopheryl acetate, 115.9 mg; riboflavin, 5 mg; D-pantothenic acid, 15 mg; cyanocobalamine, 20 ␮g. c The metabolizable energy content of Ca/Na-formate and K-diformate was assumed as 0. d Cystine + cysteine. b

tained 3.7% formic acid. Conventional swine diets contain about .005% formate from natural sources. In Exp. 1, pigs were fed according to a standard restrictive feeding scale as described by Øverland (1997) or a semi-ad libitum feeding regimen. In Exp. 2 all pigs were fed according to a semi-ad libitum feeding regime, whereas in Exp. 3 pigs were fed according to a standard restrictive feeding scale (Øverland, 1997). The feeding scale used in Exp. 1 and Exp. 3 provided a moderate feeding intensity during the growing period followed by an increasing feeding intensity during the finishing period. The feeding program provided by the feeding scale was from about 14 to 31 MJ ME/d during the growing period and from about 31 to 40 MJ ME/d during the finishing period. The semi-ad libitum feeding regimen involved providing pigs twice daily with all the feed they could consume within about 30 min without producing any leftover. All pigs were individually fed twice daily: half their ration at 0730 (± 15 min) and half at 1330 (± 15 min). Feed intake was adjusted once weekly according to each pig’s live weight. Thus, due to differences in growth

rate, pigs in the different treatment groups received different levels of feed per day. Pigs were given free access to water from nipple drinkers. Water was also provided directly in the trough during meals. A 7-d preliminary period adjusted the pigs to experimental diets and pens. Feed consumption was recorded daily. Pig weights were measured at 7-d intervals in Exp. 1 and 2 and at 14-d intervals in Exp. 3 to determine ADG, ADFI, and gain/feed (G/F). Feed refusals for each pig were recorded and subtracted from the total feed intake. Experiments 1 and 2 were split into a growing period from wk 0 to 7 of the experimental period, a finishing period from wk 8 to slaughter, and the overall period. Carcass Characteristics. The pigs were slaughtered at a commercial slaughterhouse. In Exp. 1 and 3, carcass characteristics were measured after 1 d of chilling according to procedures described by Sundstøl (1973). Dressing percentage was determined by (hot carcass weight/final weight) × 100. Live weight was monitored the day before slaughter. Percentage of carcass lean was determined commercially on the slaughter line using a GP2Q pistol (Hennessy System Ltd., Auckland, New

1878

Øverland et al.

Zealand) to measure the depth of the longissmus dorsi and the backfat thickness at two sites (between the 10th and 11th ribs, 6 cm from the midline, and behind the last rib, 8 cm from the midline). A tracing of a cross section of the cutlet behind the last rib was made on tracing paper. The total area and meat area in the cutlet were determined by planimeter (Coradi AG, Zu¨ rich, Switzerland). The fat area of the cutlet was determined by the difference between the total area and meat area. The P2 backfat thickness was measured 8 cm from the midline behind the last rib using tracing paper and a ruler. In Exp. 3, carcass characteristics were also determined by a primal cuts method (M. Røe, Norwegian

Table 2. Percentage composition and chemical content of basal diet in Exp. 3 (as-fed basis) K-diformate, % Item Barley Oats Wheat bran Fish meal Meat and bone meal Soybean meal, 45% Canola meal Fish silage Rendered fat Limestone Salt Premixa Choline chloride L-Lysine, HCl, 98 L-Threonine DL-Methionine K-Diformate Calculated content Metabolizable energy, MJ per kgb Analyzed content, g per kg Dry matter Ash CP Petroleum ether extract Crude fat Calcium Total P Na Cl K Lysine Threonine Methionine Cystinec Tryptophan Buffering capacity, mEq

0

.6

1.2

49.807 28.0 7.0 1.3 4.6 3.4 1.3 2.5 .5 .8 .4 .103 .1 .15 .03 .01 0

49.51 27.833 6.958 1.292 4.573 3.38 1.292 2.485 .497 .795 .398 .103 .099 .149 .03 .01 .596

49.216 27.668 6.917 1.285 4.545 3.36 1.285 2.47 .494 .791 .395 .102 .099 .148 .03 .01 1.186

12.1

12.0

12.0

895 48 154 54 34 7.8 6.4 2.1 4.5 5.9 8.78 5.91 3.04 3.01 1.80 522

894 50 153 54 33 7.7 6.3 2.1 4.7 7.7 8.50 5.83 2.95 2.94 1.81 552

894 54 154 57 33 7.7 6.3 2.1 4.5 9.3 8.49 5.73 2.69 2.89 1.80 686

a Provided the following amounts per kilogram of feed: Zn (ZnO) 105 mg; Fe (FeSO2ⴢH2O) 75 mg; Mn (MnO) 60 mg; Cu (CuO) 15 mg; I (Ca(IO3)2) 7.44 mg; Se (Na2SeO3), .3 mg; vitamin A, 8,400 IU; cholecalciferol, 700 IU; DL-α-tocopheryl acetate, 70 mg; riboflavin, 5 mg, D-pantothenic acid, 15 mg; cyanocobalamine, 20 ␮g; niacin 20 mg; biotin, .2 mg; folic acid, 1.5 mg; thiamin, 2 mg; pyridoxine, 3 mg. b The metabolizable energy content of Ca/Na-formate and K-diformate was assumed as 0. c Cystine + cysteine.

Meat Coop., Oslo, Norway, personal communication). After 2 to 5 d of chilling, the right side of each carcass was separated into primal cuts and skinned. The primal cuts were deboned and separated into the following fat content categories: 1, lean tissue; 2, fat content up to 6%; 3, fat content up to 23%. The neck was not deboned, and the neck and the bacon side were not separated into fat content categories. The category 1 cuts (loin, tenderloin, and inside round) were assumed to have a fixed fat content. Fat Content Category 3 was analyzed for fat content using a Scanalyser (model S2 1997 No. 8330-33, Scanbio, Aalborg Denmark). This technique is based on differences in density between the lean and fat tissue in the sample. Based on these recordings, the carcass lean percentage was calculated according to the procedures used by the Norwegian Meat Cooperative (Oslo, Norway) and the Norwegian Pig Breeders’ Association (Hamar, Norway). In Exp. 2, samples from the loin muscle were taken at slaughter for sensory and chemical analyses and samples from liver and kidney were taken for determination of formate content. In Exp. 3, the stomach of each pig was collected at slaughter for evaluation for stomach lesion scores. All pigs received a normal morning meal the day of slaughter. They were slaughtered 3 to 4 h postprandially, when there was still feed in their stomachs. The aglandular part of the stomach was visually evaluated within 2 h after slaughter. The stomachs were scored on a scale from 1 to 8: 1 being normal; 2, 3, and 4 being of increasing severity of hyperkeratosis; and 5, 6, 7, and 8 having gastric ulcerations < 2 cm2, 2 to 8 cm2, 8 to 16 cm2, and > 16 cm2, respectively (Baustad and Nafstad, 1969). The pigs were slaughtered over a period of 5 wk. One trained veterinarian performed all evaluations. Photographs were taken of some stomachs for further evaluation by an independent veterinarian when necessary. Sensory Analyses. In Exp. 2, sensory analyses were performed on the loin muscle at the Norwegian Food ˚ s, Norway. Sensory quality was Research Institute, A analyzed using Qualitative-Descriptive-Analysis method (ISO, 1985). The samples were taken from the 5th to the 14th rib of the loin region. The samples were stored in a freezer at −20°C for approximately 1 wk before the sensory analysis was conducted. Pork was first defrosted and then prepared by making 1.5-cmthick loin slices from each pig. Prior to the sensory evaluation, the samples were vacuum-packed in plastic bags with a high aroma barrier and stored 1 d at 5°C before sensory testing. The samples (3 slices from each pig per sensory quality assessor) were heated in a 72°C water bath for 40 min before they were presented to 12 trained sensory quality assessors. The samples were analyzed for intensity of the following attributes: odor intensity, odor of pork meat, off-odor, whiteness, color/ hue, color strength, flavor intensity, flavor of pork meat, sweetness, bitterness, acid/sour flavor, off-flavor, fatness, juiciness, firmness, and tenderness. The odor at-

1879


tributes were assessed immediately after opening the vacuum bags. Flavor and texture attributes were evaluated after removing the samples from the bag. The assessors evaluated the samples at their own pace using a computerized system for direct recording of data (CSA Compusense Inc., Guelph, Ontario, Canada) on a continuous scale. The computer translated the responses into scores of between 1 and 9, where 1 equals no intensity and 9 equals high intensity. The color attributes were evaluated following the Natural Colour System (NCS) (SSI, 1979). The attribute hue was evaluated from yellow/red = 1 to red = 9. Microbiologial Examination. In Exp. 3, the microbiological examination was conducted on digesta collected from eight pigs from the control and the 1.2% K-diformate group. Pairs of pigs from the same litter receiving the control or K-diformate diet were used when possible. Thus, seven littermates and two pigs from different litters were used. The microbiological sampling was performed on two occasions of 1 wk apart. Four pigs from each treatment group were sampled at each time. The digestive system was taken out on the slaughter line and the stomach and intestines were removed and transferred into plastic bags, and brought to a separate room for further preparations. The sample collection started immediately. All sampling was completed within 2 to 2.5 h after slaughter. From each pig, samples were taken from three sites in the gastrointestinal tract: Sample 1, middle of duodenum; Sample 2, jejunum, 5 m caudally from flexura duodenojejunalis; and Sample 3, rectum, about 5 cm cranially from anus. Digesta samples from about 10 cm of the gut were collected, put into clean plastic beakers and kept on ice during the 2-h transport to the Norwegian College of Veterinary Medicine, Oslo, where the samples were prepared upon arrival. The method used was based on isolating and enumerating the coliform bacteria in the digesta on selective medium [MacConkey agar, Bacto, Difco Laboratories (0140-01-0)] based on MacConkey(1905). About 1 g of digesta was sampled aseptically and diluted in 10-fold dilution series with sterile NaClwater. After spreading on the agar, they were incubated aerobically at 37°C for 24 h. The total number of coliform bacteria was counted. Chemical Analyses. In Exp. 1, 2, and 3, dry matter, CP, crude fiber, and EE of the diets were determined by standard methods (AOAC, 1990). Amino acid composition of the diet was determined by hydrolysis for 24 h at 110°C using 5 mL of 6 M HCl containing 1 mg of phenol per mL, and norleucine as internal standard. Methionine and cysteine + cystine (half-cystine) were determined as methionine sulfone and cysteic acid, respectively, following oxidation with performic acid according to the procedure of Moore (1963). Phosphorus and calcium content of the diet was analyzed by atomic absorption spectrophotometry according to methods described by AOAC (1990). Formate content of diets and of fish silage in Exp. 3 was determined by ion chromatography according to standard Norsk Hydro proce-

dures (HRE-NIT-A33E) with HPLC. In Exp. 2, formate content of kidney, liver, and muscle tissue were analyzed according to Boehringer Mannheim kit Cat. No. 979 732. The buffering capacity of diets was assessed according to Prohszka and Baron (1980) and is expressed in milliequivalents (mEq) of 1.0 N HCl required to obtain pH 3 in a 1 kg sample. Statistical Analyses. Statistical analyses were performed using the GLM procedure of SAS (1990). Each pig was the experimental unit. Results are presented as the least squares means of the pigs in each treatment, and variability in the data is expressed as standard error of the mean. Pigs were blocked by initial weight and litter. Treatment, block, and sex were included as explanatory effects. Final weight was used as a covariant in the analysis of carcass traits. Significant differences among means were determined using the least significant difference test. The chosen level of significance was 5%, and trends were defined as levels between 5% and 10%. In Exp. 3, orthogonal polynomials were used to test linear and quadratic responses of increased dietary levels of K-diformate. All calculations of bacteria counts were done on logarithmic transformation values. The data from the bacteria counts were tested for significant differences by Microsoft Excel 97 Statistical Tools; Student’s t-test paired samples for means. Animal Care. All pigs were cared for according to laws and regulations controlling experiments with live animals in Norway (i.e., the Animal Protection Act of December 20, 1974, and the Animal Protection Ordinance concerning experiments with animals of January 15, 1996).

Results and Discussion Diets. In Exp. 1 and 2, the diets were similar in chemical composition except for the contents of lysine, threonine, and methionine, which were higher in the control diet (Table 1). The diets were calculated to contain the same level of Ca, but chemical analysis revealed a higher Ca content in the Na/Ca-formate diet. The pH values of the control, Ca/Na-formate, and K-diformate diets were 5.83, 6.62, and 5.54, respectively. The pH of the drinking water was 7.30 in all experiments. In Exp. 3, the pH values of the 0%, 0.6%, and 1.2% K-diformate diets were 6.0, 5.6, and 5.4%, respectively. The control diet had a higher content of lysine, threonine, methionine, and cystein, whereas proximate content of diets was similar among treatments (Table 2). Health. In Exp. 1 and 2, no health problems related to the dietary treatments were encountered during the experiment. In Exp. 3, one pig died on d 46 on the .6% K-diformate diet. A twisted gut was revealed at necropsy. No health problems related to the dietary treatments were encountered during the whole experimental period. The pigs had a good health status. They were free from swine dysentery, scabies, and mycoplasma.

1880

Øverland et al.

Table 3. Effect of Na/Ca-formate and K-diformate on growth performance and carcass traits of growing-finishing pigs in Exp. 1a Item

Control

Na/Caformate

Initial weight, kg Final weight, kg Growth performance Days on test Growing period ADG, g ADFI, kg Gain/feed Finishing period Body weight at wk 7, kg ADG, g ADFI, kg Gain/feed Overall period ADG, g ADFI, kg Gain/feed Carcass traits Carcass weight, kg Dressing percentage Percentage carcass lean GP2Q method P2-backfat thickness, mm Meat area in cutlet, cm2 Fat area in cutlet, cm2

23.0 104.1

23.2 104.5

88.7

89.6

752x 1.69 .444x

758x 1.70 .447xy

Kdiformate

SEM

Pvalue

Limit feeding

Appetite feeding

SEM

Pvalue

23.0 105.1

.4 .5

NS NS

23.0 104.9

23.0 104.2

.3 .4

.92 .20

87.3

.8

.12

90.0

87.1

.6

.001

.02 .004

.001 .18 .005

728 1.60 .456

810 1.82 .446

7 .01 .003

.001 .001 .02

.07 .49 .16 .65

58.7 1,133 2.93 .387

62.8 1,098 3.06 .359

.6 .02 .02 .005

.001 .1 .001 .001

.02 .004

.02 .18 .27

912 2.20 .387

934 2.36 .396

7 .01 .005

.03 .001 .001

797y 1.73 .461y

9

59.8 1,118 2.97 .377

60.3 1,099 2.98 .369

62.0 1,130 3.04 .373

.7 19 .03 .006

917x 2.26 .406

911x 2.28 .401

942y 2.30 .410

75.8 72.8 54.4xy

75.9 72.5 53.5x

76.8 73.0 55.0y

.4 .3 .4

.15 .30 .04

76.1 72.6 54.8

76.2 73.0 53.8

.4 .2 .3

.80 .19 .04

11.5 45.7 17.3

11.4 45.1 17.4

11.0 47.3 16.6

.5 .9 .7

.73 .18 .71

11.3 46.3 17.2

11.4 45.7 17.1

.4 .7 .6

.88 .55 .92

8

a Values are main effect means of a 3 (diet) × 2 (feeding method) factorial arrangement of treatments. Each treatment was fed 6 barrows and 6 gilts. x,y Means in a row lacking a common subscript letter differ (P < .05).

Growth Performance and Carcass Traits. In Exp. 1, there were no significant feeding regime × diet interactions, and thus only the main effects are presented (Table 3). This means that the response in performance of adding the formate products to diets was not influenced by feeding intensity. During the growing and overall period, pigs fed according to appetite had a significantly higher ADG and ADFI but a lower G/F compared with the limit-fed pigs in Exp. 1. During the finishing period, appetite feeding tended to increase ADG, and it significantly increased ADFI and reduced G/F. Dietary addition of K-diformate improved (P < .01) ADG and G/F compared with the control and Ca/Na-formate-fed pigs during the grower period. During the finisher period, there was no significant effect of Ca/Na-formate or K-diformate on growth performance. Overall, K-diformate increased ADG (P < .02) compared with the control and Ca/Na-formate diets. There were no differences among treatments for overall ADFI or G/F. The results show that K-diformate improved weight gain of growing-finishing pigs but not feed efficiency. Adding Ca/Na-formate to diets had no effect on growth performance of pigs during the grower, finisher, or overall period. There were no significant differences among treatments for any of the carcass traits measured, suggesting that neither Ca/Na-formate nor K-diformate affected carcass quality. Øverland and Lysø (1995), also found a positive effect on

growth performance of growing-finishing pigs fed Kdiformate (1.5%) but no effect of adding Ca/Na-formate (1.44%) to conventional Norwegian diets. The lower effect on growth performance of Ca/Naformate compared with K-diformate could be due to differences in the Ca content of the diets. Boulduan et al.(1988) reported that a high Ca content may increase the buffering capacity of the diets, which results in a lower effect of the acidifiers. However, there were no major differences in buffering capacity among the diets, which were within the narrow range of 604 (control) to 647(K-diformate) mEq (Table 1). As stated previously, these numbers are based on the method described by Proha`szka and Baron (1980), expressed as milliequivalents of 1.0 N HCl required to obtain pH 3 in a 1-kg sample. If pH of 4 was used instead, K-diformate would reduce the buffering capacity of the diet due to its pKa of 3.75, whereas Ca/Na-formate would still increase it. This has been shown in studies by Mroz and colleagues (unpublished data), in which the buffering capacity was reduced from 430 mEq in the basal diet to 402 mEq in the test diet containing 1.0% K-diformate, when based on a pH of 4. Thus, the limited effect of Ca/Na-formate on growth performance might be that it acts as a buffer and not an acidifier. Another difference in the mode of action of these formates is that K-diformate provides H+ ions from its content of 35.4% formic acid, which may reduce the pH in the stomach and the duodenum.

1881


The dietary addition of .9 or 1.8% K-diformate significantly reduces the pH of the duodenal digesta of weanling pigs by about .4 pH unit (Mroz and colleagues, unpublished data). The Ca/Na-formate only provided formates, so it would have no pH-reducing effect in the gastrointestinal tract. Thus, the mode of action of Kdiformate may also be a result of a reduced pH, which may decrease bacterial growth and improve pepsin activity (Kirchgessner and Roth, 1988). Furthermore, Roth and Kirchgessner (1998) reported that the growthpromoting effect of formic acid and K-diformate was superior over that of Ca-formate and Na-formate. The effect of K-diformate on growth performance of growing-finishing pigs in Exp. 2 is shown in Table 4. During the growing period, the addition of K-diformate to diets improved G/F (P < .02) of pigs but had no effect on ADFI. Overall, K-diformate improved G/F (P < .04) of pigs by 2.8% but had no effect on ADFI. In contrast to Exp. 1, these results show a greater effect of K-diformate on feed efficiency than on live weight gain, possibly as a consequence of the greater increase in feed intake in this trial. The effect of increasing levels of K-diformate in diets on growth performance and carcass traits of pigs in Exp. 3 is shown in Table 5. During the growing period, the dietary inclusion of .6% and 1.2% K-diformate increased ADG (linear P < .001) of pigs. Potassium-diformate also increased ADFI (linear P < .001) and G/F (linear P < .09) of pigs during this period. During the finishing period, K-diformate increased ADFI (linear P < .004) but had no effect on G/F. Overall, supplementing diets with 1.2% K-diformate gave a shorter time to market (linear P < .008) and a higher ADG (linear P < .001) and ADFI (linear P < .006). The results show that adding K-diformate to diets gives a positive dosedependent effect on growth performance of growingfinishing pigs. This is in agreement with the positive responses observed with K-diformate on the growth performance of pigs in Exp. 1 and 2. The restrictive feeding scale implies that pigs were fed equal levels of ME per day according to their live weight. Thus, the greater ADFI of the pigs receiving

the K-diformate treatments was a result of the improvement in ADG. In Exp. 2, where pigs were fed according to a semi-ad libitum feeding regimen, the improvement in feed intake might have been due to the improvement in ADG and in the palatability of the diet with the Kdiformate. The K-diformate may have increased the freshness and hygenic quality of the feed. The response to K-diformate in growth performance was greater in Exp. 2, in which pigs were fed a stronger feeding intensity (semi-ad libitum) than in Exp. 1 and 3. These results suggest that the effect of K-diformate on performance is greater when pigs are fed a stronger feeding intensity. This is in agreement with Partanen and Mroz (1999), who reported a higher response to organic acids when pigs were fed according to appetite than when fed restrictively. The response in performance and carcass quality to K-diformate might have been influenced by the combination of limit feeding and only using one diet during the growing and finishing period. This implies an undersupply of nutrients, especially during the growing period. Kirchgessner et al. (1997) also reported a positive effect of K-diformate (.8%) on the growth performance of limit-fed growing-finishing pigs. In Exp. 1 and Exp. 3, the effect on growth performance was greater during the grower period than in the finisher period. This was also observed by Kirchgessner et al. (1997), suggesting that the effect of K-formate on performance was greater in pigs in the grower than in the finisher phase. The grower period is a time of fast protein gain, but this is not always achieved because of limited feed intake or nutrient availability. Thus K-diformate, by enhancing nutrient availability (Roth et al., 1998), ensures a higher protein gain. Adding K-diformate to diets reduced carcass fat percentage (linear P < .03) and fat area in the cutlet (linear P < .09) (Table 6). The K-diformate diet also increased lean percentage of ham (linear P < .01), flank (linear P < .02), loin (linear P < .09), and neck and shoulder (linear P < .09). Furthermore, K-diformate led to a numerical increase in carcass lean percentage determined by the primal cut method and by the GP2Q pistol as well as in the meat area in cutlet. The increased carcass lean

Table 4. Effect of K-diformate on growth performance of growing-finishing pigs in Exp. 2 Item Initial weight, kg Final weight, kg Days on test Growing period ADG, growing period, g ADFI, growing period, kg Gain/feed, growing period Overall period ADG, g ADFI, kg Gain/feed a

Control

K-diformate

SEM

P-valuea

24.3 81.7 65

24.3 88.4 65

1.3 3.5 —

.9 .25 —

855 1.96 .436

957 2.05 .468

43 .08 .006

.17 .50 .02

883 2.11 .419

987 2.19 .450

38 .07 .008

.13 .45 .04

Values are means for two barrows and three gilts fed semi-ad libitum.

1882

Øverland et al.

Table 5. Effect of K-diformate in diets on growth performance and carcass traits of growing-finishing pigs in Exp. 3a Contrastb

K-diformate, % Item

0

Initial weight, kg Final weight, kg Growth performance Days on test Growing period ADG, g ADFI, kg Gain/feed Finishing period Body weight, kg ADG, g ADFI, kg Gain/feed Overall period ADG, g ADFI, kg Gain/feed Carcass traits Carcass weight, kg Dressing percentage Meat area in cutlet, cm2 Fat area in cutlet, cm2 Percent carcass lean GP2Q method P2 backfat thickness, mm Carcass traits, primal cuts method Percent carcass lean Percent fat in carcass Percent lean of ham Percent lean of flank Percent lean of loin Percent lean of neck and shoulder

.6

27.0 103.9

27.2 103.9

1.2

SEM

Linear

Quadratic

27.1 104.1

.7 .6

.98 .86

.90 .90

1.2

.008

.91

.02 .01

.0001 .001 .09

.72 .99 .65

1.0 14 .020 .01

.01 .11 .004 .687

.78 .58 .86 .423

10 .02 .01

.0003 .006 .11

.80 .99 .85

89.4x

87.1x

84.6y

748x 1.85x .402

793y 1.92x .412

828y 1.99y .415

60.6x 980 2.976 .327

62.8xy 986 3.020 .324

64.3y 1014 3.057 .330

863x 2.41x .357

886x 2.45x .360

915y 2.49y .367

74.4 71.6 43.6 20.1 54.1

75.4 72.5 44.9 19.8 54.1

75.1 72.3 45.0 18.7 54.9

.5 .3 .8 .6 .4

.09 .12 .22 .09 .17

.11 .12 .51 .59 .38

14.0

13.8

13.6

.4

.55

.96

53.6 19.4x 61.0x 50.3x 58.2 56.2

53.5 19.2xy 61.1x 52.1xy 58.4 56.3

54.1 18.3y 62.4y 53.1y 59.8 57.0

.3 .3 .4 .8 .6 .3

.25 .03 .01 .02 .09 .09

.30 .41 .18 .73 .41 .38

13

a

Values are means for 32 (16 barrows and 16 gilts) limit-fed pigs. Probability of linear and quadratic contrasts of increased dietary levels of K-diformate. x,yMean in a row lacking a common subscript letter differ (P < .05). b

Table 6. Effect of K-diformate on sensory quality of loin muscle in Exp. 2a Item Odor intensity Odor of meat Acidic odor Rancid odor Off-odor Flavor intensity Flavor of meat Acidic flavor Sweetness Bitterness Rancid flavor Off-flavor Firmness Tenderness Juiciness

Control

K-diformate

SEM

P-value

6.6 4.7 3.7 1.6 3.5 6.4 5.0 4.2 2.8 3.6 1.8 3.1 4.2 6.0 5.4

6.5 4.8 3.9 1.5 3.3 6.5 5.0 4.1 2.9 3.7 1.8 3.1 4.5 5.8 5.0

.2 .2 .2 .1 .3 .1 .1 .2 .1 .1 .1 .3 .1 .2 .1

.66 .98 .51 .51 .56 .39 .84 .68 .78 .83 .58 .99 .18 .52 .10

a Values are means for five pigs fed semi-ad libitum. Sensory parameters are given on a scale from 1 to 9, where 1 is no intensity and 9 the highest intensity.

content and the decreased carcass fat content occurred despite a higher daily nutrient intake of pigs fed the diets containing K-diformate. However, the daily intake of essential nutrients such as indispensable amino acids was also higher, which may have contributed to the increased carcass lean content. The results from Exp. 3 suggest that adding K-diformate to diets for growing-finishing pigs leads to a dosedependent increase in carcass quality. These findings are in agreement with those of Kirchgessner et al. (1992a), who found that adding formic acid to diets (.6 to 2.4%) resulted in a linear increase in protein content of the carcass. Roth et al. (1998) reported an increased apparent total-tract digestibility of amino acids as well as the retention of N by weanling pigs receiving dietary K-diformate. Jongbloed and Jongbloed (1996) reported an improved apparent total-tract digestibility of N in growing pigs receiving diets containing formic acid. Thus, the increase in carcass lean observed in the present experiment could have been the result of an increase in the digestibility and retention of N and a higher overall nutrient availability associated with the addi-


tion of K-diformate in diets. Roth et al. (1996), however, found no significant decrease in side fat thickness, or increase in percentage carcass lean or meat:fat ratio in the carcass when adding .65, 1.3, or 1.95% K-diformate to pig diets. These discrepancies in the response of carcass traits to the addition of K-diformate could have been due to differences in protein level and quality of the diets as well as in pig genotype. The improvement in protein digestibility is expected to be greater with a protein source of a lower quality than a higher quality as reported by Partanene and Mroz (1999). Also, the effect on performance and carcass quality from K-diformate is expected to be greater when pigs are provided with a limited supply of protein and other nutrients. Roth et al. (1996) utilized soybean meal as a protein source and the CP content of the grower and finisher diets was 166 g and 145 g per kilogram of feed, respectively. In Exp. 3, the level of CP was 155 g per kilogram of feed during the overall period and the protein sources were meat and bone meal, fish meal, canola meal, and fish silage. The higher fed intake of the pigs receiving K-diformate in the present experiment could also have contributed to the difference in response. Roth et al. (1996) reported no differences in feed intake among the different treatment groups. Sensory Quality. Results from the sensory analyses in Exp. 2 showed no significant differences in meat characteristics between the control and K-diformatefed animals for any of the sensory parameters tested, except for muscle juiciness, in which the muscle from the control pigs tended to have a greater juiciness (Table 6). No formate was detected in the liver, kidney, or muscle tissue, indicating no accumulation of formate in these tissues. The detectable level for the method is 20 ± 1 ␮g/g. These results show that the addition of Kdiformate to diets at a level equal to .8% did not affect sensory quality of the pork or cause retention of formate in edible tissues such as muscle, liver, and kidneys. Stomach Alterations. Adding 1.2% K-diformate to diets had no negative effect on the severity of stomach keratinization or incidence of stomach ulcers of pigs when fed the diet for approximately 85 d compared with those fed the control diet. The average stomach lesion scores were 2.8 (SD ± .97), 2.7 (SD ± 1.00), and 2.5 (SD ± 1.05) in the 0%, .6%, and 1.2% K-diformate-fed pigs, respectively. No stomach ulcers (score > 5) were observed in pigs receiving the diets containing 1.2% Kdiformate, whereas 2 out of 31 pigs had a stomach ulceration score of 5 from the .6% K-diformate treatment. These results indicate that dietary levels up to 1.2% Kdiformate have no negative effect on stomach alteration scores of growing-finishing pigs fed typical Norwegian diets, in accordance with the results of Fjetland and Jarp (1998). Gastrointestinal Tract Microflora. The effect of K-diformate on the total content of coliform bacteria in various segments of the gastrointestinal tract in Exp. 3 is shown in Figure 1. Adding 1.2% K-diformate to diets decreases the total number of coliform bacteria in the

1883

Figure 1. Total coliforms in gastrointestinal tract of growing-finishing pigs. The error bars represent SEM of each treatment and sample location. Statistical significant difference between treatments on each sampling location is indicated by **P < .05 or *P < .1. duodenum (P < .03), jejunum (P < .02) and rectum (P < .10). The reduction was 1 to 2 logarithmic units in all segments of the gut. In the present study, the focus was on coliform bacteria, which belong to the enterobacterial group. Escherichia coli is by far the dominant type among the coliform bacteria in the gastrointestinal tract. The effect of K-diformate on coliform numbers can be assumed to be representative for most enteric bacteria, such as Escherichia coli and the genera Salmo¨ stling and Lindnella, Shigella, and Enterobacter (O gren, 1993). These results indicate that the growth-promoting effect of K-diformate can be partially explained by its inhibitory effect on the population of coliform bacteria in the gastrointestinal tract. Because the microbial population is reduced, the metabolic needs are reduced and the availability of dietary energy and nutrients to the host animal is increased, resulting in increased growth rate and enhanced feed efficiency. Reduced number of coliform may also result in a better health status of the pig. The primary mode of action is via an antimicrobial effect of formic acid and formate resulting from the dissociation of K-diformate in the gastrointestinal tract. It is mainly the undissociated form of the formic acid ¨ stling and Lindthat exerts the antimicrobial effect (O gren, 1993). The concentration of the undissociated acid increases with decreasing pH. Formic acid passively diffuses through the bacteria cell wall, interfering with the fine-tuned pH balance in the cytoplasm, which results in a fatal interruption of the energy balance and biochemical processes. There is also a suppression of the cells’ enzymes and nutrient transport systems, which inhibits the ability of the bacteria to multiply (Partanen and Mroz, 1999; Roth and Kirchgessner, 1998). The reduction in coliform bacteria in the present study is in agreement with Gedek et al. (1992) and Kirchgessner et al. (1992b), in which the addition of formic acid to diets resulted in a reduction in coliform bacteria in the gastrointestinal tract. Other studies have also shown a reduction of coliform bacteria in the gastrointestinal tract of adding formic acid (Eckel et al.,

1884

Øverland et al.

1992) or K-diformate (Canibe and Jensen, unpublished data) to diets for weanling pigs.

Implications Adding K-diformate to barley-oat-based swine diets improved the growth performance and increased carcass lean content of growing-finishing pigs. Under similar conditions, the addition of Ca/Na-formate had no effect. Supplementing swine diets with up to 1.2% Kdiformate did not have any negative effect on stomach ulcerations of pigs. Potassium diformate is, therefore, an effective and safe feed additive in growing-finishing pig diets.

Literature Cited AOAC. 1990. Official Methods of Analysis (15th Ed.). Association of Official Analytical Chemists, Arlington, VA. Baustad, B., and I. Nafstad. 1969. Gastric Ulcers in Swine. 4. Effects of Dietary Particle Size and Crude Fiber Contents on Ulceration. Pathol. Vet. 6:546−556. Bolduan, G., H. Jung, R. Schneider, J. Block, and B. Klenke. 1988. Die Wirkung von Propion- und Ameisensa¨ ure in der Ferkelaufzucht. J. Anim. Physiol. Anim. Nutr. 59:72−78. Eckel, B., F. X. Roth, M. Kirchgessner, and U. Eidelburger. 1992. Zum Einfluß von Ameisensa¨ ure auf die Konzentratin an Ammoniak und biogenen Aminen im Gastrointestinaltrakt. Mitteilung. Untersuchungen zur nutritiven Wirksamkeit von organischen Sa¨ uren in der Ferkelaufzucht. J. Anim. Physiol. Anim. Nutr. 67:198−205. Fjetland, O., and J. Jarp. 1998. The effect of different feed compounds on gastric ulcer prevalence and severity in slaughter pigs. In: Proc. 15th IPVS Congr., Birmingham, England. 3:398. Gedek, B., M. Kirchgessner, U. Eidelburger, S. Wiehler, A. Bott, and F. X. Roth. 1992. Zum Einfluß von Ameisensa¨ ure auf die Keimzahlen der Mikroflora und deren Zusammensetzung in verschiedenen Segmenten des Gastrointestinaltraktes. 5 Mitteilung zur nutritiven Wirksamkeit von organischen Sa¨ uren in der Ferkelaufzucht. J. Anim. Physiol. Anim. Nutr. 67:206−214. ISO. 1985. 6564-1985 Sensory Analysis-Methodology-Flavour Profile Methods (1st Ed.). International Organisation for Standardisation, Geneva, Switzerland. Jongbloed, A. W., and R. Jongbloed. 1996. The effect of organic acids in diets for growing pigs on enhancement of microbial phytase efficiacy. ID-DLO Report No. 96009, Institute for Animal Science and Health, Lelystad, The Netherlands. p 19. Kirchgessner, M., B. Gedek, F. X. Roth, and U. Eidelburger. 1992a. Zum Einfluß von Ameisensa¨ ure auf Ko¨ rperzusammensetzung und Na¨ hrstoffretention. 2. Mitteilung. Untersuchungen zur nutritiven Wirksamkeit von organischen Sa¨ uren in der Ferkelaufzucht. J. Anim. Physiol. Anim. Nutr. 67:101−110. Kirchgessner, M., B. Gedek, S. Wiehler, U. Eidelburger, and F. X. Roth. 1992b. Zum Einfluß von Ameisensa¨ ure, Calcium formiat

und Natrium-hydrogencarbonat auf die Keimzahlen der Mikroflora und deren Zusammensetzung in verschiedenen Segmenten des Gastrointestinaltraktes. J. Anim. Physiol. Anim. Nutr. 67:73−81. Kirchgessner M., B. R. Paulicks, and F. X. Roth. 1997. Effects of supplementation of diformate complexes (Formi LHS) on growth and carcass performance of piglets and fattening pigs in response to application time. Agribiol. Res. 50:1−6. Kirchgessner, M., and F. X. Roth. 1988. Ergotrope Effekte durch organische Sa¨ uren in der Ferkelaufzucht und Schweinemast. ¨ bers. Tiererna¨ hr. 16:93−108. U MacConkey, A. 1905. Lactose-fermenting bacteria in feces. J. Hyg. 5:333−378. Moore, S. 1963. On the determination of cystine as cysteic acid. J. Biol. Chem. 238:235−237. NRC. 1988. Nutrient requirements of swine (9th Ed.). National Academy Press, Washington, D.C. ¨ stling, C. E., and Lindgren, S. E. 1993. Inhibition of enterobacteria O and Listeria growth by Lactic acitic and formic acids. J. Appl. Bacteriol. 75:18−24. Øverland, M. 1997. New restricted feeding scales for growing-finishing pigs and sows in Norway. In: Proc. NJF-seminar No. 274. Energy and protein evaluation for pigs in the Nordic countries. Research Centre Foulum, Denmark. pp 86−92. Øverland M., and A. Lysø. 1995. Addition of formic acids and/or formates in diets for growing-finishing pigs. Proc. 48th Annual meeting of the EAAP, Vienna, Austria. p 6. Partanen, K., and Z. Mroz. 1999. Organic acids as an alternative for prophylactic medication of pig diets. Nutr. Res. Rev. 12:1−30. Paulicks, B. R., F. X. Roth, and M. Kirchgessner. 1996. Dose effects of potassium diformate (Formi LHS) on the performance of growing piglets. Agribiol. Res. 49:318−326. Proha`szka, L., and F. Baron. 1980. The predisposing role of high dietary protein supplies in enteropathogenic E. coli infections of weaned pigs. Zentbl. Vetmed. Reihe B. 27:222−232. Roth, F. X., and M. Kirchgessner. 1998. Organic acids as feed additives for young pigs: Nutritional and gastrointestinal effects. J. Anim. Feed Sci. 7:25−33. Roth, F. X., M. Kirchgessner, and B. R. Paulicks. 1996. Nutritive use of feed additives based on diformates in the rearing and fattening of pigs and their effects on performance. Agribiol. Res. 49:307−317. Roth, F. X., W. Windisch, and M. Kirchgessner. 1998. Effect of potassium diformate (FormiLHS) on nitrogen metabolism and nutrient digestibility in piglets at graded dietary lysine supply. Agribiol. Res. 51:167−175. SAS. 1990. SAS user’s guide: Statistics (Version 6, 4th Ed.). SAS Inst. Inc., Cary, NC. SOU. 1997. Antimicrobial Feed Additives. Report from the Commission on Antimicrobial Feed Additives Stockholm 1997, SOU 132:83−140. SSI. 1979. Color Atlas. Swedish Standard SS 01 91 02, Swedish Standard Institution, Stockholm, Sweden. Sundstøl, F. 1973. The correlation between some characters obtained from feeding experiments with Pigs. Acta Agric. Scand. 23:93−95.