1st Page

Journal of Animal Science - Animal Nutrition

Effects of prepartum dietary cation-anion difference and acidified coproducts on dry matter intake, serum calcium, and performance of dairy cows1


This article in JAS

  1. Vol. 92 No. 2, p. 666-675
    Received: Jan 31, 2013
    Accepted: Nov 25, 2013
    Published: November 24, 2014

    3 Corresponding author(s):

  1. D. J. Rezac*,
  2. E. Block,
  3. D. Weber22,
  4. M. J. Brouk* and
  5. B. J. Bradford 3
  1. Department of Animal Sciences and Industry, Kansas State University, Manhattan 66506
    Arm & Hammer Animal Nutrition, Princeton, NJ 08543


Two products designed to deliver supplemental anions were evaluated for their effects on DMI, total serum Ca, and performance of transition dairy cows relative to a control diet that did not contain supplemental anions. Diets differed in dietary cation-anion difference (DCAD) and anion source. Treatments were diets including a control (CON; DCAD +17.7 meq/100 g DM; n = 13), Bio-Chlor (BC; DCAD +2.5 meq/100 g DM; n = 14), and SoyChlor (SC, DCAD +0.4 meq/100 g DM; n = 15). Treatments began 21 d before expected calving dates and continued through parturition (mean treatment length 20.98 d); on calving, all animals received the same diet. Milk yield was measured through 21 d in milk, and milk samples were collected daily between 5 and 21 d in milk. Data were analyzed using mixed models with repeated measures. Prepartum DMI was 9.0, 8.5, and 7.5 ± 0.6 kg/d for CON, BC, and SC treatments, respectively, and tended to be lower for SC than CON (P = 0.07). Postpartum DMI and milk yields were similar among treatments. Milk protein, lactose, and urea nitrogen concentrations were highest for SC and lowest for BC, with CON being intermediate. Plasma glucose, measured on d 5, 10, and 21 postpartum, tended to be different among treatments (P = 0.06; 66.7, 57.1, and 63.8 ± 3.1 mg/dL for CON, BC, and SC, respectively). Serum total Ca concentrations did not differ among dietary treatments and only tended to change over time; values were not indicative of clinical hypocalcemia. With limited sample size, no significant effects of treatment were detected for incidence of postpartum health disorders or plasma β-hydroxybutyrate concentration. Although DMI tended to be depressed in the prepartum period by SC, this intake depression was not accompanied by negative effects on performance or health in the postpartum period. Results suggest that cows were not adequately stressed to cause hypocalcemia or that DCAD values near 0 were insufficient to improve postpartum health and performance or both.


Decreasing the dietary cation-anion difference (DCAD) of prepartum diets has a positive influence on Ca metabolism (Block, 1984; Joyce et al., 1997) and can decrease the incidence of milk fever in dairy cattle (Block, 1984; Horst et al., 1997; Charbonneau et al., 2006; Lean et al., 2006). However, meta-analysis has also shown that decreasing the DCAD can result in a decrease in DMI (Charbonneau et al., 2006). Overton and Waldron (2004) concluded feeding strategies that support increased energy supply during the close-up period are optimal for peripartum metabolic health. Thus, adopting a feeding strategy that results in greater DMI depression during the close-up period may prove detrimental to the health and performance of the cow.

To address the suboptimal DMI observed with anionic salts, products containing supplemental anions were designed with the goal of decreasing DCAD but with more acceptable palatability and thus fewer negative effects on DMI. Because these ingredients have been formulated using coproducts of various milling processes, they are referred to as acidified coproducts (ACP). Published research comparing the effects of ACP to anionic salts or effects among different ACP in transition cows is limited, but Vagnoni and Oetzel (1998) observed the lowest DMI for dry cows fed a total mixed ration (TMR) containing an ACP compared with anionic salts and a control diet. The impact of ACP on prepartum DMI therefore remains unsettled.

Our primary objective was to evaluate DMI responses to rations containing 1 of 2 commercially available ACP relative to a control ration containing no supplemental anions when fed to prepartum dairy cows. Secondary objectives included analysis of peripartum serum Ca and early lactation performance. We hypothesized that diets containing ACP would support adequate prepartum DMI, maintain serum Ca concentrations at parturition, and positively influence postpartum milk production.


Experimental Design and Treatments

All experimental procedures were approved by the Kansas State University Institutional Animal Care and Use Committee. Twenty-nine multiparous (7 in parity 3, 22 in parity 2) and 16 primiparous pregnant Holstein cows were selected from the Kansas State University Dairy Teaching and Research Unit herd, blocked by parity (primiparous vs. multiparous), and randomly assigned to 1 of 3 treatments beginning 21 d before expected parturition. Treatment diets (Table 1) differed in DCAD and source of anionic supplement and were control (CON; target DCAD +10 meq/100 g DM; n = 9 cows, n = 4 heifers), Bio-Chlor (BC; Church and Dwight Co. Inc., Princeton, NJ; target DCAD -10 meq/100 g DM; n = 8 cows, n = 6 heifers), and SoyChlor (SC; West Central Cooperative, Ralston, IA; target DCAD -10 meq/100 g DM; n = 9, cows n = 6 heifers). Although heifers are less likely to experience periparturient hypocalcemia than cows (Reinhardt et al., 2011), heifers were included because they have been observed to have more negative DMI responses to anionic supplements (Moore et al., 2000) and they are comingled with prepartum cows on some commercial dairies. Early in the study, urine pH was evaluated for randomly selected cows on each treatment to verify that anionic supplements induced urine acidification (8.00, 7.25, and 6.88 ± 0.13 for CON, BC, and SC, respectively; CON differed from other treatments at P < 0.05 by Student’s t test, n = 7).

View Full Table | Close Full ViewTable 1.

Ingredients and chemical composition of control (CON), Bio-Chlor (BC), SoyChlor (SC), and lactation diets

Item CON BC SC Lactating
Ingredient, % DM
    Corn silage 22.7 22.7 22.7 23.7
    Wheat straw 37.8 37.8 37.8
    Wet corn gluten feed1 11.2 11.2 11.2 32.2
    Alfalfa hay 12.0
    Rolled corn 10.2 10.0 10.1 19.3
    Soybean hulls 4.5 4.2 1.6 1.8
    Solvent soybean meal (48% CP) 9.5 4.8 8.0
    Mechanically extracted soybean meal2 6.1
    Blood meal 0.8 0.8 0.8
    Bio-Chlor3 5.3
    SoyChlor4 5.8
    Limestone 1.0 1.0 1.6
    Magnesium oxide 0.3 0.3 0.8
    Molasses 1.5 1.5 1.5
    Micronutrient premix5 0.5 0.5 0.5 2.5
Chemical composition, % DM
    DM, % as fed 49.7 49.6 49.5 57.4
    CP 15.0 14.9 15.2 18.3
    ADF 30.8 31.0 31.1 14.0
    NDF 49.4 49.7 49.1 30.1
    Non-fiber carbohydrate (NFC) 30.2 29.5 29.8 44.5
    EE 2.1 2.1 2.3 3.2
    Ca 0.84 0.80 0.62 0.76
    P 0.34 0.34 0.34 0.50
    Mg 0.36 0.37 0.34 0.31
    K 1.26 1.21 1.21 1.43
    Na 0.09 0.14 0.10 0.32
    Cl 0.19 0.64 0.80 0.26
    S 0.20 0.26 0.21 0.28
    NEL,7 Mcal/kg DM 1.21 1.21 1.23 1.72
    DCAD,8 mEq/100 g DM +17.7 ± 1.9 +2.5 ± 2.4 +0.4 ± 2.6 +25
1Sweet Bran, Cargill Inc., Blair, NE.
2SoyBest, Grain States Soya, West Point, NE.
3Bio-Chlor, Church and Dwight Co. Inc., Princeton, NJ. Contains (% DM) 48.6% CP, 4.9% ADF, 17.3% NDF, 17.4% NFC, 3.7% ether extract (EE), 0.09% Ca, 0.79% P, 2.1% Mg, 1.2% K, 1.5% Na, 9.1% Cl, and 3.6% S.
4SoyChlor, West Central Cooperative, Ralston, IA. Contains (% DM) 20.1% CP, 22.4% ADF, 26% NDF, 27.9% NFC, 5.1% EE, 4.5% Ca, 0.30% P, 2.8% Mg, 0.48% K, 0.04% Na, 10.3% Cl, and 0.35% S.
5CON, BC, and SC micronutrient premix consisted of 12.1% trace mineral salt, 3.7% 4-Plex (Zinpro Corp., Eden Prairie, MN), 7.3% selenium premix (0.06%), 5.0% vitamin A (30,000 IU/g), 5.0% vitamin D (30,000 IU/g), 64.3% vitamin E (44 IU/g), 0.7% ethylenediamine dihydroiodide (44.1 mg/g), and 3.6% Rumensin 80 (Elanco Animal Health, Greenfield, IN). Lactating micronutrient premix consisted of 5.2% trace mineral salt, 24.1% menhaden fish meal, 27.5% sodium bicarbonate, 3.1% MFP (Novus International Inc., St. Charles, MO), 1.9% 4-Plex (Zinpro Corp.), 1.1% selenium premix (0.06%), 0.5% vitamin A (30,000 IU/g), 0.2% vitamin D (30,000 IU/g), 10.3% vitamin E (44 IU/g), 0.03% ethylenediamine dihydroiodide (44.1 mg/g), 0.22% Rumensin 80 (Elanco Animal Health), 17.2% XP Yeast (Diamond V Mills, Cedar Rapids, IA), and 8.6% Reashure (Balchem Corp., New Hampton, NY).
6Calculated as DM – (ash + NDF + CP + EE).
7Estimated according to NRC (1989).
8Values for treatment diets are means ± SD, n = 6. The CON diet differed from both BC and SC (P < 0.001), but BC and SC did not differ (P = 0.16).

Treatments were formulated for similar CP using a combination of solvent-extracted soybean meal and soybean hulls to balance diets. In addition, limestone and magnesium oxide were used to try to minimize differences in mineral content of diets, except that Cl content was allowed to increase with ACP treatments, by design. Treatment diets were offered ad libitum beginning 21 d before expected parturition date through parturition. After parturition, all cows received a common lactation TMR (Table 1) offered ad libitum. Prepartum diets were mixed daily using a batch mixer (Precision Horizontal Batch Mixer, H. C. Davis Sons Manufacturing, Bonner Springs, KS) and were fed once daily at 1400 h. Wheat straw particle size was reduced using a bale processor (model BPX900, Vermeer Corp., Pella, IA) set for an average cut length of 7.6 cm. Dry matter content of the prepartum diets was lowered to 50% by adding water to the diets at mixing time. The lactation TMR was mixed daily in a TMR wagon (414-14B Forage Express, Roto-Mix, Dodge City, KS) and was offered at 0700 and 1400 h. Cows were housed in a tie-stall facility from the beginning of the study through 14 d postpartum and then were moved to a free-stall facility, where they continued to receive the lactation ration until 21 d postpartum.

Data and Sample Collection and Analysis

Cows were milked 3 times daily at 0200, 1000, and 1800 h; milk yield was recorded at each milking through 21 d postpartum, and milk samples were collected at all milkings between 5 and 21 d postpartum. Samples were analyzed for concentrations of fat, true protein, lactose (B-2000 Infrared Analyzer, Bentley Instruments, Chaska, MN), milk urea nitrogen (MUN; MUN spectrophotometer, Bentley Instruments), and somatic cells (SCC 500, Bentley Instruments; Heart of America DHIA, Manhattan, KS). Analyzed values for each milking were composited by day. Samples of feed ingredients were collected weekly and composited by month for analysis. Feed ingredient samples were analyzed using standard wet chemistry methods (Dairy One, Ithaca, NY) for DM, CP, NDF, ADF, ether extract (EE), Ca, P, Mg, K, Na, Cl, and S. Dry matter was determined by drying samples in a forced-air oven at 105°C for 8 h. Concentration of CP was determined by oxidation and detection of N2 using a Leco FP-528 combustion analyzer (Leco Corp, St. Joseph, MI, AOAC method 990.03). Concentration of NDF was determined (Van Soest et al., 1991) using an ANKOM A200 filter bag technique (ANKOM Technology, Macedon, NY). Concentration of ADF was determined using an ANKOM A200 filter bag technique with solutions the same as given in AOAC method 973.18 (AOAC, 2000). Crude fat concentration was determined by ether extract (AOAC method 2003.05). Concentrations of Ca, P, Mg, K, and Na were determined using a Thermo IRIS Advantage HX Inductively Coupled Plasma Radial Spectrometer (Thermo Fisher Scientific Inc., Waltham, MA) after microwave digestion. Chloride ion concentration was determined using a Brinkman Metrohm 716 Titrino Titration unit (Riverview, FL) and sulfur concentration using near-infrared reflectance spectroscopy (AAOC method 989.03). Dietary cation anion difference was calculated as [(Na + K) – (Cl + S)] meq/100 g.

Blood samples for total Ca analysis were collected from coccygeal vessels into serum Vaccutainer tubes (Becton Dickinson and Company, Franklin Lakes, NJ) at 1300 h beginning 7 d before expected parturition and continuing through 5 d postpartum. Serum samples were allowed to clot at 20°C for 60 min and then were centrifuged at 2,500 × g for 5 min at 20°C; serum was removed and frozen at -20°C until analysis. Additionally, on d 5, 10, and 21 postpartum blood samples were collected at 1300 h into EDTA-containing Vacutainers (Becton Dickinson and Company) and were centrifuged at 2,500 × g for 5 min at 20°C within 10 min of collection, and plasma was removed and frozen at -20°C until analysis for plasma glucose and β-hydroxybutyrate (BHBA). Serum Ca was determined by atomic absorption spectroscopy according to Bowers and Rains (1988). Glucose was determined in plasma samples using a commercial kit (Wako Chemicals USA, Richmond, VA) on the basis of the method of Raabo and Terkildsen (1960), and plasma BHBA was measured (Pointe Scientific, Canton, MI) according to McMurray et al. (1984).

Displaced abomasum, retained placenta, metritis, mastitis, and milk fever were diagnosed according to Kelton et al. (1998). Clinical ketosis was identified using urinary ketone reagent strips (ReliOn, Bayer Healthcare LLC, Mishawaka, IN). Ketosis was recorded when urine acetoacetate exceeded 40 mg/dL for 2 consecutive days. Animals diagnosed with clinical ketosis were treated orally with Keto-Gel (Jorgensen Laboratories Inc., Loveland, CO) once daily at 0700 h until urine acetoacetate decreased to less than 40 mg/dL. Body condition scores were evaluated by 2 trained investigators, and BW were recorded on d –21 and –7 relative to expected parturition and d 1, 10, and 21 postpartum (difference between actual and expected calving date: mean = -0.02 d, interquartile range = –7 to 2 d).

Statistical Analysis

A total of 45 animals were included in the study; however, 3 (all multiparous, 1 on each treatment) were removed because of unrelated health events (pyrexia and bovine respiratory disease). Cows with other disorders, which may have been related to treatment, were included in the analysis. Data were analyzed using the MIXED procedures of SAS (version 2001; SAS Inst. Inc., Cary, NC) with repeated measures over time by modeling the fixed effects of treatment, parity, day relative to parturition, day × treatment interaction, parity × treatment interaction, day × parity interaction, and day × parity × treatment interaction, as well as the random effect of animal. Autoregressive (AR[1]) or heterogeneous autoregressive (ARH[1]) covariance structures were used to model repeated measures over time within cow; the best-fitting structure was chosen for each variable on the basis of the Bayesian information criterion. Spatial power covariance structures were used for BW, BCS, BHBA, and glucose data because of unequal spacing of sample times. Prepartum and postpartum DMI data were analyzed separately. When treatment effects were significant, means were separated by Student’s t test. Treatment differences were declared significant at P < 0.05, and tendencies were declared at 0.05 < P < 0.10. Incidences of periparturient disorders were tested by Fisher’s exact test.


Chemical analysis of feed ingredients revealed that diets closely matched targets for NDF, non-fiber carbohydrate, and CP concentrations (Table 1), but DCAD values were substantially more cationic than the target values of +10 and -10 meq/100 g DM. Differences in DCAD values were attributed to greater than expected K concentrations in corn silage and grain mixes. Even though target DCAD values were not achieved, the difference of approximately 20 meq/100 g DM between the control and ACP diets was maintained.

Pre- and postpartum means for DMI, milk yield, and milk composition are shown in Fig. 1 and Table 2, respectively. Prepartum DMI tended to be greater for CON than SC (P = 0.07), with BC being intermediate. Postpartum DMI through 14 d in milk (DIM) did not differ by parity or treatment. Milk yield, energy-corrected milk yield, fat yield, fat content, protein yield, lactose yield, and SCC were not altered by treatment through 21 DIM; however, lactose content (P = 0.02), protein content (P = 0.03), and MUN concentration (P < 0.01) were all significantly greater for SC than for BC, with CON being intermediate (Table 2). A significant day × parity × treatment interaction was detected for milk fat content. Fat content was greater on d 15 postpartum for primiparous cows consuming CON prepartum (P < 0.05) and was greater on d 5, 11, 16, 17, and 21 for multiparous cows consuming BC prepartum (P < 0.05).

Figure 1.
Figure 1.

Dry matter intake responses to prepartum acidified coproducts. (A) Prepartum DMI for cows that consumed the control (CON), Bio-Chlor (BC; Church and Dwight Co. Inc., Princeton, NJ), or SoyChlor (SC; West Central Cooperative, Ralston, IA) diets beginning 21 d before expected calving differed by day (P < 0.01) and tended to differ among treatments (P = 0.07) but did not differ by parity (P = 0.71). Treatment × day interaction: P = 0.27; treatment × parity × day interaction: P = 0.30; pooled SEM = 0.66. (B) Postpartum DMI was affected by day (P < 0.01) but not by parity (P = 0.20) or prepartum treatment (P = 0.61). Treatment × day interaction: P = 0.51; treatment × parity × day interaction: P = 0.16; pooled SEM = 1.17.


View Full Table | Close Full ViewTable 2.

Effects of control (CON), Bio-Chlor (BC), and SoyChlor (SC) treatments on milk production and milk composition through 21 d in milk

Item CON BC SC SEM Day Parity Treatment
Milk yield, kg/d 31.9 33.9 34.6 2.4 <0.01 <0.01 0.70
ECM, kg/d 35.3 37.6 37.2 2.3 0.08 <0.01 0.74
Fat yield, kg/d 1.32 1.46 1.34 0.08 0.13 <0.01 0.41
Fat, % 4.33 4.46 4.01 0.2 <0.01 0.03 0.27
Protein yield, kg/d 1.01 1.01 1.13 0.09 <0.01 <0.01 0.46
Protein, % 3.18a,b 2.99b 3.29a 0.1 <0.01 0.55 0.03
Lactose yield, kg/d 1.52 1.59 1.68 0.13 <0.01 <0.01 0.62
Lactose, % 4.70a,b 4.66b 4.87a 0. 1 <0.01 0.08 0.02
SCC, 1,000 cells/mL 162 131 107 68 0.98 0.40 0.86
MUN, mg/dL 8.82a,b 8.57b 9.11a 0.12 <0.01 0.19 <0.01
a,bMeans within a row that do not share a superscript differ (P < 0.05).
1BioChlor is manufactured by Church and Dwight Co. Inc., Princeton, NJ. SoyChlor is manufactured by West Central Cooperative, Ralston, IA.

Treatment did not affect BCS during the peripartum period (P = 0.68), but BW tended to differ between treatments (P = 0.09, Fig. 2). Differences in BW between treatments largely reflected differences on enrollment, with no evidence of treatment × day interactions (P = 0.99). Multiparous cows also lost more BCS than primiparous cows during the study (parity × day interaction, P < 0.001).

Figure 2.
Figure 2.

Peripartum BCS and BW response to prepartum acidified coproducts. Body condition score and BW data were collected on d -21, –7, 1, 10, and 21 relative to calving. (A) Peripartum BCS did not differ between control (CON), Bio-Chlor (BC; Church and Dwight Co. Inc., Princeton, NJ), or SoyChlor (SC) diets prepartum (P = 0.68). However, BCS was affected by parity (P = 0.04) and day (P < 0.01). Treatment × day interaction: P = 0.84; treatment × parity × day interaction: P = 0.07. (B) Peripartum BW tended to be affected by prepartum treatment (P = 0.09) and was significantly affected by both parity (P < 0.01) and day (P < 0.01). Treatment × day interaction: P = 0.99; treatment × parity × day interaction: P = 0.92.


Mean plasma glucose concentrations measured on d 5, 10, and 21 postpartum tended to be greater for CON than for BC (P = 0.06; Fig. 3A), with SC being intermediate. No significant effect of time was observed on postpartum plasma glucose concentration, but a significant treatment × parity × day interaction (P < 0.01) occurred. Plasma BHBA concentrations analyzed on d 5, 10, and 21 postpartum were not altered by treatments (Fig. 3B) but differed over time (P < 0.01), with means of 9.98, 6.99, and 6.09 ± 0.69 mg/dL on d 5, 10, and 21 postpartum, respectively. Peripartum serum Ca concentrations did not differ among treatments and only tended to change over time (P = 0.07). Serum Ca concentrations from d –7 prepartum through d 5 postpartum are shown in Fig. 3.

Figure 3.
Figure 3.

Postpartum plasma glucose and β-hydroxybutyrate (BHBA) concentrations in primiparous and multiparous cows fed different prepartum diets. (A) Plasma glucose concentrations were greater for primiparous cows that consumed SoyChlor (SC; West Central Cooperative, Ralston, IA) prepartum on d 10 postpartum than for cows that consumed Bio-Chlor (BC) and the control diet (CON) prepartum (P < 0.05) but were greater for cows that consumed BC prepartum on d 21 than for cows that consumed SC and CON prepartum (P < 0.05). Postpartum plasma glucose in multiparous cows was not affected by treatment on any day. Treatment: P = 0.06; treatment × parity × day interaction: P < 0.01. (B) Postpartum plasma BHBA concentrations differed by day (P < 0.01) but not by parity (P = 0.27) or treatment (P = 0.91). Treatment × day interaction: P = 0.58; treatment × parity × day interaction: P = 0.36.


No differences were observed in frequencies of postpartum health disorders (Table 3). The most prevalent disorder was ketosis, and only 2 cows were diagnosed with clinical signs of milk fever during the study, with 1 case each for CON and BC and no cases for SC. Of the 17 animals diagnosed with ketosis, 14 were multiparous (5, 5, and 4 from CON, BC, and SC, respectively). Two primiparous cows from CON and 1 from BC were diagnosed with ketosis. All cows diagnosed with a displaced abomasum were multiparous. Retention of fetal membranes 24 h after parturition occurred in 2 cows from CON and 1 from SC.

View Full Table | Close Full ViewTable 3.

Postpartum health disorders for control (CON), Bio-Chlor (BC), and SoyChlor (SC) treatments1

Item Treatment
Cows on treatment 13 14 15
Ketosis 7 6 4
Displaced abomasum 3 1 1
Retained placenta 2 0 1
Metritis 1 0 0
Mastitis 1 1 0
Milk fever 1 1 0
Subclinical hypocalcemia2 10 8 9
Clinical hypocalcemia3 2 2 1
1No significant treatment effects detected using Fisher’s exact test. BioChlor is manufactured by Church and Dwight Co. Inc., Princeton, NJ. SoyChlor is manufactured by West Central Cooperative, Ralston, IA.
2Serum calcium 5.6 to 8.0 mg/dL (Goff, 2008) at least once in the first 72 h postpartum.
3Serum calcium <5.6 mg/dL (Goff, 2008) at least once in the first 72 h postpartum.

Dietary cation-anion difference values for BC and SC did not differ greatly from values used in other experiments. Vagnoni and Oetzel (1998) offered a ration to dry cows containing Bio-Chlor as an anion source to decrease the DCAD to –5.1 meq/100 g DM; this treatment decreased urine pH from 8.33 to 6.37 ± 0.20, demonstrating the acidogenic properties of the diet. Peterson et al. (2005) offered diets with DCAD values of -1.30 and -2.00 meq/100 g DM containing SoyChlor as an anion source, and this DCAD concentration was apparently sufficient to maintain peripartum serum Ca concentrations above the clinical hypocalcemia threshold (>5.6 mg/dL; Goff, 2008), although there was not a positive DCAD treatment for comparison.

Prepartum DMI for SC tended to be lower than CON (P = 0.09) but did not differ from BC. No data were collected to directly evaluate differences in acidogenic properties of the diets, but the degree of acidosis inflicted by a lowered DCAD has been suggested to be correlated with the extent to which DMI is reduced (Vagnoni and Oetzel, 1998). This proposed relationship is based on evidence that anionic salts that were the least acidogenic (Oetzel et al., 1991) also had minimal negative effects on intake compared with other, more acidogenic anionic salts (Oetzel and Barmore, 1993). Decreases in DMI for diets containing supplementary anions traditionally have been associated with palatability issues (Horst et al., 1997), but the potential role of physiological sensing of absorbed ions has not been assessed carefully. In the present study, DMI was lower for all treatments than observed for diets with similar DCAD (Vagnoni and Oetzel, 1998; Moore et al., 2000), and no difference in parity was observed as in Moore et al. (2000). All prepartum diets in the current study contained wheat straw at 38% of diet DM, which was expected to promote satiety. Dann et al. (2006) fed diets containing wheat straw at 26% of DM to cows from dry off to 25 d before expected parturition and observed a mean DMI of 10.4 ± 0.5 kg DM/d. Our diets contained substantially more wheat straw, which could explain the lower intakes we observed.

Mean serum Ca concentrations remained above the threshold for clinical hypocalcemia (5.6 mg/dL) throughout the peripartum period for all treatments. The ability to make inferences in this study is limited, therefore, by the lack of serum Ca response to the cationic control diet; surprisingly, CON (DCAD +17.7 meq/100 g DM) did not cause greater occurrence of hypocalcemia. A possible explanation for this observation may be the low concentration of K in these diets. Ramos-Nieves et al. (2009) formulated prepartum diets low in K with (DCAD -15 meq/100 g DM) and without (DCAD +11 meq/100 g DM) the addition of anions and concluded that the supplementation of anions to the low-K diets did not change the incidence of clinical or subclinical hypocalcemia. Potassium concentrations in our prepartum diets (Table 1) were less than the 1.29% of DM fed by Ramos-Nieves et al. (2009), suggesting that the lack of a significant decrease in serum Ca for CON may be related to the beneficial effect that low-K prepartum diets has on postpartum Ca homeostasis. Kurosaki et al. (2007) observed that a DCAD of +1.2 meq/100 g DM prevented the onset of milk fever in multiparous cows. Research has yet to elucidate an optimal prepartum DCAD, but these findings suggest that a DCAD concentration of –5 to +2.5 meq/100 g DM may be sufficient to attenuate the onset of clinical hypocalcemia at the time of parturition.

Total serum Ca did not differ among treatments or parities and only tended to differ over time (Fig. 4). It should be noted that the multiparous group was relatively young, with no cows greater than or equal to fourth parity; this likely contributed to the relative lack of acute hypocalcemia in this study (Reinhardt et al., 2011). Block (1984) reported that total serum Ca concentrations were greater for cows fed an anionic diet (DCAD = -12.9 meq/100 g DM) vs. a cationic diet (DCAD = +33.1 meq/100 g DM) on d -3, -2, -1, 0, 1, and 3 relative to parturition. Oetzel et al. (1988) observed similar results when feeding prepartum diets supplemented with ammonium salts to decrease DCAD to –7.5 meq/100 g DM. Similar to our results, Ramos-Nieves et al. (2009) did not observe a difference in mean peripartum total serum Ca concentration between treatments with differing DCAD. The lack of a dramatic decrease in serum Ca concentrations in the days around parturition suggests that cows did not undergo a large Ca stress at parturition, possibly because of some dietary or environmental factor that was independent of DCAD treatment. As stated above, cows fed low-K diets prepartum did not respond to an ACP with altered serum Ca concentrations (Ramos-Nieves et al., 2009), which was consistent with our observations. Ramos-Nieves et al. (2009) did, however, observe an effect of the ACP on peripartum serum concentrations of P and the occurrence of hypophosphatemia. Unfortunately, serum P was not measured in this study. The low concentration of the strong cation K in the diets presumably prevented a state of metabolic alkalosis, which has been shown to cause the parturient dairy cow’s inability to respond to the increased Ca demand by blunting tissue response to parathyroid hormone (Goff and Horst, 1997; Horst et al., 1997). Additionally, milk yields were not particularly high during this study (Table 2), which resulted in relatively moderate demand for milk Ca secretion and may have contributed to the lack of hypocalcemia across treatments.

Figure 4.
Figure 4.

Peripartum total serum Ca concentrations of cows fed different prepartum diets. Serum Ca concentration was measured beginning 7 d before parturition through 5 d postpartum. Serum Ca concentrations tended to decrease postpartum (P = 0.07). Pooled SEM = 0.09; BC = Bio-Chlor (Church and Dwight Co. Inc., Princeton, NJ); SC = SoyChlor ; CON = control.


Prepartum calcium intakes varied among treatments and were 75.6, 68.5, and 46.0 ± 4.3 g/d for CON, BC, and SC, respectively (P < 0.001). There are reasons to consider these differences when evaluating the responses observed herein. Calcium is a relatively strong cation (Horst et al., 1994), despite the fact that it is not included in the popular 4-mineral dietary cation-anion difference (DCAD4) equation used in designing the treatments here. Therefore, consideration of the Ca contribution to the strong ion pool would separate the treatments to a slightly greater extent than the DCAD4 equation would suggest (i.e., CON would be relatively higher and SC relatively lower). Second, because Ca is absorbed through both active and passive transport mechanisms at different dietary concentrations, it is possible that active transport mechanisms (which are more heavily regulated) would have had a differential impact across diets. Indeed, prepartum Ca intake has been associated with altered milk fever incidence (Boda, 1956; Goings et al., 1974; Green et al., 1981; Kichura et al., 1982). However, the amounts consumed by cows on all treatments described here were less than those associated with negative effects on Ca metabolism and greater than those associated with positive effects (Horst et al., 1994). It is possible that differences in dietary Ca had some impact on the outcomes of these treatments, but the Ca intake likely does not explain the relatively high postpartum serum Ca concentrations across treatments.

Publications reporting the effects of DCAD and/or anion source on production parameters are limited and offer mixed results (Block, 1984; Joyce et al., 1997; Moore et al., 2000; Roche et al., 2003; Ramos-Nieves et al., 2009). Block (1984) reported that cows fed cationic diets vs. anionic diets produced 6.8% less milk in a 305-d lactation period (P < 0.05), and when the cationic group was separated into animals diagnosed with milk fever and those who were healthy, 305-d milk yield was 14% greater (P < 0.05) for animals fed the anionic diet prepartum vs. those who were fed the cationic diet and diagnosed with milk fever. More recently, Ramos-Nieves et al. (2009) reported nearly identical milk production between early lactation cows fed prepartum diets with DCAD values of either +11 mEq/100 g of DM or -15 mEq/100 g of DM. Unfortunately, meta-analyses conducted to evaluate the effects of prepartum DCAD have not evaluated milk production responses, likely because most of the original studies did not include subsequent production data (Charbonneau et al., 2006; Lean et al., 2006). Investigators have focused mainly on the physiological effects of DCAD, aiming simply to prevent clinical and subclinical hypocalcemia. In general, little evidence shows that subclinical hypocalcemia has negative effects on productivity (Lucey et al., 1986); in fact, a recent study reported that cows with subclinical hypocalcemia sustained higher milk yields during the first month of lactation (Jawor et al., 2012).

Differences observed in milk protein concentration and MUN can most likely be explained by postpartum CP intake (2.10, 2.21, and 2.43 ± 0.24 kg CP/d for CON, BC, and SC, respectively). Although postpartum CP intake did not differ among treatments, intake for SC was numerically the greatest, which could have contributed to both increased MP supply and urea synthesis for that treatment.

No significant effects of treatment were detected for BW. Ramos-Nieves et al. (2009) observed that multiparous cows consuming a low-DCAD diet had lower BW 2 wk postpartum compared with cows consuming a control ration with a positive DCAD, but no differences in calculated energy balance were detected in that study.

Plasma glucose measurements taken on d 5, 10, and 21 postpartum revealed some treatment effects on specific days, and BC tended to be lower than CON overall. Few studies investigating the effects of prepartum DCAD concentration or source have reported postpartum glucose concentrations. Ramos-Nieves et al. (2009) observed that postpartum glucose did not differ between cows fed prepartum diets with a low DCAD or a control DCAD. No treatment-related differences in postpartum BHBA concentrations were detected; however, BC had the lowest mean postpartum plasma glucose and also had the numerically greatest postpartum plasma BHBA concentrations.

No differences were detected in the incidence of postpartum health disorders. With the number of animals included in this study, the power to detect differences in discrete variables is relatively low, and numerical differences between treatments were not great. The most prevalent disorder was ketosis, affecting 40% of cows on the study, with no significant differences among treatments. This prevalence rate is higher than observed in other trials (Goff and Horst, 1997; Joyce et al., 1997; Moore et al., 2000), which was an issue of concern. One factor to consider is that diagnosis of ketosis in this study was based on urine acetoacetate concentration, which can lead to artificially high detection rates (LeBlanc, 2010). The prevalence of ketosis in this study may also be attributed to the low postpartum DMI observed. Another possible explanation for the high ketosis rates may be that cows struggled to adapt to the abrupt change from the relatively lowly fermentable prepartum diets, with 38% wheat straw, to the lactating ration containing more highly fermentable carbohydrate sources and thus suffered from cases of subacute or acute ruminal acidosis. Acidosis can suppress ruminal fiber digestion and passage and increase propionate production, which can, in turn, decrease DMI and increase the risk of ketosis (Plaizier et al., 2008). As intuitive as this theory may seem, research by Penner et al. (2007) found that progressively increasing dietary concentrate during the dry period to allow ruminal adaptation before lactation did not prevent postpartum acidosis; in fact, this treatment increased some measures of acidosis in early lactation (when a common lactation diet was fed). These findings suggest that feeding a more fermentable diet in the weeks just before parturition may not be necessary to promote rapid adaptation to a lactation diet.

The frequency of displaced abomasum within the study population was also an issue of concern. All animals diagnosed with a displaced abomasum were in a state of concurrent ketosis, which is a common finding. It is likely that in some cases the ketosis led to appetite suppression, hypomotility, and eventual gastrointestinal atony. However, attempting to infer a direct causal relationship from ketosis to displaced abomasum or vice versa may not be valid (Anderson and Rings, 2008).

Dairy nutritionists seeking to apply these findings will, unfortunately, be hampered by the lack of clarity provided by the results. Anionic salts are known to suppress DMI prepartum, which is a major impediment to their use (Moore et al., 2000), and ACP are intended to avoid this problem. The finding that SC tended to suppress DMI compared to CON suggests that ACP may not completely prevent the hypophagia associated with anionic salts. The lack of hypocalcemia in postpartum CON cows makes it impossible to assess the efficacy of the ACP treatments in improving Ca homeostasis. It is possible that in more “challenged” animals, the ACP treatments provided at a DCAD of approximately 0 mEq/100 g DM would have proven effective at preventing hypocalcemia. On the other hand, it is equally possible that the commonly recommended target of -10 mEq/100 g DM would have been required to significantly alter Ca homeostasis. If so, higher feeding rates of the ACP would have been required, which may have had more dramatic effects on DMI. The complex tradeoffs between DMI and mild, adaptive acidosis induced by anionic diets must be considered in the context of farm-specific disease rates and management approaches. The data reported herein contribute to the limited literature on responses to ACP in transition dairy cows.


Feeding prepartum diets with a DCAD of approximately 0 meq/100 g DM containing supplemental anions provided by ACP marginally decreased prepartum DMI compared with the control. The ACP did not increase total serum Ca concentrations at the time of parturition compared with a control ration containing no supplemental anions, but few cows experienced hypocalcemia across treatments. Treatments had few effects on productivity or postpartum health. Although the DCAD of the control ration was higher than our target, the dietary concentration of the strong cation K was relatively low. The low dietary K concentration may have limited the metabolic alkalosis that is typically associated with decreased tissue sensitivity to parathyroid hormone, which could explain the relatively normal serum Ca concentrations during the peripartum period across all treatments. Close-up diets that are inherently high in K, in contrast, might benefit to a greater extent from the addition of anions to decrease DCAD. An alternative interpretation of these data is that decreased DCAD values would have supported even higher postpartum Ca concentrations. The ideal target DCAD for balancing effects on prepartum DMI and postpartum Ca homeostasis remains to be resolved.




Be the first to comment.

Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.