Influence of prepartum dietary cation-anion difference and the magnitude of calcium decline at the onset of lactation on mineral metabolism and physiological responses

The onset of lactation is characterized by substantially altered calcium (Ca) metabolism; recently, emphasis has been placed on understanding the dynamics of blood Ca in the peripartal cow in response to this change. Thus, the aim of our study was to delineate how prepartum dietary cation-anion difference (DCAD) diets and the magnitude of Ca decline at the onset of lactation altered blood Ca dynamics in the peripartu-rient cow. Thirty-two multiparous Holstein cows were blocked by parity, previous 305-d milk yield and expected parturition date, and randomly allocated to either a positive (+120 mEq/kg) or negative (−120 mEq/kg) DCAD diet from 251 d of gestation until parturition (n = 16/diet). Immediately after parturition cows were continuously infused for 24 h with (1) an intravenous solution of 10% dextrose or (2) Ca gluconate (CaGlc) to maintain blood ionized (iCa) concentrations at ~1.2 m M (normocalcemia) to form 4 treatment groups (n = 8/treatment). Blood was sampled every 6 h from 102 h before parturition until 96 h after parturition and every 30 min during 24 h continuous infusion. Cows fed a negative DCAD diet prepartum exhibited a less pronounced decline in blood iCa approaching parturition with lesser magnitude of decline relative to positive DCAD-fed cows. Cows fed a negative DCAD diet pre-partum required lower rates of CaGlc infusion to maintain normocalcemia in the 24 h postpartum relative to positive DCAD-fed cows. Infusion of CaGlc disrupted blood Ca and P dynamics in the immediate 24 h after parturition and in the days following infusion. Collectively, these data demonstrate that prepartum negative DCAD diets facilitate a more transient hypocalcemia and improve blood Ca profiles at the onset of lactation whereas CaGlc infusion disrupts mineral metabolism.


INTRODUCTION
Calcium (Ca) homeostasis is a highly coordinated process that maintains blood Ca concentrations through modulations of kidney Ca reabsorption, intestinal Ca transport, and bone resorption (Goff et al., 2002).The onset of lactation presents an extreme Ca disturbance due to the rapid and substantial increase in Ca demand by the mammary gland due to irreversible loss of Ca to milk (Ramberg et al., 1970).Thus, the periparturient cow often experiences Ca dysregulation as Ca is pulled from peripheral tissue and blood pools at a rate which cannot be maintained (Goff et al., 2002).Despite a wide range of interventions developed to support Ca homeostasis, subclinical hypocalcemia still occurs and affects up to 50% of multiparous dairy cows (Reinhardt et al., 2011).The deleterious effects of hypocalcemia in the peripartal period have been well-characterized and have clearly demonstrated the increased risks for displaced abomasum, ketosis, mastitis, metritis and decreased neutrophil oxidative burst (Martinez et al., 2014;Merriman et al., 2019;McArt and Neves, 2020;Venjakob et al., 2021).Thus, prepartum nutritional intervention strategies are commonly used practices that have been demonstrated to reduce hypocalcemic incidences and improve cow health and performance (Block et al., 1994;Lean et al., 2019).
One common dietary intervention strategy currently used is feeding a negative DCAD diet prepartum (Lean et al., 2013).Feeding a negative DCAD diet prepartum improves Ca metabolism in the peripartum cow through instigating a mild metabolic acidosis via feeding excess dietary anions (Block, 1984).The corresponding metabolic acidosis elicited allows better maintenance in blood Ca concentrations at the onset of lactation by facilitating increased Ca flux via kidney, rumen, and bone alterations to increase readily exchangeable Ca pools allowing labile Ca to be readily partitioned to meet milk Ca demand at the onset of lactation (Bushinsky et al., 1993;Block, 1984;Grünberg et al., 2011, Wilkens et al., 2016).
A transient decline in blood Ca at the onset of lactation has been suggested to be pertinent for stimulating critical adaptive responses to allow for a successful adaptation to lactation.McArt and Neves (2020) established that cows that experience a transient hypocalcemia at the onset of lactation produce more milk than prolonged-, delayed-, and normo-calcemic cows.Seely et al. (2021) further supported these findings as cows experiencing a transient hypocalcemia produced more milk despite consuming similar amounts of feed.These findings suggest thoughtful consideration of the classification and timing of hypocalcemia thresholds and efficacy of commonly used oral and intravenous Ca intervention strategies to improve blood Ca concentrations after calving.Such treatments could disrupt the homeostatic and homeorhetic adaptations to control blood Ca that may be stimulated by transient hypocalcemia.Calcium bolus and intravenous Ca treatments shortly after parturition successfully increase blood Ca acutely and provide benefits, but may alter blood Ca dynamics resulting in prolonged hypocalcemia and increase risks of metritis and morbidity of cows (Blanc et al., 2014;Martinez et al., 2016).
Therefore, the aim of our study was to determine how the prevention of a transient hypocalcemia at parturition in combination with prepartum dietary DCAD would affect blood Ca dynamics and influence mineral metabolism.We hypothesized that preventing Ca decline in the immediate 24 h after parturition would dis-rupt mineral metabolism and that this would be more robust in cows fed a positive DCAD diet prepartum.

MATERIALS AND METHODS
All experimental procedures performed in this study were approved by the College of Agriculture and Life Sciences Animal Care and Use Committee at the University of Wisconsin-Madison (protocol number A006054) and strictly followed.A sample-size calculation for 0.1 mM differences in ionized Ca (iCa) concentrations was calculated between negative and positive DCAD groups on d 0 of lactation using ionized Ca concentration data collected from cows on negative and positive DCAD diets reported in Slater et al., 2018 andRodney et al., 2018 with a type I error risk of 5% and a type II error risk of 10%, resulting in an n = 8 per treatment group.

Cows and Housing
The experiment was conducted at the Dairy Cattle Center at the University of Wisconsin-Madison and used 32 pregnant, multiparous Holstein cows ranging from second to fourth lactation, with one positive DCAD-fed cow being removed from analysis due to catheter infection (Table 1).Thus, the final experimental dataset included 31 multiparous Holstein cows with postpartum lactation number averaging 2.81 ± 0.27 (SEM).Prepartum, pregnant cows were brought to the Dairy Cattle Center from the Blaine Dairy at the University of Wisconsin-Madison and introduced to facilities at 241 d of gestation to allow for acclimation to tiestall housing and fed a standard low-energy, high-fiber, low-DCAD dry cow diet before dietary enrollment (diet not shown).From thereon until 60 DIM cows were housed in the enclosed tie-stall facility and let out daily prepartum and twice daily postpartum for

Treatments and Feeding Management
At 251 d of gestation cows (gestation length of 278 d) were blocked by parity, previous 305-d milk yield, and expected parturition date and randomly allocated to receive either a positive (+120 mEq/kg) or negative (−120 mEq/kg) prepartum dietary treatment ad libitum.Urine pH was expected to be in a range of 6.0 to 6.5.The negative DCAD was achieved by supplementing with an acidogenic protein meal (BIO-CHLOR, Church & Dwight Co. Inc.) until parturition (n = 16/ diet; Table 2).Upon parturition, a sample was taken within 5 min (represented as 0 h), and then cows were immediately started on infusion.The first cow to calve within each block, on each diet, was assigned to 24-h continuous intravenous infusion of 23% Ca gluconate solution (CaGlc; Aspen Veterinary Resources, Greeley, CO) with the goal to maintain normocalcemia (~1.2 mM ionized Ca concentrations).The subsequent pair cow within each block on the same diet received a continuous 24-h intravenous infusion of 10% dextrose solution at a rate equivalent to the CaGlc infusion received by their diet-pair.Hence, the 4 treatments beginning at parturition and then maintained throughout the remainder of the experiment were: prepartum positive DCAD and postpartum CaGlc infusion (PCa), prepartum positive DCAD and postpartum dextrose infusion (PDex), prepartum negative DCAD and postpartum CaGlc infusion (NCa), and prepartum negative DCAD and postpartum dextrose infusion (NDex).
The rate of infusion and energy equivalency were maintained equally between CaGlc and dextrose groups.Infusion volumes were matched within each diet of each block by an infusion pump (Heska Vet IV, Loveland, CO) to modulate rate of infusion equally between cows paired by infusion (CaGlc and dextrose).Infusion rates were adjusted based upon blood iCa concentrations in CaGlc infused cows on the same diet, with blood sampled every 30 min to modulate infusion rate and maintain blood Ca appropriately.Cows infused with CaGlc all began at an infusion rate of 100 mL/h (2.14 g Ca/100 mL) and then adjusted every 30 min to achieve 1.2 mM blood iCa concentrations which was determined by CG8+ cartridges using a cow-side hand-held biochemical analyzer (i-Stat System, Abbott Laboratories, Abbott Park, IL).All blood samples collected were analyzed within 1 to 2 min of collection.Dextrose cows were infused at the same rate as a CaGlc cow on the same diet within the same block, resulting in separate infusion rates for positive DCAD and negative DCAD cows.The selection of 1.2 mM iCa as normocalcemia was based upon iCa values previously reported (Rodney et al., 2018;Slater et al., 2018).Energy equivalencies in the infusions were matched to account for metabolizable energy on a 100 mL basis for the gluconic acid delivered when infusing CaGlc.The CaGlc contains 21 g per 100 mL of gluconate.The estimated energy supplied by the Ca gluconate infusion, based on Stetten and Stetten Jr. (1950), and allowing for the conversion of gluconate to glucose is 0.30 MJ and dextrose infusion is 0.29 MJ.The difference in energy availability between groups is estimated to be 0.008 MJ.Intravenous 10% dextrose was prepared aseptically with 50% Dextrose Solution (Aspen Veterinary Resources, Greeley, CO) diluted with Sterile Water (Aspen Veterinary Resources, Greeley, CO) to make 10% dextrose for infusion.
Cows were fed TMR diets once daily pre-and postpartum at 0900 h with feed intake recorded daily by farm staff.Daily feed intake was recorded with refusals measured before each morning feeding.Samples of individual feed ingredients, experimental diets and the lactating herd diet were collected weekly.Samples were dried using a forced-air drying oven as described in Toledo et al. (2017) with forage and grain DM values updated weekly to adjust as-fed inclusion rates of ration ingredients.Dry matter intake was calculated based upon feed offered and refused and the accompanying weekly DM value for the respective TMR.Prepartum and postpartum TMR samples had minerals analyzed for wet chemistry at a commercial laboratory (Dairyland Labs, Arcadia, Wisconsin).The composition of experimental diets and the lactating herd diet are presented in Table 2. Milk yield was recorded at each milking during the first 30 d after parturition.

Blood and Urine Sample Collection
Before cows starting experimental diets, baseline blood samples were collected via the coccygeal vein for 2 d.Urine was collected daily prepartum starting −21 d relative to expected parturition date and at 0, 1, 2, 3, 6, 9, 12, 15, 18, 21, 24, 27, and 30 after parturition by gently massaging the perineal area and collecting samples mid-stream in a 5-mL Eppendorf tube.Urine pH was immediately measured and recorded (B-712 LAQUA twin, Horiba Scientific, Edison, NJ), with daily pH calibrations performed before measurement and recording.Ten days before expected parturition date gas sterilized, tygon tubing, jugular catheters were inserted into both jugular veins with catheter placement, maintenance and monitoring as previously described (Slater et al., 2018) and removed at 4 DIM.During continuous 24-h infusion, one catheter was exclusively used for intravenous infusion and the opposing catheter used for blood sampling.Upon completion of the 24-h infusion the catheter used for CaGlc or dextrose infusion was not sampled from and only flushed.
From ~4 d (−102 h) before parturition blood was collected every 6 h through 4 d postpartum (+96 h).Before blood sampling from catheters 8 mL of whole blood were drawn and discarded to remove any residual heparinized saline.Whole blood was sampled from a jugular catheter every 6 h from −102 h through +96 h into 10-mL BD Vacutainer Serum (367820, BD, Franklin Lakes, NJ) and 10-mL Lithium Heparin 158 USP Units (367880, BD, Franklin Lakes, NJ) blood collection tubes and inverted gently.Immediately following inversion, whole blood was analyzed within 1 to 3 min for iCa, pH, HCO 3 , base excess and partial pressure of CO 2 (pCO 2 ), K, and Na using a cow-side hand-held biochemical analyzer (i-Stat System, Abbott Laborato-  1232 ries, Abbott Park, IL).Serum samples were maintained at room temperature and allowed to clot before being placed at 4°C until processing.Serum and plasma were harvested from vacutainer collection tubes by centrifugation at 3,000 × g for 20 min at 4°C.Samples were then divided into multiple aliquots and stored at −80°C until analysis.

Milk Samples
Composite milk samples were collected for the first 4 milkings after parturition and then during the morning and evening milking once weekly throughout the first 15 weeks of lactation to be analyzed for milk fat, protein, lactose, and SCC by a commercial laboratory (AgSource, Verona, WI).A secondary composite milk sample was also collected during the first 4 milkings after parturition into a 10 mL conical and stored at −20°C until analysis of tCa.Colostrum was harvested within 8 h of parturition.

BW and BCS
Cows were weighed before enrollment in the experiment and then twice weekly during the prepartum period, in the morning, as they went out to exercise.Body condition score was determined before enrollment and once weekly by the same trained evaluators using a 1 to 5 scale (Ferguson et al., 1994).Cows were weighed twice weekly during the postpartum period immediately after milking located in the return alley.Body condition was scored weekly postpartum as described above.

Mineral Serum and Milk Laboratory Analyses
Total serum Ca (tCa) concentrations were determined using a colorimetric Ca assay (700550, Cayman Chemical, Ann Arbor, MI) per manufacturer's instructions as previously described (Laporta et al., 2015) with serum tCa intra-and interassay coefficient of variation (CV) of 1.98% and 8.41%, respectively.Milk tCa concentration was determined by colorimetric assay (700550, Cayman Chemical, Ann Arbor, MI).Milk samples were thawed and mixed thoroughly.Then, 500 uL of 0.5 M acetic acid was added to 500 uL of sample, mixed, and centrifuged for 12 min at 13,000 × g for 12 min at 4°C.This allowed for digestion and precipitation of fat from the milk.After centrifugation, supernatant was pipetted off and diluted to a final concentration of 1:50 with ddH 2 0. The intra-and interassay CV of milk tCa were 11% and 4.09%, respectively.
Serum magnesium (Mg) and phosphorus (P) were determined using Catachem VETSPEC reagents on the ChemWell-T AutoAnalyzer (ChemWell-T, Aware-ness Technologies, Palm City, FL).All standards were within the expected calibrated ranges provided by the manufacturer during calibration events (Catachem, Oxford, CT).Samples determined by the autoanalyzer were read in cuvettes in duplicate and a reference pool sample was included and used for assay quality control (Pralle et al., 2021).Methods for Mg and P are based on the work of Ratge et al. (1986) using xylidyl blue and Amador and Urban (1972) using ammonium molybdate, respectively.Intra-assay CV never exceeded 10% for determination of blood metabolites and the interassay CV for magnesium were 15% and 8% for Mg and P, respectively.

Statistical Analysis
Data were analyzed using the MIXED procedure of SAS (version 9.4, SAS Institute Inc., Cary, NC).Fixed terms in all prepartum models were diet, time, block, and the interaction between time and diet.The SLICE command was used to evaluate prepartum interactions when P ≤ 0.10.Fixed terms in all postpartum models included the prepartum mixed model described with additional fixed effects of infusion, and the interactions between infusion and diet, infusion and time, and the 3-way interaction of infusion, diet, and time.Two separate analyses were used for blood metabolites in the postpartum period to analyze response during infusion (6 h through 24 h) and response after infusion (30 h through 96 h).Time was considered the repeated measure in all analyses and to account for autocorrelated errors the ar(1) structure was used for all analyses except urine pH postpartum, upon which the spatial power structure was used to account for unequal spacing of samples.The random statements in all models included cow(treatment).Data were analyzed for normality and when that assumption failed, data were transformed.Transformations were based on diagnostic plots and overall model fit with either rank or log-transformation performed on response variables to improve normality.If transformation was required in one analysis, it was thus transformed in the adjoining analysis to maintain consistency.The adjustment of Tukey was made for all pairwise comparisons.Least squares means and SEM are reported for each variable and statistical significance was declared at P ≤ 0.05, with tendencies for differences declared at 0.05 ≤ P ≤ 0.10.

RESULTS
Of the 31 cows included in the statistical analyses, 3 cows (1 NCa, 1 PCa, and 1 PDex) experienced clinical hypocalcemia (when a cow was unable to rise and confirmed by iCa <0.7 mM) and were treated with intravenous CaGlc infusion on d 1 of lactation.Therefore, those 3 cows were removed prematurely from the 6-h time analysis postpartum and contributed data from −102 h up until Ca treatment.Disease incidences were recorded throughout the experiment and were: NCa (1 retained placenta, 1 clinical hypocalcemia), NDex (2 retained placenta), PCa (1 clinical hypocalcemia), PDex (2 retained placenta, 3 ketosis, 1 displaced abomasum, 1 clinical hypocalcemia).Subclinical ketosis was identified and diagnosed by barn and veterinary staff via urine Ketostix and a confirmation of serum BHB >1.20 mM.However, these cows were included in the analyses for dry matter intake and milk yield.

Ionized Ca and tCa Concentrations
During the 4 d before parturition cows fed a prepartum negative DCAD diet had higher iCa concentrations relative to cows fed a positive DCAD diet (1.21 vs. 1.18 ± 0.01 mM, respectively; P = 0.03; Figure 1A).An effect of time was observed prepartum with all cows, regardless of diet, experiencing a transient decline in iCa and tCa concentrations as they approached parturition (P < 0.0001 and P < 0.0001, respectively).Moreover, from 30 h before parturition until parturition cows fed a negative DCAD diet prepartum had higher iCa concentrations than cows fed a positive DCAD diet (P < 0.05).As cows approached parturition blood iCa concentrations decreased earlier relative to parturition in positive DCAD-fed cows (positive DCAD −24 h vs. 0 h, P = 0.01) compared with negative DCAD-fed cows (negative DCAD −18 h vs. 0 h, P < 0.0001).
Immediately after parturition (0 h) and before infusion, cows fed a negative DCAD diet had greater iCa concentrations than cows fed a positive DCAD diet (1.06 vs. 0.98 ± 0.02 mM, respectively; P = 0.008) but no differences were detected in tCa concentrations (1.80 vs. 1.76 ± 0.05 P > 0.05).Infusion of CaGlc robustly increased blood iCa concentrations relative to dextrose infused cows (1.22 vs. 1.02 ± 0.02 mM, respectively; P < 0.0001) and increased blood tCa concentrations (2.19 vs. 1.77 ± 0.05 mM, respectively; P < 0.0001).An effect of time was observed (P = 0.02) with iCa concentrations decreasing at 18 h postpartum and then increasing 6 h later.Positive DCAD dextrose infused cows tended to have decreased iCa concentrations compared with NDex cows during the 24 h postpartum (P = 0.07; Figure 1A), but no differences were detected in iCa or tCa concentrations between PCa and NCa cows during the 24-h infusion period (P > 0.05).
No effect of diet was observed from 30 through 96 h after parturition (P > 0.05); however, CaGlc infused cows had decreased iCa (1.04 vs. 1.15 ± 0.02 mM, respectively; P = 0.002) and tCa concentrations (2.01 vs. 1.86 ± 0.06 mM, respectively; P = 0.04) relative to dextrose infused cows.Further, a tendency for iCa concentrations was observed for a diet by infusion by time interaction (P = 0.06), with CaGlc infused cows exhibiting a rapid decline in iCa after termination of infusion followed by a corresponding increase once iCa concentrations recovered.Positive DCAD CaGlc cows experienced a calcemic nadir 12 h after termination of infusion (36 h) and had decreased iCa concentrations relative to all other treatments (0.89 mM; P < 0.03), whereas NCa cows experienced their lowest blood iCa concentrations 18 h after termination of infusion (42 h) and had decreased iCa concentrations compared with dextrose infused cows on both diets (0.91 mM; P < 0.04).After the calcemic nadir occurred in PCa and NCa cows iCa concentrations increased in both groups for the remainder of the experiment.Moreover, PDex and NDex cows both had iCa concentrations that steadily increased from 30 to 96 h postpartum.Further, NCa cows had the lowest iCa concentrations at 90 h postpartum relative to all other treatments (P < 0.03).

Magnesium and P Concentrations
Prepartum serum Mg concentrations did not differ between diets (P > 0.05).A tendency for a diet by time effect was observed (P = 0.10) as negative DCAD-fed cows had a greater oscillation in serum Mg the 102 h before parturition (Figure 3A).No difference was observed in serum Mg concentrations due to diet (P > 0.05) during the 24-h infusion period; however, cows infused with CaGlc had lower serum Mg concentrations (P = 0.10) than dextrose infused cows.A time (P = 0.001) and an infusion by time interaction (P = 0.10) was observed during the infusion period as all cows had a decline in serum Mg concentrations, and cows infused with CaGlc exhibited a greater reduction in serum Mg concentrations during the infusion period (Figure 3A).
Serum P did not differ due to diet prepartum (Figure 3B).Regardless of diet, all cows declined in blood P concentrations as they approached parturition (P < 0.0001).No effect of diet was observed during the infusion period.The CaGlc infusion tended to elevate blood P concentrations relative to dextrose infusion (P = 0.06; Figure 3B) during the infusion period.As cows progressed through the 24-h infusion period, all cows declined in blood P concentrations, with a nadir reached at 24 h postpartum (Figure 3B).In the period after termination of infusion CaGlc infused cows tended to have lower (P = 0.07) serum P concentrations than dextrose infused cows.A tendency for a time effect was observed (P = 0.07) as serum P concentrations increased steadily from 30 to 96 h postpartum (Figure 3B).

Acid-Base Status
Cows fed a negative DCAD diet had decreased urine pH concentrations across the prepartum period (P < .Whole blood ionized Ca (iCa) concentrations of cows fed a negative (−120 mEq/kg) or positive (+120 mEq/kg) DCAD diet and infused with Ca gluconate (NCa: CaGlc infusion after negative DCAD prepartum diet; and PCa: CaGlc infusion after positive DCAD prepartum diet) or 10% dextrose (NDex: dextrose infusion after negative DCAD prepartum diet; and PDex: dextrose infusion after positive DCAD prepartum diet) for 24 h after parturition (A), infusion rates of cows fed a negative or positive DCAD diet and infused with Ca gluconate or 10% dextrose for 24 h after parturition (B), and mean averages of infusion rates (C) across the infusion period.Whole blood iCa concentrations: infusion (P < 0.0001), time (P < 0.0001), infusion × time (P < 0.0001).Infusion rate: diet (P = 0.004), time (P < 0.0001), diet × time (P = 0.01).Error bars represent standard error of the mean.1236 0.0001) relative to cows fed a positive DCAD diet (6.03 vs. 8.16 ± 0.07 pH, respectively).No differences in urine pH were detected due to diet or day (P > 0.05) postpartum, however an interaction of infusion by day was significant (P < 0.0001) as CaGlc infused cows had lower urine pH on d 0 and d 1 of lactation (Tables 3,   4, and 5).A diet by day effect occurred (P = 0.0004) with cows fed a negative DCAD diet prepartum having lower urine pH concentrations on the immediate day postpartum.

DMI and Production Parameters
Dry matter intake was not different between cows fed a negative DCAD or positive DCAD diet prepar- 1238 tum (P > 0.05).However, an effect of day was evident (P < 0.0001) with DMI decreasing as cows approached parturition (Figure 4A).Postpartum dry matter intake tended (P = 0.09) to be greater for cows fed a negative DCAD prepartum.Asignificant effect of day (P < 0.0001) was observed as DMI increased as cows progressed into lactation and a tendency (P = 0.06) for higher DMI in CaGlc infused cows (Figure 4A) was also observed.Milk yield did not differ with diet or infusion (P > 0.05); however, an infusion by DIM interaction occurred as dextrose infused cows had greater volatility in milk yield during the first 50 DIM (P = 0.004; Figure 4B), specifically in the PDex treatment.Milk yield increased in all cows (P < 0.0001) as DIM increased.
Weekly milk components did not differ due to diet or infusion for fat content, protein content, lactose content, or SCS (Table 6).A diet by infusion interaction was observed for lactose content (Table 6; P < 0.01).Calcium content in milk was not significantly different due to diet or infusion during the first 6 milkings postpartum (Table 6).Body condition score was not different due to diet pre-or postpartum (P > 0.05).Body condition score postpartum was lower (P = 0.04) in dextrose infused cows (3.27) compared with CaGlc infused cows (3.35).All cows declined in BCS as they progressed into lactation.Body weight was not different pre-or postpartum due to diet (P > 0.05).Neither diet or infusion altered BW postpartum (P > 0.05).

DISCUSSION
The onset of lactation results in dynamic changes in mineral metabolism as the peripartal cow undergoes a rapid shift in physiology to adapt to the demands of lactation.In the dairy cow the shift in Ca metabolism is exacerbated due to the massive increase in Ca requirements between late gestation and early lactation.Calcium research has sought to understand and characterize the magnitude and duration of Ca decline at the onset of lactation and factors that modulate this decline.The data herein demonstrates prepartum negative DCAD diets influence the magnitude and duration of Ca decline.Moreover, manipulating the decline in blood Ca immediately postpartum severely disrupts blood Ca dynamics.
The present data corroborates the role of negative DCAD in improving Ca metabolism as cows fed a negative DCAD diet prepartum maintained elevated blood tCa and iCa concentrations as they approached parturition and had a smaller magnitude of decline in tCa and iCa.Feeding a negative DCAD diet prepartum induced a mild metabolic acidosis evident by the prepartum reductions in urine pH, blood pH, base excess, Na, pCO 2 and HCO 3 and an increase in blood K of cows.Prepartum negative DCAD diets are a commonly used strategy to mitigate hypocalcemic risk at the onset of lactation and improve metabolism and lactation performance (Block, 1984;Martinez et al., 2018b;Rodney et al., 2018).Negative DCAD diets elicit their actions by inducing a mild metabolic acidosis via feeding excess anions (Block, 1984), which allows for improved Ca metabolism through a variety of mechanisms, some of which are increased ruminal Ca transport (Wilkens et al., 2016), increased parathyroid hormone (PTH) secretion (Lopez et al., 2002) and tissue sensitivity to PTH (Goff et al., 2014), and stimulation of osteoclast activity (Krieger et al., 1992) to collectively facilitate a readily exchangeable Ca pool.Additionally, when cows were challenged with EDTA, a nonspecific Ca chelator, cows fed a reduced DCAD were able to maintain blood Ca concentrations more readily (Giesy et al., 1997).Our data supports this collective action as cows infused with CaGlc and fed a negative DCAD diet prepartum (NCa) required lower rates of infusion to maintain normocalcemia relative to PCa cows.Thus, suggesting cows fed a negative DCAD require less support (exogenous Ca) to maintain normocalcemia during the immediate period postpartum due to prepartum acidogenic exchangeable Ca pool benefits and increased Ca flux.
Interestingly, phasic increases in infusion rate were observed in both groups of cows at different times across the 24-h period, with PCa cows exhibiting larger changes in infusion rates to maintain 1.2 mM iCa concentrations; however, both NCa and PCa exhibited similar trends in modulation of required CaGlc infusion rate increases to maintain normocalcemia across the 24-h after parturition.Upon initiation of infusion cows on both diets promptly increased iCa and reached iCa maintenance levels, with PCa cows exhibiting a slightly slower increase to iCa maintenance concentrations, despite higher rates of CaGlc infusion than NCa cows.Moreover, cows on both diets required phasic increases in infusion rates beginning at 4 to 5 h and at 18 to 19 h postpartum.Previous work using the Ca chelators EDTA and EGTA to induce subclinical hypocalcemia suggested changes in blood Ca concentrations indicate the timing of adaptive responses to blood Ca which are representative of various systems critical for Ca homeostasis (Ramberg et al.,1967;Connelly et al., 2022).Although this study maintained normocalcemia, rather than inducing hypocalcemia, an interesting overlap of phase characterization in response to the timing of blood Ca changes suggest that similar systems are being activated or deactivated in PCa and NCa cows.These similar phasic increases in CaGlc requirements to maintain normocalcemia indicate a Ca draw from physiological systems involved in Ca homeostasis at time points that are relative to parturition itself, or a deactivation process of systems previously stimulated by prepartum decline as exogenous CaGlc is supplied.Although the current study cannot definitively answer this question, we hypothesize based on the numerical increases or maintenance in iCa in NDex and PDex cows at similar time points postpartum, it is a termination of a particular physiological system or depletion of Ca stores that limit the ability to support blood Ca concentrations.Further, despite NCa requiring lower rates of CaGlc infusion across the infusion period, ~22 h after parturition during CaGlc infusion, NCa cows began to require similar infusion rates of CaGlc as PCa cows to maintain normocalcemia (Figure 2B, 2C).This observation further suggests calcemic systems activated prepartum due to metabolic acidosis which allowed for greater endogenous blood Ca maintenance are no longer present.Moreover, by 24 h postpartum urine and blood pH were increasing and thus metabolic acidosis conditions at which infusions were initiated were no longer present to the same extent.
Most studies examining blood Ca dynamics at the onset of lactation collect samples daily, with few experiments fully capturing the decline in blood Ca concentrations as cows approach parturition.Thus, due to sampling technique and time points the entirety and rate of the magnitude of Ca decline at the onset of lactation may potentially be missed and different Ca profiles may be captured and classified inaccurately.Research has suggested that blood Ca concentrations may start to decline 9 h before parturition (Megahed et al., 2018), with the present data corroborating and expanding the time at which blood Ca declines before parturition.Horst and Jorgensen (1982) also observed similar declines in blood Ca concentrations before parturition, with cows that became paretic or borderline paretic experiencing a decline in blood Ca up to 24 h before parturition, whereas nonparetic cows were normocalcemic 12 h before parturition.Interestingly, blood Ca concentrations began to decline in positive DCAD cows, that are more susceptible to hypocalcemia, ~18 h before parturition, whereas negative DCAD cows, that are less susceptible to hypocalcemia, exhibited a decline in iCa concentrations 12 h before parturition.However, regardless of timing of initiation of blood Ca declines, cows on both diets saw further subsequent declines in blood Ca concentrations at each successive 6 h time point as cows approached parturition.Moreover, this poses interesting consideration when discussing blood Ca profiles around parturition and studies that aim to evaluate the magnitude of Ca decline and rate of decline as 24-h sampling intervals or sampling times not standardized to the anticipated time of parturition postpartum may characterize the periparturient Ca profile inaccurately.The data herein, Horst and Jorgensen (1982), and Megahed et al. (2018) suggest that blood Ca declines much earlier than previously thought and may not be properly captured in studies with less intensive sampling and standardization.
Blood Ca remained decreased across the entire 24-h postpartum period in dextrose infused cows, with notable increases in iCa concentrations not occurring until ~36 h postpartum.This supports Megahed et al. (2018) who found that blood Ca remained decreased for at least 28 h after parturition.Although timing of the calcemic nadir looks relatively similar in 6-h sampling frequency in PDex and NDex, 3-min iCa analysis indicated that the calcemic nadir occurs earlier during the 24 h after parturition in PDex relative to NDex cows.However, the calcemic nadir is a relative term as blood Ca remains low the entire 24-h period.After termination of infusion iCa concentrations rapidly decreased in CaGlc infused cows with the calcemic nadir being achieved more rapidly in PCa cows (36 h) than NCa cows (42 h).Hypercalcemia due to intravenous Ca infusion results in a corresponding hypocalcemia (Blanc et al., 2014).Calcium gluconate infusion seemingly delayed the transient hypocalcemia experienced in PDex and NDex cows, with the transient hypocalcemia occurring after termination of infusion in CaGlc infused cows.Interestingly, during the corresponding transient hypocalcemia in PCa and NCa cows, cows on both diets were able to increase iCa concentrations earlier during the transient hypocalcemia experienced than PDex and NDex cows.This suggests a greater magnitude of Ca decline may stimulate more rapid and robust feedback mechanisms to improve blood Ca concentrations.It is important to consider within the discussion of cow response that variability exists between animals on biological Ca setpoint.Within the current study we assumed a set point of 1.2 mM for eucalcemia to administer infusion treatment, but variation may still exist between cow's for response due to our selected eucalcemia set-point being When evaluating the 6 h sampling frequency the present data supports previous studies demonstrating the benefits of iCa concentration maintenance at the onset of lactation due to prepartum negative DCAD diets.Cows fed a prepartum negative DCAD diet experienced a smaller magnitude of Ca decline as they approached parturition and immediately after parturition.Cows fed a positive DCAD diet exhibited a more rapid decline in blood Ca as they approached parturition relative to cows fed a negative DCAD diet.However, no differences were observed in rate of return to normocalcemia postpartum, with PDex and NDex cows increasing blood Ca concentrations in similar fashions beginning at ~30 to 36 h postpartum.From when Ca increased postpartum in PDex and NDex cows and considering the prepartum blood Ca decline, beginning at 12 h (negative DCAD) and 18 h (positive DCAD) postpartum, an approximate 2 to 3 d transient hypocalcemia occurred in all cows.A decline in blood Ca multiple hours before parturition and maintenance for ~48 h supports the physiological events required for PTH to be able to exert action, as PTH requires ~48 h of action to activate Ca support mechanisms and mobilize Ca reservoirs (Goff et al., 1986;Martín-Tereso and Verstegen, 2011) to return blood to normocalcemia.Congruently, cows fed a negative DCAD diet have been suggested to have increased PTH secretion and action (Lopez et al., 2002;Goff et al., 2014) and a more labile Ca pool available to support a return more readily to normocalcemia (Lean et al., 2014;Rodney et al., 2018).Thus, prepartum negative DCAD diets improve resistance to Ca decline at the onset of lactation and facilitate a more transient hypocalcemia at the onset of lactation relative to positive DCAD diets.
Feeding a negative DCAD diet prepartum did not alter prepartum blood P concentrations and cows on all diets experienced the classical decline in blood P as cows approach parturition and during the 24-h infusion period (Goff et al., 2002).Blood P was increased in response to CaGlc infusion during the 24-h infusion period, which corresponds to Wilms et al. (2022) who observed an increase in serum P concentrations as blood Ca increased in response to intravenous infusion of Ca to cows immediately postpartum.Parathyroid hormone, which is secreted in response to low blood Ca concentrations increases renal phosphate loss and increases salivary secretion of phosphorus (Wright et al., 1982), and a reduction in PTH secretion may contribute to the elevation in blood P in response to CaGlc infusion during the 24 h postpartum.Interestingly, in the after termination period blood P concentrations were reduced due to CaGlc infusion.The opposite effect, an increase in renal phosphate and salivary secretion due to increased PTH secretion and action could give rise to the reduced blood P concentrations.Serum Mg concentrations were lower in CaGlc infused cows than dextrose cows in the 24-h infusion period.Elevated blood Ca decreases renal reabsorption of Mg in the loop of Henle (Quamme, 1989) and could contribute to lower serum Mg concentrations.Given that both Na and K are included in the calculations for the DCAD equation [(Na + K) − (Cl + S)], we examined concentrations of Na and K in the blood.Our findings that the negative DCAD diet increased blood K and decreased blood Na are consistent with an induced metabolic acidosis.Acidemia shifts K from intracellular stores to extracellular space (Aronson and Giebisch, 2011) and the body will conserve K at the expense of Na, which supports our findings of increased blood K and decreased Na concentrations (Block, 1984).
Dry matter intake was only numerically decreased in negative DCAD-fed cows prepartum.Negative DCAD diets reduce dry matter intake due to metabolic acidosis and corresponding acid-base status changes (Zimpel et al., 2018).Prepartum urine pH, blood pH, HCO 3, base excess and pCO 2 were all significantly reduced in cows fed a negative DCAD diet, which is consistent with studies demonstrating that a mild metabolic acidosis was induced through feeding supplemental anions (Rodney et al., 2018;Zimpel et al., 2018).Postpartum DMI tended to be higher in cows fed a negative DCAD diet and infused with CaGlc, which may be due to the larger number of morbidity incidences in the positive DCAD and dextrose infused cows.This is congruent with the literature as negative DCAD diets have been shown to improve DMI postpartum (Lean et al., 2019) and supplemental Ca in the form of boluses has shown benefit in multiparous cows with low Ca at parturition and suggested to improve health (Leno et al., 2018).Milk yield was lowest in PDex cows, which could be due to both lower DMI and increased incidences of disease.However, the current study was not powered to analyze binomial outcomes, the main focus was investigating physiological interactions surrounding the periparturient period.Cows that presented with a postpartum disease in the PDex group were left in the dataset due to all corresponding diseases being potential effects of prepartum dietary DCAD treatment (Martinez et al., 2018a;Venjakob et al., 2021).We did not observe al- terations in milk composition due to dietary treatments except during the postpartum period where a DCAD × infusion effect on milk lactose content occurred.Lactose content was lowest in CaGlc infused cows, who made more milk than their dextrose infusion counterparts on the same diet.Although significantly different, large variations were observed in milk yield and components and likely of little biological relevance.

CONCLUSIONS
Negative DCAD diets improved blood Ca concentrations in the immediate days around parturition.Cows fed a negative DCAD diet experienced a more transient hypocalcemia at the onset of lactation, with a smaller magnitude of Ca decline.All cows regardless of prepartum dietary treatment exhibited a decline in iCa 12+ h before parturition.However, negative DCAD-fed cows maintained higher iCa concentrations approaching parturition and during the immediate day after.Infusion of CaGlc disrupted blood Ca and P concentrations during the infusion period and immediately after termination of infusion, with blood Ca response profiles differing based upon prepartum diet.Collectively, the data herein demonstrates that reducing prepartum DCAD reduces the magnitude of Ca decline at the onset of lactation and length of transient hypocalcemia.Moreover, interruptions via extended Ca infusion in transient hypocalcemia markedly disrupts blood Ca and P dynamics.

2
Postpartum lactation diet was fed ad libitum from the day of parturition until 60 DIM.3 Bio-Chlor is a patented prepartum fermentation product containing dried condensed extracted glutamic acid fermentation product, dried condensed corn fermentation solubles, processed grain by-products, and magnesium chloride (Arm & Hammer Animal Nutrition, Princeton, NJ).Analysis on a DM basis consisted of CP = 48.63%,K = 1.22%, S = 3.60%, Na = 1.49%, and Cl = 9.08%.4 Rumensin (Elanco Animal Health, Greenfield, IN): 16 mg/kg of DM and vitamin D at a rate of 2mEq of Na + mEq of K) − (mEq of Cl + mEq of S)].

Table 1 .
Connelly et al.: CALCIUM DECLINE AND MINERAL METABOLISM Means (±SEM) for days on treatment diet, previous lactation 305-d mature-equivalent milk production, parity, BW, BCS, and ionized Ca (iCa) at enrollment 1 Ca = calcium gluconate infusion; Dex = 10% dextrose infusion.2 Parity is before calving.1230 exercise.Postpartum cows were milked twice daily and fed the standard herd lactating cow diet.

Table 2 .
Dietary ingredients and average nutrient composition of diets fed pre-and postpartum 1Prepartum diets were fed ad libitum from 251 d of gestation until parturition and formulated for a positive (+120 mEq/kg) or negative (−120 mEq/kg) DCAD.

Table 3 .
Connelly et al.: CALCIUM DECLINE AND MINERAL METABOLISM Effect of DCAD on acid-base status and blood parameters in Holstein cows prepartum (LSM ± SEM)

Table 4 .
Effect of DCAD and infusion 1 on acid-base status on blood parameters in Holstein cows during the experimental infusion period (LSM

Table 5 .
Effect of DCAD and infusion 1 on acid-base status on blood parameters in Holstein cows after termination of infusion (LSM ± SEM)

Table 6 .
Connelly et al.: CALCIUM DECLINE AND MINERAL METABOLISM Effect of DCAD and infusion 1 on postpartum daily milk yield, weekly milk composition, and milk Ca during the first 6 milkings (LSM ± SEM) Connelly et al.: CALCIUM DECLINE AND MINERAL METABOLISM