Sodium, potassium and chloride nutrition of the lactating dairy cow
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/AA00003704/00001
 Material Information
Title: Sodium, potassium and chloride nutrition of the lactating dairy cow influence of dietary cation-anion interrelationships on acid-base status and lactational performance
Physical Description: xxi, 267 leaves : ill. ; 29 cm.
Language: English
Creator: Sanchez, William Kenneth, 1958-
Publication Date: 1992
 Subjects
Subjects / Keywords: Dairy cattle -- Feeding and feeds   ( lcsh )
Dairy cattle -- Effect of salt on   ( lcsh )
Genre: bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1992.
Bibliography: Includes bibliographical references (leaves 256-266).
Statement of Responsibility: by William Kenneth Sanchez.
General Note: Typescript.
General Note: Vita.
 Record Information
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: aleph - 001759741
notis - AJH2824
oclc - 26736944
System ID: AA00003704:00001

Full Text










SODIUM, POTASSIUM AND CHLORIDE NUTRITION OF THE LACTATING
DAIRY COW: INFLUENCE OF DIETARY CATION-ANION INTERRELATIONSHIPS ON
ACID-BASE STATUS AND LACTATIONAL PERFORMANCE











BY

WILLIAM KENNETH SANCHEZ


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1992


UrI'IVR:"ITY OF FLr.DA L!77..r S
































To my wife, who gave me a reason;
to my father, who gave me wisdom; and,
to my late brother, who gave me an example.













ACKNOWLEDGMENTS


This dissertation could not have been possible without the help of

many individuals. I am very grateful to Dr. David K. Beede, chair of my

supervisory committee. Dr. Beede has not only been my major advisor and

mentor, he has been a genuine friend. Upon moving to Gainesville, my

family and I stayed at his home; we feasted on Patti Beede's home-cooked

meals; and we enjoyed the chance to play with Ogden and visit the

horses. Dr. Beede was always willing to put his busy schedule on hold

for me, to stop and listen to a new idea, to solve a problem, or to

suggest tactfully that I get back to the lab. Whether it was

professional or personal, Dr. Beede had time to discuss it with me. I

shall always be indebted to Dr. David Beede.

I also am grateful to the other members of my supervisory

committee. Dr. Charles Staples spent many hours discussing dairy cattle

nutrition with me and provided extensive professional and personal

advice. Dr. Michael DeLorenzo spent many hours educating me. Whatever

the topic, Dr. DeLorenzo challenged me to understand it thoroughly. I

always left his office in a better mood. Both he and Dr. Staples shared

many stories about raising children while in graduate school and assured

me that it was possible.

I am very thankful to have had Dr. Richard Miles, Dr. John

Cornell, and Dr. John Bauer as additional members of my supervisory

committee. My first visit with Dr. Miles lasted over two hours. In

iii








that time he simplified the entire subject of acid-base physiology. He

transformed those negative logarithms into simple little H* ions. He

convinced me that acid-base physiology was not only a simple subject but

one directly tied to animal nutrition and of prime importance to my

research. Many of the ideas for this research came from that first

meeting. Dr. John Cornell also made a difficult subject--statistics--

much easier for me. Dr. Cornell contributed many hours to the design

and analysis of this research and gave me many hours of professional and

personal advice. Thanks to Dr. Cornell, I have gained many valuable

statistical tools that I shall use in future experimentation. Dr. John

Bauer also was very helpful while on the committee. I was extremely

fortunate to have access to the Chloridometer he provided us. I am very

grateful to Dr. Charles Wilcox who helped with much of the design and

statistical analysis of experiments in this dissertation.

I thank the Florida Dairymen for contributing milk check-off funds

which were used to pay for some of the research in this dissertation.

Thanks are also extended to the Salt-Institute and to Church and Dwight

for generously funding portions of this research.

Dr. Natzke, chair of the Dairy Science Department, also deserves

thanks for his valuable advice and for finding a line in the budget to

include my departmental assistantship. The list of other dairy science

faculty who have helped me at some point along the way can be found on

the board in the front lobby of the Dairy Science Department. None of

the faculty can be excluded from this student's acknowledgments.

Before my program began, I spent my first semester working at

North Florida Holsteins, a commercial dairy, where I learned many of the








facets of large herd dairy management. I am very grateful to Don

Bennink, owner and Mike Casey, operations manager of North Florida

Holsteins for that experience.

Estelle Hirchert, biological scientist for Dr. Beede, contributed

immensely to my program. As a Ph.D student in "her" laboratory, I

benefitted from her skillful training and cheerful service. Estelle not

only took it upon herself to organize, prepare and transport a plethora

of labelled tubes, syringes, and properly functioning laboratory

equipment to the farm during my final experiment, she even picked me up

at my home and drove us both to the dairy each Tuesday when sample

collections began at 3:30 a.m. Estelle gave up many evenings and

weekends for the research in this dissertation. I also thank Joyce

Hayen, who helped set up many of the laboratory analyses and always had

a kind word for me and the rest of the graduate students.

This research obviously could not have been conducted without a

dairy and dairy cows to work with. Dale Hissem and others at the Dairy

Research Unit need to be thanked for keeping the Dairy Research Unit

properly functioning and for making sure my cows were healthy and well

cared for. Dale and Mary Ellen Hissem generously opened their home to

me after I had spent a long day at the farm. They also gave my children

what just what they needed (i.e., time with a loving grandma and

grandpa).

I am thankful for the goodwill extended to me by my fellow

graduate students. Changzheng Wang, another dairy nutrition Ph.D.

student, helped with the experiments, provided valuable advice, and was

always available for a thought-provoking discussion. I thank Doug








McCullough, Matt Lucy and Dane Schoenbaum for making sure I left the lab

on occasion. Ted Ruiz was very helpful during the second experiment. I

was fortunate to have the skillful assistance of Irene Fiorito, German

Davolos, Doug McCullough, Craig Thomas, Jorge Savio, Carlos Garcia,

Paulette Tomlinson, Laura Lynn and Patrick Joyce during the last

experiment. Many other graduate students and post-doctoral fellows also

deserve a big thanks for their help and friendship throughout.

My wife and children and I are grateful for the many families we

had the pleasure of meeting during our stay at Tanglewood Village. I

also am grateful to Paul and Cindy Johnson, Dane Schoenbaum, Mike Moser

and Jill Davidson for making my final months in Gainesville without my

family bearable.

My beautiful wife Sandy gave me never-ending love and support

throughout my entire post-graduate education. To her I am most

grateful. She gave me everything I could ask for, including two

beautiful children, sweet little Melinda and Eric. All the pressures

vanished when I came home to them.

Although my wonderful brothers and sisters questioned my sanity,

they supported my decision to return to graduate school and were always

there when I needed them. I am very grateful to my sisters for their

constant caring and for making sure my father visited us while his

grandchildren were still very young. I thank Dario and Joann and all

the Casciato's for their love and support throughout. I also thank my

dear father. He has given me so much over the years. I especially

thank him for just being my friend. I hope to give my children as much

as my Dad gave me. Finally, I thank God.














TABLE OF CONTENTS

ACKNOWLEDGMENTS . . . . . . . . . . . . .

LIST OF TABLES . . . . . . . . . . . . .

LIST OF FIGURES . . . . . . . . . . . . .

LIST OF ABBREVIATIONS . . . . . . . . . . .

ABSTRACT . . . . . . . . . . . . . .

CHAPTERS


xii

xvii

xx

PAGE


1 INTRODUCTION . . . . . .


2 LITERATURE REVIEW . . . . . . . . . . 5

Review of Na, K and C1 metabolism . . . . . . 5
Dietary Requirements and Recommendations . . . ... 11
Influence of Cation-Anion Interrelationships on
Acid-Base Status . . . . . . . . .... 17
Nutritional Concepts Related to Cation-Anion
Interrelationships. . . . . . . . .22
Cation-Anion Interrelationships and Effects of
Cation Difference on Animal Performance . . . ... 26


3 INTERRELATIONSHIPS AMONG DIETARY SODIUM, POTASSIUM AND
CHLORIDE: EFFECTS ON ACID-BASE STATUS, MINERAL
METABOLISM AND LACTATIONAL PERFORMANCE OF DAIRY CATTLE

Introduction . . . . . . . . . . . .
Materials and Methods . . . . . . . . .
Results . . . . . . . . . . . . .
Discussion . . . . . . . . . . . .
Conclusions . . . . . . . . . . . .

4 DIETARY MIXTURES OF SODIUM BICARBONATE, SODIUM CHLORIDE AND
POTASSIUM CHLORIDE: EFFECTS ON ACID-BASE STATUS, MINERAL
METABOLISM AND LACTATIONAL PERFORMANCE OF DAIRY CATTLE .

Introduction . . . . . . . . . . . .
Materials and Methods . . . . . . . . .
Results . . . . . . . . . . . . .
Discussion . . . . . . . . . . . .
Conclusions . . . . . . . . . . . .


34

34
.35
S46
S61
S98



100

100
101
108
129
137


S 1









5 INFLUENCE OF DIETARY MACROMINERAL INTERRELATIONSHIPS AND
CATION-ANION DIFFERENCE ON LACTATIONAL PERFORMANCE: USING A
LARGE DATA SET AND EMPIRICAL MODELS TO IDENTIFY AND QUANTIFY
EFFECTS . . . . . . . . . . . .. 138

Introduction . . . . . . . . . . . 138
Materials and Methods . . . . . . . .... 140
Results and Discussion . . . . . . . . . 144
Conclusions . . . . . . . . . . . 177

6 INFLUENCE OF DIETARY POTASSIUM BY CHLORIDE INTERACTION AND
CATION-ANION DIFFERENCE ON PHYSIOLOGICAL RESPONSES OF LATE
LACTATION DAIRY CATTLE . . . . . . . .. 178

Introduction . . . . . . . . . . . 178
Materials and Methods . . . . . . . . 181
Results and Discussion . . . . . . . . 188
Discussion . . . . . . . . . . . 214

7 SUMMARY AND RECOMMENDATIONS FOR FUTURE RESEARCH. . . . 218

APPENDICES

A STATISTICAL TABLES FOR CHAPTER 3 . . . . . . .. 227

B STATISTICAL TABLES FOR CHAPTER 4 . . . . . .... 238

C STATISTICAL TABLES FOR CHAPTER 6 . . . . . . .. 244

REFERENCES . . . . . . . . . . . . . 256

BIOGRAPHICAL SKETCH . . . . . . . . . . . 267


viii













LIST OF TABLES


TABLE PAGE

3-1. Ingredient composition of basal diet. . . . . . ... 36

3-2. Analyzed chemical composition of concentrate, corn silage
and TMR (DM basis) . . . . . . . . . .. 36

3-3. Dietary concentrations of Na, K and C1 and calculated CAD of
experimental diets (% of diet DM) . . . . . ... 39

3-4. Summary of P values for effects included in reduced models.. 47

4-1. Composition of basal (control) diet . . . . .... .102

4-2. Composition and nutrient analysis of experimental diets (%
of diet DM) . . . . . . . . . . . 104

4-3. Effect of different mixtures of NaHCO3, NaCl and KC1 on dry
matter intake (DMI), milk yield (MY), 3.5% fat-corrected
milk (3.5% FCM) yield, milk fat percentage (MF), milk
protein percentage (MP), and body weight gain (BWG). . 110

4-4. Effect of different mixtures of NaHCO,, NaCl and KC1 on
blood plasma Na (PNa), K (PK), C1 (PC1), Ca (PCa) and Mg
(PMg) . . . . . . . . . . . . 113
4-5. Effect of different mixtures of NaHCO NaCl and KC1 on
whole blood Na (WBNa), K (WBK), C1 (WBC1), Ca (WBCa), and Mg
(WBMg). . . . . . . . . . ..... . 119

4-6. Effect of different mixtures of NaHCO NaCl and KCl on milk
Na (MLNa), K (MLK), C1 (MLC1), Ca (MLCa) and Mg (MLMg). . 122

4-7. Effect of different mixtures of NaHCO3, NaCl and KC1 on
whole blood hydrogen ion concentration (H'), bicarbonate
(HC03,), pCO2, anion gap (ANGAP) and base excess (BE).
Values for whole blood pH are shown for reference only. . 127

5-1. Type of study, total cow-period observations, season, forage
type, and reference of studies included in data base. . 142

5-2. Concentrations of dietary macrominerals and cation-anion
difference: mean, SEM and ranges in data base. . . .. 143








5-3. Least squares analysis of variance for dry matter intake
(DMI), milk yield, 4% fat-corrected milk yield (FCM) and
milk composition from macromineral models. . . . .. 147

5-4. Regression coefficients and standard errors of estimates
from reduced macromineral models for dry matter intake
(DMI), milk yield (MY) 4% fat-corrected milk (4% FCM) yield,
milk fat percentage (MF) and milk protein percentage (MP). 148

5-5. Least squares analysis of variance for dry matter intake
(DMI), milk yield, and 4% fat-corrected milk (4% FCM) yield
for cation-anion difference (CAD) models. . . . ... 149

5-6. Regression coefficients with standard errors of coefficient
estimates from reduced cation-anion difference (CAD) models
for dry matter intake (DMI), milk yield (MY) 4% fat-
corrected milk (4% FCM) yield. . . . . . . .. 165

5-7. Regression coefficients with standard errors of estimates
from independent data set predicted cation-anion difference
(CAD) models for dry matter intake (DMI) and milk yield
(MY) . . . . . . . . . . . . . 171

5-8. Predicted dry matter intake and milk yield from cation-anion
difference (CAD) models as compared to independent
experiments. . . . . . . . ... ..... .175

6-1. Ingredient composition of treatment diets (% of diet DM). 183

6-2. Analyzed composition of dietary treatments (% of diet DM). 189

6-3. Effects of dietary K, Cl and cation-anion difference (CAD)
on blood variables not affected by dietary treatment by h
interactions. . . . . . . . . . .. . 191

6-4. Effects of dietary K, C1 and cation-anion difference (CAD)
on urine variables not affected by dietary treatment by h
interactions . . . . . . . . .. . 196

6-5. Effects of dietary K, C1 and cation-anion difference (CAD)
on blood variables not affected by dietary treatment by d
interactions . . . . . . . . .. . 204

6-6. Effects of dietary K, C1 and cation-anion difference (CAD)
on urine variables not affected by dietary treatment by d
interactions . . . . . . . . .. . 206

6-7. Effects of dietary K, C1 and cation-anion difference (CAD)
on fecal composition and apparent mineral digestibilities
not affected by dietary treatment by d interactions. . 211








6-8. Effects of dietary K, C1 and cation-anion difference (CAD)
on dry matter intake (DMI), milk yield and composition
variables not affected by dietary treatment by d interac-
tions . . . . . . . . . . . . 213













LIST OF FIGURES


FIGURE PAGE

3-1. Three-factor central composite design used to evaluate
interrelationships among dietary Na, K and C1
concentrations. . . . . . . . .. .. 38

3-2. Response surface for DMI plotted against dietary Na and K
with C1 fixed at .8%. . . . . . . . . . 49

3-3. Response surface for DMI plotted against dietary Na and C1
with K fixed at 1.4%. . . . . . . . . .. 50

3-4. Response surface for 3.5% FCM yield plotted against dietary
Na and C1 with K fixed at 1.4%. . . . . . . 51

3-5. Response surface for 3.5% FCM yield plotted against dietary
K and C1 with Na fixed at .55% . . . . . . .. 52

3-6. Response surface for milk fat percentage (MF) plotted
against dietary Na and C1 with K fixed at 1.4% . . .. 53

3-7. Response surface for milk fat percentage (MF) plotted
against dietary Na and K with C1 fixed at .8%. . . . 54

3-8. Response surface for milk protein percentage (MP) plotted
against dietary K and C1 with Na fixed at .55%. . . 55

3-9. Response surface for body weight gain (BWG) plotted against
dietary K and C1 with Na fixed at .55%. . . . . 56

3-10. Response surface for blood bicarbonate (HC03,) plotted
against dietary C1 and Na with K fixed at 1.4%. . . 58

3-11. Response surface for blood pCO2 plotted against dietary Na
and C1 with K fixed at 1.4% . . . . . . . . 59

3-12. Response surface for blood base excess (BE) plotted against
dietary Na and C1 with K fixed at 1.4% . . . . .. 60

3-13. Response surface for plasma Na (PNa) plotted against dietary
Na and C1 with K fixed at 1.4% . . . . . . . . 62

3-14. Response surface for plasma Na (PNa) plotted against dietary
K and C1 with Na fixed at .55% . . . . . . .. 63

xii








3-15. Response surface for plasma K (PK) plotted against dietary
Na and K with C1 fixed at .8% . . . . . . .

3-16. Response surface for plasma C1 (PC1) plotted against dietary
Na and C1 with K fixed at 1.4% . . . . . . . .

3-17. Response surface for plasma Ca (PCa) plotted against dietary
Na and C1 with K fixed at 1.4% . . . . . . .

3-18. Response surface for whole blood Na (WBNa) plotted against
dietary Na and K with C1 fixed at .8% . . . . .

3-19. Response surface for whole blood K (WBK) plotted against
dietary Na and K with C1 fixed at .8% . . . . .

3-20. Response surface for whole blood Cl (WBC1) plotted against
dietary K and C1 with Na fixed at .55% . . . . . .

3-21. Response surface for whole blood Ca (WBCa) plotted against
dietary K and C1 with Na fixed at .55% . . . . . .

3-22. Response surface for whole blood Mg (WBMg) plotted against
dietary K and C1 with Na fixed at .55% . . . . . .

3-23. Response surface for milk K (MLK) plotted against dietary K
and C1 with Na fixed at .55% . . . . . . . .


3-24. Response surface for milk C1 (MLC1) plotted a
C1 and K with Na fixed at .55% . . . .

3-25. Response surface for milk Ca (MLCa) plotted a
Cl and Na with K fixed at 1.4% . . . .

3-26. Response surface for milk Mg (MLMg) plotted a
Na and Cl with K fixed at 1.4% . . . .

3-27. Dry matter intake (DMI) response to CAD . .

3-28. Yield of 3.5% FCM response to CAD . . .

3-29. Milk protein percentage (MP) response to CAD

3-30. Body weight gain (BWG) response to CAD . .

3-31. Blood bicarbonate (HC03) response to CAD .

3-32. Blood base excess (BE) response to CAD . .

3-33. Blood pCO2 response to CAD . . . . .

3-34. Plasma Cl (PC1) response to CAD . . . .


against


against


against


dietary


dietary


dietary


xiii


. . . 75

. . . 89

. . . 90

. . . 91

. . . 92

. . . 93

. . . 94

. . . 95

. . . 96







3-35. Milk C1 (MLC1) response to CAD . . . . . .... .97

4-1. Regression of milk protein (MP, %) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaC1 (...) and KC1 (---) . .. 111
4-2. Regression of body weight gain (BWG, kg/d) response to
binary (0:50:50), tertiary (33:33:33) and primary (100:0:0)
mixtures of NaHCO3 (-), NaC1 (...) and KCL (---) .... 114
4-3. Regression of plasma K (PK, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHC03 (-), NaCl (...) and KC1 (---) . . .. 115
4-4. Regression of plasma Cl (PC1, meq/1) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaCl (...) and KC1 (---) . . .. 116
4-5. Regression of plasma Ca (PCa, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHC03 (-), NaC1 (...) and KC1 (---) . . .. 117
4-6. Regression of whole blood K (WBK, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaCl (...) and KC1 (---) . . .. 120
4-7. Regression of whole blood C1 (WBC1, meq/L) response to
binary (0:50:50), tertiary (33:33:33) and primary (100:0:0)
mixtures of NaHCO3 (-), NaC1 (...) and KC1 (---) . .. 121
4-8. Regression of milk Na (MLNa, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaCl (...) and KC1 (---) . . .. 123
4-9. Regression of milk K (MLK, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaCl (...) and KC1 (---) . . .. 124
4-10. Regression of milk Cl (MLC1, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHC03 (-), NaCl (...) and KC1 (---) . . ... 125
4-11. Regression of milk Ca (MLCa, meq/L) response to binary
(0:50:50), tertiary (33:33:33) and primary (100:0:0) mix-
tures of NaHCO3 (-), NaC1 (...) and KC1 (---) . . ... 126
4-12. Regression of whole blood bicarbonate (HCO03, meq/L)
response to binary (0:50:50), tertiary (33:33:33) and primary
(100:0:0) mixtures of NaHC03 (-), NaC1 (...) and KC1 (---) 128
4-13. Regression of milk C1 (MLC1, meq/L) response to CAD . . 135








4-14. Regression of milk Mg (MLMg, meq/L) response to CAD . . 136

5-1. Response surfaces for DMI (A), 4% FCM yield (B) and milk fat
(C) plotted against dietary Na and K with all other macro-
minerals at their mean concentration. . . . . .. 150

5-2. Response surfaces for dry matter intake (DMI) (A), milk
yield (B) and 4% FCM yield (C) plotted against dietary Na
and Ca with all other macrominerals at their mean
concentration . . . . . . .. . . . 153

5-3. Response surfaces for FCM (A), milk fat (B) and milk protein
percentage (C) plotted against dietary K and Ca . . . 156

5-4. Response surfaces for dry matter intake (DMI) (A) and 4% FCM
yield (B) plotted against dietary K and C1 with all other
macrominerals at their mean concentration . . . .. 159

5-5. Response surfaces for dry matter intake (DMI) (A) and milk
yield (B) plotted against dietary Na and Mg with all other
macrominerals at their mean concentration . . . .. 166

5-6. Dry matter intake (DMI), milk yield (MY) and 4% fat-
corrected milk (4% FCM) yield response to CAD . . .. 168

5-7. Dry matter intake (DMI) (A) and milk yield (B) least squares
means and regression curves for independent data sets. 172

5-8. Dry matter intake (DMI) (A) and milk yield (B) regression
curves for independent data sets compared to predicted CAD
model regression curves. .. . . . . . . 173

5-9. Dry matter intake (DMI) (A) and milk yield (B) regression
curves from independent data sets (corrected for experiment
effects) compared to predicted CAD model regression curves. 174

6-1. Least squares means of ionized Ca during d 1 . . . .. 192

6-2. Least squares means of plasma K during d 1 . . . .. 194

6-3. Least squares means of urine Na/creatinine ratio during d 1 197

6-4. Least squares means of urine K/creatinine ratio during d 1 199

6-5. Least squares means of urine Cl/creatinine ratio during d 1 200

6-6. Least squares means of urine Cl/creatinine ratio at h 0 and
h 13 on d . . . . . . . . . . . 201

6-7. Least squares means of urine Na/creatinine ratio during d 1
through 7 . . . . . . . . . . . . . 207








6-8. Least squares means of urine Cl/creatinine ratio during d 1
through 7 . . . . . . . . . . . . . 209

6-9. Least squares means of rectal temperature during d 1 through
7 . . . . . . . . . . . . . ... 215














LIST OF ABBREVIATIONS


Excluding abbreviations of common weights and measures, the
following are used within the dissertation. Terms may or may not be
accompanied with long form descriptions in the text.

ADF = Acid detergent fiber
ANGAP = meq [(Na + K) (Cl + HC03)] in whole blood or plasma
ANOVA = Analysis of variance
ARC = Agricultural Research Council
BE = Blood base excess
BW = Body weight
BWG = Body weight gain
Ca = Calcium
Ca" = Calcium ion(s)
CaCl2 = Calcium chloride
CaCO3 = Calcium carbonate
CAD = Cation-anion difference expressed as meq (Na + K C1)/100
g diet DM
CCD = Central composite design
CF = Crude fiber
C1 = Chloride
Cl = Chloride ion(s)
Co = Cobalt
CO2 = Carbon dioxide
CP = Crude protein
Cr203 = Chromic oxide
Cu = Copper
d = Day(s)
df = Degrees of freedom
DHIA = Dairy Herd Improvement Association
DIM = Days in milk or stage of lactation
DM = Dry matter
DMI = Dry matter intake
DUA = Dietary undetermined anion
ECF = Extracellular fluid
FCM = Fat-corrected milk
Fe = Iron
GLM = General linear models
h = Hour(s)
H = Hydrogen ion(s)
[H*] = Hydrogen ion concentration
HCO = Bicarbonate ion(s)
HK:HCl = High K, high C1
HK:LC1 = High K, low C1
H20 = Water


xvii








HPO = Dibasic phosphate
H2P04 = Monobasic phosphate
I = Iodine
K = Potassium
K' = Potassium ion(s)
KC1 = Potassium chloride
K2CO, = Potassium carbonate
K Cr O = Potassium dichromate
HE = Hydrochloric acid
LK:HC1 = Low K, high C1
LK:LC1 = Low K, low Cl
LSMS = Least squares means
meq = milliequivalent(s)
MF = Milk fat
Mg = Magnesium
MgC12 = Magnesium chloride
MgO = Magnesium oxide
Mg SO4 = Magnesium sulfate
MLCa = Milk Ca
MLC1 = Milk Cl
MLK = Milk K
MLMg = Milk Mg
MLNa = Milk Na
Mn = Manganese
MP = Milk protein
MY = Milk yield
Na = Sodium
Na = Sodium ion(s)
NaCl = Sodium chloride
NaHCO, = Sodium bicarbonate
NaHPO4 = Sodium phosphate
NDF = Acid detergent fiber
NEL = Net energy of lactation
NH3 = Ammonia
NH' = Ammonium ion(s)
NH4C = Ammonium chloride
(NH4)2SO4 = Ammonium sulfate
NRC = National Research Council
OH' = Hydroxide ion
P = Phosphorous or probability. Probability is associated
with a coefficient between 0 and 1 and the < and >
symbols.
pH = Negative logarithm of [H']
PCa = Plasma Calcium
PC1 = Plasma Cl
pCO2 = partial pressure of CO2
PK = Plasma K
PMg = Plasma Mg
PNa = Plasma Na
R2 = Coefficient of determination
RBS = Red blood cell
S = Sulfur


xviii








SAS
Se
SEM
SID
Si02
TMR
UIP
USDA
WBCa
WBC1
WBK
WBMg
WBNa
WK
Yr
Zn
Zn2SO4


xix


Statistical Analysis System
Selenium
Standard error of the mean
Strong ion difference
Silicon dioxide (washed sand)
Total mixed ration
Undegradable intake protein
United States Department of Agriculture
Whole Blood Ca
Whole Blood C1
Whole Blood K
Whole Blood Mg
Whole Blood Na
Week(s)
Year(s)
Zinc
Zinc sulfate














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

SODIUM, POTASSIUM AND CHLORIDE NUTRITION OF THE LACTATING
DAIRY COW: INFLUENCE OF DIETARY CATION-ANION INTERRELATIONSHIPS ON
ACID-BASE STATUS AND LACTATIONAL PERFORMANCE

By

William Kenneth Sanchez

May, 1992

Chairman: David K. Beede
Major Department: Animal Science

Objectives were to determine the influence of dietary sodium (Na),

potassium (K), and chloride (Cl) on acid-base status and lactational

performance of dairy cattle. Experiment one was conducted to study

individual and interrelated effects of varying concentrations of Na, K,

Cl and cation-anion difference (CAD; milliequivalents (Na + K Cl)/ 100

grams diet). Five concentrations of Na (.31 to .85%), K (.86 to 1.96%),

and Cl (.32 to 1.15%) in a central composite design were fed to 48

midlactation cows. Dietary CAD ranged from +12 to +62. Dietary

concentrations of Na, K and C1 influenced lactational performance.

Optimal concentrations were above current recommendations but were

interdependent. Dietary Na spared dietary K (and vice versa).

Increasing dietary C1 was deleterious unless accompanied by an increase

in dietary Na and/or K. Optimum CAD was between +23 and +50.

Experiment two was conducted to evaluate source of Na, K and C1

(sodium bicarbonate, sodium chloride and potassium chloride). Seven

xx








diets with mixtures of these salts (1% of diet dry matter) and one

control diet were fed to 36 midlactation cows. Treatments were defined

according to a simplex-centroid mixtures design. None of the mixtures

had a significant influence.

Data from these experiments were combined with eight others for

empirical modeling. Objectives were to identify and quantify effects of

macromineral interrelationships and CAD on lactational performance.

Performance was optimal at .58% Na, .40% Mg and +38 CAD. Consistent

interrelationships between Na and K, Na and calcium (Ca), K and Ca, and

K and Cl were found.

A final experiment was conducted to determine physiological

responses to K x Cl interaction. Hourly and daily blood and urine

samples were collected. Cows were assigned to a 4 x 4 Latin square and

fed diets with low and high concentrations of K factored with low and

high concentrations of C1 (1.1 versus 1.8% K and .43 versus .91% C1).

The data suggested a physiological mechanism for K x Cl interaction was

present. The diet with high C1 and low K did not contain sufficient

cation to accompany excess C1 into the urine. Chloride was retained and

subclinical hyperchloremic metabolic acidosis developed.

The need to consider dietary macromineral interrelationships and

CAD for lactating dairy cattle was established. Results are directly

applicable to the dairy industry and should provide economic benefit.


xxi













CHAPTER 1
INTRODUCTION

The first general review of mineral metabolism in this country was

published in 1922 (Mattill and Mattill, 1922). Since then researchers

have identified as many as 36 elements that are nutritionally important

to domestic animals (NRC, 1980). Volumes have been written about the

singular effects of each element, but comparatively little is known

about their interrelated effects. Twenty years ago Jacobson et al.

(1972) suggested that mineral interrelationships are the least

understood and most promising area of mineral nutrition of dairy cattle.

With much foresight, Shohl (1939), writing about limitations of

knowledge in mineral metabolism, noted that when mineral functions are

described as the action of single elements, a degree of simplicity is

assumed erroneously. He further suggested that the relations of

elements to one another, to water, to acid-base equilibrium, and to body

functions are so complicated that such a simplistic scheme no longer was

adequate. Over 20 years ago Paquay et al. (1968) recognized the

importance of mineral interrelationships and suggested that the metabo-

lism of ration constituents depends on the other dietary constituents

nearly as much as on their own intake and sometimes even more. In spite

of these early claims, current mineral requirement guidelines (ARC,

1980; NRC, 1989) do not adequately account for potential dietary

interrelationships.








2

Designing diets to meet the nutritional needs of the high

producing dairy cow is a challenge. As a result of improved genetics,

nutrition, health and other management practices, milk yields per cow

have doubled in less than 30 years (USDA, 1990). Yet the efficiency of

milk production must continue to improve if the dairy industry is to

maintain or enhance its role as an important producer of food (NRC,

1989). As the genetic potential of our dairy cattle population

continues to increase, designing diets with the optimum concentration of

minerals becomes even more critical. Concentrations of the monovalent

macrominerals, sodium (Na), potassium (K) and chloride (Cl), are

particularly important to lactating cows. Interrelated effects among

dietary concentrations of these three macrominerals have been suspected

for some time (Shohl, 1939), but few theories have been proposed to

explain how they affect lactational performance of the dairy cow.

Shohl and Sato (1923) were first to propose that mineral

interrelationships were related to acid-base status. Shohl (1939)

included a chapter in his book entitled Cation-Anion Relationships and

proposed that maintenance of normal acid-base equilibrium required

excretion of excess dietary cations and anions. He hypothesized that

consumption of either excess mineral cations relative to anions or

excess anions relative to cations resulted in acid-base disturbances.

Once animal nutritionists began to test this hypothesis, mineral

interrelationships were found to affect numerous metabolic processes.

Leach (1979) and Mongin (1980) reviewed related literature and concluded

that mineral interrelationships had profound influences. They theorized

that for an animal to maintain its acid-base homeostasis, input and










output of acidity had to be maintained. It was shown that net acid

intake was related to the difference between dietary cations and anions.

The monovalent macromineral ions, Na, K and Cl were the most important

elements in the expression (Mongin, 1980).

Mongin (1980) was one of the first to propose a three-way

interrelationship among dietary Na, K and C1. He proposed that the sum

of Na plus K minus Cl (in meq per 100 g diet DM) could be used to

predict net acid intake. This sum commonly has been referred to as the

dietary cation-anion balance (Tucker et al., 1988a) or dietary

electrolyte balance (West et al., 1990) but Sanchez and Beede (1991)

coined the term "cation-anion difference" to represent the mathematical

calculation used more precisely and to avoid the erroneous connotation

that mineral cations truly are balanced with mineral anions in the diet.

Effects of cation-anion difference (CAD) in poultry have been

studied extensively. Feeding low (i.e., below +20) dietary CAD,

particularly from excessive mineral anions, has disturbed acid-base

status, growth rate and eggshell mineralization, and has increased the

incidence of tibial dyschondroplasia in poultry (Austic, 1988).

Weanling pig growth also has been reduced by feeding diets with

relatively low CAD (Yen et al., 1981).

Tucker et al. (1988a) were first to evaluate this concept with

lactating dairy cattle and suggested that there may be an optimal CAD

for achieving maximal intake and milk production. They compared diets

formulated with -10, 0, +10 and +20 meq/100 g CAD and reported that a

diet with +20 improved dry matter intake (DMI) by 11% and milk yield

(MY) by 9% compared to a diet with -10 CAD. Blood bicarbonate (HC03-)










increased linearly with increasing CAD indicating that higher CAD

improved acid-base status. Because lactation diets typically contain

more than +20 CAD, these results were considered more theoretical than

practical. Studies with dietary CAD concentrations above +20 were

needed. West et al. (1990) provided additional information when they

evaluated diets with +2.5, +15, +27.5 and +40 CAD. They noted that

increasing CAD from +2.5 to +27.5 improved DMI, milk yield and blood

bicarbonate concentration. No additional improvement was observed in

cows fed diets between +27.5 and +40 CAD. In another study by that

group (West et al. 1991), diets with even higher CAD (+10, +21.7, +33.4

and +45.1) were fed. Increasing CAD increased DMI and blood pH

linearly, and blood HCO3 curvilinearly.

From these studies it appears that cation-anion mineral

interrelationships have a significant influence on acid-base status and

lactational performance of dairy cattle. But many questions still

remain. Projects in this dissertation were designed to address those

questions. Overall objectives were:

(1) to determine optimal dietary concentrations of Na, K and Cl

and CAD over a wide range of concentrations;

(2) to identify and quantify the influence of Na, K and C1

interrelationships and CAD on acid-base status and

lactational performance;

(3) to determine the influence of different commercial sources

of Na, K and C1; and,

(4) to characterize physiological responses to the dietary K x

Cl interrelationship that was discovered during the project.













CHAPTER 2
LITERATURE REVIEW

Review of Na, K and C1 Metabolism
This review will focus on literature related to the hypothesis

that cation-anion mineral interrelationships influence acid-base status

and lactational performance of dairy cattle. The chemical

classifications; function, absorption and excretion; requirements and

recommendations of Na, K and Cl will be reviewed. This will be followed

by a discussion of the influence of dietary Na, K and Cl on acid-base

status and lactational performance of dairy cattle.


Chemical Classification

Nutritionists refer to Na, K and C1 as macrominerals because they

are required in greater quantities in the diet and are present at higher

concentrations in animal tissue than trace minerals (NRC, 1989).

Chemically classified as alkali metals, Na and K occur mainly as ions in

biological systems. They possess a single valency electron which is

bound weakly. A feature of C1, which is a halogen, is the propensity

for its atoms to gain electrons, especially in the formation of the

alkali salts NaCl and KC1.










Function, Absorption and Excretion

Sodium. Sodium in the mammalian body occurs primarily in the

extracellular fluid and in bone (Aitken, 1976). The exchangeable

fraction is responsible for regulating extracellular fluid volume and

acid-base status (McKeown, 1986). In addition, Na is involved in nerve

impulse transmission, muscle contraction and regulation of plasma and

cellular water and ion content (Lunn and McGuirk, 1990). Sodium also

plays a critical role in the elaborate Na-K adenosine-triphosphatase

enzyme (ATPase) responsible for creating the electrochemical gradient

required to orchestrate cellular transport.

The idea of a Na pump in the cell membrane was introduced by Dean

(1941) in his classical paper on the theories of electrolyte equilibrium

in muscle. Skou (1957) described an ATPase that was Na and K dependent.

He suggested that the enzyme might be an essential part of a Na-K pump

responsible for maintaining high K and low Na in cells. It was not

until 1979 when it was proven "beyond reasonable doubt" that the pump

was actually a Na-K pump, coupling movement of three Na' to the outside

of the cell for two K' to the inside of the cell (Neilsen, 1979).

Considerable information has since been accumulated, demonstrating that

the Na-K ATPase is a transmembrane substrate of the Na-K pump.

The Na-K pump is universal to cellular physiology and enables all

eukaryotic cells to function. The potential energy of the ionic

chemical gradient created by the Na-K pump drives cellular transport.

Other nutrients are transported by gradients set up by the Na-K pump.

Phosphate, amino acids, and glucose are moved into cells and H, Ca",










HC03", K* and C1 are moved out of cells by these gradients (Lechene,

1988).

Sodium salts are very soluble within the gastrointestinal tract

(Peeler, 1972). Most dietary sources of Na are presumed to be nearly

completely available. O'Dell and Savage (1966) indicated that acetate,

citrate and carbonate forms of Na were more effective than NaCl in

stimulating growth of chicks. Cation-anion interrelationships and not

biological availability of Na salts may have been responsible for these

effects (to be discussed in a later section). Because Na also exists in

non-exchangeable form within the crystalline fraction of bone (Edelman

et al., 1954), the Na in animal by-product feedstuffs may not be as

bioavailable as in other feed sources of Na. The ARC (1980) estimated

that 91% of consumed Na was absorbed.

A close association between Na, K and C1 excretion exists. The

kidney is the primary organ regulating excretion of these ions (Lunn and

McGuirk, 1990). Sodium is the primary effector of ion excretion and

changes in reabsorption are the primary determinants of Na excretion.

Hormone systems, including aldosterone, renin-angiotensin and arterial

natriuretic factor, work with receptors in various tissues to monitor Na

concentration, which in turn controls fluid volume, blood pressure and

renal processing of the other ions.

The specific sequences of events related to hormonal regulation

are as follows. Upon detection of reduced pressure in the renal

afferent-arterioles, reduced plasma volume and/or reduced plasma Na, the

juxtaglomerular cells of the kidney secrete rennin. Rennin is an enzyme

that cleaves a decapeptide, angiotensin I, from a plasma alpha-globulin,










angiotensinogen. A converting enzyme present in plasma then cleaves a

dipeptide from angiotensin I and forms angiotensin II. Angiotensin II,

which also is stimulated by adrenal corticotrophic hormone and excessive

plasma K, mediates the release of aldosterone by arcuate cells of the

zona glomerulus in the adrenal cortex. Aldosterone acts in the cortical

collecting tubule to increase luminal permeability to Na and K and to

increase the peritubular Na-K/ATPase activity (Kleinman and Lorenz,

1989; Lunn and McGuirk, 1990). The net effect is resorption and

retention of Na and excretion of K.

Potassium. In contrast to Na, most of the K in the mammalian body

is inside cells. Potassium is the chief intracellular cation and most

concentrated mineral element in milk (Hemken, 1983). Potassium

functions in acid-base and osmotic pressure regulation, 02 and CO2

transport, and nerve and muscle contractions and is essential to many

enzyme reactions.

There is considerable variation in red blood cell (RBC) K reported

in the literature. Values between 22 and 106 meq/kg have been reported

for bovine RBC K (Aitken, 1976; Hemken, 1983). The wide variation may

be due to the method of measuring RBC K. Potassium in the RBC has been

measured either directly by separating red cells from plasma in a

centrifuged sample of whole blood, or indirectly by estimating K in

whole blood and in plasma, measuring packed cell volume, and then

calculating K by taking the difference. The first method is usually

more accurate because indirect measures can lead to large errors.

However, centrifugation may not give complete separation of plasma from

RBC so it is necessary to wash the cells with an isotonic solution that










removes excess plasma but maintains viability of the RBC. If blood is

handled such that cells are not ruptured, washing may be the preferred

method to estimate RBC K (Aitken, 1976). There also are different ways

to express RBC K. The concentration of K in the RBC can either be

expressed as meq/kg dry cells or meq/L wet cells.

Potassium in feed occurs as simple ions, which upon consumption,

are readily solubilized within the gastrointestinal tract and are almost

completely absorbed. It was concluded by Hemken (1983) that the true

digestibility of K is relatively high (95% or higher) for most

feedstuffs. Dietary K from both inorganic and organic sources is

utilized efficiently (Peeler, 1972).

Urine is the main excretory route for dietary K. As discussed

previously this is primarily controlled by aldosterone. Blood pH also

affects urinary K excretion (McGuirk and Butler, 1980). At the onset of

an alkalotic condition, intracellular hydrogen protons are exchanged

with plasma K as part of the regulatory mechanisms that control blood

pH. In the renal tubules, a large gradient exists between intracellular

and luminal fluid (urine) K. This gradient causes K to leave the tubule

cells and enter the urine. The importance of K administration in the

treatment of metabolic alkalosis in ruminants has been demonstrated

(McGuirk and Butler, 1980). The feces, which are another route of

eliminating K, can contain both undigested and endogenous K. Paquay et

al. (1969b) estimated that 2.2 g of K per kg DMI were eliminated through

the feces.

Chloride. Chloride comprises the major anionic component of

mammalian extracellular fluid. Throughout the body, the metabolism of








10

C1 is associated intimately with Na and K function. Chloride functions

in the maintenance of acid-base equilibrium, transport of 02 and CO2,

and formation of gastric HC1 (NRC, 1989). Most texts have described the

role of C1 in maintaining ionic and fluid balance as passive to that of

Na and K. However, research with lactating dairy cattle has shown that

during C1 deficiency, the ion functions independently to mediate C1

conserving mechanisms (Fettman et al., 1984b). Cows fed a deficient

amount of Cl in the study of Fettman et al. (1984b) and in several other

studies conducted at Cornell University were able to conserve Cl by

reducing excretion in urine, feces and milk. It also was demonstrated

that cows may have a specific appetite for C1. Cows fed diets with a

deficient amount of Cl sought out and consumed more salt block than cows

fed diets with adequate amounts of C1.

As a strong ion, C1 is always dissociated in solution (Stewart,

1981). Although not extensively studied, true digestibility of C1 from

all feed sources is presumed to be very high (i.e. 90% or greater). The

ARC (1980) used 91% as the absorption efficiency for C1. Paquay et al.

(1969a) found that Cl digestibility was not influenced by Cl intake but

was negatively correlated with DMI, energy, and pentosan intake; and

positively correlated with nitrogen and K intake. Although Cl

digestibility is high, the concentration in feedstuffs varies

considerably (Adams, 1975; Coppock and Fettman, 1977).

Absorption of C1 represents an interesting phenomenon. Thirty

years ago, Sperber and Hyden (1952) showed that C1 was transported

through the rumen against a large concentration gradient (normal ruminal

fluid concentration may range from 10 to 30 meq/L, whereas plasma C1








11

normally ranges from 90 to 110 meq/L). Therefore, C1 was assumed to be

actively transported. Martens and Blume (1987) verified that C1 was

actively co-transported with Na across the ruminal wall in sheep. The

mechanism of co-transport could not be deduced. Chloride absorption in

the upper small intestine is via passive diffusion; Cl follows Na along

an electrical gradient (Coppock, 1986). In the distal ileum and large

intestine, Cl absorption is exchanged with bicarbonate secretion.

Bicarbonate ions may serve to neutralize acids produced by intestinal

fermentation (Coppock, 1986).

Studies with steers and wethers have shown that approximately 98%

of ingested C1 was excreted in the urine (Nelson et al., 1955).

However, in several studies with lactating cows, a major portion of C1

consumed was excreted in the feces (Coppock, 1986). In general,

negative ion concentration in extracellular fluid is regulated

secondarily to regulation of positive ion concentration, and when the

load exceeds maximum capability for reabsorption from the kidney, the

excess is excreted in urine (Hilwig, 1976). A reciprocal relationship

exists between Cl and HC03' ions in the kidneys (Kleinman and Lorenz,

1989). Excretion of one Cl ion is coupled to reabsorption of one HC03'

(or vice versa, depending on systemic pH; Fischer et al., 1983).


Dietary Requirements and Recommendations

Methods of Determination

Requirements and recommended allowances of dietary Na, K and C1

have been established (NRC, 1989; ARC, 1980). Both factorial and

empirical methods have been used in their determination.










Dietary requirements have been established by the factorial

method. This method involves the estimation of the amount (i.e. grams)

of an element that is required for body growth, pregnancy and lactation.

These factors are then added to the maintenance requirements needed to

offset incomplete absorption, endogenous secretion into the

gastrointestinal tract, urinary excretion, and insensible losses through

dribbled saliva, sweat, tear ducts and skin. The factorial method thus

establishes net requirements (ARC, 1980).

Dietary recommendations have been defined as the diet

concentration (i.e., percentage or parts per million) of a mineral

needed to achieve some predictable animal response (Beede, 1991). These

have been established using the empirical method, which consists of

adding graded quantities of a mineral to the diet and then measuring the

selected response.

Beede (1991) suggested that mineral requirements determined by the

factorial method are conceptually sound but have limited practical

value. This is because requirements may be determined with animals in

varied physiological states, fed with varied numbers of feeds. Dietary

recommendations include a margin of safety that accounts for variation

due to the physiological state of the animal and due to different feed

sources and combinations of feeds. The NRC (1989) provides tables for

both requirements and recommendations for minerals. The following will

provide a summary of recent dietary recommendations for Na, K and C1.










Recommended Dietary Concentrations of Na, K and C1

Sodium. The NRC (1989) recently updated mineral requirements and

made considerable changes in recommendations for dietary Na, K and Cl.

Due to the historical use of common salt (NaCl) to meet Na and C1 needs,

recommendations were reported previously as a general recommendation for

common salt (NaCl). However, an important separation was made (NRC,

1989) between the requirements for "salt" and the individual components

of salt (Na and Cl). This refinement should prove useful because

physiological needs are related to the individual ions, not the salt

molecule (which of course dissociates into Na and C1 upon consumption).

Also, because of the trend to feed dairy cows NaHCO3 as a source of Na

(instead of NaCl), this separation should help eliminate potential

underfeeding of Cl.

The NRC (1989) lists .18% as the recommended Na concentration for

lactating dairy cattle. However, this concentration may not maintain Na

balance. The NRC (1989) states that data from balance experiments

involving lactating cattle fed many different diets showed that Na

balance was negative when dietary Na was below .20% DM. Lactational

performance may be enhanced with Na concentrations above .18%. Florida

researchers found that lactating dairy cattle consumed more feed and

produced more milk when fed diets with higher than recommended

concentrations of Na. In a cool weather study, Mallonee et al. (1982)

fed three concentrations of Na: .16, .42, and .70% of diet DM. Several

cows fed .16% Na exhibited classical signs of Na deficiency within 2

weeks. Cows fed diets .42% Na consumed the most feed, whereas cows fed

diets with .7% Na produced the greatest quantities of milk (Na results










were averaged across K concentrations, another variable in the study).

Because diets were not isochloridic in that study, responses may not

have been wholly due to Na.

In a warm weather study with equal concentrations of dietary Cl

(Schneider et al., 1986), increasing dietary Na above NRC (1989)

recommendations (from .18 to .55% using either NaC1 or NaHCO3 as the

source of Na) improved MY and FCM yield. Other researchers have

reported increased MY and milk composition with increasing dietary Na,

but have attributed those improvements to ruminal buffering by HC03O

rather than to Na (Erdman et al., 1980; Rogers et al., 1982a and 1982b;

Kilmer et al., 1981). Schneider et al. (1986) suggested that

lactational responses to increasing dietary Na may be due in part to Na

per se and not wholly to HC3O_. Subsequent calculations demonstrated

that 60% of the FCM yield response was due to Na. Additional research

will be needed to better define the correct amount of Na to feed to

lactating dairy cattle.

Potassium. The NRC committee increased the recommendation for

dietary K in their most recent update (NRC, 1989). Minimum

recommendations were increased from .8 to .9% for average milk

production and to 1.0% for high production. Factors responsible for the

increase included: (1) inclusion of higher milk production categories;

(2) potential use of by-product feeds with lower concentrations and
reduced bioavailabilty of K; and (3) increased need during heat stress.

Historically, dietary K recommendations have been the subject of debate.

Recommended concentrations of dietary K (DM basis) for lactating

dairy cattle have included .5% (Ward, 1966), .7% (NRC, 1971), .8%










(Dennis et al., 1976; Dennis and Hemken, 1978; NRC, 1978; Erdman et al.,

1980), .9% (NRC, 1989), 1.0% (Linsner, 1980) and 1.2% (Bolenbaugh,

1977). Even greater concentrations have been suggested for cows

strained by heat and humidity (Mallonee et al., 1985; Schneider et al.,

1984b, 1986; West et al., 1987b). It is possible that some of the

discrepancies between published K recommendations were due to differing

dietary Cl concentrations used in the various studies. Paquay et al.

(1969b) found a close correlation between K and C1 in the urine and

suggested that adequate C1 was needed to accompany urinary excretion of

excess dietary K. Dietary C1 concentrations in most studies that

investigated K were not reported making it difficult to evaluate K x Cl

interactions. In retrospect, analyzing diets for C1 would have been

valuable.

Beede et al. (1983) reviewed the relationships of K nutrition and

heat stress in lactating cows. During heat stress, dietary K needs are

increased. Losses of K via sweat and dribbled saliva increase and are

compounded by reduced feed and K intake. Cows without the benefit of

shade responded to increasing dietary K (from .66 to 1.08%) more than

did cows under shade. Increasing dietary K (from .93 to 1.29 and 1.53%

K) improved lactational performance during hot weather in Texas (West et

al., 1987b).

As mentioned previously, alkalosis can increase urinary excretion

of K. During heat stress, cows pant to cool themselves because panting

alters alveolar ventilation and CO2 is eliminated faster than it is

produced. In addition, partial pressure of CO2 (pCO2) is lowered and

blood pH rises (Collier et al., 1982). Two factors associated with this










heat-stress-induced-alkalosis contribute to a need for increasing

dietary K. First, decreased pCO, reduces renal tubular acid secretion

and can cause compensatory loss of alkali reserve (i.e. to maintain

electroneutrality of the urine, K' replaces H'). Second, at the onset

of alkalosis, intracellular H' ions are exchanged with plasma K as part

of the regulatory mechanisms that control H. In the renal tubules, a

large gradient exists between intracellular and urinary K. This

gradient causes K to exit the tubule cells and become excreted in urine.

Chloride. The Cl recommendation recently was set at .25% of diet

DM (NRC, 1989). This guideline is based on work designed to establish

the minimal requirement of Cl and role of C1 in acid-base status

(Fettman et al., 1984b; Coppock, 1986). Fettman et al. (1984b)

concluded that the requirement for C1 was above .1%. In their published

figures, average daily milk production for cows fed the .45% C1 diet

appeared lower but was not different from cows fed the .27% C1 diet.

Low dietary C1 concentrations (.1% C1) led to subclinical primary-

hypochloremic secondary-hypokalemic metabolic acidosis as well as

reduced feed and water intake, body weight, and milk production.

Coppock (1986) suggested that about .2% Cl is required for a

lactating cow in midlactation. He suggested that .25% would be too low

during peak lactation and negative energy balance. Underwood (1981)

proposed that the Cl requirement should be substantially higher than the

Na requirement because cow milk contains more than twice as much C1 as

Na. Other factors that may affect C1 needs include type of diet, rate

of growth, pregnancy and heat stress (Coppock, 1986).









Chloride values of feeds are not always reported in feed tables,

appear highly variable, and are missing from many ration balancing

programs. Commercial feed analysis laboratories often only report feed

Cl concentrations upon special request (Coppock, 1986; Chandler, 1988).


Influence of Cation-Anion Interrelationships on Acid-Base Status


General Principles of Acid-Base Status

To understand how macromineral cation-anion interrelationships

influence acid-base status, a review of acid-base status is essential.

Concentration of H' ([H']) in extracellular fluid (ECF) is found in very

small quantities--one tenth to one hundred millionth of an eq/L

(approximately 40 neq/L, i.e., pH 7.4). This is a trillion times less

concentrated than water. Although [H'] is minuscule, [H*] is critical

to living systems for several key reasons (Stewart, 1981; Morris, 1986):


(1) because H' atoms are so small they have a large charge
density and a large electric field gradient,

(2) hydrogen bonds are important in determining molecular
structure and configuration,

(3) enzyme activity is very sensitive to [H'],

(4) [H'] turnover amounts to more than 150 mole/d in the human,
which is more than any other metabolite, and,

(5) because H20 can dissociate into H' there is an infinite
supply of H' ions.










Regulation of Acid-Base Status

The regulation of pH in body fluids ranks high among the

homeostatic priorities. Boron (1986) probably said it simplest when he

stated that "virtually every biological process is pH sensitive."

Stability of the [H'] is maintained by a number of biological buffering

mechanisms involved in various systems of the body (Hilwig, 1976).

A series of chemical buffer systems (bicarbonate-carbonic acid

buffer pair, phosphates, plasma proteins, and hemoglobin) respond to

short-term perturbations. These systems act to neutralize acids or

bases produced by tissues or derived exogenously. Buffering mechanisms

in the respiratory system act to provide the main route of elimination

of CO2 and are of prime importance in maintaining the HCO3:pCO2 ratio, a

determinant of extracellular pH. The third method of [H'] regulation is

through renal elimination of non-volatile acids and bases (Hilwig,

1976).


Electrolytes and Strong Ions

Many substances, when dissolved in solution, dissociate into ions.

These substances are chemically defined as electrolytes because they can

conduct electricity. Strong electrolytes are by definition (Stewart,

1981) completely dissociated, whereas weak electrolytes change their

degree of dissociation. Most substances that contain Na, K and Cl are

strong electrolytes and dissociate completely and rapidly in solution.

Other macromineral sources have variable dissociation constants

(Stewart, 1981). Because they arise from strong electrolytes, Na, K and










C1 are classified as strong ions and are ionized in solution. A

physical law that applies to ionic solutions is the law of electrical

neutrality (Stewart, 1981). Biological fluids must be maintained in an

electrically neutral environment. The sum of positive charges must

equal the sum of negative charges. Dietary electrolytes contribute

positive and negative charges to the body and therefore can alter

electrical balance, acid-base status, and subsequent animal

performance. Because ions react according to charge (valence) they

should be expressed in moles (or equivalents) of charge rather than

moles of atoms. This is easily demonstrated by the reaction of one mole

(1 g) of H ions with one mole (35 g) of C1 ions. The reactants have the
same number of atoms but are drastically different in weight. When

considering the influences that strong ions and cation-anion inter-

relationships have, concentrations should be converted to milli-

equivalents (meq; measure of ionic charge).


The Strong Ion Difference

Peter Stewart, the late muscle physiologist, quantified [HW] in

body fluids and demonstrated mathematically that [H] of the blood

depends only on three variables: the total weak acid; pCO2 in the

blood; and the difference between strong cations and strong anions in

blood (strong ion difference: SID). The concept of SID is becoming

widely used in veterinary medicine (Eicker, 1990) and is notably similar

to the dietary cation-anion difference concept used by animal

nutritionists (to be discussed later).










Specific Effects of Macromineral Salts on Acid-Base Status

Nutrient metabolism results in the degradation of nutrient

precursors into strong acids and bases. During normal metabolism the

flux of H* is great. In typical rations fed to dairy cattle, inorganic

cations exceed dietary inorganic anions by several meq per day. Carried

with excess dietary inorganic cations are organic anions which can be

combusted to HC03. Therefore, a diet with excess inorganic cations

relative to inorganic anions is alkaline (Austic, 1988). However, for a

dietary mineral salt to affect acid-base status it must satisfy the

following three criteria:


(1) the mineral salt must be dissociated and solubilized in the
gastrointestinal tract.;

(2) upon dissociation and solubilization it must be absorbed;
and,

(3) after absorption, any non-mineral portion must be
metabolized such that [H] in blood increases or decreases.


Although NaCl meets the first two criteria, it does not affect

acid-base status because neither Na nor Cl are metabolized. Thus NaCl

has a neutral effect on acid-base status (disregarding indirect solute

effects). Examples of mineral salts that meet these criteria include

ammonium chloride, an acidogenic agent, and sodium bicarbonate, an

alkalogenic agent. Examining their chemical reactions upon absorption

illustrates how each affect acid-base status.

Ammonium chloride (NH4CI). When NH4CI reaches the

gastrointestinal tract it first dissociates into NH4+ and Cl ions. The

NH4 ions are further metabolized into two NH3 molecules and one H' ion.









Two NH3 molecules combine with CO2 (in the urea cycle) and form urea.

The steps in the reaction are:

(1) 2 NH4C1 <------> 2 Cl H + 2 NH4

(2) 2 NH4 ------> 2 H+ + 2 NH3
(3) 2 NH3 + CO2 ------> urea + H20

Because H' ions in blood are increased (underlined in the above

reactions), NH4C1 is an acidogenic salt. The C1 ion is fixed but NH4+

is metabolized with the net addition of two H ions. The metabolism of

the NH4 portion of NH4C1 is the key reason why NH4C1 is considered an

acidogenic salt. Note that some texts refer to C1 as an acidogenic ion.
Chloride is only acidogenic when it is associated with a metabolizable

cation (or with H as in HC1). Note that (NH4)2S04 affects on acid-base

status are similar to that of NH4C1.

Sodium bicarbonate (NaHCO3. Another example of a salt that
affects acid-base status is NaHC03, an alkalogenic salt, commonly fed to

ruminants to buffer ruminal pH. Upon metabolism of NaHCO3, the Na' ion

is fixed and will not affect acid-base status. But the HC03" is

metabolized to H20 and CO2 with the net removal of one H0 ion.

(1) NaHCO3 + H20 <-----> Na+ + OH_ + HCO,3

(2) HC03O + H+ <-----> H20 + CO2 (expired in lungs).

Ammonium chloride is acidogenic and NaHCO3 is alkalogenic because
the non-mineral portion of these salts are metabolized. It is not the
minerals per se that are acidogenic or alkalogenic but rather the non-

mineral portions that contribute or consume H'.









Calcium chloride (CaC12). One additional mechanism exists to

explain how a mineral salt affects acid-base status. In this case only

a portion of the mineral is absorbed. An example of a type of salt

that functions in this manner is CaCl2. Upon consumption of CaC12, only

a small percentage of Ca is absorbed (Peeler, 1972). To maintain

neutrality in the digestive tract, unabsorbed Ca* ions combine with

HC03" to form the precipitate CaCO3, which is then excreted in the

feces.

CaCl2 + HCO,3 <-----> 2 Cl + H* + CaCO3 (precipitate).

Calcium chloride is an acidogenic salt because it increases

systemic H. However, the reason CaCl2 is acidogenic is due to

incomplete absorption of Ca. Other acidogenic salts which react and

behave similarly are MgCl2, and Mg2SO4.

Note that the pH of a salt (or feed) should not be used to predict

the effect it will have on acid-base status. For example, a chemically

neutral salt such as CaCl2 results in acidosis, whereas Na2HPO4,

chemically an acid, results in alkalosis.


Nutritional Concepts Related to Cation-Anion Interrelationships

Leach (1979) and Mongin (1980) reviewed nutritional concepts

related to cation-anion interrelationships. Historically,

nutritionists intuitively knew it was difficult to evaluate the effect

of one macromineral without considering the influences of others. Early

concepts evaluated total ash, mineral ratios, and differences among two

or more of the macrominerals.










Acid or Alkaline Ash

Nutritionists first investigated the alkalinity and acidity of the

diet under the acid- or alkaline-ash concept (Shohl, 1939). It was

recognized that human food either had an acid or alkaline ash. When

food is metabolized in the body, organic anions, such as acetate,

citrate, malate, etc., are oxidized. Inorganic cations originally

associated with these organic anions remain. Because organic anions can

buffer H* ions generated through metabolism (see next section), a food

with a large amount of organic anions (and thus inorganic cations) was

considered alkaline. The pH of the ash represented the acid or alkaline

nature of human food.


Ratios

Later concepts used to evaluate the effects of macromineral

interrelationships involved ratios of two or more minerals. Leach

(1979) listed several macromineral ratios that were used to express the

influence of macromineral interrelationships on acid-base status and

animal performance. Apparently, the use of ratios were not widespread.

A professor from the University of Florida Poultry Science Department

advises his students that mineral ratios can be misleading (R.D. Miles,

personal communication). As an example he cites the Na/Cl ratio

mentioned by Leach (1979). A typical lactating dairy cattle diet might

have about 15 meq of Na and 7.5 meq of C1 (2:1 ratio). The difference

between meq from Na and C1 is 7.5 meq (15 meq from Na minus 7.5 meq from

C1). If the meq from each ion were doubled in the diet, the ratio would

still be 2:1 but the difference would grow to 15 meq. The additional









7.5 meq Na could impact acid-base status, yet, this would not be

detected by calculating the ratio. Other researchers recognized the

shortcoming with ratios and began to use differences.


Dietary Undetermined Anion

One of the most common difference expressions used was the

"dietary undetermined anion" (DUA) (Austic, 1988). With DUA, cations

and anions are classified as separate entities. All seven macrominerals

are considered in the expression:

DUA = meq [(Na + K + Ca + Mg) (C1 + P + S)]/100 g diet DM.

Under the law of electrical neutrality (Stewart, 1981), the positive and

negative charges in the diet need to balanced electrically. If the

mineral cations in the DUA expression exceed the mineral anions, then

some "undetermined" anions must be present in the diet because by

definition and in actuality the feed is electrically neutral. These

undetermined anions presumably consist of carboxylate groups of organic

compounds, such as bicarbonate, citrate, lactate, and fatty acids

(Austic, 1988). Because dietary organic anions are precursors to HC03,,

they contribute alkali. Consequently, DUA measures alkaline potential

of the diet. Because trace mineral cations and anions do not represent

large concentrations in the diet, ionic contributions from trace

minerals are ignored in the DUA expression. A problem with including

the multivalent macrominerals (Ca, Mg, P, and S) in the DUA expression

for ruminants, relates to the variable and incomplete bioavailability of

these ions compared to Na, K and C1 (Peeler, 1972).









Cation-Anion Difference (CAD)

The expression most widely used in ruminant nutrition is the

monovalent cation-anion difference (CAD) expressed as Na + K Cl (in

meq/100 g diet DM). This expression is considered superior because it

comes closest to representing feed ions that are completely dissociated,

solubilized from their respective salts, and absorbed into the body.

To calculate CAD, mineral concentrations are first converted to

milliequivalents using the following equation:


meq/100 g = (milliqrams)(valence)
(atomic weight)


As an example, the CAD value of a diet with .18% Na, .9% K and .25% Cl

(minimum recommendation for lactating cattle; NRC, 1989) will be

calculated. There are 180 mg Na (.18% = .18 g/100 g or 180 mg/100 g),

900 mg K and 250 mg Cl per 100 g diet DM. Therefore this diet contains:

meq Na = (180 mq)(1 valence) = 7.8 meq Na
(23 g atomic weight)

meq K = (900 mq)(1 valence) = 23.1 meq K
(39 g atomic weight)

meq Cl = (250 mq)(1 valence) = 7.0 meq C1
(35.5 g atomic weight).


The next step is to sum the meq from the cations and subtract the meq

from the anions:

meq (Na + K C1) = 7.8 + 23.1 7.0 = + 23.9 meq/100 g diet DM.

Therefore, a diet for a lactating dairy cow containing the minimum

recommended concentrations of Na, K and C1 has a CAD of +23.9.








26

This method to calculate CAD is straight forward due to the single

charge present on the monovalent ions. The incomplete dissociation of

other macrominerals affects charge valency and thus acid-base status

(Dweyer et al., 1985). For example, P, a divalent anion, actually is

assigned a valence of 1.8 (instead of 2) because in blood, P exists as

both H2PO4 and HP04. Because approximately 80% exists as the divalent

anion, the valance is assumed to be 1.8 [(.8 x 2) + (.2 x 1) = 1.8]. An

assumption to the degree of dissociation (which is affected by several

factors) is critical to the estimated influence of P on acid-base

status.


Cation-Anion Interrelationships and Effects of
Cation-Anion Difference on Animal Performance

Sodium, Potassium and Chloride Interrelationships

Because Na, K and C1 function together to maintain acid-base

equilibrium, it is likely that they have interrelated effects on animal

performance. Studies have been conducted on requirements for each

mineral, but data on their interrelationships are limited.

Studies from other species suggest that optimal concentrations of

each of these macrominerals depends upon relative concentrations of the

others (Fontenot et al., 1960; Scott, 1970; Johnson and Karunajeewa

1985). Studies conducted with lactating dairy cattle on the Na x K

interaction have had variable results. Some have found that the

response to dietary Na concentration depended on the associated

concentration of dietary K (Schneider et al., 1986), whereas others have

found that response to dietary Na did not depend on dietary K (Erdman et










al., 1980; O'Connor et al., 1988). Results may have been related to

whether or not C1 concentrations were equalized among treatments.

Most published dietary Na x K interaction effects were related to

nutrient absorption which usually indicated a sparing of the two cations

for each other. In poultry and rats, additional dietary Na spared a

portion of the K requirement (Kumpost and Sullivan, 1966; Burns et al.,

1953; Grunert et al., 1950). Fontenot et al. (1960) reported that

additional dietary Na depressed K absorption in lambs. Increasing

dietary K intake in sheep resulted in an increase in fecal Na (Suttle

and Field, 1967). Scott (1970) found that high dietary K impaired

intestinal absorption of Na and low dietary K increased urinary Na

excretion in cattle. Campbell and Roberts (1965) reported that

apparent intestinal absorption of Na in heifers was impaired by high

dietary concentration of K but lower concentrations of K increased

urinary loss of Na. Scott (1967) observed that an increase in the

ruminal fluid concentration of one of these ions is accompanied by a

reciprocal decrease in the other, resulting in an almost constant meq

sum of Na plus K. Jackson et al. (1971) observed an Na x K interaction

on microbial populations in the rumen.

In lactating cattle, previous studies on the interrelationships

between dietary Na and K have not revealed a sparing effect of dietary

Na and K for each other. Erdman et al. (1980) found no benefit of

additional dietary Na (.52 vs. .31%) with either low (.42%) or adequate

(.84%) dietary K. O'Connor et al. (1988) also reported no benefit on

lactational performance due to additional dietary Na (.24 vs. .62%) with

either 1.14 or 1.59% dietary K. Chloride was not equalized across diets








28
in those studies which could explain the lack of response. Martens and

Blume (1987) observed that Na and C1 absorption in sheep was coupled by

a dual co-transport mechanism for Na*/H' and C1-/HC03' which was related

to ruminal K concentrations. An alteration in the relative amounts of

dietary Na and K thus could affect acid-base status which in turn could

affect lactational performance.

Coppock et al. (1982a) studied Na x Cl interactions with cows fed

diets with Na:Cl ratios between 1:1 and 4:1 and did not find differences

in MY or milk composition. Belgium researchers (Paquay et al., 1969)

found a positive correlation between K and C1 in the urine of cows and

suggested that the need for dietary K may be related to dietary C1.


Cation-Anion Difference

The cation-anion concepts of Shohl (1939) appear to be creeping

back into animal nutrition with the recent findings that dietary mineral

interrelationships affect acid-base homeostasis. Mongin (1980)

concluded that in order to maintain acid-base homeostasis, an animal

needs to regulate input and output of acidity. Net acid intake can be

extrapolated from the difference between dietary fixed cations and

anions (i.e. those ions that are not metabolized during digestive or

metabolic processes). Tucker et al. (1988a) discovered that acid-base

balance and lactational performance were related to dietary CAD.

Nonruminants. The bulk of research on the CAD concept has been

conducted with nonruminants. Effects of CAD on acid-base status and

various health and growth responses in poultry have been investigated

extensively. Diets with excessive mineral Cl and other anions, were










deleterious to acid-base status, growth rate, eggshell mineralization,

and incidence of tibial dyschondroplasia (Austic, 1988). Mongin (1980)

cited one study that tested various combinations of CAD on growth of

broiler chicks. That data described a uniform bell shaped curve

response to increasing CAD. Optimum performance was established at +25

to +30 CAD. Several diets were marginally deficient in Na, K or C1

which could explain some of the dramatic effects observed.

Growth responses to CAD also have been observed with swine. Yen

et al. (1981) examined the effect of CaCl2 and NaHCO3 on feed intake and

weight gain. They observed negative effects from feeding a diet with 4%

CaC12 consistent with disturbances in acid-base status. When 2.03%

NaHCO3 was added to this diet, deleterious effects were eliminated.

Patience et al. (1987) reported no difference in growth rate between 0

to +34 CAD but observed a depression with -8.5 CAD. Golz and Crenshaw

(1990), reported maximal growth in growing pigs fed a diet with +23.8

CAD but an interaction between K and C1 influenced gains to a greater

extent than CAD.

Dairy cattle. Most of the research with CAD in dairy cattle

nutrition has been derived through studying milk fever problems and dry

cow nutrition. It appears that significant progress can be made in

combating hypocalcemia by considering CAD in the dry cow diet

(Dishington, 1975; Lomba et al., 1978; Block, 1984; Oetzel, 1988; Wang

and Beede, 1992). In general, studies have shown that when dry cow

diets had negative CAD and when cows were in Ca balance, milk fever was

diminished. In one study, milk fever was prevented 92% of the time when

low CAD and high Ca was fed prepartum (Dishington, 1975). An increased










mobilization of bone Ca and enhanced absorption of Ca prepartum, which

makes greater amounts of calcium available postpartum, appeared to be

the cause.

Very few studies have examined the influence of CAD on lactating

dairy cattle. Wheeler (1981) summarized several studies designed to

investigate dietary buffers and could find no clear relationship between

dietary CAD and animal performance. However, these studies were not

designed specifically to evaluate CAD. Beef cattle did gain slightly

less when CAD was below +10 than when CAD was above +77.

Coppock (1986) also reviewed the influence of dietary CAD on

lactational performance of dairy cattle. In general it was noted that

ruminants could withstand higher CAD than poultry or swine. In

summarizing trials after the fact, he found that CAD had no influence

between +10 to +40. Escobosa et al. (1984) studied effects of feeding

either .23% NaCl, .23% NaCl plus 2.28% CaCl2, or .23% NaCl plus 1.7%

NaHCO3. Diets had -14, +20 or +35 CAD. It was found that the excess

Cl (-14 CAD) depressed feed intake and resulted in acidosis.

Tucker et al. (1988a) in Kentucky appear to have been the first to

conduct a study specifically designed to evaluate the effect of CAD on

acid-base status and lactational performance of dairy cattle. They

compared diets formulated with -10, 0, +10 or +20 CAD. Twelve

midlactation cows were assigned to three 4 x 4 Latin squares. Squares

were arranged in a split-plot. Sub-plots represented either Na, K or Cl

formulated diets and main-plots were the different CAD concentrations.

A diet with +20 improved DMI 11% and MY 9% compared with a diet with -10

CAD. Blood HCO, increased linearly with increasing CAD which indicated










an improvement in acid-base status with high CAD compared with low CAD.

Because there were no effects due to square, they concluded that

responses to increasing CAD were independent of specific Na, K or C1

effects. Because lactation diets typically contain greater CAD than

+20, these results were primarily theoretical rather than practical.

The next question that had to be answered was whether or not responses

would continue to increase with diets above +20 CAD.

West et al. (1990) in Georgia answered part of this question when

they evaluated diets with up to +40 meq/100 g diet DM. Their study used

two 4 x 4 Latin squares blocked by environmental temperature (cool vs.

hot). Separate squares included four Holstein and four Jersey cows.

Diets contained +2.5, +15, +27.5 or +40 CAD. No effect of environment

was reported but increasing CAD from +2.5 to +27.5 increased DMI, MY and

blood HC03-. These findings suggested that performance was depressed

with lower CAD. At +27.5 CAD, negative effects were overcome. Above

+27.5 CAD no additional improvement was attained.

In another study by this group (West et al. 1991), diets with even

higher CAD (+10, +21.7, +33.4 and +45.1) were fed to a total of 16

lactating dairy cows during hot weather. Source of cation (Na or K)

used to manipulate CAD also was compared. Increasing CAD increased DMI

linearly, independent of Na or K source. Yield of 3.5% FCM was not

affected by CAD or cation source. Milk fat concentration was greater

with Na- compared with K-manipulated diets (3.92 vs. 3.62%). Blood pH

increased linearly whereas blood HC03' increased curvilinearly; there

was no effect due to cation source on acid-base status. Their results

indicated that increasing CAD improved DMI and acid-base status in a










manner consistent with other studies. In general, CAD was independent

of a specific Na or K effect.

The influence of Na, K and Cl at constant CAD was evaluated by

Tucker and Hogue (1990). Diets were formulated to provide +32 CAD in

either: a basal diet (adequate in dietary Na, K and C1), a basal diet

containing an additional 1.17% NaCl, or a basal diet containing an

additional 1.56% KC1. Fifteen midlactation cows were assigned to

replicated 3 x 3 Latin squares. The KC1-fed cows consumed more DM and

had lower milk fat percentage than NaCl-fed cows, but there were no

differences in milk yield. It was concluded that dietary CAD was a more

important determinant of dietary impact on systemic acid-base status

than actual dietary concentrations of Na, K and C1.

In summary, dietary CAD appears to exert its effect on lactational

performance by altering acid-base status. When diets are below a

certain CAD they likely contain insufficient inorganic cations in

relation to inorganic anions. Amount of organic anions normally brought

to the diet by inorganic cations are decreased and therefore cannot

participate in buffering H'. Acidosis occurs and normal milk production

is compromised. From the limited data available, no specific dietary

CAD recommendation can be made, but the optimum appears to be somewhere

between +27.5 and +40. This may depend upon other factors such as the

digestibility, fermentability and acid producing potential of the diet,

dietary concentrations of other fixed ions, and rate of intake and

production capacity of the animal.

The major focus of this research was to address Na, K and Cl

nutrition of the lactating dairy cow. Determining the influence of








33

dietary Na, K and C1 interrelationships, particularly the three-way,

dietary cation-anion difference interrelationship, on lactational and

physiological responses was fundamental.














CHAPTER 3
INTERRELATIONSHIPS AMONG DIETARY SODIUM, POTASSIUM AND CHLORIDE:
EFFECTS ON ACID-BASE STATUS, MINERAL METABOLISM AND
LACTATIONAL PERFORMANCE OF DAIRY CATTLE

Introduction


Twenty years ago Jacobson et al. (1972) suggested that mineral

interrelationships are the least understood and most promising area of

dairy cattle mineral nutrition. The lactating dairy cows requirement

for each mineral has been established, but information on mineral

interrelationships is limited. Because Na, K and C1 function together

to maintain fluid balance, osmotic regulation, and acid-base equilibrium

(Kleinman and Lorenz, 1989), it is likely that they have interrelated

effects on animal performance.

Studies from other species suggest that optimal dietary

concentrations of these minerals depend on relative concentrations of

the others (Fontenot et al., 1960; Scott, 1970; Johnson and Karunajeewa

1985). For lactating dairy cattle, there have been studies that have

found Na x K interactions (Schneider et al., 1986) but there also are

studies that have found no Na x K interaction (Erdman et al., 1980;

O'Connor et al., 1988). Coppock et al. (1982a) studied Na x C1

interactions with cows fed Na:C1 ratios between 1:1 and 4:1 (percent of

DM) but did not detect differences in milk yield or milk composition.

Paquay et al. (1969b), however, found a positive correlation between K










and C1 in the urine of cows and suggested that the need for dietary K

may be related to concentration of dietary C1.

Mongin (1980) concluded that in order to maintain acid-base

homeostasis, an animal needs to regulate input and output of acidity.

Net acid intake can be extrapolated from the difference between dietary

fixed cations and anions (i.e., those ions that are not metabolized

during digestive or metabolic processes). Tucker et al. (1988a) showed

that acid-base status and lactational performance were related to fixed

cation-anion difference calculated as meq (Na + K C1)/100 g diet DM

(CA.D).

Objectives of the present study were three-fold: (1) to determine

optimal dietary concentrations of Na, K and C1, (2) to determine if

optimal dietary concentrations of Na, K and C1 were interrelated, and

(3) to determine if dietary CAD was related to lactational and

physiological responses.



Materials and Methods


Management

Forty-eight midlactation Holstein cows were fed diets (Tables 3-1

and 3-2) twice daily at 0800 and 1400 h. An electronic feeding system

(American Calan, Inc., NorthWood, NH) was used to measure daily feed

intake and refusals of individual cows. Total mixed diets (Table 3-1)

were made by combining corn silage with cottonseed hulls and concentrate

immediately before each feeding which was delivered in a mobile mixing

and feeding unit with electronic scales (American Calan, Inc.,










TABLE 3-1. Ingredient composition of basal diet.

Item Percent of DM

Corn silage 40.00
Ground yellow corn 31.00
Corn distillers dried grains 15.00
Cottonseed hulls 5.50
Hydrolyzed feather meal 2.50
Urea 1.00
Vitamin-mineral premix1 .75
Treatment mineral mixture2 4.25


1Contained Ca 29%, P 13%, Mg 2%, S 1%, Mn .22%,
Zn .33%, Cu .12%, I .007%, Se .003%, Co .0002%,
Vitamin A 110,000 IU/kg, Vitamin D3 99,000 IU/kg, and
Vitamin E 330 IU/kg.

2Contained various combinations of CaCO3, MgO, KC1,
NaC1, CaC12, Na2CO3, K2C03 and SiO2 (washed sand) in
proportions designed to obtain treatment formulations.


TABLE 3-2. Analyzed chemical composition of concentrate,
TMR (DM basis).


corn silage and


Nutrient
CP, %
ADF, %
Ca, %
P, %
Mg, %
S, %
Fe, ppm
Zn, ppm
Cu, ppm
Mn, ppm
NEL, mcal/kg3


Concentrate
Mean SEM
21.2 .28
14.8 .34
1.18 .035
.50 .012
.31 .008
.26 .002
379.0 25.0
74.5 1.76
16.1 .9
49.1 1.45
1.70 .002


Corn Silage
Mean SEM
8.65 .22
36.2 1.62
.35 .02
.21 .006
.27 .01
.15 .002
161.0 52.0
39.8 1.17
6.67 .3
31.6 3.8
1.43 .03


'Means and SEM for
composite samples taken


each component were calculated
throughout the experiment.


from analyses of


TMR calculation based on 60:40 concentrate:corn silage (DM basis);
cottonseed hulls were part of concentrate when sampled.

3Value calculated from chemical analysis.


TMR'
Mean
16.18
23.36
.85
.38
.29
.22
292.0
60.6
12.3
42.1
1.59










NorthWood, NH). Cows were fed so that 5 to 10% of feed (as-fed basis)

remained at 0630 h. Remaining feed was removed and weighed prior to

morning feeding. All cows were fed the same corn silage-alfalfa hay

based TMR during a 1 month preliminary period. Cows were housed in a

freestall barn and were randomly assigned to one of four equal sections

within the barn. Cows had access to exercise lots and drinking water at

all times. Milking was at 0500 and 1600 h.


Treatments

Each cow received a different dietary treatment in each of four

consecutive 28-d periods from early March to late June. Cows were

assigned treatments according to a partially balanced incomplete block

design. Treatments consisted of dietary combinations of Na, K and C1,

defined according to a three-factor second-order rotatable central-

composite design (CCD; Figure 3-1; Table 3-3; Khuri and Cornell, 1987).

This CCD consisted of eight treatments arranged in a 2 x 2 x 2 factorial

(high and low concentrations of Na, K and C1), plus six- 2 x 3 axial

treatments (very high and very low concentrations of Na, K and C1), and

one center point treatment (middle concentration of Na, K and C1). With

this CCD, only 15 treatments were required compared to the 27 that are

required in a standard factorial arrangement of treatments. In addition

to being more economical to run than a 3 x 3 x 3 factorial, this CCD

allowed each factor to be studied at five instead of only three

concentrations (Khuri and Cornell, 1987).

Treatment combinations of Na, K and C1 were selected by first

defining the center point (treatment 15; .55% Na, 1.4% K and .8% C1).








38













A 13


F. i

:iI ------



15 A.11
A -------- -- |^ ---------- A 1 1




*^----------i
Na



A 14






Figure 3-1. Three-factor central composite design used to evaluate
interrelationships among dietary Na, K and C1 concentrations. Center
point, factorial, and axial treatment design points are labeled,
numbered and indicated by .. C = center point (middle concentration of
Na, K and C1), F = 2 x 2 x 2 factorial points (low and high
concentration of Na, K and C1), and A = axial points (very high and very
low concentration of Na, K and C1). See Table 3-3 for dietary
concentrations.











TABLE 3-3. Dietary concentrations of Na, K, C1 and
anion difference (CAD) of experimental diets (% of


calculated cation-
diet DM).


concentration
Cl%
.5
.5
1.1
1.1
.5
1.1
1.1
.5
.3
1.33
.8
.8
.8
.8
.8


(DM basis)'
CAD2
62.0
46.7
29.8
45.1
29.3
12.3
27.7
44.6
51.3
23.1
52.0
22.4
50.0
24.4
37.2


SUnless otherwise indicated, concentrations shown were calculated
values required to provide Na, K, and C1 concentrations for a second-
order three-factor central composite design. Actual analyses were .31,
.34, .57, .74, .85; .86, 1.06, 1.42, 1.71, 1.96; .32, .50, .83, 1.08,
1.15 for Na, K and C1 respectively. Actual concentrations were used in
statistical analysis.

2 CAD = meq (Na + K C1)/100g diet DM.

3 Na and C1 concentrations intentionally were not formulated at
their very low and very high concentrations, respectively.


Treatment
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15


Na %
.75
.75
.75
.75
.35
.35
.35
.35
.55
.55
.89
.213
.55
.55
.55


Dietary
KA.
1.7
1.1
1.1
1.7
1.1
1.1
1.7
1.7
1.4
1.4
1.4
1.4
1.9
.9
1.4










Next factorial points (treatments 1 through 8) were selected at

concentrations below and above the center point (Na .2%, K .3%, C1

.3%). These eight combinations of Na, K and C1 comprised the eight

vertices of Figure 3-1. The figure is drawn in the metric of what are

called coded or design variables. These were defined as: x, = (Na -

.55)/.2, x2 = (K 1.4)/.3, x3 = (C1 .8)/.3. Note that the

concentrations for the factorial treatments were determined by setting

the values of x,, x2 and x3 equal to +1 and -1 and then solving for Na,

K and C1. Also note that the coded value for the center point equaled

0. Once the center point treatment and eight factorial treatments were

selected, axial points (treatments 9 through 14) were added to augment

the factorial design so that a second order model could be fit and

curvilinear effects could be studied. Axial points consisted of extreme

values of Na, K and C1 that were the same distance from the center point

as were the factorial points. The coded values of the axial points were

1.68. By using these coded values, factorial and axial points could

be arranged spherically around the center point making the CCD

rotatable, which is a desirable statistical property (Khuri and Cornell,

1987). When fitting axial points for each mineral (e.g., very low,

.21%, and very high, .89%, concentrations of Na) the other two minerals

were held at their middle (center point) concentration (e.g., 1.4% K and

.8% C1). Exact axial concentrations for all treatments (i.e., very low

Na and very high C1) were not formulated because of the potential

deficiency or toxicity at these concentrations (Table 3-3).

Consequently, responses at these concentrations were determined by

extrapolation.










Ingredients used to alter concentrations of Na, K and C1 in

different treatments included NaC1, KC1, CaC,, Na2CO3, and K2CO3.

Carbonate salts were used in place of bicarbonate salts because they

contributed more Na and K, and less organic anion (i.e., dietary

buffer). When a formulation choice existed, NaC1 and KC1 were chosen

over CaC1, due to the potential toxicity of CaC12 (Tucker et al, 1988b).

Calcium carbonate was used to maintain constant Ca concentration; SiO2

(washed sand) was used to maintain constant ash concentration in all

dietary treatments. Diets were formulated to be equal in all other

nutrients and energy.

The center point (treatment 15) had a total of 24 cow-period

replications, whereas the other 14 treatments had a total of 12 cow-

period replications. With the exception of treatment 15 (which had

three cows remain on that treatment throughout the entire experiment),

every treatment occurred with every other treatment within the same cow-

period sequence at least twice but no more than three times. No

treatment followed another treatment more than once and treatments were

assigned to each cow only once throughout the entire experiment. A

total of 192 cow-period observations were attainable for each dependent

variable.



Sample Collection and Analysis

Intake and milk production data for statistical analyses were

collected during the last 2 wk of each period. Two milk samples from

each cow were collected at each milking (twice daily) during the last 3

d of each period. One sample preserved with K2Cr207 was analyzed for










fat and protein concentration by DHIA Testing Laboratory (Raleigh, NC).

The unpreserved milk sample was frozen (-10 OC) for later mineral

analysis. Milk fat composition from the last six milkings (weighted to

corresponding milk weight) was used to calculate 3.5% FCM yield. Body

weights were recorded on three afternoons immediately prior to the start

of the experiment and after the p.m. milking on d 25, 26 and 27 of each

experimental period.

Corn silage was sampled three times weekly for DM. Amounts of corn

silage fed were adjusted as needed to maintain desired DM proportions of

ration components. Corn silage and batch mixes (which included 5.5%

cottonseed hulls) were sampled, dried for 48 h at 55 OC, ground through

a 2 mm screen and frozen at -10 OC for later analyses. Samples of corn

silage and concentrate were analyzed by commercial laboratory (Northeast

DHIA Forage Testing Lab, Ithaca NY) for DM, CP, ADF, Ca, P, Mg, S, Fe,

Zn, Cu, and Mn. Energy concentration (NEL) was calculated from chemical

analysis. Concentrate and corn silage Na, K and C1 were determined from

analyses of composite samples in the Dairy Nutrition Laboratory at the

University of Florida. Thawed feed samples were dried (100 OC for 24 h)

and ashed (550 oC for 4 h). Ash was dissolved in 3N HC1 solution,

diluted with deionized water and analyzed via atomic absorption

spectrophotometry (Model 5000, Perkin Elmer, Inc., Norwalk, CT). Feed

samples for C1 analysis were dissolved in 25 ml of a .4N HNO3 40%

glacial acetic acid solution, shaken vigorously for 1 h and centrifuged

at 12,000 x g for 10 min. Supernatant was harvested and analyzed for C1

by coulometric titration (Model 4-2500, Haake Buchler Instruments, Inc.,

Saddlebrook, NJ; Cotlove, 1963).








43

On the morning of d 28 of each period, immediately following milking

but prior to feeding, blood samples were collected from the jugular vein

of each cow. A 5 ml sample was collected anaerobically into plastic

syringes coated with ammonium heparin (200 U/ml). This blood sample was

kept on ice and analyzed for blood pH and partial pressure of CO2 (pCO,)

within 2 h of collection (Model 1304 pH/blood gas analyzer,

Instrumentation Labs, Lexington, MA). Blood bicarbonate (HC03) was

calculated from blood pH and pCO2 according to the Henderson-Hasselbach

equation (Kleinman and Lorenz, 1989).

Another 25 ml blood sample was collected from each cow in 14-mi

plastic syringes coated with ammonium heparin (20 U/ml) and decanted

into two 14-mi plastic tubes. One tube of blood was centrifuged (2500 x

g) for 20 min. Plasma was harvested, transferred to 7-ml plastic tubes,

kept on ice for 4 h and then frozen (-10 OC) for later mineral analyses.

The other tube of whole blood was kept on ice for 4 h and then frozen

(-10 OC) for later mineral analyses. Plasma concentrations of Ca, Mg,

K and Na were analyzed after thawed samples were deproteinized with 10%

TCA, vortexed, centrifuged at 2500 x g for 10 min, and diluted with

deionized water. Milk samples were thawed and pooled within cow-period

prior to analysis. Whole blood and milk concentrations of Ca, Mg, K and

Na were analyzed after wet digestion of 1 ml of sample in 3 ml of

concentrated HNO3, heated in 30 ml glass tubes for 20 min, and diluted

with deionized water. Atomic absorption analysis for plasma, whole

blood and milk Ca, Mg, K and Na were analyzed utilizing the same method

as for feeds. Whole blood and milk C1 were determined after acid-zinc

sulfate protein precipitation (somogyi precipitate, Cotlove, 1963).










Plasma C1 concentration was determined by coulometric titration

(Cotlove, 1963).

Original plans were to collect blood and urine from a subgroup of 24

cows, but due to a time delay in urine collection and the need for the

rapid analysis of blood gasses, urine sampling was abandoned in favor of

collecting additional blood samples. Therefore, 24 blood samples were

collected in period one and 48 were collected in periods two through

four.


Statistical Analysis

Least squares ANOVA general linear model (GLM) procedures of SAS

(1985) were used. The analysis was conducted in two stages. In stage

one, sources of variation included effects of cow, period, dietary

treatment and residual error. Least squares means (LSMS) for treatments

were calculated and then used as observations in stage two.

In stage two, treatment effects were partitioned into linear,

quadratic and two-way interaction terms expressed in model form and used

to fit a complete second order response surface. The mathematical model

(or second-order polynomial) fitted was: Y = B0 + Bx, + B2x2 + B3x3 +

B1,x,2 + B22X22 + 33X32 + 12XX2 + 13Bxx B + B23X2X3 + E, where Y =

dependent variable (or observed response variable), B0 = constant (or

intercept); B,, 82, B3 are linear coefficients or parameters for Na, K

and Cl; B,, B22, B33 are quadratic coefficients for Na, K and Cl; B12'

B13, B23 are crossproduct or interaction coefficients; and e = residual
effect. For convenience, the chemical abbreviations (Na, K and C1) were

substituted for x,, x2, and x,, respectively, in reported models.










The LSMS were analyzed by the fitting of these multiple regression

equations using PROC REG (SAS, 1985). Higher order terms that did not

attain a significance level of P = .15 were dropped from the model using

the method of Maximum R2 (SAS, 1985). This resulted in simpler, reduced

model forms in all cases. The maximum R2 method was used previously in

a CCD used to study the effects of dietary Na, K, C1 on growing pigs

(Golz and Crenshaw, 1990). To verify the utility of using LSMS instead

of the individual cow-period observations, parameter estimates (for all

response variables) from reduced models were compared with those that

included all cow-period observations (with cow and period terms plus the

reduced set of continuous linear, quadratic and two-way Na, K and C1

interaction terms). Parameter estimates were the same in both analyses

(LSMS and cow-period models). Because probability values may change

after model reduction, a moderate significance level (P < .15) was

chosen to prevent removal of variables that might contribute to the

predictive power. For several response variables, a significant effect

(P < .15) in the LSMS model was not significant (P > .15) in the cow-

period models. These effects were included in reduced models,

nonetheless. Linear effects of Na, K and C1 variables always were

included in each model (independent of level of probability).

Probability values and standard errors of the coefficient estimates were

generated from fitting the reduced model form in the cow-period models.

Intercept and parameter estimates from LSMS-reduced models were used to

generate response surfaces; R2 values were from LSMS models.

If a reduced model contained a negative quadratic mineral term there

was a concentration of that mineral within the experimental region that








46

maximized the response. The concentration that maximized a response was

determined by setting the first partial derivative of the quadratic

equation equal to zero and then solving. Concentrations of other

minerals in the quadratic equation were set at their mean concentration

for this calculation.

Results



Intake, Milk Yield, Milk Composition, and Body Weight Gain

Figures 3-2 through 3-9 present dry matter intake (DMI), 3.5% FCM

yield, milk composition and body weight change response surface plots

generated from reduced models. Surface plots were helpful to visualize

the simultaneous change in concentrations of two minerals while holding

the third mineral at its mean concentration. Yield of 3.5% FCM response

was essentially equivalent to actual milk yield (MY) response (mean and

SEM = 21.8 and .60; 21.5 and .67, respectively), therefore only the 3.5%

FCM yield is presented. Listed in Table 3-4 are the probability (P)

values (significance levels) associated with the coefficient estimates

retained in the reduced models for all response variables. Only those

effects with P < .1 were considered significant and are listed. When a

two-way interaction was present, linear and curvilinear effects were not

interpreted. When a curvilinear effect was present, linear effects were

not interpreted.

Dry matter intake was affected by interactions between Na and K and

between Na and C1. Increasing dietary Na increased DMI only with low

dietary K (Figure 3-2). Increasing dietary Na increased DMI only with





























S O * O
: 0 0


000 .000 *0


401 o o :oo0 oo M C o


* * *0 **.. -
- : : : ; :






: : : :


: :0: :
. o


C'4


.000 -0 * 0 0






. . . . . . .






0 0 .0 .0 .0 . . .


C.J 10 0 C a -
0 O0 000
V V


O 0 0 c 0,
0 '*-- f^ (0 0 ro a (0 -' .1 0 00 0


0- p- m mmo to c 00
v V





S000 0000 0 0 r -
V V









-' Zh- (3 O n3- OZ) OO O
S .- 0. :X u. w. Li L (.3 u t co Cm c cQ -.J -
0 L, X. m. = ma a t a- CL CL C0. 32 _E 3 3 n E: X: IE x


0 C 0 0 *v v-
X V V


x 10








48

high dietary C1 (Figure 3-3). When dietary Na was increased and not C1

(or vice versa) DMI declined.

Yield of 3.5% FCM increased with increasing dietary Na (independent

of dietary K and C1 concentration; Figure 3-4). Response to dietary K

depended on the concentration of C1 (Figure 3-5). Although linear x

quadratic interaction terms were not included in the initial

mathematical models and thus were not tested statistically, there

appeared to be an interaction between the quadratic Cl and linear K

term. The shape of the quadratic response to C1 (Figures 3-5) was

different for differing concentrations of K in the diet. Maximum 3.5%

FCM yield response to dietary Cl depended upon the accompanying

concentration of K.

Milk fat percentage was affected by a dietary Na x K interaction

(Table 3-4) but this effect was overshadowed by the quadratic effects of

Na, K, and C1. Because the MF regression included all three (Na, K and

C1) quadratic terms whose coefficients were negative, it was conjectured

that a maximum MF value existed within the experimental region. Figures

3-6 and 3-7 seem to support this conjecture. In fact, an estimated

maximum MF value of 3.6% was discovered with .60% Na, 1.34% K and .69%

C1. This concentration of K may not have been optimal because the

quadratic K term had a P value of .28 in the reduced model (Table 3-4).

Had the K2 term been removed, the Na x K effect likely would have been

more influential, which in turn would have affected the optimal Na

concentration.
















DMI (kg/d)


23-' 1.7

2214

21 /

20 I 0,9
,21 .35 .55 75 ,89
Na C%)


Figure 3-2. Response surface for DMI plotted against dietary Na and
K with C1 fixed at .8%. Reduced model with SE for each coefficient in
parentheses: DMI = 19.51 + 3.95 (3.56) Na + 3.39 (1.21) K 3.40 (1.37)
C1 5.28 (2.12) Na x K + 6.43 (2.38) Na x C1. R2 = .55. Mean and SEM
for DMI = 22.5 and .38 kg/d















DMI (kg/d)


25

24

23

22

21


.21


,55
Na Co)I


'3

11
Go\




U)


/,


Figure 3-3. Response surface for DMI plotted against dietary Na and
Cl with K fixed at 1.4%. Reduced model with SE for each coefficient in
parentheses is: DMI = 19.51 + 3.95 (3.56) Na + 3.39 (1.21) K 3.40
(1.37) C1 5.28 (2.12) Na x K + 6.43 (2.38) Na x C1. R = .55. Mean
and SEM for DMI = 22.5 and .38 kg/d.
















3.5% FCM Yield (kg/d)


21 .35 .55


1 3
1 3


8 y

C -


,89


Na %D


Figure 3-4. Response surface for 3.5% FCM (3.5% FCM) yield plotted
against dietary Na and C1 with K fixed at 1.4%. Reduced model with SE
for each coefficient in parentheses: 3.5% FCM yield = 21.41 + 2.19 (.97)
Na 3.02 (2.09) K + 2.36 (5.27) C1 5.68 ( 3.04) Cl2 + 4.09 (2.51) K
x C1. R2 = .64. Mean and SEM for 3.5% FCM = 21.5 and .67 kg/d.
















3.5% FCM Yield (kg/d)


1 3


0o\
.8 U


0.9 1,1 1.4 1,7
K (%o


Figure 3-5. Response surface for 3.5% FCM (3.5% FCM) yield plotted
against dietary K and Cl with Na fixed at .55%. Reduced model with SE
for each coefficient in parentheses: 3.5% FCM yield = 21.41 + 2.19 (.97)
Na 3.02 (2.09) K + 2.36 (5.27) C1 5.68 ( 3.04) C12 + 4.09 (2.51) K
x Cl. R = .64. Mean and SEM for 3.5% FCM = 21.5 and .67 kg/d.















Milk Fat (%)


3.4


3.2


3,0


1 00





0


.21 ,35 ,55 75 ,89
Na (%


Figure 3-6. Response surface for milk fat percentage (MF) plotted
against dietary Na and C1 with K fixed at 1.4%. Reduced model with SE
for each coefficient in parentheses: MF = 1.74 + 2.71 (1.03) Na + 1.04
(.67) K + .97 (.56) C1 1.40 (.80) Na2 .24 (.23) K2 .70 (.37) C12 -
.72 (.37) Na x K. R = .87. Mean and SEM for MF = 3.45 and .07 %.
















Milk Fat (%)


3,6


3.4


3,2


3,0


.21


7L


,35 ,55 .75


Na [%)


Figure 3-7. Response surface for milk fat percentage (MF) plotted
against dietary Na and K with C1 fixed at .8%. Reduced model with SE
for each coefficient in parentheses is: MF = 1.74 + 2.71 (1.03) Na +
1.04 (.67) K + .97 (.56) C1 1.40 (.80) Na2 .24 (.23) K .70 (.37)
C12 .72 (.37) Na x K. R2 .87. Mean and SEM for MF = 3.45 and .07
%.















Milk Protein (%)


0.9 11 ,


1,4 1,7 1.9


Figure 3-8. Response surface for milk protein percentage (MP)
plotted against dietary K and C1 with Na fixed at .55%. Reduced model
with SE for each coefficient in parentheses: MP = 3.56 .02 (.07) Na -
.50 (.15) K .92 (.25) C1 + .65 (.18) K x C1. R2 = .53. Mean and SEM
for MP = 2.83 and .05 %.


2.0

1 ,5

1.0

0,5

0.0


.3

1
















Body Weight Gain (kg/d)


29.'


28


27


26
0,3 0.5


0.8 1,1
SI [o%)


Figure 3-9. Response surface for body
against dietary K and C1 with Na fixed at
for each coefficient in parentheses: BWG =
(.98) K 4.73 (1.67) C1 + 3.43 (1.18) K x
for BWG = .90 and .30 kg/d.


weight gain (BWG) plotted
.55%. Reduced model with SE
4.25 + .63 (.46) Na 2.68
C1. R2 = .38. Mean and SEM


.75 (
55
55 U


35


1.3









Acid-Base Status

Figures 3-10 through 3-12 show the response surface plots for blood,

HC03 pCO2, and base-excess. Values for pH were transformed to H'

concentration [H'] prior to generating LSMS. Least squares means for H'

responses were then transformed back to pH for ease in interpretation,

but pH data were not evaluated statistically (Murphy, 1982).

Although there appeared to be a quadratic effect of C1 on blood

HC03" (Figure 3-10), the coefficient estimates for the C12 term was

nonsignificant (P = .11). There was a significant negative linear

effect of dietary C1 on blood HCO,3 (Table 3-4). Blood pCO2 was

affected by Na x C1 interaction. Blood pCO2 decreased with increasing

concentrations of dietary Na when dietary C1 concentration was low, but

increased with increasing dietary Na when dietary C1 concentration was

high (Figure 3-11). Blood base excess (BE) responded quadratically to

increasing dietary Na and was maximum at .60% Na (Figure 3-12). There

was no effect of dietary Na, K or C1 on blood [H'] or anion gap

(calculated as meq [(Na + K) (C1 + HCO3)]/L in plasma) (P >.1) Mean

and SEM for [H'] and anion gap were 48.81 and 1.53; 5.72 and 3.29,

respectively.


Mineral Metabolism

Figures 3-13 through 3-26 show the response surface plots for

plasma; whole blood; and milk Na, K, C1, Ca, and Mg responses. The only

significant interactions were for plasma Na (Na x C1 and K x C1; Figures

3-13 and 3-14) and whole blood Mg (K x C1; Figure 3-22).















Blood Bicarbonate (meq/L)


1.4 1,7 1.9
K C%D


Figure 3-10. Response surface for blood
against dietary C1 and Na with K fixed at 1.
for each coefficient in parentheses: HCO =
(.43) K 7.63 (3.71) C1 + 4.25 (2.40) C 2.
HC03 = 27.33 and .51 meq/L.


bicarbonate (HCO,3) plotted
.4%. Reduced model with SE
29.33 + .96 (.76) Na + .40
R = .49. Mean and SEM for


3.0

2.9

2.8

2,7

2,6


1.3


1 o\0
8 U


0


0.9 1.1














Blood pCO2 (mm Hg)


60


55


50


45


1. 3

1 00
8 L/


(3


21 .35 .55 .75
Na Co


Figure 3-11. Response surface for blood pCO2 plotted against dietary
Na and Cl with K fixed at 1.4%. Reduced model with SE for each
coefficient in parentheses: pCO2 = 61.83 18.90 (10.49) Na + 1.12
(1.76) K 11.62 (7.15) C1 + 21.15 (12.44) NaCl. R2 = .34. Mean and
SEM for pCO2 = 53.00 and 2.14 mm Hg.
















Blood Base Excess (meq/L)


.21 .35 ,55
Na C%)


Figure 3-12. Response surface for blood base
against dietary Na and C1 with K fixed at 1.4%.
for each coefficient in parentheses: BE = -2.41
(.48) K .96 (.56) C1 10.92 (5.75) Na2. R2 =
BE = .71 and .57 meq/L.


excess (BE) plotted
Reduced model with SE
+ 13.03 (6.38) Na + .25
.43. Mean and SEM for


2.0


-1.0


0.0


-1,0


103


o\










Plasma K (Figure 3-15), plasma Ca (Figure 3-17), and whole blood K
(Figure 3-19) responded quadratically to increasing dietary Na. Whole

blood C1 (Figure 3-20) and milk K (Figure 3-23) responded quadratically

to increasing dietary K. Milk K was maximized with 1.42% K. Plasma C1

(Figure 3-16), whole blood Ca (Figure 3-21), and milk Cl (Figure 3-24)

responded quadratically to increasing dietary C1. Plasma Cl was maximal

at .81% C1.

There were few linear effects of dietary Na, K and Cl (independent

from interaction effects) on mineral metabolism. Whole blood Na (Figure

3-18) decreased as dietary K increased. Milk Ca (Figure 3-25) decreased

with increasing dietary C1 and milk Mg decreased with increasing dietary

Na (Figure 3-26). Plasma Mg and milk Na were not affected by dietary

Na, K or C1 (P > .1). Mean and SEM were 2.12 and .11; 23.17 and 1.18

for plasma Na and milk Mg, respectively.


Discussion


In this study response surface techniques were used to plot and

interpret acid-base status, mineral metabolism and lactational

performance responses to dietary Na, K and C1. These methods were

useful in helping to gain a cognizance of singular and interaction

effects among dietary Na, K and C1.














Plasma Na (meq/L)


130


125


120


115


.3

10

3


0


21 .35 ,55 75 .89
Na C%D


Figure 3-13. Response surface for plasma Na (PNa) plotted against
dietary Na and C1 with K fixed at 1.4%. Reduced model with SE for each
coefficient in parentheses: PNa = 124.89 + 33.41 (15.31) Na 14.62
(9.57) K 7.88 (18.90) C1 37.13 (18.15) Na x C1 + 20.58 (11.43) K x
C1. R = .64. Mean and SEM for PNa = 123.27 and 3.14 meq/L.















Plasma Na (meq/L)


130


125


120


115
0,9 1.1


. 3

1 9
3,

3 L_!


1.4 1,7


Figure 3-14. Response surface for plasma Na (PNa) plotted against
dietary K and C1 with Na fixed at .55%. Reduced model with SE for each
coefficient in parentheses is: PNa = 124.89 + 33.41 (15.31) Na 14.62
(9.57) K 7.88 (18.90) C1 37.13 (18.15) Na x C1 + 20.58 (11.43) K x
C1. R2 = .64. Mean and SEM for PNa = 123.27 and 3.14 meq/L.















Plasma K (meq/L)


21 .35 .55 .75 ,89
Na C%


Figure 3-15. Response surface for plasma K (PK) plotted against
dietary Na and K with C1 fixed at .8%. Reduced model with SE for each
coefficient in parentheses: PK = 6.08 4.22 (1.65) Na + .12 (.12) K +
.07 (.14) C1 + 3.50 (1.49) Na2. R2 = .47. Mean and SEM for PK = 5.15
and .15 meq/L.


5.8

5.6

5.4

5,2


7 0
00
u

















Plasma Cl (meq/L)


.35 .55 .75
Na C%D


Figure 3-16. Response surface for
dietary Na and C1 with K fixed at 1.4%
coefficient in parentheses: PC1 = 91.0!
19.56 (7.80) C1 12.01 (5.04) C12 R2
95.25 and 1.07 meq/L.


plasma C1 (PC1) plotted against
. Reduced model with SE for each
S- 3.53 (1.60) Na .72 (.90) K +
= .50. Mean and SEM for PC1 =


.3


o\0


.21
















Plasma Ca (meq/L)


5,4

5.2

5,0


.3

1


21 .35 .55 .75
Na C%]


Figure 3-17. Response surface for plasma Ca (PCa) plotted against
dietary Na and C1 with K fixed at 1.4%. Reduced model with SE for each
coefficient in parentheses: PCa = 5.65 3.95 (2.21) Na + .09 (.17) K +
.17 (.19) C1 + 3.55 (1.99) Na. R2 = .30. Mean and SEM for PCa = 4.92
and .20 meq/L.















Whole Blood Na (meq/L)


.75
00
.55

35 flr


1.1 1.4 1,7 1.9
K C%]


Figure 3-18. Response surface for whole blood Na (WBNa) plotted
against dietary Na and K with C1 fixed at .8%. Reduced model with SE
for each coefficient in parentheses: WBNa = 91.47 + .71 (2.05) Na 2.58
(1.16) K .81 (1.34) Cl. R = .33. Mean and SEM for WBNa = 87.60 1.38
meq/L.


88

87

86


0,9
















Whole Blood K (meq/L)


.21


70


.35 .55 75
Na E%]


figure 3-19. Response surface for whole blood K (WBK) plotted
against dietary Na and K with C1 fixed at .8%. Reduced model with SE
for each coefficient in parentheses: WBK = 16.03 12.01 (4.82) Na -
4.86 (3.31) K + .17 (.40) C1 + 10.54 (4.36) Na2 + 1.93 (1.18) K. R =
.58.. Mean and SEM for WBK = 10.23 and .42 meq/L.















Whole Blood C1


86

84

82

80


0,9 1.1


8
00


"1----0, 3
1 4 1 7 1.9
K C%o


Figure 3-20. Response surface for whole blood C1 (WBC1) plotted
against dietary K and C1 with Na fixed at .55%. Reduced model with SE
for each coefficient in parentheses: WBC1 = 88.09 .70 (2.22) Na -
16.13 R12.57) K + 13.43 (7.94) C1 + 8.18 (4.21) Na2 8.18 (5.55) K x
C1. R = .53. Mean and SEM for WBC1 = 83.50 and 1.51 meq/L.


(meq/L)















Whole Blood Ca (meq/L)


2.4

2,3

2.2

2.1

2.0


0,9


.3


o\



U>


1.1 1,4 1,7 1.9


K C%D


Figure 3-21. Response surface for whole blood Ca (WBCa) plotted
against dietary K and C1 with Na fixed at .55%. Reduced model with SE
for each coefficient in parentheses: WBCa = 2.00 .07 (.09) Na + .35
(.18) K .41 (.53) C1 + .57 (.27) C12 .34 (.22) K x C1. R2 = .57.
Mean and SEM for WBCa = 2.15 and .06 meq/L.















Whole Blood Mg (meq/L)


1.70

1,65

1. 60

1.55

1,50


. 3

10
^ u


1.1 1.4 1.7 1.9
K C%D


Figure 3-22. Response surface for whole blood Mg (WBMg) plotted
against dietary K and C1 with Na fixed at .55%. Reduced model with SE
for each coefficient in parentheses: WBMg = 1.89 .30 (.54) Na .06
(.17) K .35 (.22) C1 + .55 (.41) Na2 .28 (.21) Na x K + .25 (.15)
K x C1. R2 = .76. Mean and SEM for WBMg = 1.61 and .04 meq/L.








































1.1 1.4 1.7 9
K C%


Figure 3-23. Response surface
dietary K and C1 with Na fixed at
coefficient in parentheses: MLK =
+ 1.50 (.67) C1 4.36 (1.86) K.
33.44 and .67 meq/L.


for milk K (MLK) plotted against
.55%. Reduced model with SE for each
24.16 .38 (1.00) Na + 12.34 (.22) K
R2 = .45. Mean and SEM for MLK =


Milk K (meq/L)


3

1 ^
o\
Li















Milk Cl (meq/L)


36

34 .

32

30 -

28
0,3


.9

7 (
o\


0.5 0.8 1. 1.3


Cl [C%


Figure 3-24. Response surface for milk C1 (MLC1) plotted against
dietary C1 and K with Na fixed at .55%. Reduced model with SE for each
coefficient in parentheses: MLC1 = 38.42 2.50 (1.82) Na .52 (1.04)
K 19.46 (8.72) C1 + 13.84 (5.69) C12 R2 = .50. Mean and SEM for
MLC1 = 30.53 and 1.23 meq/L.
















Milk Ca (meq/L)


0o
/.75

.55

35 <


C [)


Figure 3-25. Response surface for milk Ca (MLCa) plotted against
dietary C1 and Na with K fixed at 1.4%. Reduced model with SE for each
coefficient in parentheses: MLCa = 67.76 + 16.42 (26.57) Na 17.08
(8.51) K 22.17 (10.73) C1 32.05 (19.99) Na2 + 12.52 (10.59) Na x K +
11.96 (7.55) K x C1. R = .85. Mean and SEM for MLCa = 49.59 and 2.04
meq/L.














Milk Mg (meq/L)


.21


3

011
U


.35 .55 .75
Na C%)


Figure 3-26. Response surface for milk Mg (MLMg) plotted against
dietary Na and C1 with K fixed at 1.4%. Reduced model with SE for each
coefficient in parentheses: MLMg = 9.28 1.12 (.42) Na + .39 (.24) K +
.48 (.28) C1. R = .48. Mean and SEM for MLMg = 9.58 and .28 meq/L.


10.5

10 0

9.5


8.5










Determining optimal dietary concentrations of Na, K and Cl was an

objective. However, milk fat was the only response variable that was

maximized by a single concentration of Na, K and C1 within the range of

dietary concentrations used. Most of the responses were influenced by

interrelationships among dietary Na, K and Cl. The importance of

interrelated effects among Na, K and C1 was a key finding in this study.

These will be addressed following a discussion of the linear and

quadratic effects of each mineral.


Linear and Quadratic Effects of Dietary Sodium

Regression models indicated that dietary Na in the range of .21 to

.89% had a positive linear effect on 3.5% FCM yield independent of

dietary concentrations of K and C1. Mallonee et al. (1982) reported an

increase in DMI and MY with increasing dietary Na (from .16 to .70% Na).

Because diets were not isochloridic in that study, responses may not

have been attributable wholly to Na. In a study with equal

concentrations of dietary C1 (Schneider et al., 1986), it was reported

that increasing dietary Na above NRC (1978) recommendations (from .18 to

.55% using either NaC1 or NaHCO3 as the source of Na) improved MY and

FCM yield. In agreement with the findings in the present study, they

found that low concentrations of Na (.18%) limited MY and FCM yield

production.

The positive MF response to increasing dietary Na from .21 to .55%

is in accord with earlier work (Rogers et al., 1982; Escobosa et al.,

1984; Kilmer et al. 1981; and Schneider et al., 1986). Although diets

in this study were not equalized in carbonate content, they were not fat








77
depressive (MF mean 3.45%) which supports the suggestion by Schneider et

al. (1986) that lactational response to dietary Na may be due partly to

Na moiety per se and not wholly to buffering effects. Carbonate salts

were used to manipulate dietary Na and K because they have less

buffering potential (Herod et al., 1978) and contribute more Na and K

than bicarbonate salts.

In contrast to Rogers et al. (1982a) a positive linear effect of

dietary Na on MP was not observed in this study. Rogers et al. (1982a)

reported that cows consuming 2% NaHCO3 produced more total milk protein

than cows fed a basal diet without NaHCO3.

In agreement with findings in the present experiment, Nestor et al.

(1988) found a similar decline in serum K with increasing dietary Na.

In contrast to the current report, Escobosa et al. (1984) and O'connor

et al. (1988) reported higher concentrations of K in blood plasma of

cows fed increasing dietary amounts of Na (from sodium bicarbonate and

sodium chloride, respectively). Erdman et al. (1980) reported that

increasing dietary Na from .31 to .52% with .42% K elevated serum K but

not with .84% K. The lack of a linear effect of dietary Na on plasma K

may have been because diets were not isochloridic in those studies.

However, Schneider et al. (1986) equalized C1 concentrations and saw no

effect of sodium bicarbonate on plasma K. The reason for the negative

linear effect of dietary Na on milk Mg observed in this study is

unknown. O'Connor et al. (1988) saw no effect of dietary Na (.24 versus

.62%) on milk Mg.

Plasma K, Ca, whole blood K and base excess responded quadratically

to dietary Na. Base excess in blood was maximal at .60% dietary Na.










Although blood H+, HC03 and anion gap were not affected directly by

dietary Na concentration (P > .1), single blood samples may not have

been sufficient to detect differences. Also, maintenance of

physiological acid-base status is high on the list of homeostatic

priorities (Kronfeld, 1979). Nonetheless perturbations in acid-base

status and mineral metabolism were evident. Lower BE and greater plasma

C1, plasma Ca, plasma K, and whole blood K at low (.3 and .35%) vs.

middle (.55%) concentration of Na indicated that additional dietary Na

was needed to offset subclinical acidosis. Blood base excess, which

represents the metabolic component of blood H' (Kleinman and Lorenz,

1989), is negatively correlated with metabolic acidosis. Plasma C1 is

positively correlated with metabolic acidosis (Lunn and McGuirk, 1990).

Flux of K into red blood cells and reattachment of Ca to bone and plasma

proteins as H+ stress lessened could explain the decline in plasma K and

Ca. Removal of K and Ca from plasma would be necessary to maintain

electrical neutrality because plasma C1 concentrations linearly

decreased as dietary Na increased to .55%. Urine samples and more blood

samples would be needed to completely understand the physiological

response to dietary Na. However, the direct effect of dietary Na on

3.5% FCM [independent of dietary K and/or C1] coupled with its probable

role in acid-base status indicates the need to reevaluate Na

recommendations for optimal lactational performance of dairy cattle.


Linear and Quadratic Effects of Dietary Potassium

The only linear or curvilinear effects of K (independent of dietary

Na and C1) were on whole blood Ca, whole blood C1 and milk K. Other










responses apparently were masked by Na x K and K x C1 interactions.

These interactions may help resolve discrepancies among published

reports on requirements and optimal allowances of dietary K. Reports on

the requirements or recommendations of dietary K for lactating dairy

cattle have been a subject of controversy. Reports have included .5%

(Ward, 1966), .7% (NRC, 1971), .8% (Dennis et al., 1976; Dennis and

Hemken, 1978; NRC, 1978; Erdman et al., 1980), .9% (NRC, 1989), 1.0%

(Linsner, 1980) and 1.2% Bolenbaugh (1977). Even greater requirements

have been reported for cows in heat stress (Mallonee et al., 1985;

Schneider et al., 1984b; Schneider et al., 1986; West et al., 1987b).

The K x C1 interaction in this study indicates that responses to

dietary K concentration depended upon dietary C1 concentration. It is

probable that some of the discrepancies among published K

recommendations are due to differing dietary C1 concentrations used.

Dietary C1 concentrations were reported infrequently in previous studies

which in retrospect would have been very valuable information.

Increased dietary K led to a linear decrease in whole blood Na which

likely represented a reduction in red blood cell Na considering that

dietary K did not affect plasma Na. Increasing dietary K also elevated

whole blood Ca. Because plasma Ca was not affected by dietary K, the

rise in whole blood Ca probably was due to increased red blood cell Ca.

Increasing dietary K elevated milk K (maximum at 1.4% K) and decreased

milk Ca which suggests that the mammary gland is involved in the

homeostasis of body K.




University of Florida Home Page
© 2004 - 2011 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated May 24, 2011 - Version 3.0.0 - mvs