Reproductive efficiency of swine as influenced by feeding fructose during lactation or glutathione during early gestation


Material Information

Reproductive efficiency of swine as influenced by feeding fructose during lactation or glutathione during early gestation
Physical Description:
xii, 354 leaves : ill. ; 28 cm.
Campbell, Wendy Jo, 1960-
Publication Date:


Subjects / Keywords:
Swine -- Reproduction   ( lcsh )
Swine -- Feeding and feeds   ( lcsh )
Fructose   ( lcsh )
Glutathione   ( lcsh )
Animal Science thesis Ph. D
Dissertations, Academic -- Animal Science -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1988.
Bibliography: leaves 280-353.
Statement of Responsibility:
by Wendy Jo Campbell.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001079358
oclc - 19087371
notis - AFG4290
sobekcm - AA00004794_00001
System ID:

Full Text







The author dedicates this dissertation to all of the swine who so

enthusiastically cooperated during this project while receiving such

little recognition.


The professional guidance, time, and patience of Dr. Fuller W.

Bazer are gratefully acknowledged. The author also extends

appreciation to Dr. W. C. Buhi, Dr. R. M. Shireman, Dr. D. C. Sharp

and Dr. W. W. Thatcher for their advice, encouragement and efforts

throughout this program. For providing the author with a listening

ear and helping hand, sincere gratitude is extended to Dr. J. H.

Brendemuhl who often helped motivate the author. Appreciation is

also extended to Dane Bernis for mixing diets and for providing both

practical and humorous information, and to him and his wife, Lynda,

for their friendship. To other faculty members who were helpful, the

author extends her gratitude for their concern during her education.

The author has been blessed with numerous friends who often

assisted with projects and elevated the quality of her life. For

example, F. and A. Fliss have been a constant source of encouragement

and assistance. Likewise, for the cooperation of her laboratory

companions, C. Ashworth, M. Murray, J. Dore, J. Vallet, K. Young, J.

Harney, D. Dubois, M. Dones-Smith, M. Mirando, T. Ott, L. Smith and

their spouses and(or) friends, the author is grateful. Thanks is

also extended to W. Grubaugh, Dr. K. Bachman, Dr. C. J. Wilcox, Dr.

R. Miller and P. Miles for their technical assistance. Without the

help of L. Joe Padgett and K. Corbett, work at the swine unit would

have been unbearable. The friendship of L. and R. Lawrence, M.


McGuire, A. Williams, B. and M. Schwingel, T. Dawson, K. Bailey, S.

and V. Williams, P. and R. Miles, S. Black, J., D. and B. Yates and

T. and S. TenBroeck will be remembered long after the completion of

this dissertation.

The author extends deep appreciation to J. Weithenauer, for

providing a comfortable home when the author's husband got tired of

her and for unending humor and encouragement. For their constant

love and support, the author will be forever indebted to her family.

Lastly and certainly not the least, the author expresses heartfelt

gratitude to her husband, Donnie Ray, for his blend of affection and

humor without which the author would have had less of a reason to

pursue her goals.



ACKNOWLEDGEMENTS .. ...... . ... .iii

LIST OF TABLES . . .viii

LIST OF FIGURES . . .. . x

ABSTRACT . . . .. xi




[Introduction: Reproductive Efficiency . 4
Control of the Estrous Cycle in Swine . 7
Maternal Recognition of Pregnancy and Factors Associated
with the Establishment of Pregnancy . .. 16
Embryonic/Fetal Mortality . .. 28
Effects of Nutrition or Manipulation of Metabolism of Sows
During Gestation on Embryonic Survival, Litter Size and
Sow and Litter Performance During Lactation. .50
Influences of Nutrition Prior to Breeding or During Early
Gestation on Embryonic Survival. .. .52
Effects of Nutrition During Mid- or Late-Gestation on
Piglet Birth Weight and Survival. . ... 54
VEffects of Manipulating Sow Metabolism During Gestation
on Piglet Survival and Growth . 61
Control and Initiation of Parturition . .63
Physiology and Characteristics of Lactation in Swine. 75
vInfluence of Nutrition on Sow Performance During Lactation. 83
vInfluence of Nutrition on Post Weaning Sow Performance. 92
Lactation . .. 92
Post Weaning. . . 101
vFactors Influencing Piglet Survival . ... 103
Factors Associated with Anestrous During and After
Lactation in Sows . .... .. .107
Metabolism of Glutathione . ... .116
Functions of Glutathione. . .. 126
Previous Research Associating Glutathione With Factors
Affecting Embryonic Growth . ... 142
Metabolic Influences of Dietary Fructose .. 148
Occurence and Utilization . .. 148
General Physiological Roles of Fructose .. 149

Influences of Fructose in Swine in Comparison to
Other Species . . 150
Digestion, Absorption and Transport . .. 157
General Metabolism. . . .. 159
Implications in Diabetes. . ... 164
Glucose Tolerance as Affected by Dietary Fructose 166
Effects on Other Glucoregulatory Hormones .. 173
Effects on Lipogenesis and Blood Lipids .. 175
Effects on Uric and Lactic Acid Production. ... 179
Effects on Mineral Status . .. 180
Effects on Amino Acid Absorption. . ... 181
Effects on Ethanol Metabolism . .. 181
Previous Research Designed to Assess Effects of Fructose
on Sow and Litter Performance . ... .183

SURVIVAL. . . 196
Introduction. . . 196
Materials and Methods . .. 198
Results . . . 201
Discussion . . 201

Introduction. . . .. 207

Materials and Methods . .
Animals, Dietary Treatments and Collection of
PerfnrmanrPe Data

Catheterization and Bloo
Glucose Tolerance Test.

Processing of
Assay Procedur
Fructose. .

Blood Samp
es for Pla
. .

d Collect

les .
sma Sampl
. .

ion Protocol

es .

. .. ., .
. .
. .. .
. .
. .
.o o o. .
. .. .
. .
. .. .
. .
. .
. .
o ..o .

Growth hormone. .
Glucose . .
Nonesterified fatty acids .
Insulin . .
Assay Procedures for Milk Samples
Milk fat. . .
Lactose . .
Protein . .
Statistical Analyses. .
Results . .
Litter Performance. .
Litter size . .
Average piglet weight .
Litter weight . .

Sow Performance . .
Sow weight change and feed consumption during
lactation . . .
Postweaning interval to estrus. .
Milk yield. . . .
Milk composition. . .

. 210

S. 210
S. 214
. 215
. 215
. 216
. 216
S. 216
. 216
S. 217
. 217
S. 218
S. 218
. 218
S. 218
S. 218
. 221
. 221
S. 221
. 221
. 224
. 224

Plasma Constituents. ... . .. .231
Fructose. . . ... ... .231
Glucose . . .. 235
Insulin . . .. .. 237
Nonesterified fatty acids . ... .246
Growth hormone. . . .. 251
Response to glucose infusion. . ... 251
Discussion. . . ... .257


LITERATURE CITED. . . ... .280














. 199

. 202

S. 211

S. 213

. 222

S. 223

. 225

S. 226


OR DEXTROSE BY DAY. . . ... 232


OR DEXTROSE BY DAY. . . ... 238


DEXTROSE BY DAY ....................






OF LACTATION. . . ... .252



Figure Page

4-1 Percentage of sows that returned to estrus ... .229

4-2 Concentrations of fructose in plasma after ingestion of
Dextrose or HFCS . .... .234

4-3 Concentrations of glucose in plasma after ingestion of
Dextrose or HFCS . .... .240

4-4 Concentrations of insulin in plasma after ingestion of
Dextrose or HFCS . .... .244

4-5 Concentrations of nonesterified fatty acids in plasma
after ingestion of Dextrose or HFCS. . ... 249

4-6 Concentrations of glucose in plasma in response to
glucose infusion for each day and treatment. ... 254

4-7 Concentrations of insulin in plasma in response to
glucose infusion for each day and treatment. ... 256

4-8 Glucose clearance for each treatment and day ...... 259

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



Wendy Jo Campbell

August, 1988

Chairman: Fuller W. Bazer
Major Department: Animal Science

Supplemental glutathione (GSH) was fed to gilts on d 2 through 29

in a corn soybean meal-based diet; 0 (n = 5) or 10 (n = 5) mg GSH/kg

body weight (BW) per d. Embryonic survival, uterine, placental and

embryo weights, activities of glutamic-oxalacetic and glutamic

pyruvic transaminase in plasma, number of corpora lutea, number of

live embryos and total embryos and allantoic fluid and amnionic fluid

volumes were not influenced (P > .05) by dietary GSH on d 30.

In a second experiment, 45 multiparous (M) and 36 primiparous (P)

crossbred sows were fed, on a metabolic BW basis, from d 0 though d

28 of lactation, isonitrogenous and isocaloric corn-soybean

meal-based diets containing either 28% high fructose corn syrup or

22% dextrose (D). Plasma of sows fed fructose (F) contained: (1)

lower (P < .05) concentrations of glucose (G) postprandially; (2)

lower (P < .05) insulin preprandially and (3) higher (P < .05) F pre-

and post- prandially. Concentrations of growth hormone (GH) and


nonesterified fatty acids in plasma and response to G challenge were

not altered (P > .05) by treatment. However, postprandially, GH in

plasma of P sows tended (P = .09) to be greater than for M sows.

Milk yield was not altered by treatment. Milk from D sows had higher

(P < .05) concentrations of lactose and milk from F sows had more (P

< .05) fat. Treatment did not affect (P > .05) weight loss of sows

or the postweaning interval to estrus. However, P sows lost more (P

< .05) weight than M sows and produced milk with greater (P = .06)

concentrations of protein. Litter size (LS) was higher (P < .05) for

sows fed F, and F had a greater effect (P < .05) over time on litter

weight than D. However, weights of piglets and litters at weaning

were similar (P > .05) between treatments. Sows fed F have lower

concentrations of G in plasma initially, but conversion of F to G

results in similar performance except for a slight advantage for F on

piglet survival.


Reproductive efficiency in swine is not represented by any one

trait but rather the net sum of animal performance during various

stages of reproductive life. Reproductive efficiency is partially

determined by embryonic survival during gestation and by performance

of sows and their litters during lactation. These two major areas

greatly influence number and weight of piglets weaned for that

interval. However, another relevant factor is reproductive

performance of sows after weaning which influences number of piglets

produced per year. Various factors impact upon reproductive

performance of swine. One of the more important elements required

for production is nutrition; however, its specific interaction with

reproduction is not well defined. Because producers can easily

implement new feeding regimes, it is an area through which knowledge

gained through research can yield rapid benefits. Objectives of this

dissertation were to clarify effects of specific dietary ingredients

on embryonic survival and lactational and rebreeding performance in


In the first experiment, specific objectives were to determine

effects of supplementation of glutathione (GSH) in diets during early

gestation on factors associated with concepts development and

embryonic survival. Because up to one-third of potential piglets are

lost due to embryonic mortality during gestation, factors associated

with survival are relevant. In swine and other species in which



multiple offspring are born, it is intriguing that certain embryos

survive while others perish within the same uterine environment.

This has led to the suggestion that embryos which survive are able to

more effectively compete for nutrients, growth factors or conceptus-

endometrial contact and(or) secretions during early blastocyst

expansion. Glutathione has been associated with numerous

physiological events. Most relevant is its apparent involvement in

in vitro growth of the ovum and stimulation of ovum growth and

uterine proliferation in rabbits. Concentrations of GSH also

increase in pig conceptuses during early growth and differentiation.

Therefore, it was hypothesized that exogenous GSH may provide

additional stimuli for growth in conceptuses and, if so, allow less

advanced embryos to survive. Since dietary ingredients provide the

most practical method of administering exogenous compounds, GSH was

incorporated into diets of gestating swine. Objectives included

evaluating effects of dietary GSH on embryonic survival, weights of

uteri, embryos and placentae, numbers of corpora lutea and live, dead

and total embryos, activities of glutamic-oxaloacetic and glutamic-

pyruvic transaminase and volumes of allantoic and amnionic fluid.

Measurements were obtained at d 30 of gestation, because most

embryonic losses occur prior to this time.

Milk production of sows is one of the primary factors limiting

piglet growth. In addition, primiparous sows often experience

anestrus after weaning which can decrease number of piglets born per

sow per year. Therefore, it is of interest to determine the optimum

and cost-effective dietary composition which will alter metabolism of

sows to allow maximum milk production while not hindering subsequent

reproductive performance after weaning. For numerous reasons,

effects of replacing part of the dietary energy source of sows with

high fructose corn syrup (HFCS) during gestation and(or) lactation

have been inconclusive. However, promising effects of dietary

fructose were reported on milk production of sows during lactation

and weights of litters at weaning, while negative effects on sow

weight change during lactation may be detrimental during the

postweaning interval to estrus. These effects have not been

confirmed and have been controversial. Effects of ingestion of

fructose in numerous species often depends upon other dietary

ingredients, metabolic status of animals prior to experimentation and

duration of fructose consumption. In some studies fructose resulted

in increased concentrations of insulin and glucose in plasma, while

opposite effects have been reported following other experiments.

Therefore, objectives of one aspect of this study were to further

evaluate effects of replacing part of the dietary energy source of

lactating sows with high fructose corn syrup and dextrose.

Production traits of interest were: (1) litter size and litter weight

gain throughout a 4 wk lactation; (2) sow milk production, weight

loss and interval to estrus after weaning; (3) plasma metabolic

indices, including concentrations of glucose, fructose, growth

hormone, insulin and nonesterified fatty acids in plasma pre- and

post-prandially and (4) response to glucose challenge.


Introduction: Reproductive Efficiency

Factors affecting reproduction in pigs and management

considerations to enhance their reproductive efficiency have been

reviewed (Johnson, 1984; Terqui and Legault, 1984; Dziuk and Bellows,

1983; Yen et al., 1987). As Johnson (1984) indicated: "reproductive

efficiency is a function of age at puberty, fertility (the percentage

of breeding herd females that conceive), fecundity (the number of

offspring per pregnancy) and offspring survival to weaning." Terqui

and Legault (1984) stated that the number of offspring weaned per dam

is one of the best estimate of reproductive efficiency in all species

and that this has been called "numerical productivity." In swine,

numerical productivity (Pn) of sows can be expressed as [0'(1 -

em)(l pm)/I] x 365, where "0" is ovulation, rate, "em" is embryonic

mortality rate from ovulation to parturition, "pm" is mortality rate

from birth to weaning and "I" is the interval between successive

parturitions in days. Terqui and Legault (1984) indicated that "I"

represents the interval from weaning to fertilization in lactating

sows. They suggested that numerical productivity should also include

age at first parturition (ml) and the interval between last

parturition and culling of the female (m2). Therefore, their

estimate of productivity would be: P'n [N x O'(1 em)(l -

pm)]/[mI + (N 1) x I + m2] x 365, where "N" represents the

average number of reproductive cycles at culling. Although

reproductive efficiency can be defined in numerous ways, e.g., "Sow

Productivity Index" (Yen et al., 1987), all methods estimate number

of pigs produced within the sow's reproductive life. Goals of

producers include maximizing piglets per litter and litters per sow

per year. Factors which influence reproductive efficiency include

genetics, environmental conditions, e.g., photoperiod, temperature,

social interactions, and nutrition.

Age at puberty has moderate heritability (.30) in swine

(Cunningham et al., 1974) and is influenced most by environmental

factors. Tess et al. (1983) determined that the most important

economic consideration in reproductive efficiency of swine is number

of live pigs per litter at birth. This can be influenced by

ovulation and fertility rates, embryonic mortality and losses at

parturition. Although ovulation rate can be increased by selection

(Cunnigham et al., 1979) or administrations of exogenous hormones

(Webel and Day, 1982), it is balanced by decreased embryonic survival-

(Cunningham et al., 1979). Fertility is not a major problem in swine

with the exception of summer or heat-induced infertility (Wetteman

and Bazer, 1985), which can be overcome to a large degree with proper

management. Likewise, parturition is not a major complication in

swine, but can be successfully induced (First and Lohse, 1984) to

increase management efficiency; however, it has limited usefulness

for decreasing the interval between farrowings. Genetic selection

for reproductive efficiency has limited potential, due to low

heritabilities for reproductive traits (Johnson, 1984; Terqui and

Legault, 1984) and other production traits are important in meat

producing animals. Nutritional factors may be exploited to maximize

reproductive efficiency. However, optimum nutritional regimens for

the various stages of sows reproductive cycles have not been

determined. Supplementation of certain nutrients, e.g., riboflavin,

during gestation may increase embryonic/fetal survival to term (F. W.

Bazer and M. T. Zavy, personal communication). Manipulation of the

metabolic status of sows during gestation by dietary or

pharmacological agents which induce a diabetic state may increase

fetal energy reserves and enhance survival at birth (Britt, 1986).

Nothing is more vital to neonatal piglets than availability of

immunoglobulins and energy from colostrum at birth. Therefore,

nutrition during gestation and(or) lactation has a significant

influence on piglet survival and growth by influencing milk

production. Similarly, nutritional status influences weight loss and

body condition of sows and, therefore, their ability to rebreed after

weaning which, in turn, affects the interval between parturition.

Scientists are continually searching for factors which will

enhance reproductive efficiency. These approaches enhance our

understanding of physiological mechanisms associated with

reproduction and suggest management protocols to enhance reproductive

efficiency. Nutrition is definitely not a solitary factor, but as

innovative goals are realized, larger litters will be produced and

nutritional requirements will need to be modified to complement

advances in other disciplines that maximize reproductive efficiency

in swine. This review summarizes our knowledge of factors affecting

reproductive efficiency in swine.

Control of the Estrous Cycle in Swine

Estrous cyclicity in all mammalian species is dependent upon

proper interactions between the components of the hypothalamic-

pituitary-gonadal axis and the uterus (Anderson, 1974b). Appropriate

morphological modifications and altered secretary functions of these

reproductive tissues allow the sexually mature female pig to display

polyestrous activity (Anderson, 1974b) under domestic conditions and

to complete one estrous cycle approximately every 21 3 days

(Anderson, 1974b; Sorenson, 1979). Although the pig is not a

seasonal breeder, seasonal variation does exist (Ledwitz-Rigby and

Rigby, 1987; Claus and Weiler, 1985). Numerous reviews concerning

reproductive physiology of swine have been published (Anderson,

1974b; Hansel and Convey, 1983; Foxcroft and Van de Wiel, 1982).

This review is not intended to describe mechanisms associated with

follicular growth, intrafollicular communication, selection of

ovulatory follicles, ovulation and other specific events. Rather,

this review is to provide a general overview of the reproductive

biology of female pigs.

The 21-day-estrous cycle in swine is commonly divided into four

phases: estrus, metestrus, diestrus and proestrus (Anderson, 1974b).

The female is sexually receptive to a boar for approximately 24 to 66

h during estrus and the first day of "estrus" is referred to as day

(d) 0 of the estrous cycle. At estrus, concentrations of estrogen,

follicle stimulating hormone (FSH) and progesterone in plasma are low

(Anderson, 1974b). Approximately 38 to 42 h after onset of estrus,

multiple Graffian follicles, selected by a yet undefined mechanism,

ovulate. Ovulation itself results from a complex physiological

cascade of events (Murdoch, 1985) initiated primarily by the

ovulatory surge of luteinizing hormone (LH). The LH surge is induced

by the positive feedback of estrogen produced from growing follicles

on the hypothalamic-pituitary axis. Concentrations of LH in plasma

decline rapidly following the LH surge. Metestrus begins as the

female ceases to be sexually receptive and is characterized by low

levels of estrogen, FSH and LH, but increasing concentrations of

progesterone in plasma produced by developing corpora lutea (CL).

During diestrus (d 3 to 4 though d 15 to 16), progesterone secretion

is maximal while concentrations of LH, FSH and estrogen in plasma

remain low. In nonpregant pigs, the CL begin to regress between d 15

to 16 due to prostaglandin F 2-alpha (PGF) secretion by the uterus.

Therefore, progesterone secretion by CL declines and there is

initiation of the proestrous phase. Simultaneously, preovulatory

follicles begin to develop and produce estrogen so that

concentrations in plasma peak around d 18. The elevated estrogen

levels decline to low levels before the subsequent estrous period

begins, but are elevated sufficiently to initiate the ovulatory surge

of LH and initiate a new estrous cycle. Some of these events will

now be discussed in more detail. The luteal phase diestruss),

terminated by the mechanism of luteolysis, the early follicular

phase, late follicular phase and pre-ovulatory period (proestrus and

estrus) and the periovulatory period (including the early luteal

phase (metestrus) will be discussed in this order as suggested by

Foxcroft and Van de Wiel (1982).

Maintenance of luteal function and thus, continued progesterone

production (Foxcroft and Van de Wiel, 1982), blocks the final stages

of follicular growth and steroidogenesis and controls the length of

the estrous cycle. A rise in progesterone generally occurs on d 3 to

4 of the estrous cycle (Van de Wiel et al., 1981). Although the

pattern of secretion may be altered by season (Perotti et al., 1979),

progesterone secretion, as measured in the ovarian vein (Gomes et

al., 1965), is maximal on d 10 to 12, declines slowly to d 13 to 15

and then falls rapidly. However, Masuda et al. (1967) reported

maximal progesterone production on d 8 in the ovarian vein. It is

possible that pigs can sequester progesterone in body fat and return

progesterone into the circulation after luteolysis resulting in a

latent fall in progesterone, and thus, longer effects of progesterone

on the hypothalamic-pituitary axis (Foxcroft and Van de Wiel, 1982).

Whether there is a physiological requirement for a luteotrophin other

than LH in swine is not clear. Prolactin has been demonstrated to

stimulate progesterone secretion by luteinized pig granulosa cells

(Veldhuis et al., 1980) and isolated luteal cells in vitro

(Gregoraszczuk, 1983; Grinwich et al., 1983). Recently, Bramley and

Menzies (1987) characterized receptor numbers for prolactin in the

porcine corpus luteum. Prolactin receptors were low during the

periovulatory and early luteal period, increased in the mid-luteal

phase and declined in the late luteal phase. In pregnant swine, as

gestation advanced, prolactin receptor concentrations in CL

increased. It is known that exogenous estrogens elevate plasma

progesterone levels, prolong CL lifespan (Ford et al., 1982) and

stimulate progesterone production by porcine granulosa cells in vitro

(Goldenburg et al., 1972). However, whether estrogen and prolactin


play necessary and direct luteotrophic roles endogenously is not


Luteolysis in pigs is strongly associated with production of PGF

by the uterus (see Bazer et al., 1982). Moeljono et al. (1976, 1977)

proposed that PGF is the luteolysin in swine and reported increasing

PGF levels in utero-ovarian vein plasma coincident with luteal

regression (Gleeson et al., 1974; Moeljono et al., 1976; Ford et al.,

1982). Exogenous PGF was luteolytic in gilts in which CL were

maintained beyond their normal life span by estradiol treatment

(Kraeling et al., 1975) or hysterectomy (Moeljono et al., 1976).

However, PGF can only exert a luteolytic effect after d 12 (Diehl and

Day, 1974; Moeljono et al., 1976). To reach the ovary,

prostaglandins produced by the uterus can be transferred to the

ovarian artery and possibly into the lymphatic circulation (Kotwica,

1980). Since exogenous gonadotropins cannot maintain CL of the cycle

in the presence of an intact uterus (Anderson, 1966), it is suggested

that PGF inactivates some component of the LH stimulatory mechanism

that is effective in maintaining luteal function in hysterectomized

females. Thus, PGF overrides the luteotrophic effect of LH by

blocking the LH-adenylate cyclase pathway. It appears that a loss of

LH-receptor mediated mechanisms within luteal tissue results in the

decline in progesterone synthesis (Foxcroft and Van de Wiel, 1982).

Evidence of coincident increases in prolactin and decreases in

progesterone (Van de Wiel et al., 1981) and indications that

prolactin inhibits progesterone secretion by porcine granulosa cells

(Rolland et al., 1976), have led to speculation that prolactin plays

a role in luteolysis. However, the initial fall in progesterone and


increases in estradiol and prolactin occur simultaneously, so it is

difficult to establish exact cause and effect relationships.

However, the majority of literature indicates that PGF is the

luteolysin which, in the absence of a concepts, overrides

luteotrophic effects of LH and(or) prolactin (see Bazer et al., 1982;

Moeljono et al., 1976).

Resumption of follicular growth and production of estrogens are

initiated following luteolysis. The factor responsible for

initiating onset of increased estrogen production by follicles during

the late luteal/early follicular phase of the cycle is not known.

Whether progesterone blocks ovarian follicular development directly

(thus, removal of progesterone may involve removal of an inhibitory

influence on intra-ovarian factors necessary for initiation of late

follicular growth) or by acting on the hypothalamic-pituitary axis to

inhibit an appropriate pattern of LH:FSH production is not clear.

Whatever the exact mechanism, as the luteal phase ends, certain large

follicles which had previously begun development have increased: (1)

granulosa LH receptor numbers (100 X); (2) LH-stimulated cAMP

production; (3) ability of granulosa cells to secrete progesterone

and convert androgen to estrogens and (4) responsiveness to FSH to

stimulate androgen aromatization in granulosa cells (Anderson et al.,

1979; Leung et al., 1979; Schwartz-Kripner and Channing, 1979).

These changes may be similar to those proposed for rats (Armstrong

and Dorrington, 1977; Richards et al., 1978). Foxcroft and Van de

Wiel (1982) suggest that these follicular responses involve estrogen-

dependent changes in receptors for LH and FSH in granulosa cells and

induction of aromatase enzymes intracellularly. This would allow for

LH-induced increases in androgen production by theca cells followed

by increases in aromatization of androgens to estrogens by granulosa

cells. It is important to note that, unlike sheep and cows, pig

granulosa cells produce estradiol in the absence of detectable

androgens. This is possibly due to stored androgens of thecal origin

in the granulosa cells (Evans et al., 1981). Additionally, porcine

thecal cells, as well as granulosa cells, can aromatize androgens to

estradiol. Although the exact mechanism is not clear, Graffian

follicles develop and estrogen production increases in the early

follicular phase. As a result, during the late follicular phase and

pre-ovulatory period, concentrations of circulating estrogens are

elevated between d 18 to 20 (Van de Wiel et al., 1981). Estrogens

exert a positive feedback on the hypothalamus in pigs (Edwards, 1980)

and, in vivo, the pre-ovulatory surge of LH and FSH occurs

approximately 50 to 55 h after the rise in estrogens. However, the

rise in estradiol initially results in suppression of episodic LH

release and, later, inhibition of FSH production which results in low

gonadotrophin levels immediately prior to their surge (Vandalem et

al., 1979; Foxcroft, 1978; Van de Wiel et al., 1981). The rise in

plasma LH and its characteristic preovulatory surge 40 to 48 h prior

to ovulation, with a duration of approximately 20 h concident with

the onset of estrus, has been confirmed by numerous researchers

(Liptrap and Raeside, 1966; Niswender et al., 1970; Henricks et al.,

1972; Rayford et al., 1971; Parvizi et al., 1976; Vandalem et al.,

1979; Edwards, 1980; Van de Wiel et al., 1981). However, the

relationship between the onset of behavioral estrus and timing of the

LH surge is variable (Foxcroft et al., unpublished data in Foxcroft

and Van de Wiel, 1982). Van de Wiel et al. (1981) reported maximum

LH levels of 6 ng/ml during the LH surge versus 3 ng/ml associated

with tonic secretion of LH during the luteal phase. Thus, the LH

surge of the pig is not characterized by a discrete LH surge, but one

of rather long duration (Foxcroft and Van de Wiel, 1982). The

response of FSH to estradiol positive feedback is variable. However,

it appears that estrogen directly induces a surge of FSH release

(Elsaesser and Foxcroft, 1978; Edwards, 1980) concident with the LH

surge, but of lower magnitude. As a result of the LH surge,

luteinization of the theca and granulosa cells occurs, estradiol

production declines rapidly, progesterone synthesis gradually

increases and ovulation occurs (Catt et al., 1979; Hunzicker-Dunn et

al., 1979). The requirement of the FSH surge for these phenomena is

questionable (Edwards, 1980), since periovulatory events have been

observed in early weaned sows which did not experience an FSH surge.

Ovulation generally occurs near the end of d 2 of the estrous

cycle when LH levels are low and FSH levels are high (Rayford et al.,

1974; Vandalem et al., 1979; Edwards, 1980; Van de Wiel et al.,

1981). Whether this rise in FSH is due to removal of inhibin at the

time of ovulation (Channing, 1979) or declining estradiol (Edwards,

1980) is not certain. Foxcroft and Van de Wiel (1982) suggest that

inhibin would specifically regulate FSH secretion. Furthermore,

whether or not the increase in FSH is important for recruitment of

follicles destined to ovulate during the subsequent estrous period

has not been resolved for pigs. Increasing concentrations of

progesterone in the postovulatory period are associated with a

decline in FSH (Vandalem et al., 1979; Edwards, 1980, Van de Wiel et

al., 1981), and simultaneously, development of episodic LH release

with increasing amplitude and decreasing frequency. Wilfinger (1974)

and Van de Wiel et al. (1981) also reported increased prolactin

secretion at this time with no known role, but with a possible

function in expression of behavioral estrus. Although numerous

patterns of hormonal secretion have been characterized, the exact

mechanisms) responsible for follicular recruitment have not been


Attempts to control length of the luteal or follicular phase of

the cycle pharmacologically to facilitate management of labor and

increase reproductive efficiency have led to inconsistent results

(for review, see Webel and Day, 1982; Day, 1984). Currently, the

most promising approach towards effective regulation of the estrous

cycle in swine involves the use of synthetic steroids with progesta-

tional activity, e.g., altrenogest. Altrenogest has been effective

in synchronizing estrus to a 4-day period in approximately 95% of

gilts which have already exhibited estrus by administering altreno-

gest (15 to 20 mg/day) orally for 18 consecutive days (Redmer and

Day, 1981). Other regimens and doses have been tested (see Webel,

1978; Webel and Day, 1982; Day, 1984), and treatment of young gilts

with altrenogest appears to be a potentially useful management tool.

Genetic selection, pharmacological agents and elevating the

nutritional plane of energy (flushing) (Anderson and Melampy, 1972;

Den Hartog and Van Kempen, 1980; Dziuk and Bellows, 1983) prior to

ovulation have all been utilized to increase ovulation rates in

gilts. It is firmly established that increased feed or energy intake

prior to mating will increase ovulation rate, and the optimum intake

period has been reported to be in excess of 11 d (Den Hartog and Van

Kempen, 1980) or as little as 8 to 10 d (Anderson and Melampy,

1972). However, whether flushing increases ovulation rate beyond

normal expectations, or reverses nutritional inhibition due to

previous low levels of feeding is not certain (Anderson and Melampy,

1972). For a review of management recommendations (genetic,

nutritional and age considerations) of gilts aimed at improving

ovulation rates refer to Christenson (1986) or Aherne and Kirkwood


The mechanism whereby increased dietary energy elevates ovulation

rate is not known. However, neither increased dietary energy or

exogenous insulin is necessarily accompanied by changes in

gonadotropins or estradiol (Cox et al., 1987). Cox et al. (1987)

suggest the increased ovulation rate induced by diet and insulin may

be due to recruitment of more follicles or lessening of atresia, thus

providing more preovulatory follicles. Direct effects of insulin on

the ovaries have been reported, including promotion of progesterone

production by granulosa (May and Schomberg, 1981) and luteal cells

(Ladenheim et al., 1984) in culture, and androstenedione production

by thecal cells (Barbieri et al., 1983). In addition, insulin and

insulin receptors have been identified in the vicinity of

gonadotropin releasing hormone-producing neurons (Havrankova et al.,

1978) in the hypothalamus. The effects of high energy and insulin on

LH secretion were not clear in the study of Cox et al. (1987). In

experiment 1, both dietary energy and insulin were positively related

with number of LH pulses and ovulation rate. However, in experiment

2 no treatment differences were detected. In a separate report by

Flowers et al. (1986), FSH and LH were positively influenced by

flushing between 2 and 5 d prior to estrus.

Thus, the effects of flushing and of insulin are not well

defined, but any or all components of the hypothalamo-hypophyseal-

ovarian axis may be affected by insulin (Cox et al., 1987). Since

increased ovulation rate is accompanied by decreased embryonic/fetal

survival (Johnson et al., 1985), it may be more practical to select

for larger litters and attempt to reduce prenatal and postnatal

losses in pigs rather than attempt to increase ovulation rates.

Maternal Recognition of Pregnancy
and Factors Associated with the Establishment of Pregnancy

Under normal conditions, the luteolysin (presumed to be PGF; see

Bazer et al., 1982) in the pig causes regression of the CL, beginning

around d 15 in nonpregnant pigs. This allows for recurrent estrous

cycles. However, should the female become pregnant, the developing

conceptuses must signal their existence to the uterine endometrium in

order to maintain functional CL and endometrial development and

secretary activity which will permit a gestation period of 114 to 115

d to follow. As reviewed by Bazer et al. (1982), the concepts is

believed to initiate luteostatic mechanisms via the production of

estrogens (also see Flint et al., 1979a). The exocrine-endocrine

theory (Bazer and Thatcher, 1977) suggests that under the influence

of estrogen produced by the conceptuses, PGF and other components of

endometrial origin, including necessary histotrophic factors, are

directed into the uterine lumen exocrinee direction) (Marengo et al.,

1986). This prevents release of the luteolysin towards the uterine

vasculature (endocrine direction). Thus, establishment and

maintenance of pregnancy are initiated.

Probably the single most important requirement for establishment

of pregnancy is maintenance of CL. Loss of CL function at any stage

of gestation leads to abortion within 24 to 36 h (Belt et al., 1971)

since CL are required for progesterone production. Concentrations of

progesterone in maternal plasma during pregnancy reach 30 to 40 ng/ml

on d 12 to 14, decline to 10 to 25 ng/ml by d 25 (Guthrie et al.,

1974; Robertson and King, 1974; Knight et al., 1977), and remain

fairly constant until d 100. Then, progesterone levels decline

gradually prior to parturition at which time they decrease abruptly

to <1 ng/ml. However, DeHoff et al. (1986) reported that concentra-

tions of progesterone peak at d 45 (35 ng/ml), decline at d 60 and

then remained farily constant until a rapid periparturient decrease.

The CL of pigs appear to be autonomous until d 14, and then

require only basal LH support until d 40 to 50. After that time, a

possible role for prolactin in CL maintenance has been suggested (see

Heap et al., 1973). However, based on later studies (Whitacre and

Threlfall, 1981; R. R. Kraeling, unpublished data, cited by Bazer et

al., 1982), it appears that prolactin may not be necessary for CL

maintenance after d 70 of gestation. Recently, Bramley and Menzies

(1987) suggested that prolactin may be luteotrophic during pregnancy

as indicated by increasing numbers of prolactin receptors as gesta-

tion advances. Whatever the luteotropic agents may be, it is impera-

tive that they function properly to ensure successful pregnancy.

Following fertilization, embryos move through the oviducts and

into the uterine horns at about the 4-cell stage, approximately 48 h

after ovulation (Hunter, 1974). Transport through the oviduct is

entirely a maternal function involving flow of secretions, movement

of cilia, peristaltic contractions, and the actions of gonadal

steroids and prostaglandins (Hunter, 1974). Pig embryos fail to

develop beyond the early blastocyst stage if confined to the oviduct

(Murray et al., 1971; Pope and Day, 1972). Therefore, it is a

requirement that the embryos reach the uterine environment.

Having passed through the oviduct, porcine embryos temporarily

are free to move throughout the uterus (Dziuk et al., 1964), but are

actively undergoing morphological changes. Niemann and Elsaesser

(1986) suggested that estradiol-17B is necessary for

morula-blastocyst transformation in vitro and that the effect is

mediated through specific binding sites. Niemann and Elsaesser

(1986) tested in vitro effects of withdrawing estradiol on

development of porcine morula using an antiestrogen or an antiserum

to estradiol. These compounds reduced rate of blastocyst formation,

and this reduction could be partially reversed by addition of

estradiol-17B to the culture medium. In a subsequent study, Niemann

and Elsaesser (1987) confirmed that estradiol-17B is a potentially

important factor in morula-blastocyst transformation in vitro. The

effects of estrone on transformation were less evident. Whether in

vitro effects of estradiol on transformation is physiologically

relevant remains to be determined.

Embryos reach the blastocyst stage by d 5 and hatch from the zona

pellucida between d 6 and 7. After hatching, blastocysts expand from

0.5 to 1 mm diameter to 2 to 6 mm diameter on d 10. Then, between d

10 to 12, blastocysts undergo the most rapid elongation phase,

changing from spherical (3 to 10 mm diameter), to tubular (10 to 50

mm long) and filamentous (>100 mm long) forms, resulting in lengths

of 700 to 1000 mm by d 16 of pregnancy (Perry and Rowlands, 1962;

Anderson, 1978). Geisert et al. (1982a) reported that initial

elongation (from 9-10 mm to 100-200 mm) is due to remodeling of

existing trophectoderm. The final elongation (100-200 mm to

800-1,000 mm; d 14 to 16) involves cellular hyperplasia. Albertini

et al. (1987) have begun to investigate the role of actin

cytoskeleton reorganization during premiplantation development of pig

embryos from d 1 through d 8 of gestation. Their experiment was

based on the premise that actin filaments may be involved in

contractile events during compaction (Johnson and Maro, 1984) and

differentiation of trophectoderm (Lehtonen and Bradley, 1980; Reima

and Lehtonen, 1985). Although inconclusive, their results suggest

that actin is important in mechanisms responsible for blastocyst

elongation. Failure of blastocyst elongation can lead to embryonic

death early in gestation or submaximal placental development (Perry

and Rowlands, 1962; Knight et al., 1977).

Although, pig embryos are initially freely mobile, elongated

conceptuses cannot undergo further spacing so they lie end-to-end

along the length of the uterine horns in preparation for

implantation. Spacing of spherical conceptuses is a necessary

requirement since they must compete for a limited amount of

endometrial surface area to obtain support from the endometrial

secretions during placentation. Scheerboom et al. (1987) measured

uterine motility in sows during the estrous cycle and early

pregnancy. They suggested that increased frequency of myoelectrical

complexes (episodes of activity and subsequent interval of

inactivity) may be involved in the process of intra-uterine migration

of conceptuses. The interval between myoelectrical complexes

decreased on d 12 of gestation. They also reported that myometrial

activity of low magnitude and high frequency was present during d 8

to 9 of gestation. Pope et al. (1982b) had previously indicated that

contractions of the myometrium are involved in the process of embryo

spacing. Pope et al. (1982b) suspended uterine strips in baths,

during the period of intrauterine migration (d 12 of gestation), and

showed increased myometrial activity in comparison to uterine strips

excised on d 6 to 9 of gestation.

In a subsequent study, Pope et al. (1986a) attempted to examine

the process of intrauterine migration of porcine embryos by studying

a model involving migration of estradiol- and cholesterol-releasing

beads in utero. The presence of estradiol-releasing beads resulted

in activity that induced adjacent beads (estradiol and cholesterol

beads) within the uterus to migrate posteriorly from the uterine

tips. However, cholesterol releasing-beads failed to migrate as far

as estradiol-impregnated beads. The possibility that adjacent

embryos migrate as a result of activated myometrial function was

examined in a second experiment. The results were not clear, but

estradiol release locally tended to induce a random distribution of

cholesterol- impregnated beads. In a third experiment (Pope et al.,

1986a), however, estradiol tended to induce anterior migration when

cholesterol-containing beads were placed at the base of the uterine

horn. Pope et al. (1986a) could not conclude whether the uterus was

responding differentially depending on the site of estrogenic action

or whether the anterior, posterior migrations were due to random

intermixing. None-the-less, it appears that estrogen is involved in

embryo migration and the waves of uterine contractions. However,

Scheerboom et al. (1987) suggested the myometrial activity is

dependent on ovarian hormones and independent of the blastocyst

because of similar responses in cyclic and pregnant gilts.

Pig conceptuses are known to produce estrogens while undergoing

elongation (Flint et al., 1979a). In addition, Fischer et al. (1985)

demonstrated that estradiol production is initiated by large (7 mm)

spherical conceptuses and then estrone and estradiol are produced by

tubular conceptuses in amounts greater than are produced by d 12

filamentous conceptuses. Fischer et al. (1985) also demonstrated

that [3H]-progesterone could be converted to estrogens. Their

results agree with data obtained from analyses of uterine flushings

and uteroovarian vein plasma for estrogens. In addition, Fischer et

al. (1985) reported that both the concepts and pregnant endometrium

converted progesterone to estrogens. The initial increase in

estrogen content in uterine flushings (Geisert et al., 1982b) occurs

when conceptuses reach approximately 10 mm in diameter (late

spherical, d 11.5), which coincides with the time of maternal

recognition of pregnancy (Dhindsa and Dziuk, 1968; Bazer and

Thatcher, 1977; Flint et al., 1979a) and just precedes initiation of

trophectoderm elongation. Based on measurements of estrone sulfate

in maternal plasma, estrogen production by pig conceptuses is

triphasic with the first peak occurring between d 10 to 12 (Stoner et

al., 1981). Maternal estrone sulfate increases from d 16 (60 pg/ml)

to d 30 (3 ng/ml) (Robertson and King, 1974; Stoner et al., 1981),

decreases to d 46 (35 pg/ml), increases slightly to d 60, and then

increases steadily to parturition (3 ng/ml) (Robertson and King,

1974). Knight et al. (1977) reported the same pattern for

unconjugated estrone and estradiol in peripheral and uterine vein

plasma and amniotic and allantoic fluids. However, Kensinger et al.

(1986) reported higher estradiol values than Knight et al. (1977) and

concentrations similar to Robertson and King (1974). Robertson et

al. (1985) measured estrogens in fetal fluids and maternal blood and

urine and confirmed previous estimates. DeHoff et al. (1986) also

reported patterns of estradiol, estrone and their sulfated forms

similar to those described above. In addition, Robertson et al.

(1985) suggested that the second rise in estrogen concentrations (d

30) reflects estrogen synthesis by the fetus rather than the

trophoblast. Kensinger et al. (1986) and Knight et al. (1977)

indicated that conjugated estrogens at d 30 and during late pregnancy

are primarily of feto-placental origin. Furthermore, during the last

third of gestation, maternal concentrations of estrogen sulfate and

estrone are directly affected by the number of conceptuses present.

According to the exocrine-endocrine theory (Bazer and Thatcher,

1977), failure of conceptuses to produce adequate estrogens can

result in regression of CL and loss of pregnancy, since PGF will be

allowed to be secreted in an endocrine direction. That estrogens

themselves and not prostaglandin E1 or E2 are responsible for

movement of PGF into the uterine lumen was demonstrated by Marengo et

al. (1986).

While, conceptuses are undergoing extensive remodelling and

initiating steroid synthesis, the endometrium is actively preparing

for gestation. Similar to conceptuses, the uterine horns and

endometrium elongate (Perry and Rowlands, 1962) from 190 cm on d 3 to

a length of 360 cm at d 13 to 18. Some aspects of early embryonic

development have been reviewed by Allen (1984) and Bazer and First

(1983). On approximately d 13 of gestation, porcine conceptuses

begin physical attachment, seen histologically as discrete areas of

cell-to-cell contact, which is maintained by the adhesive properties

of endometrial gland exocrine secretions (Crombie, 1970).

Although the mechanisms) involved in attachment of blastocysts

to the uterine epithelium is not well understood, Morgan et al.

(1987a,b) suggested that estrogen is involved. Morgan et al. (1987a)

speculated that ultrastructural and biochemical changes occur in the

plasmalemma of the uterine surface epithelium in response to estrogen

for blastocyst attachment. Schlafke and Enders (1975) previously

suggested that attachment is likely mediated by such changes in the

plasma membrane surface of the trophoblast and uterine specific

epithelium. Alteration of protein and polysaccharide composition

occurs on the apical surface of the rabbit endometrium at the time of

implantation (Anderson et al., 1986). Alterations of glycoprotein

coatings on both the trophoblast and epithelial cell plasma membrane

have been suggested by other researchers to be responsible for close

apposition necessary for interdigitation of microvilli (Enders and

Schlafke, 1974; Jenkinson and Searle, 1977).

Keys et al. (1986) utilized a technique involving systemic

injection of Evans blue and localization of endometrial fluorescence

to determine extravascular content of the dye, and thus, endometrial

vascular permeability in the pig. Only in pregnant gilts, at d 12 of

gestation, was a well-defined area of endometrial fluorescence

detected. The response appeared in conjunction with blastocyst

elongation and was consistently confined to areas of concepts

membranes. The coincidence of increased uterine vascular

permeability at the site of attachment and elevated blood flow,

suggest enhanced transport of nutrients toward the concepts and

movement of conceptus-induced products into the maternal

circulation. Furthermore, it appears that embryonic factors act in a

localized manner to increase uterine vascular permeability

selectively during attachment. Histamines (Dey and Johnson, 1980),

prostaglandins, e.g., PGE (Davis et al., 1983) and estrogens (King

and Ackerley, 1985) of concepts origin have all been suggested to be

involved in this response (Keys et al., 1986).

From d 16, a microvillous attachment develops between trophoblast

and endometrial epithelium and extends rapidly to cover almost the

entire chorion by d 22 (Bjorkman, 1965). Expansion and development

of the allantois rapidly occurs between d 18 and 30 of gestation.

Fetal and maternal surfaces undergo complex folding to increase

surface area, and areolae develop on the chorioallantois at points

where endometrial glands open into the uterine lumen. Additional

folding of the chorion occurs in each areolus to increase surface

area and to allow for uterine gland secretions to accumulate

(Crombie, 1970). Fusion of chorion and allantois occurs between d 30

and d 60 of pregnancy, and by d 60 to 70, placental development is

complete. Because placentation in the pig is superficial with no

invasion of maternal uterine endometrium, transfer of histotroph from

the endometrial surface occurs at least through the second trimester

of gestation (Bazer and First, 1983). Maintenance of pregnancy

requires successful development of the placenta, so the embryo-fetal

units can receive nutrients, wastes can be removed and the concepts

will be protected from the maternal immune system.

Because of the noninvasive nature of placentation in the pig, it

is likely that a high proportion of the nutrients of the trophoblast

are secreted by endometrial glands and surface epithelium (Amoroso,

1952; Dantzer et al., 1981). Numerous reviews of uterine secretions

during pregnancy and blastocyst-endometrial interactions are

available (Bazer and First, 1983; Flint et al., 1982; Bazer and

Roberts, 1983; Bazer et al., 1986; Roberts and Bazer, 1988).

Fazleabas et al. (1985) demonstrated that, depending upon the

physiological purpose of a protein, some proteins (uteroferrin) are

secreted by uterine gland epithelium, while other proteins (plasmin

inhibitor) are secreted by the surface epithelium. Thus, the pig

epithelium is regionally differentiated. In addition, results of

Fazleabas et al. (1985) confirmed previous reports (Knight et al.,

1973, 1974a,b) that uterine secretary proteins are under

progestational control. Furthermore, while secretion of proteins by

the pregnant uterus is induced by progesterone, estradiol modulates

the magnitude of secretion of progesterone-induced proteins.

Included among the list of endometrial secretary proteins,

induced at least partially by progesterone (Knight et al., 1973) and

produced in early pregnancy (Zavy et al., 1977), are three enzymes:

uteroferrin, lysozyme and leucine aminopeptidase. These enzymes have

been shown to accumulate in allantoic fluid after d 30 of pregnancy

(Bazer et al., 1975; Roberts et al., 1976). Based on

immunoflourescent studies, these and certain other compounds of

endometrial origin can pass into the fetus via the chorio-allantoic

areolae (Chen et al., 1975). Uteroferrin is involved in the

transport of iron from uterine glands to the concepts (Roberts and

Bazer, 1980; Bazer et al., 1981b; Ducsay et al., 1986). Other

transport proteins such as retinol and retinoic acid binding

proteins, hydrolytic enzymes and regulatory proteins have been found

in uterine secretions of ovariectomized pigs injected with

progesterone. A high content of riboflavin in uterine flushings has

been reported during the initial period of blastocyst expansion

(Moffatt et al., 1980; Murray et al., 1980). Riboflavin appears to

increase embryonic survival when administered orally during early

gestation (F. W. Bazer and M. T. Zavy, personal communication).

Glucose, fructose and ascorbic acid are also found in uterine

flushings and increase in the pregnant uterus as gestation progresses

(Zavy et al., 1982). Thus, the allantoic fluid of pigs contains

substantial quantities of nutrients, including proteins detected in

uterine flushings, glucose, fructose, water, minerals, electrolytes

and vitamins (see Bazer et al., 1981a). The placental areolae appear

to transport these nutrients to the fetus, by fluid-phase pinocytosis

(Friess et al., 1981) or by an active transport process, e.g., water

and glucose, in order for a successful gestation to occur. So, the

allantois and amnion do not simply play a passive role in protection

of the fetus and expansion of the chorioallantoic membranes. Because

of the roles and composition of the fetal fluids, they are active in

the mechanisms associated with embryonic survival (see Bazer and

First, 1983). Thus, cooperative interactions of the endometrium and

placenta are required for the developing embryo/fetus to receive

adequate nutritional/metabolic support.

In addition to requirements for maintenance of pregnancy,

maternal immunological tolerance of the antigenically dissimilar

concepts must occur (for review, see Koch, 1985). Although the

concepts possesses antigens on its surface which are accessible to

and frequently elicit responses from the maternal immune system,

pregnancy normally progesses with no adverse effects. Although the

mechanism of immunological tolerance is not well defined, Beer and

Billingham (1979) suggested that the following events occur during

pregnancy: (1) excess antigen shedding by trophoblast may lead to

blocking of T and B cell immune responses; (2) antigen-antibody

complex production during pregnancy may block T and B cell immune

response and (3) suppressor T lymphocyte production may be enhanced

during pregnancy. Data of Murray et al. (1978) suggest that uterine

secretions of pigs contain an immunosuppressive factor. Masters et

al. (1983) indicated that a large glycoprotein (Mr>660,000) secreted

by pig blastocysts may coat the concepts and prevent maternal

lymphocytes from recognizing concepts antigens. Recently, Murray et

al. (1987) identified a high molecular weight glycoprotein

(MW>660,000) purified from d 16 porcine concepts culture medium that

had potent immunosuppressive effects. This was evidenced by

suppression of phytohemagglutinin and mixed lymphocyte-induced

blastogenesis of lymphocytes. In addition, the authors suggested

that this high molecular weight glycoprotein caused immunosuppressive

effects on anticonceptus immune responses by the mother while leaving

the systemic immune system unaffected.

As mentioned previously, pregnancy usually progresses with no

adverse affects from the antigenic incompatability. In some cases,

antigenic disparity might be advantageous for implantation and

continued intrauterine development (Gill and Repetti, 1979). Several

investigators have reported that number of conceptuses, placental

weights and fetal growth rates are higher in allogeneic versus

syngeneic matings (Billington, 1964; James, 1965, 1967; Beer and

Billingham, 1974; Beer et al., 1975).

In summary, numerous interactions between the concepts and

maternal endometrium are required to assure proper growth,

implantation, nutritional support and immunological acceptance of the

concepts. In spite of all the mechanisms designed to assure

maintenance of pregnancy, as will be reviewed in the next section,

the incidence of embryonic mortality is high in pigs.

Embryonic/Fetal Mortality

Litter size is determined by ovulation rate, fertilization rate,

concepts deaths during gestation and fetal deaths during the

perinatal period. Although ovulation rate can be increased by

nutritional means (Anderson and Melampy, 1972; Den Hartog and Van

Kempen, 1980; Dziuk and Bellows, 1983), greater embryonic death

losses during gestation generally occur so that litter size at birth

changes very little (Legault, 1978). In addition, as the incidence

of ovulation is increased via genetic selection or by gonadotropic

induction, less viable ova are produced (Koenig et al., 1986).

Developing hyperprolific lines such as Meishan pigs have led to

larger litters (Legault et al., 1981). However, an extensive system

of backcrossing with sows of more desirable growth and carcass traits

is required to enhance overall production successfully (Legault and

Caritez, 1983). In addition, the availability of hyperprolific

animals is limited. Heritability estimates of live pigs per litter

at birth range from .1 to .18 (Young et al., 1976; Legault; 1983).

Thus, selection for ovulation rate or litter size is not completely

practical. Therefore, the primary focus of this discussion will

concern embryonic mortality.

The majority of pregnancy wastage occurs before or during

embryogenesis and is referred to as embryonic mortality (Flint et

al., 1982). Rates of embryonic mortality, calculated by counting

embryos and using numbers of CL as a measure of the number of eggs

ovulated, usually range from 20 to 45% (see Flint et al., 1982). In

a study utilizing superovulated cows, some follicles were observed to

rupture and luteinize, but oocytes were not released (Monniaux et

al., 1983). Furthermore, fertility rates as well as the quality of

follicles and ovulations can also contribute to higher embryonic

mortality (see Terqui and Legault, 1984). Therefore, the number of

CL may not always represent the number of eggs shed (Terqui and

Legault, 1984) and embryonic mortality may be over-estimated.

However, in pigs, embryonic mortality estimates are usually in good

agreement with the number of CL and substantial loss due to embryonic

mortality does occur (Flint et al., 1982).

Numerous factors are associated with embryonic loss, including

chromosomal aberrations (Bishop, 1964). Based on karyotypes of

blastocysts recovered from pig uteri, McFeely (1967) and Day and

Polge (1968) suggested that 8 to 12% of fertilized eggs are abnormal

due primarily to polyspermy (which is likely to be lethal). Van der

Hoeven et al. (1985) indicated that this high percentage of

chromosomal abnormalities has not been confirmed. In agreement with

Van der Hoeven et al. (1985), Boyd (1965) reported that genetic

abnormalities account for only a small percentage (0 to 4%) of

embryonic mortality in domestic animals. Most losses due to

chromosomal abnormalities were reported to occur after d 10 (McFeely,

1967). Van der Hoeven et al. (1985) reported the karyotypes of 3- or

4-day-old pig embryos after short in vitro culture and confirmed that

a low frequency of chromosomal abnormalities exists in pig

conceptuses. This was true for both 10-day-old pig blastocysts

(Dolch and Chrisman, 1981) and 3- or 4-day-old embryos in which

hatching had not yet taken place (Van der Hoeven et al., 1985). Of

115 embryos cultured (3 to 4-day-old embryos), 80 were judged to be

of normal karyotypes. In contrast, Koenig et al. (1986) reported

25.1% of ova (8 to 32 h post-ovulation) are chromosomally abnormal.

Percentages of immature ova (70% of which never resumed meiosis)

ranged from 8.2 to 15.7%. Thus, Koenig et al. (1986) estimated that

a substantial portion of embryonic loss in swine could be accounted

for by chromosomal abnormalities and(or) ovulation of immature ova.

It is probable that certain boars (Perry and Rowlands, 1962) and

inbreeding (Squires et al., 1952; Rampacek et al., 1975) are

associated with greater losses due to genetic factors. As indicated

by Bazer and First (1983), because the genotype of the fetus and its

membranes is determined by both the dam and sire genotypes,

production of protein and steroid hormones by the placenta will be

unique. Since these hormones influence concepts development and

survival in utero, certain sire-dam matings will have higher levels

of fertility and larger litter sizes. However, the effectiveness of

genetic selection for prolificacy in pigs at this time is limited

(for review see, Bichard and David, 1985; Johnson et al., 1985,

Legault, 1985).

Genes affecting early embryonic development have been identified

by Goldbard and Warner (1982). They reported that an H-2 associated

gene, named Fed (preimplantation embryo development) controlled

either fast or slow embryonic development in mice. Warner et al.

(1984) speculated that a slower rate of development was associated

with more embryonic mortality and indicated that this was true for

outbred strains of mice (e.g., CF1, a strain in which a population of

mixed morulas and blastocysts often indicates morulas are moribund)

(C. Warner, unpublished observations cited in Warner et al., 1984).

Furthermore, Warner et al. (1984) speculated that the Ped gene may be

biologically critical. For example, they raised the possibility that

more advanced pig embryos are the ones that survive to birth. In

addition, they wished to pursue whether domestic species have a

homologue to the mouse MHC associated Ped gene. Warner et al. (1986)

have since demonstrated the presence of MHC antigens on 2 to 4 cell

stage and blastocyst stage pig embryos. In addition, Conley et al.

(1987) reported that litters from either sows or boars with the d

haplotype were larger than for all other matings. The authors

concluded that although ovulation rate was higher in SLA d/d sows

(swine MHC, haplotype d), an additional effect of the d haplotype on

embryonic survival was indicated by a significant effect of paternal

genotype on litter size. Thus, the Ped gene initially detected in

mice appears to be present in pigs as well. Recently, Warner et al.

(1988) indicated that the Ped gene product may be the Qa-2 antigen.

In addition, they indicated that although there is no known function

for any of the Q-region genes, Qa-2 antigens would be expected to be

expressed on preimplantation embryos possessing Qa-2a alleles but

not Qa-2b alleles (Warner et al., 1987). Strains of mice

expressing the Qa-2 antigen produce embryos with the slow Ped allele

(Warner et al., 1988). Precisely how the Ped gene regulates

embryonic growth and(or) maternal immunological tolerance remains to

be determined.

It is recognized that pubertal gilts have greater embryonic

losses than multi-estrous pigs (Warnick et al., 1951; MacPherson et

al., 1977). Uterine capacity does not appear to be the contributing

factor. Up to d 30 of gestation, the gilt can accommodate

approximately twice as many fetuses as are normally present (Dziuk,

1968; Rampacek et al., 1975). Archibong et al. (1987) determined

that changes in the ratio of systemic levels of estrogen and

progesterone may be related to early embryonic mortality in gilts

bred at pubertal estrus. Embryonic survival in the study by

Archibong et al. (1987) was 78.1 vs 95.4% on d 15 and 66.7 vs 89.4%

on d 30 of gestation in gilts bred at the first vs the third estrus,

respectively. Although serum estrogen and progesterone were not

significantly different, the ratio of progesterone:estrogen was

higher on d 15 (439 vs 210) and d 30 (597 vs 179), but lower on d 3

(187 vs 444) of gestation for gilts bred at their first or third

estrus, respectively. It is possible that the low

progesterone:estrogen ratio on d 3 may have altered uterine secretary

activity and resulted in early embryonic death. The higher steroid

ratios at d 15 and 30 are probably due to a reduced number of

surviving conceptuses and reduced estrogen production by the placenta

after d 12 of gestation. Although further investigation will be

required, the altered steroid ratios may be involved in embryonic

loss in gilts.

It is well known that progesterone is required for the

establishment and maintenance of pregnancy in the pig. Furthermore,

it has been suggested that a deficiency of progesterone leads to

embryonic loss. Although it is generally assumed that the maternal

hormone profile is always appropriate for establishment of pregnancy,

Ashworth et al. (1984) suggest that naturally occurring levels of

progesterone can be inappropriate and as a result embryonic loss

occurs. Because of this, experiments have been conducted utilizing

progesterone supplementation as a protective measure (Flint et al.,

1982). Results have been inconsistent. Heap et al. (1984) suggest

that progesterone therapy reduces early embryonic loss in certain

circumstances, e.g., high environmental temperatures. However, Heap

et al. (1984) indicate that designing the appropriate dose and time

frame for supplementation is difficult because of the lack of

knowledge of the ways that progesterone supports early embryonic

survival. In addition, Heap et al. (1984) emphasize that it is not

yet known whether progesterone plays an obligatory or a permissive

role. These authors attempted to clarify the role of progesterone in

embryonic survival by utilizing passive immunization against

progesterone in early pregnancy in mice and reported multiple actions

of the antibody (Heap et al., 1984). When anti-progesterone

monoclonal antibody was given repeatedly to BALB/c mice as five daily

injections from the day of mating, implantation was blocked in all

six mice. Furthermore, few of the embryos of treated dams progressed

to the blastocyst stage and there was no indication of delayed

implantation. No remnants of degenerating conceptuses were detected

at autopsy on d 30 after copulation. The authors indicated that the

primary effect of the antibody was to arrest cleavage of embryos.

This effect was stated to occur during activation of the embryonic

genome. Based on these findings, Heap et al. (1984) stated that

progesterone production in the first 3 d after fertilization is

essential for normal embryonic development. Furthermore, they

indicated that one effect of progesterone is to provide a tubal

environment in which the embryos can develop at normal rates. Their

work is supported by results of Cline et al. (1977) who found that

estrogen dominated females had embryos with retarded development.

However, if progesterone was also supplied, embryonic development

proceeded at a normal rate. Thus, the possibility remains, as

indicated by Ashworth et al. (1984), that insufficient progesterone

production due to inadequate CL function can lead to increased

embryonic losses.

Other researchers have taken the approach that a combination of

exogenous progesterone and estrone will have a more beneficial effect

than progesterone alone on embryonic survival (Sheridan et al.,

1986). Knight et al. (1973) established that a 2000:1 ratio of

progesterone:estrogen was effective in inducing a synergistic effect

on uterine protein secretion in pigs. Although Knight et al. (1974a)

did not find a significant effect on embryonic survival following

steroid injections from d 4 to 15 of gestation, mean placental weight

and allantoic fluid volume increased when measured at d 40 of gesta-

tion. Knight et al. (1974b) also demonstrated a negative effect of

doses higher than 50 mg progesterone and 25 ug estradiol per 45.4 kg

BW on uterine protein secretions. Since then, Wildt et al. (1976)

and de Sa et al. (1981) reported increased litter size following

combined injections of progesterone and estrone between d 14 to 23

and d 16 to 17 of gestation, respectively. However, since primarily

multiparous sows were in the treated group and primiparous were in

the control group, these results may be misleading. Recently,

Sheridan et al. (1986) reported results of administration of proges-

terone (25 mg) and estrone (12.5 ug) on both d 16 and 17 or on either

d 16 or 17 of gestation. Gestation length was reduced and litter

sizes were larger at birth for steroid-treated sows. However, subse-

quent litter size at weaning and interval from weaning to rebreeding

were not altered by treatment. Their results support the theory that

injection of progesterone and estrone at a 2000:1 ratio in

quantities less than 50 mg progesterone and 25 ug estrone/45.4 kg BW

will effectively decrease embryonic mortality. Although Sheridan et

al. (1986) did not determine uterine secretary response or placental

growth in response to exogeneous steroid treatment, they suggested

that due to increased protein secretion or placental growth rate

previously reported by Knight et al. (1973), nutrient availability to

embryos increased and embryonic survival was enhanced.

In many countries, a period of lowered fertility exists in the

summer (see Claus and Weiler, 1985). Ledwitz-Rigby and Rigby (1987)

suggest that part of the decreased reproductive capacity of pigs in

summer may be due to decreased capacity for follicular development

within the ovary as a result of an altered endocrine signal. They

suggest that granulosa cell responsiveness to signals is altered by

photoperiod and(or) heat (such as decreased FSH; Mauget, 1982) which

results in a their decreased ability to secrete progesterone.

Conception rate is decreased in summer and tends to be elevated in

autumn and winter (Keindorf and Plescher, 1981; Britt et al., 1983;

Dobao et al., 1983; Mattioli et al., 1983). Mattioli et al. (1983)

and Love (1981) reported that a higher incidence of embryonic loss

occurs in the summer months (June-September). Problems due to timing

of artificial insemination or mating may be responsible for reduced

conception rates. It is known that duration of estrus is prolonged

during the summer (Signoret, 1967; Nauk and Sekrii, 1983). Reduced

farrowing rates due to embryonic loss has been reported (Stork, 1979;

Mattioli et al., 1983) with the highest incidence of abortions

detected in September/October for sows mated between June and

September. Litter size is higher after mating in autumn and winter

compared to summer (Tomes and Nielsen, 1979; Bevier and Backstrom,

1980; Keindorf and Plescher, 1981; Dobao et al., 1983; Mattioli et

al., 1983; Noguera et al., 1983; Aalbers et al., 1984). According to

Bevier and Backstrom (1980) and Mattioli et al. (1983), litter size

for sows mated in the summer is usually reduced by 1 piglet or more

compared to that for sows mated in autumn or winter. Concentrations

of progesterone in early pregnancy (dl0 to 60) are lower (21.5 vs

32.4 to 40.3 ng progesterone/ml plasma) in August-September than in

all other months (Beilanski and Kremer, 1983). Seasonal influences

on the weaning-to-estrus interval may be removed by imposing

conditions of decreased photoperiod in summer (Claus et al., 1984).

However, effects of reduced photoperiod on litter size, embryonic

mortality and conception rates have not been reported. It appears

that sensitivity to photoperiodic stimuli present in wild pigs

(Mauget, 1982) may not have disappeared completely in domestic pigs

(Claus and Weiler, 1985). However, Claus and Weiler (1985) indicate

that interactions between photoperiod and embryonic/fetal survival

need to be clarified. In addition, this author suggests that effects

now attributed to photoperiod are also due, in part, to temperature


Environmental factors such as elevated ambient temperature are

also known to influence embryonic mortality (Wettemann et al., 1984;

Wettemann and Bazer, 1985). Although the following discussion deals

with effects of heat stress, it is relevant to the overall discussion

on embryonic mortality. For example, altered endocrine function in a

non-heat stressed sow may also lead to the loss of embryos. Thus, an

overall understanding of factors associated with embryonic survival

can be gained by studying this model.

Wettemann et al. (1984) indicated that heat stress of male swine

results in a decrease in fertilization rate. Furthermore, increased

embryonic mortality occurs when gilts are artificially inseminated

with extended and stored semen from heat stressed boars. However,

when heat stressed boars are naturally mated, conception rate, but

not litter size, is altered (Wettemann et al., 1976, 1979). Heat

stress of gilts prior to breeding did not influence ovulation rate or

embryonic survival (Edwards et al., 1968). However, heat stress

during the intervals from 1 to 15 days (Edwards et al., 1968) or 0 to

8 days (Omtvedt et al., 1971) after breeding reduced conception rate

and litter size when measured at d 30. Kreider et al. (1978)

reported that plasma progesterone increased and estradiol decreased

following heat stress of gilts from estrus until d 8 of gestation.

When gilts were exposed to elevated ambient temperatures from d 8

to 16 after breeding, decreased conception rates and reduced litter

size were detected at d 30 of gestation (Omtvedt et al., 1971). In

addition, Omtvedt et al. (1971) suggested that the implantation

period may be the most susceptible to heat stress. Wettemann et al.

(1984) reported effects of heat stress between d 8 to 16 after

breeding on concentrations of progesterone, corticoids and estradiol

in plasma. Luteal function, as measured by progesterone production,

was not altered between days 9 through 28 of pregnancy by heat stress

in those gilts that maintained pregnancy. However, in heat

stressed, nonpregnant gilts on d 13 to 19 after estrus, concentra-

tions of progesterone were reduced in comparison to nonpregnant

control gilts and all pregnant gilts. Yet, by d 25, progesterone in

plasma of heat stressed, pregnant gilts had not decreased to

concentrations usually present at d 25 of pregnancy. Wettemann et

al. (1984) suggested that prolonged luteal regression occurred in

these gilts following maternal recognition of pregnancy and subse-

quent embryonic mortality. Polge et al. (1966) previously reported

an extended luteal phase and estrus on approximately d 26 after

breeding in gilts with less than 4 embryos on d 12 of pregnancy.

Concentrations of estradiol in plasma in gilts heat stressed from

d 8 to 16 after breeding were greater compared to control nonpregnant

and all pregnant gilts on d 10 to 12 after estrus. Thus, the

increase in estradiol proceeded the decrease in progesterone.

Wettemann et al. (1984) suggested that because the increase in

estradiol occurred in heat stressed gilts prior to the normal

initiation of blastocyst synthesis of estrogens, improper timing of

steroidogenesis may have caused detrimental effects on concepts


Support for this theory has been generated by Morgan et al.

(1987a). Exogenous estradiol treatment on d 9 of pregnancy resulted

in advanced uterine secretary response. This was evidenced as

increased calcium, protein and acid phosphatase levels in uterine

flushings of pregnant gilts unilaterally hysterectomized on d 11.

Following removal of the remaining uterine horn on d 12, increased

protein and acid phosphatase and decreased calcium and PGF were

reported in the estradiol-treated gilts. However, in a second

experiment (Morgan et al., 1987a), exogenous estradiol followed by

hysterectomy at d 16 revealed no differences in calcium, PGF or

protein and decreased acid phosphatase. Blastocysts were fragmented

and degenerating on d 16 in these estradiol-treated gilts, although

they appeared normal in the previous trial at d 12. Collectively

these data suggest that early estradiol exposure in the uterus

advances uterine secretary response and, although blastocyst

elongation occurs, blastocysts fail to survive to d 16. Geisert et

al. (1987) previously indicated that loss of progesterone support was

not responsible for embryonic mortality observed in estradiol-

treated gilts. In addition, estradiol is not directly embryocidal

(Pope and First, 1985) since gilts maintain pregnancy when estrogen

is administered after d 11. Therefore, Morgan et al. (1987a) suggest

that failure of blastocyst attachment is the result of premature

estradiol action on uterine secretions and(or) surface proteins.

Thus, as Wettemann et al. (1984) suggested, alteration of the

hormonal sequence which influences the endometrium is detrimental to

embryonic survival.

Wettemann et al. (1984) also studied the response to exogenous

adrenocorticotropin hormone (ACTH) in heat stressed (from d 8 through

d 16 after breeding) versus control pregnant and nonpregnant gilts.

Concentrations of corticoids and progesterone (during the first 2 h

after ACTH) were reduced in heat stressed nonpregnant gilts. Marple

et al. (1972) previously reported that gilts and barrows exposed to

elevated ambient temperatures had increased ACTH concentrations in

plasma. The combined results suggest that heat stress may alter

adrenal steroid secretion. Wettemann et al. (1984) indicate that

endogenously, excessive ACTH may have the same effect on steroids as

seen following exogenous ACTH treatment. Therefore, altered adrenal

activity may decrease embryonic survival either by lowering

concentrations of necessary steroids or by stimulating greater

production of steroids which have detrimental effects. Two lines of

evidence support their suggestion. First, Tilton et al. (1972)

reported that adrenalectomy may prevent embryonic mortality induced

by heat stress of sheep. Secondly, Howarth and Hawk (1968) reported

that injection of ewes with hydrocortisone acetate resulted in

decreased embryonic survival.

Wettemann et al. (1984) also reported the effects of heat stress

from d 8 through d 16 after breeding on embryonic and uterine

responses. The amount of embryonic tissue present at d 16 was less

for heat stressed gilts. However, protein synthetic ability, as

measured by amount of labelled leucine incorporated into

macromolecules by concepts tissue, was not altered when

incorporation was adjusted for dry weight of concepts tissue.

Similarly, on d 16, total protein and PGF content per uterine flush

were similar for heat stressed and control pregnant gilts. In

addition, uterine endometrial function at d 16 after estrus was not

altered by heat stress, as indicated by similar dry weight of tissue

and incorporation of leucine into macromolecules by uterine

explants. Furthermore, steroidogenesis by conceptuses and uteri were

not altered by heat stress, as measured by total amounts of estrogens

in uterine flushes on d 16 after breeding.

Significant effects of heat stress on concentrations of the

metabolite of PGF, 13,14-dihydro-15-keto-prostaglandin (PGFM),

composition of blood gases and blood flow to the uterus were not

detected (Wettemann et al., 1984). However, concentrations of

prolactin in plasma were elevated in gilts exposed to higher ambient

temperature following breeding (d 8 to 16) (Wettemann and Bazer,

1985). The authors indicated that the precise role of these factors

in embryonic mortality was not readily apparent (Wettemann and Bazer,

1985), but could not rule out the possibility of their importance.

Putney et al. (1987) indicated that increased embryonic mortality

in cows may result from alterations in signals between the concepts

and maternal endometrium necessary for maintenance of pregnancy.

Heat shock (39 C for 6 h; 43 C for 18h) reduced (54.2%) protein

secretion of conceptuses, and in particular, decreased (74.3%)

concepts secretion of bovine trophoblastic protein-1 in comparison

to control (39 C for 24 h) conceptuses. However, protein secretion

by endometrial explants was not altered. In addition, a 1225%

increase in endometrial PGF2alpha secretion was detected in

response to heat shock. Endometrial PGE2 production and secretion

of prostaglandins by concepts tissue were not significantly

altered. An analagous situation to that for heat shock in the bovine

(Putney et al., 1987) may occur in porcine conceptus-endometrial

interactions in response to heat stress.

Elevated ambient temperatures during d 53 to 61 of gestation did

not influence conception rate, litter size, piglet birth weights or

survival rates (Omtvedt et al., 1971). However, heat stress during

later gestation (d 102 to 110 of gestation) increased fetal mortality

(Omtvedt et al., 1971). Wettemann and Bazer (1985) emphasize that

during late gestation the requirements of the fetal-placental unit

for estrogen, nutrients, and gaseous exchange are great. Therefore,

conditions leading to deficiencies during late gestation may result

in fetal death. In contrast, during mid-gestation, heat stress may

not be detrimental because the fetus is only about 10% of birth

weight and fetal-placental steroidogenic function (Knight et al.,

1977) is minimal compared to late gestation (Wettemann and Bazer,


In summary, Wettemann et al. (1984) demonstrated that exposure of

swine to elevated ambient temperature during early gestation leads to

lower conception rates and litter sizes. Wettemann et al. (1984)

emphasized that concentrations of steroid hormones control uterine

secretions (Knight et al., 1973; Bazer, 1975) and can alter transport

of ova, embryos and sperm (Chang, 1966; Hawk and Conley, 1971; Hawk,

1975). Therefore, it is possible that heat stress alters conception

rate or embryonic survival indirectly via the endocrine system or

through direct effects on the gametes or embryos (Wettemann et al.,

1984). Although these studies were conducted to determine effects of

elevated ambient temperature, it is possible that factors associated

with embryonic mortality in heat stressed animals may be induced by

other conditions and also result in embryonic loss.

The effects of migration, distribution and spacing of pig embryos

on pregnancy and fetal survival have been reviewed by Dzuik (1985).

Dziuk (1985) indicated that embryonic survival is not influenced by

uterine space prior to d 30. It is essential that a minimum of 4

conceptuses occupy the uterus at d 12 for pregnancy to be maintained

(Polge et al., 1966). Furthermore, the proportion of the uterus that

is occupied is inversely related to the probability that pregnancy

will continue (Dhindsa and Dziuk, 1968). Proper spacing must occur

prior to d 12 (Dziuk, 1985). Furthermore, prenatal growth and

survival of fetuses is greatly influenced by spacing and migration of

embryos (Webel and Dziuk, 1974; Anderson and Parker, 1976; Pope and

First, 1985). Because some embryos die between d 13 and 40 of

gestation, gaps between fetuses result since attachment takes place

at d 12 (Dziuk, 1985). In addition, the distribution and spacing of

embryos is not apparently uniform throughout the length of the

uterus. At d 40 of gestation, the greatest space available to each

fetus was at the tip of the horn and near the uterine body, with the

least available space in the middle of each horn (Hentzel et al.,

unpublished results, cited by Dziuk, 1985). Dziuk (1985) suggested

that this spacing results from death and partial resorption of

embryos initially crowded near the uterine body. Although uterine

space available to each embryo had little influence on embryonic

survival prior to d 25 to 30, in later gestation (d 70) few fetuses

survived in the crowded segments according to Dziuk (1968). Dziuk

(1985) suggested that optimum conditions for litter size, fetal

growth and survival involve uniform distribution of embryos with each

occupying 20 cm of space. He also emphasized that it is imperative

that migration, spacing and distribution be equitable very early in

gestation to maximize survival.

Koch (1985) discussed immunological implications of pregnancy in

the pig. Because cells of all higher vertebrates possess

histompatibility antigens on their surface, some mechanism has

evolved to allow for immunological acceptance of the concepts by the

maternal immune system (Koch, 1985). It has been demonstrated that

the uterus is not a privileged site for survival of conceptuses in

various species including the pig, by continued development of

trophoblastic tissue when transplanted to ectopic sites (Samuel,

1971). In pigs, the major histocompatability complex is the swine

lymphocyte antigen (SLA) complex (Vaiman et al., 1970; Viza et al.,

1970). The SLA complex is responsible for discrimination of self and

non-self and must be dealt with for successful pregnancies to occur

in swine. Although spermatozoa, seminal plasma and tissues of the

concepts represent sources of antigenic material, embryonic tissues

are considered to be the primary immunological challenge (Koch,

1985). One hypothesis proposed to explain the non-rejection of the

feto-placental unit is the masking of transplantation antigens on the

concepts. A general suppression of the immune response during

pregnancy has also been suggested. In addition, antigenic disparity

between the mother and concepts in certain circumstances may be

beneficial for implantation and continued intrauterine development

(Gill and Repetti, 1979). However, Koch (1985) indicates that the

precise nature of the maternal immune response towards the concepts

is not known. None-the-less, it is imperative that the proper

immunological relationship exist between the fetus and mother or

embryonic mortality will result.

Ford and Stice (1985) have evidence of interactions between pig

conceptuses and uterus affecting uterine blood flow previously

described by Ford and Christenson (1979). Results from numerous

studies suggest that pig conceptuses, through catechol estrogen

production, increases arterial distensibility locally, resulting in

increased baseline blood flow to each fetal-placental unit. However,

contractility of the uterine arterial vasculature to adrenergic

agonists or other vasoactive agents is not affected. This allows

conceptuses to maintain locally elevated blood flow required for

their survival. Simultaneously, the maternal system responds to

life-threatening stimuli by transiently rerouting blood flow away

from the uterus and towards other vascular supplies necessary for

maternal survival (Ford and Stice, 1985). A malfunction in this

system could result in embryonic/fetal death.

As reviewed by Bazer and First (1983), the developing concepts

may, in part, control its own destiny via the synthesis of certain

compounds. It had previously been postulated that litter size is, in

part, limited by the availability of an essential biochemical factor

(Bazer, 1975). Most embryonic mortality in pigs has been reported to

occur prior to d 12 of gestation (Perry and Rowlands, 1962). Bazer

and First (1983) indicate that loss of pregnancy may be a function of

concepts rather than maternal deficiencies. This may explain why

some embryos survive while others die within the same uterine

environment in polytocous species (Bazer and First, 1983). Thus, if

blastocysts do not develop and elongate at an adequate rate, they

fail to produce compounds critical to their own survival (Bazer and

First, 1983).

One of these necessary signals in the pig is estrogen, which is

produced by the pig blastocyst beginning on about d 11 of gestation

(Geisert et al., 1982b). Estrogen is required to ensure CL

maintenance (Bazer and Thatcher, 1977), and nutrient availability

through stimulation of histotroph secretion (Geisert et al., 1982b)

and increased uterine blood flow (Ford and Christenson, 1979).

Numerous reviews describe substances produced by the endometrium

and blastocyst which may allow interactions between the concepts and

endometrium required for embryo survival (Flint et al., 1982; Bazer

and Roberts, 1983; Bazer et al., 1986). At this time, there is no

conclusive immunoelectophoretic or chromatographic evidence for

pregnancy-specific endometrial proteins involved in stimulating

blastocyst growth. However, Flint (1981) suggests that blastocyst

estrogens control secretion of endometrial embryotrophic factors

which should be considered further. If growth promoting substances

are secreted by the endometrium, competition for growth factors may

be involved in deciding which embryos are lost (Flint et al., 1982).

Therefore, conceptuses which elongate first may obtain a higher

proportion of available growth-supporting substances and deprive

those elongating later (Flint et al., 1982).

There is considerable variation in blastocyst size during the

time of elongation (Anderson, 1978). As reviewed by Wilmut et al.

(1985), relatively retarded embryos are much less likely to survive

than synchronous embryos or embryos advanced by up to 24 h (Webel et

al., 1970; Polge, 1982). Thus, factors that advance embryonic growth

or are inherent in an embryo make the advanced embryo much mores

likely to survive in the uterine environment of a single sow (Wilmut

et al., 1986).

Pope et al. (1982a) reported that more developed embryos have an

advantage over less developed embryos when a 2 d range in age and

development was tested. More recently, Pope et al. (1986b) examined

this in embryos within a closer range of development. Utilizing

asynchronous superinduction to test the range of blastocyst

development on embryonic survival in swine, Pope et al. (1986b)

reported the following. When the proportion of less developed

embryos within a litter increased, subsequent mortality also

increased during the first 30 d of gestation. Thus, naturally

occurring variation in stage of embryonic development might explain

some of the loss. This is in agreement with conclusion of Wilmut et

al. (1986) who reviewed factors contributing to embryonic loss and

indicated that variation in stage of embryonic development can be

sufficient to prejudice survival in polytocous species. In another

study lending support to this theory, Morgan et al. (1987b) tested

effects of estradiol treatment on d 11 in synchronous and

asynchronous (blastocysts 24 h behind recepient uterus) gilts and

compared to control asynchronous gilts. Embryos were recovered in

uterine flushings on d 14. Only in the asynchronous,

estradiol-treated group did blastocysts fail to develop. Acid

phosphatase activity was higher and amount of calcium recovered was

lower in both the asynchronous and synchronous estradiol-treated

gilts. Thus, estradiol advanced uterine secretary activity. Morgan

et al. (1987b) indicated that this effect was embryocidal to embryos

that are behind in their development, since neither asynchrony or

estradiol itself resulted in embryonic loss. Furthermore, results of

Morgan et al. (1987b) indicate that blastocysts left to develop at

their own rate stimulate uterine environmental changes at the onset

of elongation and allow blastocyst development to proceed normally.

Thus, the uterus appears to play a permissive role with blastocysts

regulating appropriate uterine responses (Morgan et al., 1987b).

Morgan and co-workers (1987b) also indicated that immature

blastocysts may not possess receptors to recognize changes in

regulatory proteins of uterine origin either released by the

endometrium or exposed on the uterine surface epithelium after

estrogen stimulation. Therefore, immature blastocysts do not undergo

developmental changes required for survival. In summary, three

possibilities have been suggested to account for embryonic loss in

gilts with advanced uterine secretion due to estradiol

administration: (1) inability of blastocysts to elongate; (2) altered

uterine surface epithelium and (3) a direct embryocidal effect of the

uterine environment. These studies clearly demonstrate that

blastocysts must attain a degree of morphological maturation before

they can respond to and survive in the altered uterine environment

normally induced by estrogen on d 11 to 12 of gestation. Although

these conditions were experimentally induced (Morgan et al., 1987b),

there is normally a high degree of variation in morphological

maturity between conceptuses within the same uterus. Failure of

immature conceptuses to survive the change in intrauterine

environment induced by faster developing conceptuses may be a signi-

ficant factor contributing to embryonic death losses in pigs. When

conceptuses within the same uterus are similar in development, each

may have a greater ability to respond to the uterine environment and

greater chance for survival. Evidence for this was obtained from a

study of concepts development in prolific Chinese Meishan gilts

(Bazer, 1987). In comparison to less prolific breeds, Meishan have

less variation in concepts development between d 8 to 14 of gesta-

tion and rate of development between d 11 and 12 of gestation was

faster. These results suggest that minimizing variation among con-

ceptuses may be part of the mechanism by which Meishan pigs produce

larger litters. The larger litters do not result from higher ovula-

tion rates, but factors associated with higher embryonic survival.

In summary, neither fertilization failure (Robertson et al.,

1951a; Squires et al., 1952; Day et al., 1959; Spies et al., 1959;

Perry and Rowlands, 1962) or chromosomal aberrations (McFeeley, 1967;

Lupse, 1973), insufficient luteal development (Webel et al., 1975) or

insufficient uterine space (Webel and Dziuk, 1974) explain all of the

30% loss of potential embryos common to swine. In addition, no

single factor can explain embryonic loss or be manipulated to

adequately alter embryonic survival (Bazer and First, 1983).

Genetic, nutritional, environmental, disease and endocrine factors

all can lead to embryonic/fetal death. In a later section,

nutritional factors affecting embryonic/fetal loss, litter size and

postnatal survival will be reviewed.

Effects of Nutrition or Manipulation of Metabolism of Sows
During Gestation on Embryonic Survival. Litter Size
and Sow and Litter Performance During Lactation

Numerous factors must be considered when formulating a gestation

diet, i.e., feeding recommendations for gilts are much different than

those for sows. Gilts are now bred at young ages and leaner pork is

being produced resulting in lower body fat at breeding. Brooks

(1982) suggested that because the loss of body fat of sows during the

first 2 or 3 farrowings is great, either body fat of gilts should be

increased prior to the first farrowing to provide fat for later

depletion or sows should be fed to avoid body fat depletion. Brooks

(1982) indicated that increasing body fat prior to the first

farrowing is not practical for two reasons: (1) increased feed intake

during pregnancy reduced voluntary feed intake during lactation (Dean

and Tribble, 1961; Salmon-Legagneur and Rerat, 1962; Baker et al.,

1969) and (2) among gilts that receive the same gestation allowance

there is a tendency for those which farrow at heavier weights to lose

more weight during lactation (Brooks and Cole, 1973; Brooks and

Smith, 1980). Thus, it is more practical to maintain body fat by

providing adequate feed allowances from the first farrowing onward.

This may reduce the problem of lower ovulation rates frequently

encountered in gilts having become catabolic during lactation

(Brooks, 1982).

Cole (1982) indicated that sows do not reach reproductive

maturity until about their fourth parity. The most prolific period

will extend beyond that. Therefore, nutritional considerations

should be for more than just short-term. Cole (1982) emphasized that

nutrition impacts on reproductive performance of sows in the

following ways: (1) short term reproductive performance evaluated by

how nutrition during pregnancy affects litter size, lactation and

weights of piglets and sows at weaning; (2) performance during the

subsequent pregnancy affected by nutrition during lactation and (3)

long term reproductive performance as influenced by nutrition in

early stages of their reproductive lives. Thus, long term plans

should be developed. In general, diets are designed to conserve body

condition during lactation. Since greater weight gains during

pregnancy result in greater losses in lactation, limited weight gains

in pregnancy are desirable (see Cole, 1982).

Nutrition of the breeding pig is not the only consideration,

however. Since piglets are born with limited energy reserves, diets

fed during gestation are formulated to increase fetal glycogen

stores. In addition, attempts are being made to reduce embryonic

death losses through altered nutrition. Because of these different

goals, determination of nutrient requirements for sows is difficult.

Lewis and Reese (1986) listed additional factors which make

formulating diets for sows difficult. First, sows reproduce well

during periods of nutrient deficiency by drawing on their own body

reserves. Pregnants sows have considerably greater feed efficiency

than nonpregnant females (Salmon-Legagneur and Rerat, 1962). Second,

large numbers of sows must be utilized to draw conclusions because

variability in reproductive traits is great. Third, carry-over

effects from pregnancy to lactation and to the next gestation must be

considered. None-the-less, nutrient requirements for sows during

gestation are being determined which will allow maximum reproductive


Influences of Nutrition Prior to Breeding or During Early Gestation
on Embryonic Survival

Feeding recommendations during early pregnancy are not well

established. In a number of cases, embryonic survival has not been

associated with feeding level (Cole, 1982). Dutt and Chaney (1968)

reported higher embryonic survival in sows fed reduced dietary energy

from mating to d 20 of gestation. However, ranges for feed intake

were too broad (1.25 to 4.1 kg/d) to provide practical information.

Dyck and Strain (1980) confirmed reports of higher embryonic survival

when feed intake was reduced from 2.5 to 1.5 kg/d from mating to d 10

of gestation. However, conception rates were reduced in gilts fed

less from mating to d 30 of gestation. As reviewed by Christenson

(1986), continuous ad libitum feeding of gilts before breeding and

during early gestation increased embryonic mortality in comparison to

pigs on restricted feeding (Robertson et al., 1951b; Self et al.,

1955; Gossett and Sorenson, 1959; Haines et al., 1959; Sorenson et

al., 1961). However, a more recent study of nutritional effects on

embryonic survival indicated that actual embryo numbers present were

similar between groups (Den Hartog and Van Kempen, 1980). Other

researchers failed to confirm an effect of high energy intake on

early embryonic survival in multiparous sows (McGillivray et al.,

1964; Christenson and Zimmerman, 1966; Heap et al., 1967; Toplis et

al., 1983). Toplis et al. (1983) indicated that differences in

embryonic survival previously reported following increased feed

intake were primarily found in primiparous sows and may result from

"flushing" and increased ovulation rate (Anderson and Melampy, 1972;

Den Hartog and Van Kempen, 1980). It is not clear by what

mechanisms) dietary energy levels exert their effects on embryonic

survival in swine other than through increased ovulation rate. Bazer

et al. (1968) demonstrated that rates of embryonic survival of gilts

fed ad libitum for 14 d prior to breeding were reduced not due to

embryo viability or overcrowding, but rather to some uterine factor

which limited litter size. In trials in which embryonic survival was

decreased, reduced concentrations of progesterone in plasma may be

part of the cause. Increased feed or protein intake during early

gestation in dairy cows or gilts has been associated with decreased

concentrations of progesterone in plasma (Jordan and Swanson, 1979;

Dyck et al., 1980).

Evaluating embryonic mortality between d 20 to 30 of gestation

may lead to improper recommendations. Etienne et al. (1983) reported

levels of embryonic mortality of 18 to 33%. However, at d 105,

differences between treatments for fetal mortality were reversed, and

thus, there were no net differences. In summary, nutritional effects

on embryonic survival are not clear. Few studies have led to

conclusive results, so recommendations other than avoiding severe

deficiencies or excesses during early gestation are difficult.

Christenson (1986) recommended that producers restrict feed after

mating, in part, to reduce feed costs and also to decrease embryonic

death losses.

Recently, through a series of cooperative experiments, (F. W.

Bazer and M. T. Zavy, personal communication) effects of

supplementing gestation diets with riboflavin were observed during

early gestation. Endogenous riboflavin increases in uterine

flushings of cyclic and pregnant gilts between d 6 and 8 after onset

of estrus and then decreases to undetectable levels by d 11.

Exogenous riboflavin increased levels of riboflavin in uterine

flushings. Pregnant gilts receiving 100 mg riboflavin per day on d 4

to 10 of gestation had increased numbers of live embryos, higher

embryonic survival and greater allantoic fluid volumes at d 30. In a

subsequent trial with 48 control and 51 riboflavin treated

primiparous sows, total number of piglets and piglets alive at birth

tended (P < .11, P < .09, respectively) to be higher and piglets

alive at d 21 and 42 were greater (P < .05). At d 21 and d 42, total

weight of piglets was greater for riboflavin fed sows.

Effects of Nutrition During Mid- or Late-Gestation on Piglet Birth
Weight and Survival

Numerous authors have reported effects of severe feed deprivation

during late gestation (Pond et al., 1968a,b; 1986; Ezekwe and Opoku,

1986; Speer, 1982). Speer (1982) reviewed evidence that low protein

(.5 to 2%), protein-free diets or complete inanition resulted in no

adverse effects on embryonic/fetal development or litter size,

although birth weights were in some cases decreased. Pond et al.

(1986) demonstrated that primiparous gilts can maintain pregnancy

during feed restriction. Primiparous sows were restricted to

one-third of the recommended (NRC, 1979) feed allowance from d 84

through d 108 of gestation. Although effects on fetuses warrant

further research (Pond et al., 1986), litter size was not affected

and piglet birth weight was decreased by 13% in gilts fed the

restricted level (.6 kg/d) (Pond et al., 1986). These results

confirm previous reports by Pond et al. (1968a,b; 1969) in which

protein free diets were fed for various intervals during gestation

and pregnant gilts utilized body stores of nutrients to meet fetal

demands. Recently, Ezekwe and Opoku (1986) demonstrated that feed

deprivation (complete starvation for 7 or 14 d prior to farrowing)

did not cause permanent growth retardation of progeny. Litter size,

birth weight, weaning weight, mortality at birth and survival to

weaning of piglets were not altered by starvation of sows. Thus,

feed deprivation for short periods may reduce feed costs. However,

long term effects or effects in the subsequent reproductive phase

were not determined. In other studies, severe protein and(or) energy

restriction during both gestation and lactation decreased the number

of sows returning to estrus after weaning, increased the interval to

estrus and decreased ovulation rates (Holden et al., 1968; Svajgr et

al., 1972; Pike and Boaz, 1972; Hovell and MacPherson, 1977).

Amounts of protein required during gestation are greatly

influenced by the level of protein available during the subsequent

lactation. Mahan and Mangan (1975) indicated that sows fed 9, 13 or

17% protein diets during gestation performed similarly during

lactation if lactation protein levels were adequate. When 18%

protein location diets were supplied, no differences in sow or litter

performance were detected. However, if 12% protein was fed during

lactation, litter weights increased most for sows fed the higher

protein levels during gestation. Mahan (1977) reported no difference

in litter size between sows fed gestation/lactation diets of 14/15

versus 8.5/20 percent protein. However, milk production was less for

the 8.5/20% fed sows. Greenhalgh et al. (1977) fed either 9, 11, 13

or 15% protein during gestation followed by 13 or 17% protein during

lactation. Number of total and live piglets born increased, but not

significantly, with ingestion of greater protein during gestation.

Greenhalgh et al. (1980) indicated no difference in piglet weight or

litter size at birth due to diets containing 9 or 11% protein during

gestation followed by 13, 15, 17 or 19% protein during lactation.

Feeding either 9 or 15% protein during gestation and 12, 16 or 20%

protein during lactation increased average piglet weight and number

of piglets born, but differences were not significant (NCR-42

Committee on Swine Nutrition, 1978). In summary, sows can withstand

protein deprivation during pregnancy without adverse effects on

litter size at birth. This may be due to the greater ability of

pregnant versus nonpregnant sows to retain nitrogen

(Salmon-Legagneur, 1965; Heap and Lodge, 1967). However, results are

controversial for the effects of long term protein deprivation on

litter size and piglet weight at birth. In addition, protein

deprivation may decrease milk production of sows, but not if adequate

nutrients are supplied during lactation. O'Grady (1985) indicated

that there is agreement that reproductive performance (number, weight

and composition of piglets born/litter) is maximized when 140 to 180

g crude protein/sow is provided during pregnancy and amino acid

requirements are met. Estimated requirements for lysine vary from .4

to .65% of the diet. Estimates of requirements for other amino acids

during pregnancy are also available (Cole, 1982; NRC, 1979). Because

different energy sources have different amino acid values and(or)

digestibilities, levels of protein and specific amino acids required

will vary (Maxwell et al., 1987).

Substantial evidence indicates that piglet survival during

lactation is positively correlated with piglet birth weight. Hall et

al. (1984) clearly demonstrated the importance of piglet birth weight

on survival in a study representing over 10,000 pigs. For piglets

averaging .45, .68, .91, 1.14, 1.36 or 1.59 kg at birth, the survival

rates (%) were 16, 39, 59, 74, 86 and 95%, respectively. Therefore,

during the last trimester of gestation, researchers have attempted to

increase fetal fat reserves and(or) piglet weight by altering the

diet and(or) metabolism of sows. Swine are born with extremely

limited energy stores in the form of adipose tissue lipids (Mersmann,

1974). Ramsay et al. (1987) reported that central endocrine

regulation has an important function in the development of porcine

fetal adipose tissue metabolism. On d 45 of gestation, spinal

cautery or decapitation was performed in fetuses within one horn,

with the other horn serving as a control. Fetuses were removed on d

110 of gestation and subcutaneous adipose tissue was incubated to

assess metabolic responses to insulin on glucose oxidation or

lipolytic response induced by norepinephrine (NE) bitartrate and

measured by glycerol release. Only decapitated fetal adipose tissue

demonstrated an insulin stimulation of glucose oxidation and

lipogensis. However, NE stimulated lipolysis in fat from cauterized

and control, but not decapitated fetuses. Ramsay et al. (1987)

concluded that the absence of a NE-stimulated lipolytic response in

decapitated fetal tissue demonstrated that central endocrine

regulatory factors control fetal adipose tissue lipolysis.

Furthermore, they indicated that fetal pig adipose tissue is a

consequence of beta-receptor development in response to systemic

hormones controlled by central endocrine regulatory factors. The

inability of spinal cautery to alter glucose metabolism or

responsiveness to insulin stimulation in fetal adipose tissue

indicated that central neural mechanisms did not affect glucose

metabolism in fetal adipose tissue. Rather, changes in adipose

tissue of decapitated pigs reflected changes in fetal endocrinology.

Kasser et al. (1983) suggested that the absence of or decrease in

insulin-anatognistic hormones in decapitated pigs during gestation

permitted development of insulin sensitivity in fetal adipose tissue.

Furthermore, Walton and Etherton (1986) reported that GH antagonized

insulin sensitivity of swine adipose tissue. Martin et al. (1985)

found that decapitatation decreased GH concentrations which would

alter insulin responsiveness or receptor concentrations. Either

factor resulted in increased body lipid deposition in decapitated

fetuses (Kasser et al., 1983). Therefore, promotion of fetal lipid

storage may depend on the ability to stimulate or inhibit release of

central endocrine regulatory factors (Ramsay et al., 1987). This

fact should be considered when attempting to increase fetal energy

reserves during gestation.

In a series of experiments, Cline and co-workers reported that ad

libitum feeding with or without 5% soybean oil from d 105 of

gestation until parturition did not alter survival of piglets from

birth to weaning (Massarotti and Cline, 1984; Sterling and Cline,

1986; Lima and Cline, 1987). However, ad libitum feeding and soybean

oil inclusion in the diet increased milk production and heavier

weaning weights of piglets were sometimes detected. Cromwell et al.

(1982) increased feed intake (1.82 to 3.18 kg/d) of sows from d 90

until term and reported increased (+.05 kg) average piglet weight and

greater piglet survival (85.4 vs 87.8%). However, Lewis and Reese

(1986) summarized results of numerous experiments and reported little

or no benefit of additional feed intake late in gestation on piglet

survival. They concluded that when piglet birth weights average

greater than 1.4 kg, they will not likely be increased by dietary

factors. Cromwell (1980) presented results of increased energy

intake of sows from over 5000 litters indicating that piglet birth

weights increased by approximately .02 and .05 kg in gilt and sow

litters, respectively for each additional .05 kg of feed consumed.

Although it is desirable to increase birth weights of piglets, it

is necessary to increase sow feed consumption greatly to elicit small

increases in piglet birth weights. Most researchers recommend

pregnancy weight gains of 10 to 40 kg (O'Grady, 1985). Since, as

indicated previously, maternal weight gains during pregnancy are

deleterious during the subsequent lactation, an optimum level of feed

intake for both fetal and maternal weight gains should be used.

Feeding levels should be adjusted for age, body condition, ambient

temperature, housing, expected low average birth weights and current

feed costs. Diets for gestating sows (with BW of 140 kg) are

recommended to provide 25 Mega Joules (MJ) digestible energy (DE)/d,

e.g., 1.8 kg/d of a corn soybean meal diet.

In addition to providing more energy by increasing overall

consumption, researchers have reported beneficial effects of adding

fat to diets of sows during late gestation and(or) lactation (see

Pettigrew, 1981; Moser, 1985). Moser and Lewis (1980) summarized

results of numerous experiments, performed in the 1970s and 1980s,

which evaluated effects of fat added to sow diets on piglet

survival. Type of fat (corn oil, soybean oil, tallow, lard), level

of fat (2 to 40%) and number of days of feeding fat (5 to 135)

varied. As reviewed by Moser (1985), almost all data indicated that

fat content in colostrum and milk of sows was increased in sows fed

fat. Controversial results were obtained for piglet survival and

litter size at weaning. However, Moser and Lewis (1980) reported

little or no benefit of dietary fat on litter size at birth and

average piglet weight at birth or weaning when data from all

experiments were summarized and weighted for total number of litters

in each experiment. A 0.3 piglet increase in litter size at weaning

was detected. This 0.3 piglet/litter improvement would offset feed

cost increases due to dietary fat (Moser, 1985). However, Moser

(1985) emphasizes that some experiments detected little or no

response to dietary fat on litter size. He also indicated that the

type of dietary fat is relatively unimportant although animal fats

are usually less expensive. Herds with 80% piglet survival will have

increased piglet survival more often than herds with greater than 80%

preweaning survival rates. Cieslak et al. (1983) reported that

survival rates were 10% higher for piglets that weighed 700 to 1100

gm at birth when sows were fed gestation diets supplemented with

fat. Survival of piglets that weighed greater than 1100 gm was not

affected significantly by maternal diet. Results of other studies

confirmed that there is a greater response to added fat under

conditions that predispose to low birth weights (Seerley et al.,

1974; Boyd et al., 1978a; Seerley et al., 1981). Based on previous

results, Moser (1985) suggested that a minimum of 7.5% fat should be

added and indicated that little advantage is afforded with more than

15% fat. Furthermore, he stated that it is more beneficial to add

fat either in late gestation or lactation rather than during both

periods. This is in agreement with Pettigrew (1981) who suggested

that at least 1 kg of fat should be fed to the sow by feeding a diet

with 10% added fat for at least 7 d prior to farrowing.

Fat supplementation can enhance piglet survival in at least two

ways. First, fetal liver glycogen is elevated (Boyd, 1978b; Seerley

et al., 1974) and piglets tend to have higher blood glucose

concentrations for up to 24 h after birth (Seerley et al., 1974; Boyd

et al., 1978b; 1981; Parsons, 1979). Second, greater milk yield and

milk fat resulting from maternal dietary fat consumption provide a

greater opportunity for newborn piglets to obtain energy (see Moser,

1985). Thus, providing fat to sows during late gestation may be a

practical way to enhance survival.

Effects of Manipulating Sow Metabolism During Gestation on Piglet
Survival and Growth

In addition to altering dietary components, changes in sows'

metabolism have been induced to make nutrients more available to the

fetus. Hausman et al. (1982) induced diabetes in sows on d 78 of

gestation or fasted sows during the last 20 d of gestation. Piglets

from diabetic and fasted sows had more subcutaneous adipose tissue,

but fetal weights did not differ due to treatment. Kasser et al.

(1982) reported that glucose concentrations in maternal blood were

higher after diabetes was induced. In addition, fetuses from

diabetic sows had higher insulin and triiodothyronine and lower

growth hormone and glucagon concentrations in circulation. Survival

rates of piglets fasted for 47 to 60 h were greater in litters of

diabetic or fasted sows in comparison to litters of control sows.

Piglets from sows with diabetes induced by streptozotocin were

reported to have higher liver glycogen and lipid content in

comparison to controls (Ezekwe et al., 1984). Ezekwe et al. (1984)

also reported a positive correlation between the severity of maternal

diabetes and body fat of piglet carcasses. Replacement of 15% of the

metabolizable energy in diets of sows with either 1,3 butanediol or a

mixture of acetate and lactate (1:1 molar) during the last 24 h of

gestation did not alter piglet birth weight (Spence et al., 1985).

However, increased liver glycogen, total glucogen/gm BW and ability

to maintain blood glucose levels during a 36 h fast were detected in

piglets from sows fed either the synthetic ketogenic, butanediol, or

the actetate-lactate diets.

Spence et al. (1984a) examined effects of exogenous growth

hormone injected (sc) at a dose of 24 IU/d from d 100 of gestation

through d 21 of lactation. No difference was reported for glucose or

insulin in maternal plasma, but plasma free fatty acids in GH-treated

sows were elevated two to 2.5-fold. Spence et al. (1984b) reported

that glycogen levels were higher, but not significantly, in piglets

from sows injected with GH. However, fasted piglets from sows

treated with GH had higher plasma glucose. Fat concentrations in

colostrum increased on d 13 of lactation for sows treated with GH,

but were not different on d 20. Milk yields at 2 wk were not differ-

ent due to treatments, but were 15% higher at 3 wk of lactation for

sows receiving GH. Feed intake decreased by 22% and backfat loss was

greater for GH-treated sows. Spence et al. (1984b) suggested that

the adverse effect of GH may be due to lipolytic effects. Kveragas

et al. (1986) injected (sc) GH at a dose of 20 IU/d for 21 d prior to

parturition. Both glucose and insulin in plasma were elevated in

treated sows, suggesting insulin resistance. Free fatty acids were

also greater in plasma of sows receiving GH. Piglets from treated

sows had greater concentrations of blood glucose following fasting,

more total body lipids at birth and increased concentrations of free

fatty acids in plasma. No differences due to treatment were detected

for birth weight, number born alive or dead or survival of piglets to

d 21 of lactation. Kveragas et al. (1986) reported that a

diabetogenic state was induced in gestating sows as indicated by the

higher concentrations of plasma glucose. Furthermore, a treatment x

diet interaction was detected, so the authors indicated that dietary

energy source should be considered when evaluating effects of

exogenous GH. Injection of GH can be fatal in sows. One sow died 3

d after treatment, while three sows died within hours after or before

parturition and exhibited respiratory distress. Thus, the proper

dosage of GH for sows during gestation and lactation warrants further


Control and Initiation of Parturition

There are numerous reviews concerning parturition in swine

(Dziuk, 1979; First et al., 1982; Taverne, 1982; Bazer and First,

1983; First and Lohse, 1984; Guthrie, 1985) and factors influencing

piglet survival (Leman et al., 1979; Dziuk, 1979). Similar to other

reproductive processes, events at parturition can reduce potential

litter size. At least six percent of piglets are born dead (Randall,

1972; Sprecher et al., 1974; Leman et al., 1979). Furthermore, the

last piglet in each uterine horn has less than a 50% chance of

survival, according to Bevier and Dziuk (1976). Leman et al. (1979)

indicated that 20 to 25% of piglets died before weaning and that most

of this loss occurred during the first 3 days of the piglets' lives.

Furthermore, the incidence of stillbirths and perinatal deaths

increased as litter size increased and as the duration of parturition

increased (Leman et al., 1979). Randall (1972) concluded that piglet

survival at birth was mainly influenced by fetal hypoxia. Temporary

hypoxia during farrowing may cause permanent damage, since English

and Smith (1975) reported that piglets dying before 21 days post

partum had higher blood lactate at birth than those that survived.

By understanding mechanisms controlling parturition, ways of reducing

neonatal death losses may be developed (First et al., 1982; Bazer and

First, 1983). For example, more precise methods of pharmacologically

inducing parturition may be developed. The presence of an attendant

during parturition to assist with difficult deliveries (Hammond and

Matty, 1980) or to resuscitate piglets (Milosavljevic et al., 1972)

has been reported to save up to one additional piglet per litter. In

addition, some potential benefits of controlling time of paturition

precisely are: (1) more efficient use of farrowing facilities; (2)

occurence of births during regular working hours; (3) efficient

cross-fostering of piglets; (4) better synchronization of estrus in

sows after weaning, and (5) more uniform age and weight of piglets in

litters (Guthrie, 1985).

Hormonal changes leading to birth involve final maturation of the

fetus, termination of pregnancy, expansion of the birth canal,

synthesis of milk and the ability to eject milk (Bazer and First,

1983). Because pregnancy is maintained in the presence of functional

CL in pigs, regression of CL and loss of progesterone is the primary

requirement for parturition (First, 1979; First et al., 1982).

However, the exact nature of the initial signal for parturition and

its mode of action are not certain (First et al., 1982). It is

somewhat ironic that a product (estrogen) of the concepts is

required for establishment and maintenance of pregnancy (Bazer and

Thatcher, 1977); a product of the concepts also initiates

parturition (First et al., 1982). Cortisol clearly causes increased

PGF resulting in regression of CL. How this is accomplished has not

been determined but a proposed mechanism will be reviewed later

(First et al., 1982).

The fetal brain is required for initiation of parturition, since

decapitated (Stryker and Dziuk, 1975; Coggins and First, 1977) or

hypophysectomized (Bosc et al., 1974) pig fetuses will not initiate

parturition. Furthermore, increased glucocorticoid production by the

fetal adrenal is crucial to induction of parturition. Adrenal

atrophy in fetal pigs due to hypophysectomy (Bosc et al., 1974) or

decapitation (Stryker and Dziuk, 1975) prevents parturition at term.

In sheep, fetal adrenalectomy blocks initiation of parturition (Drost

and Holm, 1968; Liggins, 1969). Administration of dexamethasone to

fetuses (North et al., 1973) or the sow (North et al., 1973; First

and Staigmiller, 1973; Huhn et al., 1976; Coggins and First, 1977;

Huhn et al., 1978; Huhn and Kiupel, 1979) induces parturition which

is followed by live birth and milk ejection. In addition, exogenous

ACTH administration to pig fetuses during the last 10 d of gestation

will cause fetal adrenal cortex hypertrophy and subsequent premature

parturition (Bosc, 1973). The porcine fetal adrenal undergoes

hyperplasia and this increases its ability to synthesize

glucocorticoids late in gestation (Bosc, 1973; Fevre et al., 1975;

Dvorak, 1972; Lohse and First, 1981). At d 113, the relative

proportion of cortisol in comparison to corticosterone production

increases in the fetal adrenal (Lohse and First, 1981). These data

and additional information reviewed by First et al. (1982) suggest

that the increased adrenal cortex weight, due to increased mitosis is

responsible for increased cortisol production from d 89 to 105.

However, the following increase in cortisol (d 105 to 113) is

believed to be due to changes in steroid ratio (First et al., 1982).

The mechanism by which the fetal brain controls the fetal adrenal

has not been defined (First et al., 1982). First et al. (1982)

suggested that control of adrenal cortisol production by the fetal

pituitary may not be the only way in which the fetal brain regulates

the initiation of parturition. First et al. (1982) indicated that,

via an action of the fetal brain, receptors for glucocorticoids must

be developed to allow cortisol to induce parturition. The placenta

may be the target organ for glucocorticoid action prior to

parturition as suggested for other species (Linzell and Heap, 1968;

Anderson et al., 1975; Flint et al., 1975). In rabbits, when fetuses

were removed on or before d 25 of gestation, parturition at term did

not occur and could not be induced by dexamethasone. However, when

fetuses were removed on or after d 26, placentae were delivered on

the expected day of parturition (d 32) (Chiboka et al., 1977). In

pigs, however, live or dead fetal tissue must be present for

pregnancy to be maintained (Chiboka et al., 1976). Whatever the

precise mechanism is, it is believed that glucocorticoids induce

parturition by increasing PGF secretion from the uterus which results

in luteolysis (Nara and First, 1978; 1981a). How fetal cortisol

increases uterine PGF synthesis and how the fetal brain regulates the

adrenal are not known.

Based on available evidence, First et al. (1982) proposed the

following sequence of events which lead to parturition in pigs. The

fetal hypothalamus stimulates ACTH production from the fetal

pituitary. Then either a hypothalamic factor or ACTH stimulates the

adrenal cortex to release cortisol. Cortisol or another factor of

hypothalamic origin then acts on the placenta. By some mechanism,

the placenta stimulates uterine PGF production. This may involve

increased 17-alpha-hydroxylase due to cortisol stimulation and

hydroxylase-induced estradiol production which in turn increases PGF

production. Prostaglandins can increase posterior pituitary oxytocin

secretion (Ellendorff et al., 1979) or stimulate progesterone and

relaxin secretion from the CL (Sherwood et al., 1976). In addition,

PGF increases oxytocin receptors allowing for an increased pulsing of

oxytocin and PGF. Then, the combined effects of oxytocin, PGF and

declining progesterone result in myometrial contractions, while

relaxin dilates the cervix to facilitate delivery of the piglets. It

is possible that the maternal central nervous system provides the

final signal for initiation of parturition in sows (Guthrie, 1985).

Once signalled to begin, parturition proceeds via coordinated

rhythmic contractions of uterine smooth muscle, involuntary

contractions of abdominal muscles and softening or opening of the

birth canal (First and Lohse, 1984). In pigs, myometrial activity

during late gestation consists of irregular episodes of prolonged

activity in those uterine segments containing a fetus, while those

parts of the uterus which are unoccupied by a concepts are

relatively inactive (Taverne et al., 1979a). Oxytocin is low (1.3

uU/ml plasma) at this time (Forsling et al., 1979), although oxytocin

receptors are present in the pig uterus during late gestation (Soloff

and Swartz, 1975). Not until 9 to 4 h before birth, does myometrial

activity increase in all parts of the uterus. This increase in

activity of the myometrium occurs when concentrations of oxytocin

increase in peripheral plasma. The release of oxytocin seems to be

related to a lower progesterone:estrogen ratio, due to decreasing

progesterone and increasing estrogen (Forsling et al., 1979). In

addition, the frequency of uterine contractions during delivery is

positively correlated with oxytocin concentrations (Taverne et al.,

1979a). First and Lohse (1984) indicated that uterine contractions

are likely initiated by increased intracellular calcium in myometrial

smooth muscle and by formation of gap junctions. Gap junctions

provide low resistance coupling for communication between cells

(Peracchia, 1980). An increase in intracellular free calcium causes

formation of a calcium-calmodulin complex. This complex binds to and

activates myosin light-chain kinase, which phosphorylates myosin, and

thus allows myosin to interact with actin and cause contraction

(Adelstein and Eisenberg, 1980). Although the factors regulating

calcium flux and gap juntion formation are not well understood (Bazer

and First, 1983), prostaglandins may play a role. Prostaglandins

liberate calcium from intracellular binding sites and thus effect

tonic contraction. In addition, prostaglandins cause slow membrane

depolarization that increases the frequency of phasic contractions

(Liggins, 1979). The PGF may cause release of oxytocin in swine as

well (Ellendorff et al., 1979). As reviewed by Bazer and First

(1983) and First and Lohse (1984), oxytocin lowers the threshold for

initiation of potential activity and also directly influences the

rate of calcium influx.

Multiple implantation of intrauterine catheters (Zerobin, 1968)

or surface electrodes in swine (Taverne et al., 1979b) provided

evidence that myometrial contractions in all segments of the uterine

horn are synchronized during delivery. These contractions were

initiated at both ends of the uterine horn and subsequently

propagated in either a tubocervical or cervicotubal direction.

Contraction waves returned in an opposite direction upon reaching the

end of the horn (Taverne et al., 1979c; Taverne, 1982). However,

cervicotubal contractions stopped when the horn was empty and seemed

related to the presence of piglets close to the cervix. Taverne

(1982) suggested that this mechanism shortens the distance travelled

by succeeding piglets and prevents piglets from "piling up" near the

cervix. At the cervix, placental attachments could be dissociated

and cause fetal anoxia and death. In support of this idea, piglets

dying during birth are predominantly the last ones born (Randall,

1972) and usually the last piglet in each horn (Bevier and Dziuk,

1976). Although myometrial smooth muscle contractions are

synchronized, expulsion of piglets from the two uterine horns is

random (Dziuk and Harmon, 1969; Taverne et al., 1977).

In addition to uterine contractions, distension and softening of

cervical connective tissue must occur (Bazer and First, 1983; First

and Lohse, 1984). Little is known about the mechanism of cervical

softening in domestic animals (Bazer and First, 1983). However, the

involvement of relaxin on cervical distension before parturition has

been demonstrated in swine (Kertiles and Anderson, 1979; Nara et al.,

1982). Sherwood (1982) reviewed structure, source and function of

relaxin in swine. The principle source of relaxin in swine is CL

(Fevold et al., 1930; Hisaw and Zarrow, 1948; Nara et al., 1982).

The CL store and concentrate relaxin throughout pregnancy (Anderson

et al., 1973a,b; Kendall et al., 1978; Fields and Fields, 1985), but

this is not dependent upon conceptuses, since CL of hysterectomized

gilts also accumulate relaxin through d 124 (Anderson et al.,

1982a). Unmated gilts with CL prolonged by exogenous estrogen

(pseudopregnant) also produce and accumulate relaxin in their CL

(Anderson et al., 1973b). Relaxin is secreted or released by luteal

cells of CL (Belt et al., 1971; Kendall et al., 1978) as sustained

surges starting approximately 2 d prior to birth. Maximum values for

relaxin in maternal plasma are at 14 to 22 h prior to parturition

(Sherwood, 1982). The mode of relaxin release is not clear, but

there is evidence that secretion or release of relaxin is induced by

PGF directly (Sherwood et al., 1979; Nara and First, 1981b) or

indirectly through PGF-induced oxytocin release and subsequent

oxytocin action on relaxin secretion. There is some evidence that

timing of relaxin release may be dependent upon lifespan of CL

(Anderson et al., 1982a). Felder et al. (1986) reported that abrupt

shifts in both relaxin and progesterone secretion on d 111 to 113 in

hysterectomized and pregnant gilts. Thus, conceptus/fetal tissues

are not required for the surge of relaxin release coincident with

decreased progesterone. They also suggested that autonomous

regulation within the ovary or from the central nervous system and

pituitary gland may control relaxin and progesterone secretion. In a

recent study, Taylor and Clark (1988) provide evidence that relaxin

inhibits its own release from porcine luteal cells. Thus,

autoregulation may restrain relaxin secretion until the initiation of

labor. In a previous study, Taylor and Clark (1987) suggested that

prostacyclin may act alone or in combination with other prostanoids

to cause relaxin release in the pregnant pig.

Dilatation of the cervix is prematurely induced by injection of

relaxin in late pregnant gilts (Kertiles and Anderson, 1979).

Furthermore, removal of CL on d 110 of pregnancy results in difficult

delivery and a prolonged period of parturition. Relaxin replacement

to gilts without CL led to premature delivery, but short duration of

parturition with cervical dilatation (Kertiles and Anderson, 1979).

Zarrow et al. (1956) reported that relaxin caused dilatation of the

cervix, increased water content and depolymerization of cervical

glycoproteins when injected into estrogen-treated ovariectomized

nonpregnant gilts. Nara et al. (1982) found that in the absence of

the source of relaxin, delivery was prolonged and there was a high

frequency of stillborn piglets. Exogenous relaxin decreased the

duration of parturition and increased the frequency of live births

comparable to that for normal ovarian intact gilts. In addition to

inducing cervical distension, relaxin may also prevent premature

labor during late pregnancy by suppressing uterine contractions

(Porter, 1979; Taverne et al., 1979b). In summary, relaxin is

involved in cervical distensibility in swine. However, the effects

of relaxin are dependent upon prior exposure to estrogen or

progesterone followed by estrogen (Sherwood, 1982). The possibility

exists that estrogen, progesterone, prostaglandins, oxytocin and

other hormones are also involved in cervical distension (First and

Lohse, 1984; Sherwood, 1982).

Since pharmacological induction of parturition can benefit

management schemes and potentially decrease losses at birth,

protocols have been developed. Administration of exogenous PGF or

one of its analogues, alone, or in combination with oxytocin, is the

most effective and widely accepted method used today.

Chantaraprateep et al. (1986) reported a high number of live piglets

per litter in sows given 175 ug estrumate, a PGF analogue, on d 111

to 113 of gestation (im or iv). The authors attributed the live

piglet advantage to the presence of an attendant. Although birth

weights were less for treated sows, differences were not detected for

individual piglet weaning weights. As First and Bosc (1979)

indicated, the advantage of PGF or its analogue is that these

compounds induce normal parturition, as well as, parturition-related

events including initiation of lactation and expulsion of the

placenta (First and Bosc, 1979). When administered after d 110 of

gestation, most experiments show no significant difference between

treated and control sows in duration of labor, frequency of piglets

stillborn, birth weight and survival to weaning or weaning weight, as

reviewed by First et al. (1982). There is a slightly greater birth

weight for piglets from control sows, however. The doses of PGF used

in these studies ranged from 7.5 to 12.5 mg/sow (im) (Ehnvall et al.,

1976; Backstrom et al., 1976; King et al., 1979; Hagner et al., 1979;

Huhn et al., 1980). It is critical that parturition not be induced

prior to d 110 since birth weights and piglet survival will be

significantly reduced (First and Lohse, 1984).

When oxytocin is given after injection of PGF, delivery occurs in

approximately 27.2 + 2 h and variation in time of delivery is

significantly reduced compared to when PGF is given alone (Welk and

First, 1979). Dial et al. (1987) attempted to determine the optimal

dose and interval between PGF and oxytocin. They based their study

on the fact that although oxytocin administration 16 to 24 h after

PGF administration is effective in initiating parturition within 3 to

6 h, uterine intertia increases (Holtz et al., 1983; Dial, 1984; Welp

et al., 1984). Thus, increased perinatal mortality can result due to

the interrupted delivery. In one trial, sows were given 20 U

oxytocin 16, 20, or 24 h after PGF (Lutalyse) injection (10 mg on d

112, 113 or 114 d of gestation). There was a tendency for

intrapartum complications to be less frequent with the greatest

interval between injections. However, the results indicated that

synchrony of farrowing was greatest when the interval between

injections was reduced. Similarly, although degree of synchrony

increased with 30 U of oxytocin in comparison to lower doses, the

incidence of problems requiring manual intervention also rose.

Another approach to induce parturition in swine involves the use

of epostane, a competitive inhibitor of 3B-hydroxysteroid

dehydrogenase (3B-HSD) (Martin et al., 1987). On d 109 of gestation,

sows received either oral doses or subcutaneous injections of

epostane on a body weight basis. Farrowing occurred 31 to 77 h

later, depending on the dose. With the higher doses, farrowing

occurred within 31 to 33 h. The duration of farrowing, proportion of

piglets born live, birth weights, weaning weights, proportion of

piglets born live that were weaned and interval from weaning to first

postpartum estrus were not influenced by treatment. The major

influence of epostane was suggested to be through its action on

3B-HSD and subsequent removal of progesterone. Martin et al. (1987)

concluded that inhibitors of 3B-HSD could be successfully used to

induce farrowing without adversely affecting the sow or litter.

Another method to induce parturition has been reported by Guthrie

et al. (1987). Their method was developed in an attempt to reduce

the incidence of spontaneous farrowing. First et al. (1982) reported

that only 50 to 60% of sows actually begin delivery during a 10 h

interval on the expected day of parturition, because some sows farrow

before the scheduled injection of PGF. By feeding an orally active

progesterone agonist, altrenogest, early unscheduled farrowings can

be prevented (Varley et al., 1985). Therefore, Guthrie et al. (1987)

combined the advantages of altrenogest to delay parturition and PGF

to induce parturition. Oral administration of altrenogest twice

daily on four days beginning on d 109 or 110 prevented early

parturition in all sows. Injection of PGF had no adverse effects

except on lactation in 3 of 62 sows and did successfully induce

parturition. Therefore, Guthrie et al. (1987) concluded that the

combination of altrenogest followed by PGF allows better control of

the time of farrowing. However, they suggest addition of compounds

such as oxytocin or relaxin to their protocol may reduce variation in

time from administration of PGF to birth of the first piglet and

duration of parturition.

In summary, factors associated with parturition are being

elucidated and utilized to manipulate initiation of farrowing to the

advantage of the producer. However, additional experiments may lead

to more optimal doses of "inducing" agents and times for

pharmacological intervention.

Physiology and Characteristics of Lactation in Swine

As reviewed by Collier et al. (1984), a classic example of both

homeostasis and homoerhesis is the "redirection of nutrient flow that

occurs at parturition as site of maternal delivery of nutrients to

offspring shifts from placental transfer in the uterus to milk

synthesis, secretion and removal at the mammary gland." Lactation

provides piglets with nutrition and passive immunity against

pathogenic infections and is dependent upon: (1) supply and uptake of

metabolites from blood; (2) amount of secretary tissue present and

(3) stimuli for milk removal. Because all milk precursors are

derived from blood, blood flow to mammary glands as a precentage of

cardiac output increases during lactogenesis (Linzell, 1974) thereby

increasing nutrient delivery to and uptake by mammary glands. In

addition, adipose tissue adapts to demands of lactation under

homeorhetic regulation by mobilizing fatty acids (see Collier et al.,

1984). Four separate stages of lactation include: (1) mammary growth

(mammogenesis); (2) inititiation of copious milk secretion

(lactogenesis); (3) maintenance of milk secretion and (4) weaning

(mammary gland involution) (Hartmann et al., 1984).

Although it is essential that adequate fetal and prepuberal

mammary gland development occur, this review will only consider

pregnant and lactating sows. Hormones known to stimulate mammary

gland growth in swine are estrogens, progesterone and prolactin

(DeHoff et al., 1986); however, general metabolic hormones, e.g.,

insulin, thyroid hormones and corticoids, also affect mammogenesis.

Kensinger et al. (1982) indicated that the majority of mammary duct

growth of sows occurs prior to d 60 and Hartmann et al. (1984)

indicated this growth begins around d 45. Concentrations of

circulating progesterone and prolactin remain fairly constant, while

estradiol-17B increases as lobulo-alveolar development is initiated

and concentrations of mammary DNA and RNA increase (Hacker and Hill,

1972; Cowie et al., 1980; Kensinger et al., 1982). Lobulo-alveolar

development occurs between d 60 and 105, with the fastest rate of

mammogenesis occurring between d 75 and 90 (Kensinger et al., 1982).

Lyons (1958) indicated that estradiol, GH and adrenal steroids are

required for duct growth, while progesterone and prolactin, in

addition to the previous hormones, are required for lobulo-alveolar

growth, and prolactin and adrenal steroids are necessary for milk


Lactogenesis occurs in two stages: (1) stage I, which occurs

between d 90 to 105 in pigs, is associated with increased synthesis

of fatty acids, distension of alveoli and fat droplet accumulation in

epithelial cells and (2) stage II, extending from d 112 of gestation

through parturition, which is associated with increased

concentrations of RNA and additional increases in production of fatty

acids and CO2 (Kensinger et al., 1982). In sows, alveolar

secretions begin to accumulate about 2 d prior to parturition (Cross

et al., 1958) and, normally, mammary secretion can only be expressed

within a few hours prior to parturition. Lactogenesis (stage II) has

been described by numerous researchers as the time when milk can

first be expressed; however, as Hartmann et al. (1984) suggest,

analysis of lactose may be a more sensitive indicator for the onset

of lactation (Willcox et al., 1983). Two theories regarding the

mechanism for the onset of lactogenesis are: (1) mammary glands are

released from inhibition by progesterone or (2) positive stimuli of

prolactin and glucocorticoids act upon the mammary gland (Kuhn, 1977;

Fulkerson, 1979; Cowie et al., 1980). Administration of progesterone

delays onset of lactation until its withdrawal (Gooneratne et al.,

1979; Taverne et al., 1982). However, it is likely that progesterone

withdrawal at parturition is not solely responsible for lactogenesis,

but is coupled to stimulatory effects of prolactin and cortisol

(Hartmann et al., 1984). In addition, declining concentrations of

relaxin may be required for lactogenesis (Cowie, 1969), as reviewed

by Hartmann et al. (1984). Tucker (1985) stated that the minimum

requirement for lactogenesis involves secretion of prolactin,

adrenocorticotrophic hormone (which stimulates secretion of

glucocorticoids), and estrogens and decreased progesterone. Cortisol

induces differentiation of rough endoplasmic reticulum and the Golgi

appartus of epithelial cells during midpregnancy in mice, which

permits prolactin to induce synthesis of milk proteins (Tucker,

1985). Prolactin controls expression of the casein gene and other

genes, and its effects are amplified by insulin and glucocorticoids

and inhibited by progesterone (Tucker, 1985). Growth hormone appears

to act by partitioning available energy away from body tissues toward

milk production (Forsyth, 1986). Improper hormonal stimulation has

been implicated in development of mastitis-metritis-agalactia, which

results in partial or complete failure of lactation (Martin and

Threlfall, 1970).


Because alterations in milk composition cause changes in body

composition and growth of piglets (Williams, 1976), composition of

milk may be manipulated by dietary or pharmacological means to

advantageously influence piglet survival. Fat, protein and lactose

constitute approximately 60, 22 and 18%, respectively, of total

energy content in sow milk (Hartmann et al., 1984). Synthesis of

milk fat occurs in cytosolic secretary cells, and lipids accumulate

into droplets in vesicular cells of the basal cytoplasm which move to

the apex of cells and are then extruded (Cowie, 1984). Proteins in

milk are synthesized in endoplasmic reticulum, pass into the Golgi

apparatus where caseins undergo phosphorylation, form micelles within

vesicles and move to the cell apex for membrane fusion and subsequent

discharge of vesicle contents into the alveolar lumen (Cowie, 1984).

Lactose synthesis occurs in the Golgi in association with proteins.

The final step in lactose synthesis involves lactose synthase which

is made up of two proteins: (1) alpha-lactalbumin which is

synthesized in endoplasmic reticulum and (2) galactosyl-transferase

which is synthesized in the Golgi (Cowie, 1984). Lactose synthesis

produces equimolar H3PO4 intracellularly and, in order to

neutralize H+, net anion transport occurs (Jenness, 1986). Jenness

(1986) indicated that lactose synthesis and associated anion

transport regulate swelling of Golgi vesicles which, along with rate

of passage of vesicles through cells, determines volume of milk

(aqueous) secreted. Secreted protein is diluted to its final

concentration in Golgi vesicles and dilution of fat globules occurs

in alveoli (Jenness, 1986). Since volume is regulated by lactose

synthesis, fat and protein content are inversely correlated with

content of lactose in milk (Davies et al., 1983). Functions of

lactose and fat include supplying energy, while protein is a source

of amino acids for growth (Jenness, 1986). The B-l,4-galactosidic

link of lactose promotes calcium absorption and, since specific

enzymes are required to hydrolyze this bond, it may prevent

intestinal fermentation (Jenness, 1986).

Linzell et al. (1969) and Spincer et al. (1969) estimated that

mammary glands of sows utilize 50 and 31%, respectively, of glucose

entering mammary circulation and glucose is the major source of milk

lactose, glycerol and mammary CO2 (Linzell et al., 1969). Although

acetate has been demonstrated to be preferentially utilized as a

substrate, it is present in low concentrations in plasma of pigs, and

therefore, glucose utilization predominates (Linzell et al., 1969;

Spincer et al., 1969; Bauman et al., 1970). Furthermore, although

glucose and acetate can be use to synthesize fatty acids de novo,

fatty acids derived from plasma triglycerides are the major source of

fatty acids in milk (Hartmann et al., 1984). Free amino acids in

plasma give rise to milk proteins. More complete reviews of

biosynthesis of milk are available (Mepham, 1983; Larson, 1985;

Jenness, 1985).

Maternal antibodies secreted into colostrum and absorbed intact

by piglets during the first 24 h after birth (Lecce and Morgan, 1962)

are critical, since piglets are born agammaglobulinaemic.

Accumulation of immunoglobulins in colostrum occurs rapidly over the

last 2 d of pregnancy, and most of these are derived from maternal

serum (Bourne and Curtis, 1973). Colostrum contains trypsin

inhibitor (Jensen and Pedersen, 1979) which protects immunoglobulins

from intestinal proteolysis. Concentrations of all immunoglobulins

in mammary secretions fall rapidly during the first 24 h after

parturition (Curtis and Bourne, 1971). As reviewed in a separate

section, piglets are extremely susceptible to hypoglycemia (Edwards,

1972), so during the first few hours post-partum, intake of colostrum

is critical. Although colostrum contains 586 to 628 kJ energy/100 ml

(Salmon-Legagneur and Guegen, 1962), its energy is limited since it

contains immunoglobulins which will be absorbed intact and not

contribute to energy. Differences in composition of milk and

colostrum are due to the rapid decline in immunoglobulins and casein

during the first few days after parturition and increased

concentrations of lactose (Brent et al., 1973). Klobasa et al.

(1987) reported that concentrations of IgG, IgM and IgA in colostrum

declined by 50% during the first 12 h after parturition. Since

immunoglobulins, especially IgG, are the predominant proteins in

colostrum, total protein content during the first 12 h of lactation

also declined by nearly 50%. Total solids concomitantly declined

during the first day of lactation; however, fat and lactose content

increased at this time. At birth, composition of sows' colostrum was

25.6% total solids, 5.0% fat, 3.1% lactose, 15.7% total protein and

14.3% whey protein. In contrast, at 21 d of lactation, composition

of milk was estimated to be 18.7% total solids, 6.6% fat, 5.8%

lactose, 5.2% total protein and 2.8% whey protein (Klobasa et al.,

1987), which agrees with estimates of milk composition obtained on d

22 of lactation by White and Campbell (1984). Similar composition of

sows' milk (percent by weight) has been reported by others to be

81.2% water, 6.8% fat, 2.8% casein, 2.0% whey protein, 5.5% lactose

and 1.0% ash, which would supply approximately 102 kcal energy/g of

milk (Jenness and Sloan, 1970). Long (1961) estimated concentrations

(g/100g) of ash, fat, lactose, and protein in milk of sows with

values of .6, 8.2, 4.8, and 5.8, respectively.

Hartmann et al. (1984) indicated that with increased litter size

from highly selected sows and decreased neonatal deaths in

technologically sophisticated piggeries, "milk production is now one

of the most important factors limiting piglet growth and, ultimately,

pig production." Lewis et al. (1978) concluded that variability in

milk yield among sows accounts for approximately 33% of the

variability in weight gain of piglets. Milk yield is determined by

epithelial cell number and by secretary activity of cells (Forsyth,

1986). Regulation of milk yield is influenced by suckling intensity

of piglets postpartum (Blaxter, 1961) and possibly by prepartum

factors. For example, Elsley (1971) demonstrated that milk yield of

sows is related to litter size, but it is not clear whether this

regulation is entirely due to suckling and milk removal or whether

development during pregnancy may be involved (see Forsyth, 1986).

However, greater fetal numbers were not associated with additional

increases in mammary development (Kensinger et al., 1980). Milk

yield can be assessed by: (1) differences in piglet weights or sow

weights before and after suckling (Braude, 1954; Lewis et al., 1978);

(2) milking sows by hand or machine at intervals following oxytocin

injection (Hughes and Hart, 1935; Hartmann and Pond, 1960; Linzell et

al., 1969); (3) isotope dilution (Yang et al., 1980) and (4) isotope

transfer (See Oftedal, 1984). Estimates of daily milk yield of sows

between d 7 and 28 of lactation range from 5 to 13 kg/d (White and

Campbell, 1984; White et al., 1984a; Coffey et al., 1987). English

et al. (1982) and White and Campbell (1984) detected greatest milk

yield at approximately 3 wk of lactation. Yield during an 8 wk

lactation has been estimated to average 5.8 kg/d (Smith, 1952; Barber

et al., 1955; Smith, 1959; Hartman et al., 1962). Peak yields

occurred between the third and fifth wk and yields slowly declined by

the ninth or 10th wk (Allen and Lasley, 1960). Thus, early weaning

can result in the inability of sows to reach peak production. Yield

is also influenced by parity; yield increases from the first to third

lactation (Smith, 1959; Van Spaendonck, 1967). In addition, sows in

good condition produce more milk than thin sows even when thin sows

are fed extra energy during early lactation (Klaver et al., 1981).

As previously mentioned, milk yield is a linear function of rate of

blood flow to mammary glands (Kronfeld, 1969). Increasing

photoperiod from 8 to 16 h per d increased milk production by 20%,

litter weight at 21 d by 13% and piglet survival by 10% (Mabry et

al., 1982).

Suckling is required to maintain lactation; unsuckled mammary

glands involute rapidly (Cross et al., 1958; Martin et al., 1978).

Suckling maintains lactation, in part, through its stimulation of

prolactin (Bevers et al., 1978), a necessary hormone for maintenance

of lactation (Schams, 1976). During lactation, concentrations of

prolactin in serum are between 3 and 8 ng/ml compared to 1 to 2 ng/ml

after weaning (Bevers et al., 1978; Mulloy and Malven, 1979).

Administration of bromocryptine, a dopamine agonist, inhibits

prolactin secretion and prevents milk secretion in swine. In

addition to its requirement for milk synthesis, suckling behavior


is required for milk ejection through stimulation of oxytocin

release. It is generally accepted that a discrete release of

oxytocin is responsible for milk flow at each suckling (Cross,

1977). However, relaxin may help regulate the duration of milk

release (Afele et al., 1979). Milk is ejected from alveoli by

myoepithelial cell contractions in response to oxytocin which leads

to increased intramammary pressure (Cowie, 1984). Estimates of

suckling frequency vary from 12 to 18 times per 24 h (Hughes and

Varley, 1980) to once every 55 to 65 min for 2 to 3-wk-old piglets

(Whittemore and Fraser, 1974).

Influence of Nutrition on Sow Performance During Lactation

Nutrition of sows during lactation is designed to maximize milk

production and minimize body weight and fat losses. However,

nutrition during lactation also impacts short- and long-term post-

weaning reproductive efficiency (see Britt et al., 1985). Nutrient

requirements and composition of the diet during lactation are

influenced by nutrient intake during gestation, weight and condition

of the sow at onset of lactation, as well as milk yield and

composition which are influenced by litter size, seasonal or

temperature variations and stage and duration of lactation. In

addition, O'Grady (1985) indicated that energy and protein allowances

for lactation must be related to expected life of sows and, if

multiple litters are expected, allowances must permit maintenance of

body fat reserves. If protein intake is inadequate (O'Grady and

Hanrahan, 1975), sows will draw upon their own reserves to meet

demands for secretion of milk proteins and consequently lose

weight. Often, long term effects of nutrition and weight loss

during lactation on sow productivity have not been investigated.

Numerous reviews have addressed nutritional needs of lactating sows

(Cole, 1982; O'Grady, 1985; Aherne and Kirkwood, 1985; Britt, 1986).

Relative to those of pregnancy, nutritional requirements to meet

the demands of milk production are high (Cole, 1982). Energy

consumption of sows during lactation influences: (1) litter size

weaned; (2) days to estrus after weaning and (3) subsequent litter

size (Lewis and Reese, 1986). Evaluating effects of increased

energy, specifically, is difficult in numerous experiments, because

greater intakes of all nutrients occurred. In addition, carry-over

effects of energy and(or) feed intake during gestation influence

energy intake during lactation and vice versa (Salmon-Legagneur and

Rerat, 1962). Sows that consume the most feed during pregnancy lose

the most weight during lactation (Salmon-Legagneur and Rerat, 1962;

Sterling and Cline, 1986). However, if feed intake is increased

during the latter part of pregnancy (d 105 of gestation through

lactation), lactational performance may be improved without having a

detrimental effect on sow weight loss (Sterling and Cline, 1986).

Mahan and Mangan (1975) provided evidence of an interaction between

protein level fed throughout gestation and voluntary intake of gilts

during lactation. If high levels of dietary protein (18%) were

supplied during lactation, feed intake during lactation was

independent of dietary protein level fed during gestation. However,

if a -12% crude protein diet was fed during lactation, feed intake

during lactation was linearly related to level of protein fed during

gestation. Results obtained over three or four reproductive cycles

by O'Grady and Hanrahan (1975) confirmed those findings. Results of

a larger study indicated that it may be most beneficial to production

to feed diets (13% crude protein and .6% lysine) which provide

similar protein levels during pregnancy and lactation (Greenhalgh et

al., 1977).

Quality and quantity of milk produced are the two major factors

influencing dietary protein requirements during lactation. Cole

(1982) indicated that milk yield is affected by litter size and stage

of lactation which, in turn, affects the amount of energy required

for lactation. Changes in milk composition are most evident early in

lactation when production changes from colostrum to milk. Protein

content of milk declines over the first 10 to 14 d and fat production

increases during the first 3 d. Thereafter, secretion of milk

protein gradually increases and fat synthesis decreases (see Cole,

1982). While milk yield influences the amount of nutrients required,

plane of nutrition influences milk production especially during the

first 3 wk of lactation (Lodge, 1972). O'Grady (1985) indicated that

the greatest requirement for dietary protein is within the first 2 wk

of lactation. Research comparisons and recommendations have been

complicated because length of lactation periods have varied from two

to eight weeks in available literature. O'Grady (1985), for example,

calculated dietary protein requirements after consideration of

protein in milk and milk yield, as well as estimated feed intake and

stage of lactation.

Litter size weaned was increased only slightly when metabolizable

energy (ME) [>14 Mcal/(sow d)] intake during lactation was

increased for sows ranging in parity from 1 to 4 and lactation

periods of 4 to 8 wk (Elsley et al., 1968; Hitchcock et al., 1971;

O'Grady et al., 1973; Adam and Shearer, 1975; Reese et al., 1982a;

King and Williams, 1984a). When litter size is increased, greater

energy consumption by sows may be accompanied by greater milk yields

due to sows consuming higher levels of energy. Shurson et al. (1986)

reported that greater energy intake due to inclusion of 10% fat in

the lactation diet increased litter size (8.9 vs 8.4) and litter

weight (56.1 vs 48.9 kg) at 21 d of age. Sow milk yield was

increased by 13.1% for sows consuming fat and this may have been

responsible for increased litter size at weaning. Sterling and Cline

(1986) indicated that ad libitum feeding in late gestation (d 105 to

farrowing) followed by ad libitum feeding of a diet containing 5%

soybean oil during lactation improved milk yield. Feed intake during

lactation and birth weight of piglets were not influenced by ad

libitum feeding of diets containing soybean oil. In spite of

improved milk yields for sows fed ad libitum diets containing soybean

oil, number of piglets born alive was less for ad libitum fed sows.

Nelssen et al. (1985a) utilized 146 primiparous sows to

investigate effects of intakes of 10, 12, or 14 Meal ME/(sow d)

during a 28 d lactation. Sow weight and backfat loss decreased

linearly as energy intake increased. No differences in litter size

at d 14 or at weaning were detected. However, on d 14, pig weights

increased (P < .05) and litter weights tended (P = .13) to increase

linearly as energy intake of sows increased. At weaning, pig weights

and litter weights increased (P < .05) as energy intake of sows

increased. The authors concluded that intakes of 10 Meal ME/(sow

d) decreased sow and litter performance, while little difference in

performance was detected from feeding 12 or 14 Meal ME/(sow d).

Therefore, NRC (1979) recommendations of 12.8 Meal ME/(sow d)

appear to be adequate for lactating sows. However, carry-over

effects to subsequent farrowings were not determined. Dulohery et

al. (1986a,b) reported effects of increased energy during lactation

[8 vs 16 Meal ME/(sow d)] and subsequent postweaning [5.75 vs

11.5 Meal ME/(sow d)] intervals on sow and litter performance in

first and second parity sows during 21 d lactations. Litter weight

gain and pig weaning weights were increased and sow weights and

backfat losses were reduced by increased dietary energy.

Seerley et al. (1974) were the first to report that

supplementation of sow diets with lipids (corn oil) had beneficial

effects on piglet performance. Piglet survival to d 21 was increased

by adding corn oil to sows' diets from d 109 through parturition due,

in part, to increased glycogen stores for piglets at birth.

Subsequently, numerous researchers reported that adding fat to sow

diets during late gestation and early lactation improved survival

rates for newborn pigs (Moser and Lewis, 1980; Pettigrew, 1981;

Moser, 1985). Milk yield and(or) composition is improved by addition

of lipids to sow diets (Friend, 1974; Kruse et al., 1977; Coffey et

al., 1982). There is evidence that addition of fat to sow diets

confers advantages on piglet survival that are not obtained when

equivalent amounts of ME are from other energy sources. However,

piglet survival in response to fat supplementation has been varied

and somewhat inconsistent.

Moser (1984) reported that addition of fat to lactation diets of

sows increased ME intake an average of 6.5% over standard lactation

diets. In sow herds which consume an average of 3.0 kg/d of a

corn-soybean meal diet, fat supplementation may be a desirable method

of increasing energy intake by lactating sows (Cox et al., 1983a).

However, for sow herds in which consumption is greater than 4.5 kg/d,

fat supplementation (7% added tallow) resulted in only a slight

improvement in litter weights at weaning (Moser et al., 1985a,b).

Thus, addition of dietary fat during lactation is most effective

during periods of hot weather or when feed intake is reduced (Cox et

al., 1983a). Moser (1985) summarized results of numerous experiments

involving supplementation of diets with fat before and(or) after

farrowing. Type of fat varied from corn and soybean oils to tallow

and lard, while level of fat varied from 2 to 40% and the duration of

feeding fat varied from 5 to 135 d. Moser (1985) indicated that

dietary fat was more beneficial during either gestation or lactation,

rather than during both intervals in terms of number of piglets

weaned per litter. Dietary fat during lactation did not benefit

piglet survival as much as when it was consumed during gestation only

or during both gestation and lactation. Piglet survival appears to

be enhanced by increased fat content of colostrum and(or) milk

following dietary fat supplementation for sows.

Nelssen et al. (1985b) indicated that replacement of cornstarch

with tallow was not beneficial in diets for primiparous sows during

lactation when energy intake was restricted. Unlike situations in

poultry where an extracaloric effect of fat was reported (Sibbald et

al., 1962; Vermeersch and Vanschoubroek, 1968; Jensen et all, 1970;

Gomez and Polin, 1974; Whitehead and Fisher, 1975), lactational

performance of sows on a low level of energy intake was not improved

by replacement of carbohydrate with fat on an isocaloric ME basis.