Effect of day of the estrous cycle at the initiation of a timed artificial

insemination protocol on reproductive responses in dairy heifers1

F. Moreira*, R. L. de la Sota†, T. Diaz†, and W. W. Thatcher*,2

*Department of Dairy and Poultry Sciences, University of Florida, Gainesville 32611-0920;

†Facultad de Ciencias Veterinarias, Universidad Nacional de la Plata, Argentina; and
‡Facultad de Ciencias Veterinarias, Universidad Central de Venezuela, Maracai, Venezuela

ABSTRACT: Our objectives were to identify stages of
the estrous cycle at which initiation of a timed artificial
insemination (Ovsynch/TAI) protocol may reduce preg-
nancy rates and to monitor ovarian follicle dynamics
and corpus luteum development after initiation of the
Ovsynch/TAI protocol at different stages of the cycle.
Cycling Holstein heifers (n = 24) were injected twice
with prostaglandin F2α to induce estrus and were
scanned by ovarian ultrasonography to determine the
day of ovulation (d 0). Heifers were assigned to initiate
the Ovsynch/TAI protocol at d 2 (n = 5), 5 (n = 5), 10
(n =4 ), 15 (n = 5), or 18 (n = 5) of the cycle. The Ovsynch/
TAI was initiated with an injection of gonadotropin-
releasing hormone agonist followed 7 d later with an
injection of prostaglandin F2α. At 36 h after injection
of prostaglandin F2α, heifers were injected with gonado-
tropin-releasing hormone agonist and inseminated 16
h later. Heifers were scanned daily during the Ovsynch/
TAI protocol and every other day after insemination
until 16 d later. Blood samples were collected daily
starting at the 1st day heifers were scanned and contin-
ued until 16 d after insemination. Initiation of the Ov-
synch/TAI protocol at d 15 of the estrous cycle caused
heifers to ovulate prior to insemination. A shortened
return to estrus (< 16 d) was caused by ovulation failure
to the second gonadotropin-releasing hormone injec-
tion, by incomplete regression of the corpus luteum,
and by short life-span of the induced corpus luteum.
Day of the cycle in which the Ovsynch/TAI protocol
is initiated affects dynamics of follicular development,
plasma progesterone profiles, and occurrence of prema-
ture ovulation. Size of the pre-ovulatory follicle was
associated positively with subsequent progesterone
concentrations following insemination.

Key Words: Heifers, Artificial Insemination

2000 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2000. 78:1568–1576

Introduction dom during the estrous cycle, which causes ovulation or

A timed artificial insemination protocol (Ovsynch/
TAI) was devised so that cows could be inseminated
without estrus detection (Pursley et al., 1995; Schmitt
et al., 1996b). Such a protocol consists of an injection of
gonadotropin-releasing hormone (GnRH) given at ran-
1
Authors express their appreciation to Dale Hissem and to the
staff at the Dairy Research Unit for managing the experimental
heifers. Our gratitude is extended to Shane Brooks and Jennifer
Trout for their help during the bleeding and ultrasonography. Many
thanks go to Jesse J. Johnson for running the radioimmunoassays.
Lutalyse and Receptal were donated by Pharmacia-Upjohn, Kalama-
zoo, MI, and Hoechst-Roussel, Sommerville, NJ. This research was
supported by the USDA-Cooperative States Research Service-Bina-
tional Agricultural Research and Development Grant No. 94-34339-
1212 and by the Florida Milk Checkoff Program.This is Florida Agric.
Exp. Sta. Journal Series No. R-07337.
2 24 h instead of 48 h after the injection of PGF2α (Schmitt
Correspondence: P. O. Box 110920 (phone: (352) 392-5590; fax:
(352) 392-5595; E-mail: Thatcher@dps.ufl.edu).
Received July 29, 1999.
Accepted December 17, 1999.
luteinization of large follicles present in the ovary and
synchronizes the recruitment of a new follicular wave
(Thatcher et al., 1989; Macmillan and Thatcher, 1991).
At 7 d following injection of GnRH, an injection of prosta-
glandin F2α (PGF2α) induces regression of the corpus
luteum (CL) and allows for final maturation of the syn-
chronized dominant follicle (Schmitt et al., 1996b). At
48 h after injection of PGF2α, a second injection of GnRH
synchronizes ovulation of the dominant follicle, which
occurs approximately 28 h later (Pursley et al., 1995).
The tight synchrony of ovulation allows for a timed arti-
ficial insemination at approximately 16 h after the sec-
ond injection of GnRH.
Heifers submitted to the Ovsynch/TAI protocol had a
high incidence of estrus at ≤ 16 d following insemination,
and the problem of shortened return to estrus was exac-
erbated when the second GnRH injection was given at
et al., 1996b). Approximately 22.5% of heifers submitted
to the Ovsynch/TAI protocol were observed in estrus at
< 39 h after PGF2α injection (Schmitt et al., 1996b). Such

1568

Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809

by Universite Laval user
on 22 April 2018
 Day of the cycle affects a timed AI protocol
Table 1. Expected ovarian characteristics and plasma
progesterone concentrations at different
stages of the estrous cyclea
Presence of Presence Plasma
Day of the dominant of active progesterone
estrous cycle follicle corpus luteum concentration
Day 2 No No
Day 5 Yes Yes
Day 10 No Yes
Day 15 Yes Yes
Day 18 Yes No
a
Based on data from Ginther et al., 1989.
observations led to the hypothesis that stage of the es-
trous cycle in which the Ovsynch/TAI protocol is initiated
may influence efficacy of the synchronization protocol.
Our objective was to monitor follicle dynamics and
CL development in dairy heifers after initiation of the
Ovsynch/TAI protocol at different stages of the estrous
cycle that may correlate with occurrence of premature
ovulations, ovulation failure, and shortened return to
estrus.
Materials and Methods
Cycling Holstein heifers (n = 33) maintained at the
University of Florida Dairy Research Unit (Hague) were
injected twice with PGF2α (Lutalyse, Pharmacia-Upjohn
Co., MI; 25 mg, i.m.) given 11 d apart for estrus synchro-
nization. Detection of estrus was performed twice daily
(at 0600 and 1800) for 7 d after the second injection of
PGF2α. Heifers were examined daily after injection of
PGF2α with an Aloka 500-V ultrasound device equipped
with a 7.5-MHz linear-array transrectal transducer
(Aloka Co. Ltd., Japan) to confirm the day of ovulation
(d 0 of the cycle). Heifers detected in estrus and with
a confirmed ovulation (n = 25) were assigned to five
treatment groups. One heifer became sick after initiation
of the experiment and its data were discarded from the
analysis. The Ovsynch/TAI protocol was initiated at d 2
(group Day 2, n = 5), d 5 (group Day 5, n = 5), d 10 (group
Day 10, n = 4), d 15 (group Day 15, n = 5), and d 18
(group Day 18, n = 5) after ovulation. These five different
days of the estrous cycle were chosen as being most
representative of different physiological stages of the
estrous cycle according to previous observations of two-
and three-wave cycles in dairy heifers (Ginther et al.,
1989). A summary of the differences among stages is
represented in Table 1.
The Ovsynch/TAI protocol was initiated at d 2, 5, 10,
15, or 18 of the cycle, with an injection of a gonadotropin-
releasing hormone agonist (GnRHa; buserelin acetate,
Receptal, Hoechst-Roussel Agri-Vet, Sommerville, NJ; 8
µg, i.m.) followed 7 d later with an injection of PGF2α
(25 mg; i.m.) and a second GnRHa injection given 36 h
after PGF2α. The interval between injection of PGF2α
and the second injection of GnRH was reduced from 48
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
1569
to 36 h for two reasons: such a reduction would increase
the incidence of shortened returns to estrus, which would
enhance chances to detect experimental differences, and
could potentially reduce the problem of heifers being
detected in estrus prior to the second injection of GnRH.
Heifers were inseminated 16 h after the second injection
of GnRHa. Semen of two sires of proven fertility was
Low
used in this experiment, and all inseminations were per-
Rising
High formed by one technician. Heifers were observed twice
High daily for estrus for 16 d after insemination. At 21 d after
Low insemination, all heifers were re-synchronized for an
eventual second service with a GnRHa injection. Heifers
were scanned by ultrasonography for pregnancy diagno-
sis at 28 d after insemination. Heifers diagnosed preg-
nant were examined by rectal palpation at 45 d after
insemination to confirm pregnancy. Heifers diagnosed
nonpregnant at ultrasonography received an injection
of PGF2α and were re-inseminated at detected estrus.
Heifers from groups Day 2, Day 5, and Day 10 were
scanned using ultrasonography starting 2 d before the
first GnRHa injection and then daily throughout the
Ovsynch/TAI protocol until insemination. Heifers from
groups Day 15 and Day 18 started daily ultrasonography
examinations at d 10 of the estrous cycle to determine
the number of follicular waves with examinations termi-
nating at insemination. Following insemination, all heif-
ers were scanned every other day until 16 d, and again
at 21, 24, and 28 d after insemination. Ovarian struc-
tures such as follicles and CL were measured using on-
screen calipers and their relative positions were recorded
on follicular maps drawn during examination. Follicles
were categorized according to their diameter as Class I
(2 to 5 mm), Class II (6 to 9 mm), or Class III (> 9 mm).
The dominant follicle was the one that was at least 2
mm greater than other follicles (Sirois and Fortune,
1990). Day of emergence of the dominant follicle was
considered the day it was first classified as a Class II
follicle (> 5 mm; Diaz et al., 1998). Diameter of the CL
was estimated by averaging two measurements of CL
diameter at right angles to each other. In case there
were two CL present in either one or in both ovaries,
their diameters were added for quantitative analysis.
Day of onset of CL regression was considered the day
when plasma progesterone (P4) dropped to 50% of the
average concentrations for a 2-d period in the respective
luteal phases and continued to decrease thereafter.
Blood samples were collected daily immediately prior
to each ultrasound scanning during the Ovsynch/TAI
protocol and then daily until 16 d after insemination.
Additional blood samples were collected at 21, 24, and 28
d after insemination. Samples were collected by jugular
venipuncture into evacuated heparinized tubes, placed
in an ice bath, and centrifuged (3,000 × g for 30 min)
within 15 min of collection. Plasma was separated and
stored at −20°C until it was assayed for P4 by single-
antibody radioimmunoassay (Knickerbocker et al.,
1986). Sensitivity of the P4 assay was .3 ng/mL. Intraas-
say and interassay coefficients of variation were 12.25%
and 10.62%, respectively.
 1570
Analysis of data was performed using the method of
least squares ANOVA in the general linear model proce-
dure of SAS (1988). The experiment was divided into
two phases: the synchronization period, which started
at the first injection of GnRHa (d 0 of the synchronization
period) and ended 9 d later at insemination, and the
postsynchronization period, which began at insemina-
tion (d 0 of the postsynchronization period) and contin-
ued until heifers were diagnosed for pregnancy. Single
measurement variables such as dominant follicle and
CL size at different experimental days were analyzed
for treatment effects. Comparisons among means were
performed using pre-established orthogonal contrasts:
1) group Day 2 compared to groups Day 5, Day 10, Day
15, and Day 18; 2) group Day 5 compared to groups Day
10, Day 15, and Day 18; 3) group Day 10 compared to
groups Day 15 and Day 18; and 4) group Day 15 com-
pared to group Day 18. Variables involving repeated
measurements, such as number of Class I, Class II, and
Class III follicles as well as growth of the dominant
follicle, CL size, and plasma P4, were analyzed by homo-
geneity of regression procedures (Wilcox et al., 1990).
Statistical models included effects of treatment, cow
nested within treatment, and experimental day as a con-
tinuous variable. Regression curves were then analyzed
according to the above described orthogonal contrasts to
examine differences among treatment groups. Differ-
ences were considered significant at a probability value
of .05 or less.
Results and Discussion
Synchronization Period
Size of the largest follicle, size of the CL, and plasma
progesterone concentrations at the first injection of
GnRHa (d 0 of the synchronization period) are listed in
Table 2. Ovulation rates following the first injection of
GnRHa were 0% for group Day 2 (0/5), 100% for group
Day 5 (5/5), 25% for group Day 10 (1/4), 60% for group
Day 15 (3/5), and 100% for group Day 18 (5/5). Therefore,
frequencies of ovulation following the first GnRHa injec-
tion were lower (P < .01) for groups Day 2 and Day 10
than for groups Day 5, Day 15, and Day 18. The overall
ovulation rate to the first GnRHa injection was 58.3%
(14/24). Results obtained at the first GnRHa injection
Table 2. Least squares means and standard errors for the size of the largest follicle,
size of the corpus luteum(CL), and plasma progesterone (P4)
concentrations at d 0 of the synchronization period
Day of the
estrous cycle n Largest follicle, mm
Day 2 5 4.6
Day 5 5 10.0
Day 10 4 12.5
Day 15 5 11.0
Day 18 5 12.2
Different superscripts within column indicate significance (P < .01).
a,b,c,d
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
Moreira et al.
agree with expected results described in Table 1. Heifers
from group Day 2 did not have an established dominant
follicle, as denoted by the small size of the largest follicle
at d 0 of the synchronization period (4.6 ± .7 mm), failed
to ovulate to the first GnRHa injection, and had low
concentrations of plasma P4 (.8 ± 1.3 ng/mL). In all heif-
ers from group Day 5, plasma P4 concentrations were in
the ascending phase and the identified dominant follicle
ovulated upon GnRHa injection. This is in agreement
with previous observations that demonstrated a high
ovulation rate to a GnRHa or human chorionic gonado-
tropin injection at d 5 of the cycle (Schmitt et al., 1996a,
Diaz et al., 1998). Heifers from group Day 10 were in-
jected with GnRHa at emergence of the second wave
follicle as indicated by the low rate of ovulation (25%, 1/
4). Although heifers from group Day 10 had large follicles
at d 0 of the synchronization period, only one heifer
ovulated a second-wave dominant follicle that was 9 mm
in diameter. The other three large follicles were probably
first-wave dominant follicles undergoing atresia that did
not respond to GnRHa treatment. Plasma P4 concentra-
tions were high (14.1± 1.4 ng/mL) for heifers from group
Day 10, indicating the presence of an active CL. Simi-
larly, group Day 15 heifers also had high plasma P4
concentrations at d 0 of the synchronization period (14.0
± 1.3 ng/mL). A second-wave dominant follicle was iden-
tified at the first injection of GnRHa for all group Day 15
heifers. However, two heifers had a delayed emergence of
the dominant follicle and were probably too early in their
development to respond to the injection of GnRHa. This
should be expected because the emergence and selection
of the second-wave follicle is more variable than that of
the first-wave dominant follicle. Heifers from group Day
18 were injected with the first dose of GnRHa during the
proestrus phase, as indicated by the low concentration of
plasma P4 (1.6 ± 1.3 ng/mL) and by the high ovulation
rate (100%, 5/5).
The numbers of Class I, II, and III follicles between
the first injection of GnRHa (d 0) and the day of insemi-
nation (d 9) were analyzed by homogeneity of regression,
and the regression curves are represented in Figures 1a,
1b, and 1c. The regression curve for Class I follicles from
group Day 2 differed from the pooled regression curve
of the other groups (P < .01). Also, a difference between
the regression curve for group Day 5 and the pooled curve
for groups Day 10, Day 15, and Day 18 was detected (P
CL size, mm Plasma P4, ng/mL
± .7a
3.6 ± 1.1a
.8 ± 1.3a
± .7b 19.9 ± 1.1c 6.2 ± 1.3b
± .8c 24.1 ± 1.2d 14.1 ± 1.4c
± .7b,c 20.0 ± 1.1c 14.0 ± 1.3c
± .7c 15.7 ± 1.1b 1.6 ± .3a
 Day of the cycle affects a timed AI protocol
< .01). There was no difference (P > .10) among groups
Day 10, Day 15, and Day 18. Regression curves of the
number of Class II follicles differed when group Day 5
Figure 1. Regression curves of (a) Class I (pooled SE =
.47; R2 = .26), (b) Class II (pooled SE = .13; R2 = .13), and
(c) Class III (pooled SE = .43; R2= .45) follicles during the
synchronization period for groups Day 2 ( ), Day 5
(– –) , Day 10 (- - - -), Day 15 (. . . . .), and Day 18 (- . - . -).
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
1571
was compared to the pooled regression curves for group
Day 10, Day 15, and Day 18 (P < .05). No other significant
contrasts were observed among groups for Class II folli-
cles. Differences among all group comparisons were ob-
served when regression curves for the number of Class
III follicles were compared (P < .01).
During the synchronization period, the number of
Class I, Class II, and Class III follicles followed the classi-
cal process of recruitment, selection, and dominance
(Savio et al., 1988; Sirois and Fortune, 1988). The num-
ber of Class I and Class II follicles was high during
the recruitment phase and subsequently decreased upon
establishment of the dominant follicle. Differences de-
tected were due to the stage of the estrous cycle in which
the Ovsynch/TAI protocol was initiated and to the per-
centage of heifers ovulating after the first injection of
GnRHa. For instance, heifers in group Day 2 already
were in a spontaneous recruitment phase and the num-
ber of Class I and Class II follicles was high initially
during the early synchronization period but subse-
quently decreased with selection of the first-wave domi-
nant follicle. Because emergence of the dominant follicle
for group Day 2 heifers was already in progress, there
was an earlier decrease in the number of Class I and
Class II follicles, which led to a different follicular devel-
opment pattern compared to that of the other groups.
Heifers from group Day 5 had a substantial number of
Class I follicles at d 0 of the synchronization period but
few Class II follicles. Upon ovulation of the first-wave
dominant follicle, the number of Class I and Class II
follicles increased and only started to decrease later dur-
ing the synchronization period. At that point, there was
an increase in the number of Class III follicles, indicating
that a dominant follicle was selected and inhibited
growth of smaller follicles.
Analyses of the regression curves for growth of the
synchronized dominant follicle during the synchroniza-
tion period indicated significant differences (P < .01)
among all contrasts of treatment groups (Figure 2).
Emergence of the dominant follicle occurred earlier (P
< .02) during the synchronization period for group Day
2 (1.0 ± .6 d) than for groups Day 5, Day 10, Day 15,
and Day 18 (3.0 ± .6 d, 2.2 ± .7 d, 2.8 ± .6 d, and 3.0 ± .6
d, respectively). In addition, the number of days between
when the dominant follicle was first classified as a Class
III follicle until insemination was greater (P < .05) for
group Day 2 (6.4 ± .6 d) than for groups Day 5, Day 10,
Day 15, and Day 18 (5.0 ± .6 d, 5.5 ± .7 d, 4.6 ± .6 d, and
4.8 ± .6 d, respectively). Size of the dominant follicle at
experimental d 7 was greater (P < .01) for group Day 2
(13.8 ± .5 mm) compared to groups Day 5, Day 10, Day
15, and Day 18 (10.8 ± .5 mm, 11.7 ± .6 mm, 11.0 ± .5
mm, and 11.4 ± .5 mm, respectively). No differences (P <
.10) in the size of the dominant follicle among treatment
groups were detected at experimental d 8 and d 9.
Synchronization of the emergence of a new follicular
wave following the first injection of GnRHa was obtained
in groups Day 5, Day 10, Day 15, and Day 18. Heifers
in group Day 2 failed to respond to the first injection of
 1572
Figure 2. Regression curves of the growth of the domi-
nant follicle during the synchronization period (pooled
SE = .25 mm; R2=.82) for groups Day 2 ( ), Day 5
(– –) , Day 10 (- - - -), Day 15 (. . . . .), and Day 18 (- . - . -).
GnRHa and were already in the process of recruitment
of the dominant follicle at the initiation of the Ovsynch/
TAI protocol. As a consequence, the dominant follicle
from group Day 2 heifers reached diameters greater than
9 mm earlier during the synchronization period than in
other groups and were in a plateau phase at the time of
ovulation (Figure 2). It is possible that the synchronized
dominant follicle from heifers initiating the Ovsynch/
TAI protocol at d 2 is undergoing early stages of atresia
at the time of the second GnRH injection. Thus, initiation
of the Ovsynch/TAI protocol at metestrus may compro-
mise quality of the pre-ovulatory follicle and subsequent
developmental competence of the oocyte.
Plasma P4 concentrations differed markedly (P < .01)
among groups during the synchronization period (Figure
3). Plasma P4 concentrations for group Day 15 heifers
started to decrease prior to the injection of PGF2α and
were approximately 1.0 ng/mL at experimental d 7. On-
set of CL regression occurred earlier (P < .01) for group
Day 15 (3.2 ± .5 d) than for the other experimental
groups, in which CL regression occurred after the injec-
tion of PGF2α at experimental d 8. Dairy heifers initiate
CL regression between d 16.5 and 19.2 of the cycle
(Ginther et al., 1989). Thus, initiation of CL regression
at 3.2 ± .5 d after initiation of the Ovsynch/TAI protocol
in group Day 15 heifers probably was due to a normal
endogenous release of endometrial PGF2α, and both orig-
inal and accessory CL were responsive to PGF2α. Prema-
ture regression of the CL occurred in heifers that ovu-
lated (3/3) or did not ovulate (2/2) following the first
injection of GnRHa in the Day 15 group. Heifers in which
an accessory CL was induced at d 10 of the cycle, re-
gressed both original and accessory CL upon injection
of PGF2α 2, 4, or 6 d later (Howard and Britt, 1990). It
was concluded that an accessory CL induced during high
P4 concentrations was able to respond to an injection of
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
Moreira et al.
PGF2α at earlier stages of luteal development. As a re-
sult, all heifers (5/5) from group Day 15 were observed
in estrus prior to the second injection of GnRHa at exper-
imental d 9, whereas no estrous behavior was observed
in the other groups. Thus, incidence of estrus prior to
experimental d 9 was greater for group Day 15 heifers
than for other groups (P < .01). In addition, 60% (3/5) of
the heifers from group Day 15 ovulated prior to the
second injection of GnRHa, whereas no premature ovula-
tions were detected for the other groups (P < .01). There-
fore, heifers from group Day 15 underwent premature
regression of the CL during the synchronization period
and were observed in estrus prior to the second injection
of GnRHa. An asynchrony between insemination and
ovulation occurred in three of the five heifers.
Regression analyses also indicated that heifers initiat-
ing the Ovsynch/TAI protocol at d 18 had lower P4 during
the synchronization period than the other treatment
groups (P < .01). Pregnancy rates are higher when P4
exposure in the luteal phase prior to insemination is
higher (Fonseca et al., 1983). Thus, initiation of the Ov-
synch/TAI protocol at proestrus may reduce pregnancy
rates due to suboptimal P4 exposure. Two heifers from
group Day 18 had plasma P4 concentrations > 3.0 ng/
mL after injection of PGF2α and were considered to have
had incomplete CL regression. Incomplete CL regression
did not occur in other treatment groups and, therefore,
was more frequent for group Day 18 heifers (P < .01). A
possible explanation for such an observation is due to
the fact that, in group Day 18 heifers, all CL at injection
of PGF2α were induced accessory CL that had developed
under a low progesterone environment. Such CL may
have been on the borderline of being responsive to an
injection of PGF2α (Watts and Fuquay, 1985; Twagira-
mungu et al., 1995). Incomplete CL regression following
Figure 3. Regression curves of plasma progesterone
concentrations during the synchronization period
(pooled SE = .40 ng/mL; R2 = .73) for groups Day 2 ( ),
Day 5 (– –) , Day 10 (- - - -), Day 15 (. . . . .), and Day 18
(- . - . -).
 Day of the cycle affects a timed AI protocol 1573
the injection of PGF2α in the Ovsynch/TAI protocol has
been associated with lower pregnancy rates (Moreira et
al., 2000).
Excluding the three heifers from the Day 15 group that
underwent a premature ovulation prior to the second
injection of GnRHa, the frequency of ovulation after the
second injection of GnRHa was 80% for groups Day 2
and Day 5 (4/5 for both groups) and 100% for groups
Day 10, Day 15, and Day 18 (4/4, 2/2, and 5/5, respec-
tively). Only two heifers, one from group Day 2 and one
from group Day 5, failed to ovulate in response to the
second injection of GnRHa. The number of days the dom-
inant follicles of these two heifers were classified as Class
III follicles until the day of insemination was greater (P <
.01) than for heifers from other treatment groups, which
ovulated in response to the second injection of GnRHa
(7.5 ±.9 d > 5.0 ± .3 d). The overall ovulation rate in
response to the second injection of GnRHa was 90.4%
(19/21).

Postsynchronization Period
Heifers observed in estrus at < 16 d after insemination
were considered to have had a shortened return to es-
trus. Incidence of shortened return to estrus intervals
was considered one of the reasons for conception failure
in heifers submitted to the Ovsynch/TAI protocol
(Schmitt et al., 1996b). Overall, 20.8% (5/24) of heifers
had a shortened return to estrus, but that was not associ-
ated with treatment groups. Shortened return to estrus
was attributed to three causes: failure of the CL to com-
pletely regress after PGF2α injection (n = 1; group Day
18), ovulation failure following the second injection of
GnRHa (n = 2; groups Day 2 and Day 5), and short life-
span of the CL induced by the second injection of GnRHa
(n = 2; groups Day 2 and Day 15).
Figure 4a depicts follicular development and P4 pro-
files of a shortened return to estrus due to incomplete
CL regression after injection of PGF2α in a heifer from
group Day 18. Upon injection of the first dose of GnRHa,
there was an ovulation followed by the recruitment of a
new dominant follicle (DF1). A new CL was induced after
ovulation as observed ultrasonographically and by the
rising concentrations of plasma P4. Plasma P4 concentra-
tions decreased after injection of PGF2α (experimental d
7) to approximately 2.0 ng/mL 24 h later. Ultrasonogra-
phy scanning indicated that the CL induced by the first Figure 4. Individual plasma progesterone (●) and de-
injection of GnRHa did not regress completely after injec- velopment of the synchronized dominant follicle (䊏) and
tion of PGF2α, which explains the rapid increase in of the second dominant follicle (䊐) for three heifers with
plasma P4 after experimental d 8. On experimental d 8, shortened return to estrus due to (a) incomplete CL re-
injection of GnRHa induced ovulation of DF1 and a new gression after injection PGF2α (group Day 18), (b) ovula-
follicular wave was initiated with recruitment of a sec- tion failure following the second injection of GnRHa
ond dominant follicle (DF2) and concurrent development (group Day 2), and (c) short life-span of the CL induced
of an accessory CL. At experimental d 16, both CL initi- after second injection of GnRHa (group Day 2). OV = ovu-
ated regression and plasma P4 concentrations dropped lation.
sharply to 1.3 ng/mL at experimental d 18. As a conse-
quence, that particular heifer was observed in estrus
at experimental d 18 (8 d after insemination) and had
ovulated the DF2 by the following day.

Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809

by Universite Laval user
on 22 April 2018
 1574
Figure 5. Regression curves of the growth of the domi-
nant follicle (DF2) during the postsynchronization period
(pooled SE = .35 mm; R2 =.91) for groups Day 2 ( ),
Day 5 (– –) , Day 10 (- - - -), Day 15 (. . . . .), and Day 18
(- . - . -).
Two heifers failed to ovulate following the second injec-
tion of GnRHa (groups Day 2 and Day 5). Figure 4b
depicts a heifer from group Day 2 that had a follicular
development and plasma P4 profile similar to that of the
heifer from group Day 5. Upon injection of GnRHa at
experimental d 0, a dominant follicle was recruited
(DF1). After injection of PGF2α at experimental d 7, the
CL regressed and plasma P4 decreased rapidly. However,
injection of GnRHa at experimental d 8 did not induce
the ovulation of the DF1, which became a persistent
follicle under the influence of a low P4 environment.
Associated with delayed recruitment of a second-wave
follicle (DF2), a spontaneous estrus occurred at experi-
mental d 15 (6 d after insemination) and ovulated at
experimental d 16.
A heifer from group Day 2 that had a shortened return
to estrus due to a reduced CL life-span is represented
in Figure 4c. Follicular development and plasma P4 pro-
file for the Day 2 heifer was similar to that of the other
heifer from group Day 15, which also had a short CL
life-span. Following ovulation of the dominant follicle
(DF1) recruited by the first injection of GnRHa, a new
CL was formed. However, the induced CL initiated re-
gression at experimental d 17 and had plasma P4 concen-
trations lower than 1.0 ng/mL at experimental d 18.
Premature regression of the induced CL caused the
heifer to express estrus at experimental d 19 (10 d after
insemination) and the second-wave dominant follicle
(DF2) ovulated at experimental d 20.
Heifers with shortened return to estrus (n = 5) were
excluded from the analyses of the post synchronization
period for plasma P4 and growth of the second-wave
dominant follicle. The dominant follicle of groups Day 2
and Day 5 had a greater rate of growth from insemina-
tion until 7 d later compared to heifers from groups Day
10, Day 15, and Day 18 (P < .05; Figure 5). There were
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
Moreira et al.
no differences (P > .10) among treatment groups on the
day of emergence of the second-wave dominant follicle
after insemination (Day 2 = 11.0 ± .9 d, Day 5 = 11.2 ±
.8 d, Day 10 = 11.5 ± .8 d, Day 15 = 12.5 ± .8 d, and d
18 = 12.6 ± .9 d).
The increase in plasma P4 following insemination until
7 d later was lower for group Day 2 heifers than for
other treatment groups (P < .05; Figure 6). No differences
were observed when plasma P4 concentrations were eval-
uated from insemination until 16 d later among treat-
ment groups (P > .10). As observed above, the first injec-
tion of GnRHa failed to synchronize the emergence of a
new follicular wave in heifers from group Day 2. The
fact that recruitment of the dominant follicle occurred
earlier for group Day 2 may have resulted in the presence
of a follicle in the initial stages of atresia at the time of
the second GnRHa injection. Therefore, ovulation of an
aged follicle caused by initiation of the Ovsynch/TAI pro-
tocol at metestrus may affect subsequent functional com-
petence of the CL. Whereas some studies found no effect
of P4 concentrations following insemination on preg-
nancy rates (Pritchard et al., 1994), other researchers
observed that the rise in plasma P4 concentrations fol-
lowing ovulation is correlated positively to conception
rates (Maurer and Echternkamp, 1982; Butler et al.,
1996). Hence, initiation of the Ovsynch/TAI protocol at
metestrus may have a deleterious effect on pregnancy
rates.
Although final size of the preovulatory dominant folli-
cle recruited during the synchronization period did not
differ among treatment groups, size of the preovulatory
follicle influenced subsequent plasma P4 concentrations
during the post synchronization period, as characterized
by regression analyses (Figure 7). Preovulatory follicles
were classified according to their size at the last day
Figure 6. Regression curves of plasma progesterone
concentrations during the postsynchronization period
(pooled SE = .27 ng/mL; R2 = .82) for groups Day 2 ( ),
Day 5 (– –) , Day 10 (- - - -), Day 15 (. . . . .), and Day 18
(- . - . -).
 Day of the cycle affects a timed AI protocol
Figure 7. Regression curves of plasma progesterone
concentrations during the postsynchronization period
(pooled SE = .19 ng/mL; R2 = .73) for preovulatory follicles
≤ 12 mm ( ), 13 to 14 mm (– –), and ≥ 15 mm (- - - -)
in diameter.
follicles were detected at ultrasonography. Preovulatory
follicles were classified as ≤ 12 mm (n = 6), between
13 and 14 mm (n = 6), ≥ 15 mm (n = 7). Ovulation of
preovulatory follicles ≤ 12 mm resulted in lower plasma
P4 concentrations (P < .01) than ovulation of follicles >
12 mm. Also, plasma P4 concentrations were lower (P <
.03) after ovulation of follicles 13 to 14 mm in size com-
pared to follicles ≥ 15 mm. Vasconcelos et al. (1997)
reported in lactating dairy cows that initiation of the
Ovsynch/TAI protocol at metestrus, late diestrus, and
proestrus resulted in a larger synchronized dominant
follicle at injection of PGF2α (d 7) and GnRH (d 9). Fur-
thermore, lower pregnancy rates to the Ovsynch/TAI
protocol were associated with ovulation of large follicles.
Results from the present experiment with dairy heifers
contradict these prior observations. As indicated above,
size of the synchronized dominant follicle was greater
at d 7 of the synchronization period for heifers in group
Day 2 than for other groups (P < .01), but there was no
difference in sizes of the synchronized dominant follicle
at d 8 or 9 of the synchronization period among treatment
groups. In addition, ovulation of follicles ≥ to 13 mm
probably resulted in a more robust CL compared to
smaller follicles because plasma P4 concentrations were
higher (P < .01) during the subsequent luteal phase.
The injection of GnRHa at 21 d after insemination
induced ovulation in 52.6% of the heifers (10/19). Among
heifers diagnosed as pregnant at 28 d after insemination
(9/24), six heifers ovulated to the injection of GnRHa at
d 21, whereas three heifers did not respond to the same
GnRHa injection. There was no difference (P > .10) in
mean plasma P4 concentrations at 28 d after insemina-
tion between pregnant heifers ovulating or not after the
injection of GnRHa at d 21 (17.3 ± 2.5 ng/mL and 16.3
± 1.9 ng/mL, respectively).
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
1575
Overall pregnancy rate was 37.5% (9/24), as diagnosed
at 28 d after insemination. Pregnancy rate for groups
Day 2, Day 5, Day 10, Day 15, and Day 18 were 40% (2/
5), 20% (1/5), 75% (3/4), 0% (0/5), and 60% (3/5), respec-
tively. All heifers indicated pregnant by ultrasonography
were confirmed pregnant at rectal palpation at 45 d
after insemination. The number of heifers per treatment
group was too small to produce reliable estimates of
potential differences in pregnancy rates among the dif-
ferent stages of the estrous cycle in which the Ovsynch/
TAI protocol was initiated.
Heifers not diagnosed pregnant by ultrasonography
and resynchronized with an injection of PGF2α (n = 10)
were observed for estrus during the following 7 d. Five
heifers (50.0 %) were detected in estrus at either two (n
= 4) or three (n = 1) days after PGF2α injection and were
inseminated. Three heifers conceived to the resynchro-
nized estrus, which resulted in a 30.0% pregnancy rate
(3/10) and a 60.0% conception rate (3/5) after resynchro-
nization. Such a resynchronization system needs to be
further tested with a larger number of animals to pro-
duce reliable estimates of pregnancy and conception
rates.
Results demonstrated that day of the cycle at initiation
of the Ovsynch/TAI protocol affected the synchronization
program and may have influenced subsequent preg-
nancy rates. Among the 5 days of the estrous cycle at
which the Ovsynch/TAI protocol was initiated, d 2, d 15,
and d 18 may be considered less suitable to result in
appropriate conception to the timed artificial insemina-
tion service for different reasons. The interpretation that
initiation of the Ovsynch/TAI protocol at metestrus (d
2), late diestrus (d 15), and proestrus (d 18) phases of
the estrous cycle may affect subsequent pregnancy rates
is in agreement with results from a large field experi-
ment in which pregnancy rates obtained in cows initiat-
ing the Ovsynch/TAI protocol during the first 3 d or after
d 13 of the estrous cycle were lower than those in cows
initiating the Ovsynch/TAI protocol during early luteal
phase (Vasconcelos et al., 1997).
The Ovsynch/TAI protocol has been compared to sev-
eral different control treatments in lactating dairy cows,
and results indicated that pregnancy rates were similar
or greater for the Ovsynch/TAI cows (Burke et al., 1996;
Pursley et al., 1997a,b). However, in dairy heifers, re-
sults have been contradictory; some researchers detected
no differences among treatment groups (Schmitt et al.,
1996b), whereas others observed a decrease in fertility
for heifers receiving the Ovsynch/TAI protocol (Pursley
et al., 1997b). A reduction in pregnancy rates of dairy
heifers submitted to the Ovsynch/TAI protocol was de-
tected when the program was compared to pregnancy
rates obtained after three consecutive synchronization
treatments with PGF2α (Pursley et al., 1997b). Such re-
sults could be due to the fact that estrus detection is
higher and fertility is greater in dairy heifers than in
lactating cows. Nonetheless, if pregnancy rates obtained
after the initial synchronization with PGF2α (28.2%)
were compared to pregnancy rates obtained by the Ov-
 1576 Moreira et al.
synch/TAI service (35.1%), there would be difference be-
tween the two groups (Pursley et al., 1997b). However,
there are differences in follicular dynamics between
dairy heifers and lactating cows. Heifers have a faster
rate of follicular growth than lactating cows (Pursley et
al., 1995), and a high frequency of three-wave cycles was
observed in heifers (Savio et al., 1988). Those factors
may reduce the frequency of heifers that successfully
synchronize the emergence of a dominant follicle follow-
ing the first injection of GnRH of the Ovsynch/TAI proto-
col, as previously proposed (Pursley et al., 1997b). None-
theless, optimization of this timed artificial insemination
system in heifers may lead to significant increases in
pregnancy rates when applied to lactating dairy cows.
Implications
Stage of the estrous cycle at which synchronization
is initiated influences reproductive responses the timed
artificial insemination (Ovsynch/TAI) protocol. Interpre-
tation of the results obtained leads to the hypothesis
that the stage of the cycle that seems to be the most
appropriate for producing greater pregnancy rates to
the Ovsynch/TAI protocol is the early luteal phase (i.e.,
between d 5 and 10 of the estrous cycle). Future experi-
ments to increase pregnancy rates to the Ovsynch/TAI
protocol may focus on two regulatory components: 1)sup-
plementation of exogenous Progesterone during the syn-
chronization period to avoid premature ovulation and
asynchrony of insemination and 2) presynchronization
prior to initiation of the Ovsynch/TAI protocol to target
cows in early luteal phases.
Literature Cited
Burke, J. M., R. L. de la Sota, C. A. Risco, C. R. Staples, E. J. P.
Schmitt, and W. W. Thatcher. 1996. Evaluation of timed insemi-
nation using a gonadotropin-releasing hormone agonist in lactat-
ing dairy cows. J. Dairy Sci. 79:1385–1393.
Butler, W. R., J. J. Calaman, and S. W. Beam. 1996. Plasma and milk
urea nitrogen in relation to pregnancy rate in lactating dairy
cattle. J. Anim. Sci. 74:858–865.
Diaz, T., E. J.-P. Schmitt, R. L. De La Sota, M.-J. Thatcher, and
W. W. Thatcher. 1998. Human chorionic gonadotropin-induced
alterations in ovarian follicular dynamics during the estrous cycle
of heifers. J. Anim. Sci. 76:1929–1936.
Fonseca, F. A., J. H. Britt, B. T. Daniel, J. C. Wilk, and A. H. Rakes.
1983. Reproductive traits of Holsteins and Jerseys. Effects of age,
milk yield, and clinical abnormalities on involution of cervix and
uterus, ovulation, estrous cycles, detection of estrus, conception
rate, and days open. J. Dairy Sci. 66:1128–1147.
Ginther, O. J., L. Knopf, and J. P. Kastelic. 1989. Temporal associations
among ovarian events in cattle during oestrus cycles with two
and three follicular waves. J. Reprod. Fertil. 87:223–230.
Howard, H. J., and J. H. Britt. 1990. Prostaglandin F-2α causes regres-
sion of an hCG-induced corpus luteum before day 5 of its lifespan
in cattle. J. Reprod. Fertil. 90:245–253.
Knickerbocker, J. J., W. W. Thatcher, F. W. Bazer, M. Drost, D. H.
Barron, K. B. Fincher, and R. M. Roberts. 1986. Proteins secreted
Downloaded from https://academic.oup.com/jas/article-abstract/78/6/1568/4625809
by Universite Laval user
on 22 April 2018
by Day-16 to 18 conceptuses extended corpus luteum function in
cows. J. Reprod. Fertil. 77:381–391.
Macmillan, K. L., and W. W. Thatcher. 1991. Effects of an agonist of
gonadotropin-releasing hormone on ovarian follicles in cattle.
Biol. Reprod. 45:883–889.
Maurer, R. R., and S. E. Echternkamp. 1982. Hormonal asynchrony
and embryonic development. Theriogenology 17:11–22.
Moreira, F., C. A. Risco, M. F. A. Pires, J. D. Ambrose, M. Drost, and
W. W. Thatcher. 2000. Use of bovine somatotropin in lactating
dairy cows receiving timed artificial insemination. J. Dairy Sci.
83:1245–1255.
Pritchard, J. Y., F. N. Schrick, and E. K. Inskeep. 1994. Relationship
of pregnancy rate to peripheral concentrations of progesterone
and estradiol in beef cows. Theriogenology 42:247–259.
Pursley, J. R., M. R. Kosorok, and M. C. Wiltbank. 1997a. Reproductive
management of lactating dairy cows using synchronization of
ovulation. J. Dairy Sci. 80:301–306.
Pursley, J. R., M. O. Mee, and M. C. Wiltbank. 1995. Synchronization
of ovulation in dairy cows using PGF2α and GnRH. Theriogenology
44:915–923.
Pursley, J. R., M. C. Wiltbank, J. S. Stevenson, J. S. Ottobre, H.
A. Garverick, and L. L. Anderson. 1997b. Pregnancy rates per
artificial insemination for cows and heifers inseminated at a syn-
chronized ovulation or synchronized estrus. J. Dairy Sci.
80:295–300.
SAS. 1988. In: SAS/STAT User’s guide (Release 6.03 Ed.). SAS Insti-
tute Inc. Cary, NC.
Savio, J. D., L. Keenan, M. P. Boland and J. F. Roche. 1988. Pattern
of growth of dominant follicles during the estrous cycle of heifers.
J. Reprod. Fertil. 83:663–671.
Schmitt, E. J.-P., T. Diaz, C M. Barros, R. L. de la Sota, M. Drost, E.
W. Fredriksson, C. R. Staples, R. Thorner, and W. W. Thatcher.
1996a. Differential response of the luteal phase and fertility in
cattle following ovulation of the first wave dominant follicle with
human chorionic gonadotropin or an agonist of gonadotropin-
releasing hormone. J. Anim. Sci. 74:1074–1083.
Schmitt, E. J.-P., T. Diaz, M. Drost, and W. W. Thatcher. 1996b. Use
of a gonadotropin-releasing hormone agonist or human chorionic
gonadotropin for timed insemination in cattle. J. Anim. Sci.
74:1084–1091.
Sirois, J., and J. Fortune. 1988. Ovarian follicular dynamics during
the estrous cycle monitored by real time ultrasonography. Biol.
Reprod. 39:308–317.
Sirois, J. and J. Fortune. 1990. Lengthening the bovine estrous cycle
with low levels of exogenous progesterone: a model for studying
ovarian follicular dominance. Endocrinology 127:916–925.
Thatcher, W. W., K. L. Macmillan, P. J. Hansen, and M. Drost. 1989.
Concepts for the regulation of corpus luteum function by the
conceptus and ovarian follicles to improve fertility. Theriogenol-
ogy 31:149–164.
Twagiramungu H., L. A. Guilbault, and J. J. Dufour. 1995. Synchroni-
zation of ovarian follicular waves with a gonadotropin-releasing
hormone agonist to increase the precision of estrus in cattle: A
review. J. Anim. Sci. 73:3141–3151.
Vasconcelos, J. L. M., R. W. Silcox, J. R. Pursley, and M. C. Wiltbank.
1997. Effect of day of estrous cycle on synchronization of ovulation
using GnRH. J. Anim. Sci. 75(Suppl.1):89 (Abstr.).
Watts, T. L., and J. W. Fuquay. 1985. Response and fertility of dairy
heifers following injection with prostaglandin F2α during early,
middle or late diestrus. Theriogenology 18:191–197.
Wilcox, C. J., W. W. Thatcher, and F. G. Martin. 1990. Statistical
analysis of repeated measurements in physiology experiments.
In:Proc. Final Research Co-ordination Meeting of the FAO/IACA/
ARCAL III Regional Network for Improving the Reproductive
Management of Meat- and Milk-producing Livestock in Latin
America with the aid of Radioimmunoassay. pp141–151. Interna-
tional Atomic Energy Agency, Vienna, Austria.