Factors affecting conception rate after artificial insemination and pregnancy loss in lactating dairy cows

Animal Reproduction Science 84 (2004) 239–255

Factors affecting conception rate after artificial insemination and pregnancy loss in lactating dairy cows Ricardo C. Chebel, José E.P. Santos∗ , James P. Reynolds, Ronaldo L.A. Cerri, Sérgio O. Juchem, Michael Overton Veterinary Medicine Teaching and Research Center, University of California, Davis, 18830 Road 112 Tulare, CA 93274, USA Received 17 September 2003; received in revised form 30 December 2003; accepted 31 December 2003

Abstract Objectives were to determine factors associated with conception rate (CR) and pregnancy loss (PL) in high producing lactating Holstein cows. In Study 1, CR was evaluated in 7633 artificial inseminations (AI) of 3161 dairy cows in two dairy farms. Pregnancy diagnosis was performed by palpation per rectum 39 ± 3 days after AI. Environmental temperature was recorded at different intervals prior to and after AI. In Study 2, 1465 pregnancies from 1393 cows diagnosed at 31 ± 3 days after AI by ultrasonography on three dairy farms were re-examined 14 days later to determine PL. Temperature ≥29 ◦ C was considered to be heat stress (HS). Exposure to HS was defined as following: NH, no heat stress; HS1, exposure to at least 1 day of maximum temperature ≥29 ◦ C and average daily maximum temperature (ADMT) <29 ◦ C; and HS2, exposure to ADMT ≥29 ◦ C. In Study 1, exposure of cows to HS1 and HS2 from 50 to 20 prior to AI was associated with reduced CR compared to cows not exposed to HS (28.8, 23.0, and 31.3%, respectively). Post-AI HS was not associated with CR. Cows inseminated following estrus detection or timed AI had similar CR. As the number of AI increased, CR decreased. Multiparous cows had lower CR than primiparous cows, and occurrence of milk fever and retained placenta was associated with decreased CR. In Study 2, PL was not associated with exposure to HS either prior to or after AI. Cows diagnosed with clinical mastitis experienced increased PL, but parity, number of AI, AI protocol, milk production, and days postpartum at AI were not associated with PL. In conclusion, CR was affected by HS prior to AI, parity, number of AI, and postparturient diseases, whereas PL was affected by clinical mastitis. © 2004 Elsevier B.V. All rights reserved. Keywords: Conception rate; Pregnancy loss; Dairy cows; Heat stress; Artificial insemination

∗ Corresponding author. Tel.: +1-559-688-1731; fax: +1-559-686-4231. E-mail address: [email protected] (J.E.P. Santos).

0378-4320/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.anireprosci.2003.12.012

240

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

1. Introduction Conception rate (CR), estrus detection, and pregnancy loss (PL) are important factors determining the reproductive performance of dairy herds. Conception rate and estrus detection provide the values for calculation of pregnancy rate, and increases in pregnancy rates result in expected additional income per cow per year. Conversely, losses of pregnancy are costly. It has been estimated that the cost of a pregnancy lost is on average US$ 640.00 (Thurmond et al., 1990). Reproductive efficiency as measured by CR, estrus detection rate, and days open in lactating dairy cows has declined, and that has been associated with a steady increase in average milk production per cow per year (Lucy, 2001). Simultaneously, increased milk production has been accompanied by a higher incidence of postparturient disorders, which, in turn, are detrimental to fertility (Gröhn and Rajala-Schultz, 2000). Many factors influence CR: among them is cyclicity, energy balance, heat stress (HS), parity, milk production, diet, and diseases (Cartmill et al., 2001a; Gröhn and Rajala-Schultz, 2000; Hansen and Arechiga, 1999; Lucy, 2001; Moreira et al., 2001; Santos et al., 2004a). However, less characterized are the factors that affect PL in high producing lactating dairy cows, with little data available regarding the effect of different reproductive management protocols on PL. In an extensive review of the literature, Lucy (2001) suggested that PL in lactating dairy cattle has increased, along with greater adoption of fixed-time AI protocols. However, no direct comparisons were available to substantiate it. The advent of ultrasonography has allowed accurate pregnancy diagnosis as early as 25 day after AI in dairy cattle (Fricke, 2002), which facilitates the study of late embryonic mortality after the period of maternal recognition of pregnancy. Studies utilizing embryo transfer and early pregnancy diagnosis indicate that less than 50% of the viable embryos establish pregnancy by 27–30 day after ovulation in lactating dairy cows (Drost et al., 1999; Sartori et al., 2003). Several studies with lactating dairy cows have indicated that late embryonic losses, between 27 and 31 and 41–45 days after AI range from 10 to 21% (Chebel et al., 2003; Moreira et al., 2001; Santos et al., 2001, 2004b). In some cases, pregnancy losses have been reported to be higher than 21% (Cartmill et al., 2001a) or even 60.5% in cows inseminated at fixed time during HS (Cartmill et al., 2001b). Dairy cows’ tolerance to high temperatures is diminished during lactation due to increased internal metabolic heat production associated with high feed intake and milk synthesis. The reproductive performance of lactating cows under high environmental temperature is compromised because of the deleterious effect of HS on fertilization and embryo survival (Hansen and Arechiga, 1999; Wolfenson et al., 2000). Oocyte and embryo at early stages are extremely sensitive to HS, while embryos 3 days and older are thought to be more resistant (Hansen and Arechiga, 1999). Drost et al. (1999) demonstrated that bypassing fertilization and early embryonic development during HS with the transfer of frozen embryos to dairy cows improved CR when compared to AI. On the other hand, when HS was not present embryo transfer resulted in similar CR to AI (Sartori et al., 2003). These results indicate that transfer of 7-day-old embryos from cows not exposed to HS are more capable of establishing pregnancy in cows exposed to HS than AI. The objectives of the present study were to evaluate factors associated with CR and PL in high producing lactating Holstein cows. Among the factors evaluated were exposure to

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

241

HS, insemination protocol, parity, mastitis, postparturient diseases, milk yield, number of AI, and days in milk (DIM) at AI, when controlling for site and technician.

2. Materials and methods 2.1. Study 1 2.1.1. Study design, animals, and housing This was a retrospective study evaluating the effect of factors affecting CR in 7633 AI from a total of 3161 lactating Holstein cows (1864 multiparous and 1297 primiparous) included in the study. Animals were originated from two commercial dairy farms located in the Central Valley of California. In both sites, cows were housed in free stall barns. Each dairy housing facility was virtually identical in design, size, and number of animals. All pens were equipped with intermittent water sprinklers and fans in an attempt to alleviate HS from May to September. The cooling system was activated once the ambient temperature reached 26.7 ◦ C. Cows were fed a diet formulated to meet the nutrient requirements established by the NRC (2001) for Holstein cows weighing 650 kg and producing 40 kg per day of milk with 3.5% fat and 3.2% true protein. Within each site, the same total mixed ration was fed to all lactating cows after 3–4 weeks postpartum. The study period lasted 40 weeks, from April 2001 to January 2002. 2.1.2. Reproductive management As part of the regular reproductive management in site 1 all animals received two injections of 25 mg PGF2␣ (Lutalyse® , Dinoprost Tromethamine, Pharmacia Animal Health, Kalamazoo, MI, USA) at 33 ± 3 and 47 ± 3 days after calving. After the second injection of PGF2␣ cows were observed for signs of estrus once daily, in the morning, by tail chalking (Macmillan et al., 1988) using colored paint sticks (All-weather Paintstik Livestock Marker, LA-CO Industries, Chicago, IL). Cows detected in estrus were inseminated in the same morning, and those cows not found in estrus and not inseminated by 61 ± 3 DIM were enrolled in the Ovsynch protocol (Pursley et al., 1997). Cows that returned to estrus prior to pregnancy diagnosis were re-inseminated upon estrus detection and considered to be non-pregnant from previous AI. Non-pregnant cows at palpation per rectum 39 ± 3 days after AI were re-synchronized with the Ovsynch protocol and inseminated at fixed time. In site 2, all animals received an injection of 25 mg PGF2␣ at 33 ± 3 DIM. Fourteen days later, cows received an injection of 100 ␮g GnRH (Cystorelin® , Gonadorelin Diacetate Tetrahydrate, Merial Ltd., Athens, GA, USA), followed 7 days later by an injection of PGF2␣ . Cows were observed for signs of estrus as described previously and inseminated once daily. Cows not displaying signs of estrus in a period of 7 day and those found non-pregnant at pregnancy examination received another injection of GnRH followed 7 days later by PGF2␣ , and insemination was performed upon estrus detection. 2.1.3. Environment temperature and HS classification Daily ambient temperature was obtained from the National Weather Service Forecast Office in Hanford, California, located approximately 25 miles from the two sites. The

242

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

maximum, minimum, and average daily temperatures were recorded for the entire study period. Average daily maximum temperature (ADMT) was calculated and the number of days exposed to maximum temperatures above 29 ◦ C was recorded for the following 3 periods: from day −50 to day −20 (day 0 = day of AI), from day −20 to AI, and from AI to pregnancy diagnosis (day 39 ± 3). Heat stress was classified for each of the periods described above according to the exposure or not to maximum temperature above 29 ◦ C and the ADMT as follows: no exposure to HS (NH), no single daily maximum temperature was above 29 ◦ C; heat stress 1 (HS1), exposure to at least one day of maximum temperature equal to or higher than 29 ◦ C, but ADMT below 29 ◦ C; and heat stress 2 (HS2), ADMT equal to or higher than 29 ◦ C. 2.1.4. Milk production and diagnosis of clinical mastitis Cows were milked thrice and twice daily for sites 1 and 2, respectively, and milk yields were recorded for individual cows once monthly during the official California Dairy Herd Improvement Association test. Individual milk samples were also collected from consecutive milkings during the same sampling day, then composited and analyzed for SCC, fat and true protein concentrations (Foss 303 Milk-O-Scan® ; Foss Foods Inc., Eden Prairie, MN) at the DHIA Laboratory in Tulare (site 1) or Fresno (site 2), California. All cows were examined for signs of clinical mastitis by the herd personnel immediately prior to every milking. Clinical mastitis cases were characterized in all sites by the presence of abnormal milk or by signs of inflammation in one or more quarters. Affected cows were treated by intramammary infusion of antibiotics according to treatment protocols established by the herd veterinarian. A new case of mastitis was defined for the same cow when a different quarter was affected, or a period of 21 days had elapsed since the previous diagnosis (Hillerton and Kliem, 2002). Mastitis events were recorded by the farm personnel with computer software used for data management (Dairy Comp 305, Valley Ag Software, Tulare, CA). Mastitis events that occurred between the day of AI and pregnancy diagnosis were recorded for each animal enrolled in the study. 2.1.5. Postparturient diseases and lameness As part of the herd health management of the dairies involved in this study, postparturient diseases and lameness were recorded by farm personnel as described for mastitis. Retrospectively, occurrence of retained placenta, milk fever, and displaced abomasum was collected for every cow. Episodes of lameness that had occurred between 30 days prior to AI and the day of pregnancy diagnosis were also recorded. 2.2. Study 2 2.2.1. Study design, animals and housing In this prospective observational study, pregnancy data from 1393 lactating Holstein cows (602 primiparous and 791 multiparous) in three dairy herds were utilized to determine factors associated with PL. In all three sites, animals were housed in free stall barns with pens equipped with intermittent water sprinklers and fans in an attempt to alleviate HS from May to September as described previously. In sites 1 and 3, cows were fed twice daily and

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

243

in site 2 once daily. Within each site, cows were fed the same total mixed ration formulated for Holstein cows weighing 650 kg and producing 40 kg per day of milk with 3.5% fat and 3.2% true protein (NRC, 2001). All three sites were located in the Central Valley of California, and the study period was from September 1999 to December 2001. 2.2.2. Reproductive management Sites 1 and 2 followed the same reproductive program described for site 1 in Study 1, and site 3 followed the program described for site 2 in Study 1. However, pregnancy was diagnosed by ultrasonography (Ultrascan 50, Alliance Medical, Smithville, MO) at 31 ± 3 days after AI and reconfirmed by palpation per rectum 14 days later. 2.2.3. Environment temperature and HS classification Environmental temperature was obtained as described previously for Study 1. However, the periods analyzed were as follow: day −50 to −20 (day 0 = day of AI), day −20 to AI, AI to ultrasonography, and ultrasonography to palpation per rectum. The ADMT was calculated for each of the periods described above and the number of days in which the cows were exposed to temperatures above 29 ◦ C was recorded. Initially, classification according to HS exposure had been done as described in Study 1. However, after analyzing the data and observing no difference between HS1 and HS2, it was decided to divide HS exposure into two groups instead of three. Therefore, in Study 2 cows were classified as follows: NH, no heat stress, meaning exposure to no single daily maximum temperature above 29 ◦ C; or HS, exposure to at least one day of maximum temperature equal to or higher than 29 ◦ C. 2.2.4. Milk production and diagnosis of clinical mastitis In all three sites, yields of milk and milk components were obtained once monthly as described for Study 1. The definition and diagnosis of clinical cases of mastitis were the same as described for Study 1. In each of the three sites, treatment of mastitis was according to protocols prescribed by the herd veterinarian. Mastitis events that occurred between the day of AI and pregnancy reconfirmation were recorded for each animal enrolled in the study. 2.3. Statistical analyses Reproductive data were analyzed by logistic regression (Allison, 1999) with the LOGISTIC procedure of the SAS (2001) program, using a backward stepwise multivariate logistic model with variables being continuously removed from the model by the Wald statistic criterion if the significance was greater than 0.20. In Study 1, the mathematical model for analyses of CR included the effects of site, AI technician, parity, DIM at AI, number of AI, insemination protocol (insemination at detected estrus versus timed AI), milk yield, exposure to HS, postparturient diseases, mastitis, lameness, and interactions. In Study 2, the mathematical model for analyses of PL included the effects of site, AI technician, parity, DIM at AI, number of AI, insemination protocol (insemination at detected estrus versus timed AI), milk yield, exposure to HS, mastitis, and interactions. Continuous data were analyzed by ANOVA (Littell et al., 2002) using the GLM procedure of the SAS (2001) program. Descriptive statistics were evaluated by the SAS (2001) program.

244

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

The odds ratio (OR) estimates and the 95% confidence intervals (CI) from the logistic regression were obtained for each variable included in the statistical model. For continuous variables, an OR significantly higher than 1 indicates an increased risk for conception or PL, and an OR significantly lower than 1 indicates a reduced risk for conception or PL. For categorical variables, one class of each variable was considered as the reference, and an OR significantly higher (or lower) than 1 for any other class of this variable was indicative of an increase (or reduced) risk for conception or PL when compared to the class used as reference. Regression analyses between the OR for CR or PL and number of AI, milk yield categorized in 5 kg increments, and DIM at AI categorized in 10-day increments were evaluated using the regression procedure of MINITAB (2000) to determine the fitted line plot that best described these relationships. Orthogonal polynomials with linear, quadratic, and cubic relationships were evaluated. Data are presented as proportion and least square means. Treatment differences with P ≤ 0.05 were considered significant and 0.05 < P ≤ 0.10 were considered a tendency.

3. Results 3.1. Study 1 In Study 1, after evaluating 7633 AI from lactating Holstein cows, the CR obtained was 25.42%. Descriptive statistics for DIM, daily milk yield, and number of AI are depicted in Table 1. As expected, because of the study design, the ADMT for each level of HS to which cows were exposed in the three periods analyzed differed, with cows in HS2 exposed to the highest ADMT (Table 2). Exposure of cows to HS1 and HS2 from −50 to −20 days prior to AI was associated with reduced CR when compared to cows not exposed to HS (P < 0.001; Table 2). Although CR were numerically lower for cows exposed to HS2 than HS1, the OR and 95% CI were similar, indicating that 1 day of exposure to temperature ≥ 29 ◦ C was sufficient to compromise CR similarly to cows exposed to ADMT ≥29 ◦ C. However, exposure to HS from day −20 to AI, and from AI to pregnancy diagnosis was not associated with reduced CR. Primiparous cows had higher CR than multiparous cows (27.4 versus 24.1%; P < 0.01). The type of insemination protocol, timed AI or AI upon estrus detection, resulted in similar CR (P = 0.92; Table 3). A total of 1459 pregnancies were obtained in 5688 AI performed upon estrus detection, while 481 cows became pregnant of 1945 submitted to the Ovsynch protocol. Table 1 Descriptive statistics of milk yield, days in milk (DIM), and number of artificial inseminations (AI) at AI for lactating cows (Study 1) Item

Mean

Median

S.E.M.

S.D.

Milk (kg per day) DIM Number of AI

37.3 164.8 3.7

37.0 126.0 3.0

0.08 1.36 0.04

6.7 118.7 3.3

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

245

Table 2 Logistic regression model for conception rate (CR) as affected by exposure to heat stress in different days relative to artificial insemination (AI) (Study 1) Day relative to AIa

CR (%)

OR

CI (95%)

−50 to −20 (ADMT) NH (14.3 ◦ C) HS1 (22.3 ◦ C) HS2 (34.0 ◦ C)

31.3 28.8 23.0

1 0.69 0.67

0.55, 0.87 0.54, 0.83

−20 to AI NH (18.6 ◦ C) HS1 (24.5 ◦ C) HS2 (33.6 ◦ C)

30.8 27.6 23.8

1 0.99 1.11

0.80, 1.22 0.85, 1.47

AI to 45 NH (16.8 ◦ C) HS1 (24.5 ◦ C) HS2 (33.5 ◦ C)

26.0 25.7 25.2

1 1.02 1.19

0.81, 1.29 0.96, 1.48

P-value <0.001

0.50

0.11

ADMT: average daily maximum temperature. a NH: no day with temperature ≥29 ◦ C; HS1: at least 1 day with temperature ≥29 ◦ C; HS2: mean daily maximum temperature ≥29 ◦ C. Table 3 Logistic regression model of factors associated with conception rate (CR) in lactating dairy cows (Study 1) Itema

CR (%)

OR

CI (95%)

P-value

Parity Primiparous Multiparous

27.4 24.1

1 0.87

0.78, 0.97

<0.01

Insemination protocol Estrus detection Timed AI

25.7 24.7

1 1.00

0.87, 1.17

0.92

Mastitis Yes No

24.0 25.5

1 1.10

0.79, 1.54

0.58

Retained Placenta Yes No

21.4 25.8

1 1.20

0.99, 1.47

0.07

Milk Fever Yes No

12.9 25.5

1 2.25

1.05, 4.78

0.04

Displaced Abomasum Yes No

22.9 25.5

1 1.13

0.77, 1.66

0.54

Lameness Yes No

25.0 23.5

1 0.98

0.65, 1.48

0.93

a Clinical mastitis occurring between AI and pregnancy diagnosis; clinical lameness diagnosed between 30 days prior to AI and the day of pregnancy diagnosis.

246

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

1.0

Odds for conception

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0

1

2

3

4

5

6

7

8

9

10

Number of AI Fig. 1. Relationship between odds ratio for conception and number of artificial inseminations (AI). Odds = 1.04313 − 0.0696424 × number of AI; r 2 = 0.96. Effect of number of AI according to the multivariate logistic regression model: P < 0.0001 (Study 1).

Occurrence of milk fever was associated with decreased CR, while the occurrence of retained placenta tended to be associated with decreased CR (Table 3). Cows not experiencing milk fever were 2.25 times more likely to conceive than those with milk fever. Likewise, cows that did not experience retained placenta were 1.2 times more likely to conceive than those experiencing retained placenta. Occurrence of displaced abomasum during the first 30 days postpartum, mastitis from the day of AI to pregnancy diagnosis, and lameness from 30 days prior to AI to the day of pregnancy diagnosis were not associated with significant impacts on CR. Based upon the logistic regression, number of AI was associated with CR of lactating cows (P < 0.001), and as the number of AI increased, CR decreased. A negative linear relationship between the OR for conception and number of AI was observed, and cows inseminated more than 9 times were 77% less likely to conceive than animals inseminated once (Fig. 1). An increase in 1 AI performed was associated with a decline in the OR for conception of approximately 7.0% units. Daily milk yield and DIM at AI were not associated with changes in CR of dairy cows (Figs. 2 and 3). Although analysis of daily milk yield was not the goal of this study, production was significantly decreased when cows experienced cases of postpartum diseases (39.7 versus 38.4 kg per day; P < 0.01). Furthermore, the negative effects of postpartum diseases were more pronounced in primiparous than multiparous cows based upon the significant interaction between parity and postpartum disease on milk yield (P < 0.001). 3.2. Study 2 The overall PL from day 31 to 45 after AI was 12.5% (183/1465). Descriptive statistics is depicted in Table 4 for daily milk yield, DIM at AI, and number of AI. As in Study 1, the

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

247

Odds for conception

1.1

1.0

0.9

0.8 20

25

30

35

40

45

50

55

Milk, kg/d Fig. 2. Relationship between odds ratio for conception and milk production. Odds = 0.74475+0.0059333×milk; r 2 = 0.60. Effect of milk yield on conception according to the multivariate logistic regression model: P = 0.39 (Study 1). 1.3

Odds for conception

1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0

100

200

300

400

DIM

Fig. 3. Relationship between odds ratio for conception and days in milk (DIM) at insemination. Odds = 0.986354 − 0.0035116 × DIM + 0.0000086 × DIM2 ; r 2 = 0.30. Effect of DIM on conception according to the multivariate logistic regression model: P = 0.72 (Study 1). Table 4 Descriptive statistics of milk yield, days in milk (DIM), and number of artificial insemination (AI) at AI for lactating cows (Study 2) Item

Mean

Median

S.E.M.

S.D.

Milk (kg PER per day) DIM Number of AI

37.7 131.2 2.8

37.3 100.0 2.0

0.19 2.39 0.06

7.1 91.6 2.4

248

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

Table 5 Logistic regression model for pregnancy loss (PL) as affected by exposure to heat stress in different days relative to artificial insemination (AI) (Study 2) Day relative to AIa

PL (%)

OR

CI (95%)

P-value

−50 to −20 (ADMT) NH (15.1 ◦ C) HS (30.2 ◦ C)

11.3 12.8

1 1.03

0.63, 1.49

0.88

−20 to AI NH (17.4 ◦ C) HS (30.1 ◦ C)

10.3 13.0

1 1.02

0.58, 1.81

0.93

AI to 31 ± 3 NH (19.0 ◦ C) HS (29.6 ◦ C)

8.9 13.5

1 1.07

0.54, 2.13

0.84

31 ± 3 to 45 ± 3 NH (16.0 ◦ C) HS (31.1 ◦ C)

10.0 13.5

1 1.26

0.86, 1.86

0.24

ADMT: average daily maximum temperature. a H: no day with temperature ≥ 29 ◦ C; HS: at least 1 day with temperature ≥ 29 ◦ C or average daily maximum temperature ≥ 29 ◦ C.

ADMT for cows exposed to HS was higher than that for cows in NH during the 4 periods in this study (Table 5). Exposure to HS prior to the initial pregnancy diagnosis by ultrasonography at any moment relative to AI was not associated with PL in lactating dairy cows (P > 0.10; Table 5). This is demonstrated by the similar OR for PL between NH and HS exposed cows for all periods up to the day of ultrasonography. Similarly, exposure to HS between pregnancy diagnoses was not associated with changes in PL (OR = 1.26; CI: 0.86, 1.86; P = 0.24). Primiparous and multiparous cows experienced similar PL (Table 6). Cows inseminated upon estrus detection had similar PL compared to cows inseminated following timed AI. The occurrence of clinical mastitis from the day of AI to pregnancy reconfirmation was Table 6 Logistic regression model of factors associated with pregnancy loss (PL) in lactating dairy cows (Study 2) Itema

PL (%)

OR

CI (95%)

P-value

Parity Primiparous Multiparous

12.4 12.6

1 1.07

0.75, 1.54

0.70

Insemination protocol Timed AI Estrus detection

10.4 13.2

1 1.21

0.82, 1.79

0.33

Mastitisa No Yes

12.3 22.6

1 2.80

1.16, 6.78

0.02

a

Clinical mastitis occurring between the day of AI and day 45 ± 3 after AI.

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

249

Odds for pregnancy loss

2.0

1.5

1.0

1

2

3 4 Number of AI

5

6

Fig. 4. Relationship between odds ratio for pregnancy loss and number of artificial insemination. Odds = 1.14428 − 0.406407 × number of AI + 0.0761071 × number of AI2 ; r 2 = 0.79. Effect of number of AI on pregnancy loss according to the multivariate logistic regression model: P = 0.13 (Study 2).

associated with increased PL (OR = 2.80; CI: 1.16, 6.78; P = 0.02). Cows experiencing clinical mastitis were 2.80 times more likely to lose their pregnancy than those not experiencing mastitis. Although cows with more than 3 AI seem to have experienced greater PL (Fig. 4), increasing DIM at AI was not significantly associated with PL (P = 0.13). Similarly, the multivariate logistic model indicated no significant effect of daily milk yield and number of AI on PL, and Figs. 5 and 6 describe the relationship between those parameters and the OR for PL. 1.7

Odds for pregnancy loss

1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 25

35

45

55

Milk, kg/d Fig. 5. Relationship between odds ratio for pregnancy loss and milk production. Odds = 4.52586 − 0.164355 × milk + 0.0018652 × milk 2 ; r 2 = 0.72. Effect of milk yield on pregnancy loss according to the multivariate logistic regression model: P = 0.48 (Study 2).

250

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

Odds for pregnancy loss

1.2 1.1 1.0 0.9 0.8 0.7 50

100

150

200

DIM at AI Fig. 6. Relationship between odds ratio for pregnancy loss and days in milk (DIM) at artificial insemination (AI). Odds = 0.874489 − 0.00005617 × DIM; r 2 = 0.04. Effect of DIM at AI on pregnancy loss according to the multivariate logistic regression model: P = 0.90 (Study 2).

4. Discussion In Study 1, the average 305 day milk yield for multiparous cows was 39.3 kg per day, whereas primiparous cows averaged 34.5 kg per day (P < 0.001), which characterizes them as high producing dairy cows. Higher milk production makes dairy cows more susceptible to the negative effects of HS on reproductive performance (Hansen and Arechiga, 1999). Exposure to HS from −50 to −20 days prior to AI was associated with reduced CR. High producing cows have less than optimal thermoregulation as a result of the increased heat production from high feed intake and milk production. The thermoneutral temperature for lactating dairy cows is thought to be lower than 26 ◦ C, although it can vary according to the relative humidity (Kadzere et al., 2002). Nevertheless, cows exposed to temperatures above 26 ◦ C have elevated respiration rate and rectal temperature, and this may result in impaired metabolism and reproductive performance (Kadzere et al., 2002). Previous studies have shown that cows exposed to HS during the summer have lower quality oocytes, and this effect is carried out through the autumn (Roth et al., 2001a). Heat stress not only affects quality of oocytes, but also increases the number of degenerate thecal and granulosa cells, compromises steroidogenesis, and might reduce the progesterone production by the CL formed after the ovulation of the compromised follicle, therefore reducing embryo quality and development (Hansen and Arechiga, 1999; Wolfenson et al., 2000; Roth et al., 2001a; Roth et al., 2001b). Collectively, these effects could have contributed to the negative effect of HS on CR of dairy cows observed in Study 1. An interesting aspect of this study is that HS was associated with CR, but only when it occurred from −50 to −20 day before AI. It is known that folliculogenesis in ruminants is a long process (Lussier et al., 1987), and insults early in the development of an oocyte might influence future ability to conceive. Lussier et al. (1987) indicated that a period equivalent to 2

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

251

estrous cycles is required for a follicle to grow through the antral phase, from 0.13 mm to preovulatory size. Therefore, the oocyte and the follicular cells in the ovulatory follicle can potentially be affected by insults that occurred 40–50 days prior to ovulation. Roth et al. (2001a) demonstrated that continuous follicle aspiration of oocytes during autumn, when HS is no longer present, improved quality of previously damaged oocytes from summer HS. The same group also demonstrated a delayed effect of HS on steroidogenesis in preovulatory follicles (Roth et al., 2001b), which might explain the negative association between exposure to HS approximately two estrous cycles prior to AI and CR in lactating dairy cows. Multiparous cows were 13% less likely to conceive than primiparous cows, and this might be partially explained by the higher incidence of postparturient diseases compared to primiparous cows (14.9% versus 6.2%; P < 0.001). Diseases during the early postpartum period are known to affect reproductive performance of lactating dairy cows (Gröhn and Rajala-Schultz, 2000). Therefore, it is possible that older cows experienced lower CR because they were at a higher risk for periparturient problems known to affect fertility. Indeed, we demonstrated an association between CR and retained placenta and milk fever, and cows experiencing those diseases were less likely to conceive than those not experiencing them. Multiparous cows had higher milk yield than primiparous cows, which might have increased the energy demands for milk synthesis, therefore affecting the energy status of cows. In the past decades there has been a consistent increase in average milk production per cow, accompanied by a simultaneous decrease in CR. Lucy (2001), described factors associated with impaired reproductive performance in the modern dairy cow. He demonstrated evidence that the increase in rolling herd average in the last 30 years for dairies served by the Dairy Herd Improvement Association was associated with reduced CR, which has declined from 55% to around 40% in the same period. Our logistic regression model indicated that milk production was not significantly associated with CR. In fact, when the OR for conception was evaluated relative to milk production, it was clear that the OR for conception increased as milk yield increased. When cows were subjected to a timed AI protocol, those with milk production above the mean milk yield early in lactation had higher CR than those with production below the mean milk yield (Peters and Pursley, 2002). It is possible that the relationship observed by Lucy (2001) is a function of the higher incidence of postparturient problems and greater susceptibility to HS, among other factors, in high producing dairy cows, and not necessarily the higher milk production per se affecting CR. Although not evaluated by Peters and Pursley (2002), it is possible that cows with milk production below the mean milk yield experienced more postparturient problems, which not only reduce milk yield (Santos et al., 2004b), but also reduce reproductive performance of lactating dairy cows (Gröhn and Rajala-Schultz, 2000; Santos et al., 2004b). After controlling for several factors, our data indicate that, in the two herds evaluated, milk production was not significantly associated with CR. Therefore, the lower fertility experienced by high producing cows might be related to other factors associated with high milk yield such as postparturient diseases, as suggested previously by Gröhn and Rajala-Schultz (2000). Studies comparing insemination protocols with fixed time AI or insemination at detected estrus indicate mixed results on CR. Cartmill et al. (2001b) observed similar CR for lactating cows subjected to the Ovsynch timed-AI protocol and cows subjected to AI upon estrus detection following the Selectsynch protocol (GnRH followed 7 day by

252

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

a PGF2␣ ). Santos et al. (2004a) indicated that cows inseminated at detected estrus following the Presynch/Selectsynch protocol had higher CR than those inseminated at fixed time following the Presynch/Ovsynch protocol. Likewise, Jobst et al. (2000) demonstrated that cows inseminated at detected estrus induced by PGF2␣ or after spontaneous estrus had higher CR than cows inseminated following timed AI with the Ovysnch protocol. On the other hand, Cerri et al. (2003) observed an increase in CR for cows inseminated at fixed time following the Presynch/Heatsynch protocol compared to cows inseminated upon estrus detection after the Presynch/Selectsynch protocol. In the current study, lactating cows inseminated after observed estrus (5688) had similar CR to cows inseminated following the Ovsynch protocol (1945). An interesting feature of the current study is that only 28.2% of the cows inseminated following the Ovsynch protocol were subjected to a pre-synchronization protocol, indicating that timed AI for resynchronization of non-pregnant cows can be used with no detrimental effect on CR when compared to insemination following induced or spontaneous estrus. Insemination of cows after the second PGF2␣ injection of the pre-synchronization protocol might have in fact reduced CR to the Ovsynch protocol at first AI. Approximately 10–25% of lactating cows are anovulatory by 60–70 DIM (Cerri et al., 2003; Moreira et al., 2001; Santos et al., 2004a), and anovulatory cows have lower CR than cyclic cows (Cerri et al., 2003; Moreira et al., 2001; Santos et al., 2004a). By inseminating cows that displayed signs of estrus after the second PGF2␣ injection of the pre-synchronization protocol might have concentrated more anovulatory cows in the Ovsynch at first postpartum AI. Accuracy of estrus detection and lack of synchronization of ovulation can affect CR when cows are subjected to AI following estrus detection and the Ovsynch protocol, respectively. Santos et al. (2004a) observed that 17 of 254 (6.7%) cycling cows observed in estrus based on removed tail chalk had plasma progesterone > 1.0 ng/ml. Increasing the number of AI was associated with reduced CR, and this was clearly demonstrated by the 7.0% unit decline in the OR for conception with an increase in 1 AI. Because CR was highest at first AI, it is possible that cows with increased number of AI could have become a selected group of lower fertility, which further compromised CR. Furthermore, 9.8% of the cows inseminated only once experienced postparturient diseases, whereas 17.0% of the cows with 6 or more AI experienced the same diseases, and diseases affect CR (Gröhn and Rajala-Schultz, 2000). In addition to those factors, during the first AI when cows were subjected to the Ovsynch protocol, they were also pre-synchronized with 2 injections of PGF2␣ , which improves CR of cows inseminated at fixed time (Moreira et al., 2001). Pregnancy loss ranged from 7.3 to 15.3% across sites. These PL between 31 and 45 days after AI are in accordance with previous studies conducted by our group (Cerri et al., 2003; Chebel et al., 2003; Santos et al., 2001, 2004a), as well as studies conducted by others across the US (Cartmill et al., 2001a; Moreira et al., 2001). In some cases, PL has been reported to be higher than what was observed in the current study, but the interval from the initial and final pregnancy diagnosis was longer and more variable (Cartmill et al., 2001a). In one experiment, lactating cows inseminated at fixed time and exposed to HS experienced an extremely high PL (60.5%) between 27 and 30 and 40–50 days after AI (Cartmill et al., 2001b). Cows exposed to HS experienced similar PL to cows exposed to NH. Previous studies have shown that animals receiving embryos produced under normal environment condition

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

253

have higher CR than animals that undergo AI during the hot season, indicating that embryos exposed to HS in the early stages of development are more sensitive to it than embryos 5 days old or older (Drost et al., 1999). Even though HS prior to and early after AI may affect embryonic development, our data indicates that HS was not associated with PL in 31-day-old fetus. At 31 days of age, the fetus has gone through most of the critical events of its development that are more susceptible to thermal stress (Hansen and Arechiga, 1999). Therefore, once a healthy fetus is diagnosed by ultrasound at 31 days of life it is unlikely that preceding HS would affect its maintenance in the following 14 days. In one study (Cartmill et al., 2001b), lactating cows inseminated at fixed time following the Ovsynch protocol and exposed to HS experienced a 60.5% PL between 27 and 30 and 40–50 days after AI, which was higher than cows not exposed to no HS. We observed no interaction between HS and insemination protocol on PL. Insemination protocol was not associated with PL. Cows inseminated upon estrus detection experienced 13.2% PL and cows inseminated after a timed AI protocol experienced 10.4% PL. Previously, our group demonstrated that cows inseminated upon estrus detection following the Presynch/Selectsynch protocol, or at fixed time following the Presynch/Heatsynch (Cerri et al., 2003) or the Presynch/Ovysnch (Santos et al., 2004a) protocols experienced similar losses of pregnancy in the first 14 or 28 days after the initial pregnancy diagnosis on day 31 postinsemination. In Study 2, only 36.2% of the cows inseminated following the timed AI protocol had been pre-synchronized with 2 injections of PGF2␣ , indicating that timed AI protocols can be utilized for resynchronization of cows without the use of pre-synchronization with no deleterious effect on PL. Occurrence of clinical mastitis from the day of AI to the day of pregnancy reconfirmation was associated with increased risk for PL. Mastitis is generally of bacterial origin, and mastitis caused by both Gram + and Gram − bacteria have been associated with increased PL in lactating dairy cows (Santos et al., 2004b). Similar to our findings, others have also demonstrated an association between clinical mastitis and abortions in lactating dairy cows (Risco et al., 1999). In lactating dairy cows, not only clinical, but also subclinical mastitis seems to negatively influence reproductive performance (Schrick et al., 2001). When clinical or subclinical mastitis occurs prior to or immediately after insemination, lactating cows were less likely to conceive and more likely to experience PL (Santos et al., 2004b; Schrick et al., 2001).

5. Conclusion The high producing dairy cow is sensitive to HS. Cows exposed to HS, as characterized by at least 1 day of maximum temperature above 29 ◦ C, prior to AI had lower CR than cows not exposed to HS. The impacts of HS on CR might be associated to the disruptive effect of high ambient and body temperatures on follicle and oocyte competence, and subsequent fertilization and embryonic development. However, once the pregnancy is established at day 31 after AI, previous or subsequent exposure to HS seems to have little effect on pregnancy maintenance, at least in the following 14 days. The widespread adoption of synchronization protocols that allow for fixed time AI have raised concerns regarding CR and PL. We observed that insemination of cows upon detection of estrus or following the

254

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

Ovsynch protocol resulted in similar CR and PL. Milk production was not significantly associated with CR or PL in lactating dairy cows. Increasing number of AI was associated with reduced CR. Multiparous cows had lower CR than primiparous cows, but parity was not associated with PL. Postparturient diseases such as milk fever and retained placenta were associated with lower CR, and cows experiencing mastitis were more likely to lose their pregnancy between 31 and 45 days after AI. Reproductive programs to improve CR and minimize PL in lactating dairy cows should focus on reducing occurrence of postparturient diseases and mastitis, as well as minimizing exposure to HS prior to and immediately after the moment of insemination.

Acknowledgements The authors thank the collaborating dairies, their owners and staff (Corcoran State Prison Dairy, Rancho Teresita Dairy, and Souza Dairy). References Allison, P.D., 1999. Logistic regression using the SAS® system: Theory and application. SAS Institute Inc., NC, USA. Cartmill, J.A., El Zarkouny, S.Z., Hensley, B.A., Lamb, G.C., Stevenson, J.S., 2001a. Stage of cycle, incidence, and timing of ovulation, and pregnancy rates in dairy cattle after three timed breeding protocols. J. Dairy Sci. 84, 1051–1059. Cartmill, J.A., El-Zarkouny, S.Z., Hensley, B.A., Rozell, T.G., Smith, J.F., Stevenson, J.S., 2001b. An alternative AI breeding protocol for dairy cows exposed to elevated ambient temperatures before or after calving or both. J. Dairy Sci. 84, 799–806. Cerri, R.L.A., Galvão, K.N., Juchem, S.O., Chebel, R.C., Santos, J.E.P., 2003. Timed AI (TAI) with estradiol cypionate (ECP) or insemination at detected estrus in lactating dairy cows. J. Dairy Sci. 86 (Suppl. 1), 181 (Abstract). Chebel, R.C., Santos, J.E.P., Cerri, R.L.A., Galvao, K.N., Juchem, S.O., Thatcher, W.W., 2003. Effect of resynchronization with GnRH on day 21 after artificial insemination on pregnancy rate and pregnancy loss in lactating dairy cows. Theriogenology 60, 1389–1399. Drost, M., Ambrose, J.D., Thatcher, M.J., Cantrell, C.K., Wolfsdorf, K.E., Hasler, J.F., Thatcher, W.W., 1999. Conception rates after artificial insemination or embryo transfer in lactating dairy cows during summer in Florida. Theriogenology 52, 1161–1167. Fricke, P.M., 2002. Scanning the future—ultrasonography as a reproductive management tool for dairy cattle. J. Dairy Sci. 85, 1918–1926. Gröhn, Y.T., Rajala-Schultz, P.J., 2000. Epidemiology of reproductive performance in dairy cows. Anim. Reprod. Sci. 60-61, 605–614. Hansen, P.J., Arechiga, C.F., 1999. Strategies for managing reproduction in the heat-stressed dairy cow. J. Anim. Sci. 77 (Suppl. 2), E37–E50. Hillerton, J.E., Kliem, K.E., 2002. Effective of treatment of Streptococcus uberis clinical mastitis to minimize the use of antibiotics. J. Dairy Sci. 85, 1009–1014. Jobst, S.M., Nebel, R.L., McGilliard, M.L., Peizer, K.D., 2000. Evaluation of reproductive performance in lactating dairy cows with prostaglandin F2␣ , gonadotropin-releasing hormone, and timed artificial insemination. J. Dairy Sci. 83, 2366–2372. Kadzere, C.T., Murphy, M.R., Silanikove, N., Maltz, E., 2002. Heat stress in lactating dairy cows: a review. Livestock Prod. Sci. 77, 59–91. Littell, R.C., Stroup, W.W., Freund, R.J., 2002. SAS for linear models, 4th Edition. SAS Institute Inc., NC, USA.

R.C. Chebel et al. / Animal Reproduction Science 84 (2004) 239–255

255

Lucy, M.C., 2001. Reproductive loss in high-producing dairy cattle: where will it end? J. Dairy Sci. 84, 1277–1293. Lussier, J.G., Matton, P., Dufour, J.J., 1987. Growth rates of follicles in the ovary of the cow. J. Reprod. Fert. 81, 301–307. Macmillan, K.L., Taufa, V.K., Barnes, D.R., Day, A.M., Henry, R., 1988. Detecting estrus in synchronized heifers using tailpaint and aerosol raddle. Theriogenology 30, 1099–1114. MINITAB, 2000. MINITAB® Reference Manual Release 13.32. MINITAB Inc., State College, PA, USA. Moreira, F., Orlandi, C., Risco, C.A., Mattos, R., Lopes, F., Thatcher, W.W., 2001. Effects of presynchronization and bovine somatotropin on pregnancy rates to a time artificial insemination protocol in lactating dairy cows. J. Dairy Sci. 84, 1646–1659. NRC, 2001. Nutrient Requirements of Dairy Cattle. 7th Rev. Ed., Natl. Acad. Sci., Washington, DC, USA. Peters, M.W., Pursley, J.R., 2002. Fertility of lactating dairy cows treated with Ovsynch after presynchronization injections of PGF2␣ and GnRH. J. Dairy Sci. 85, 2403–2406. Pursley, J.R., Kosorok, M.R., Wiltbank, M.C., 1997. Reproductive management of lactating dairy cows using synchronization of ovulation. J. Dairy Sci. 80, 301–306. Risco, C.A., Donovan, G.A., Hernandez, J., 1999. Clinical mastitis associated with abortion in dairy cows. J. Dairy Sci. 82, 1684–1689. Roth, Z., Arav, A., Bor, A., Zeron, Y., Braw-Tal, R., Wolfenson, D., 2001a. Improvement of quality of oocytes collected in the autumn by enhanced removal of impaired follicles from previously heat-stressed cows. Reproduction 122, 737–744. Roth, Z., Meidan, R., Shaham Albalancy, A., Braw-Tal, R., Wolfenson, D., 2001b. Delayed effect of heat stress on steroid production in medium size and preovulatory bovine follicles. Reproduction 121, 745–751. Santos, J.E.P., Juchem, S.O., Cerri, R.L.A., Galvão, K.N., Chebel, R.C., Thatcher, W.W., Dei, C., Bilby, C., 2004a. Effect of bST and reproductive management on reproductive and lactational performance of Holstein dairy cows. J. Dairy Sci. 87, In Press. Santos, J.E.P., Cerri, R.L.A., Ballou, M.A., Higginbotham, G.E., Kirk, J.H., 2004b. Effect of timing of first clinical mastitis occurrence on lactational and reproductive performance of Holstein dairy cows. Anim. Reprod. Sci. 80, In Press. Santos, J.E.P., Thatcher, W.W., Pool, L., Overton, M.W., 2001. Effect of human chorionic gonadotropin on luteal function and reproductive performance of high producing lactating Holstein dairy cows. J. Anim. Sci. 79, 2881–2894. SAS, 2001. SAS/STAT® User’s guide (Release 8.2). SAS Inst. Inc., Cary, NC. Sartori, R., Gumen, A., Guenther, J.N., Souza, A.H., Wiltbank, M.C., 2003. Comparison of artificial insemination (AI) versus embryo transfer (ET) in lactating dairy cows. J. Dairy Sci. 86 (Suppl. 1), 238 (Abstract). Schrick, F.N., Hockett, M.E., Saxton, A.M., Lewis, M.J., Dowlen, H.H., Oliver, S.P., 2001. Influence of subclinical mastitis during early lactation on reproductive parameters. J. Dairy Sci. 84, 1407–1412. Thurmond, M.C., Picanso, J.P., Jameson, C.M., 1990. Considerations for use of descriptive epidemiology to investigate fetal loss in dairy cows. J. Am. Vet. Med. Assoc. 197, 1305–1312. Wolfenson, D., Roth, Z., Meidan, R., 2000. Impaired reproduction in heat-stressed cattle: basic and applied aspects. Anim. Reprod. Sci. 60/61, 535–547.