Energy Balance and Size and Number of Ovarian Follicles Detected by Ultrasonography in Early Postpartum Dairy Cows' M. C. LUCY, C. R. STAPLES, F. M. MICHEL, and W. W. THATCHER' Dairy Science Department Institute of Food and Agricultural Sciences University of Florida Gainesville 32611 ABSTRACT
ular and ovarian function in postpartum lactating dairy cows. (Key words: follicle, energy balance, cows)
Multiparous Holstein cows (n = 52) were fed one of six diets consisting of a totally mixed ration (corn silage, corn grain, soybean meal, dried distillers grains, and whole cottonseed) plus either alfalfa hay, alfalfa cubes, or bermudagrass hay fed chopped as a component in the mixed ration or separate as long hay. predicted energy balance was calculated from DM intake, milk yield and composition, and BW. On d 25 postpartum, ovarian status was programmed by injecting 25 mg of prostaglandin F% and treating cows for 15 d with an intravaginal device containing 1.9 g progesterone. Before d 25, number of class 1 follicles (3 to 5 mm; detected by ultrasonography) decreased with increasing days postpartum, and number of class 3 (10 to 15 mm) and class 4 (>15 mm) follicles increased The number of class 1 and 2 follicles (6 to 9 mm) decreased with increasing energy balance, and number of class 3 follicles increased with energy balance. Before d 25, predicted energy balance explained treatment differences in the number of follicles within each size class. After d 25, energy balance did not affect the average number of follicles per cow, but diet affected the number of follicles within each class. F'redicted energy balance and dietary treatments influenced number of follicles at different times after calving. These results identify the importance of diet and energy balance to follic-
Abbreviation key: CFV = calculated follicular volume, CIDR-B = controlled internal drug releass-bovine, CL = corpus luteum, EB = energy balance, GnRH = gonadotropin-releasing hormone, LH = luteinizing hormone, MP = milk production, NE = net energy, NEB = net energy balance, NE1 = net energy intake, NE, = net energy required for body maintenance, PEB = predicted energy balance. INTRODUCTION
Received February 19, 1990. Accepted September 4, 1990. 'Florida Agricultural Experiment Station Journal Series Number R-00517. +o whom reprint requests should be made. 1991 J Dairy Sci 7 4 4 7 3 4 8 2
The importance of postpartum energy balance (EB) on the recrudescence of normal ovarian cycles in high producing dairy cattle has been recognized in several studies (5). Energy balance has been defined as the difference between the net energy (NJ3) intake of the animal minus the NE required for maintenance and the milk secreted. Dairy cattle undergo an energy deficit in early lactation because maximum milk production is attained prior to maximum feed consumption (3,7). This situation leads to a compensatory response (homeorhesis; 3) involving adipose tissue (increased lipolysis), liver (increased gluconeogenesis and glycogenolysis), muscle (mobilization of protein reserves), and bone (mineral mobilization). Eventually, increased gastrointestinal capacity and activity (turnover) lead to increased energy intake and a positive EB. Calculated EB may reach -15 Mcal/d in some postpartum cows, and an excess of 50 d of lactation may be required to achieve an energy intake that allows for positive EB (4, 31). Deficiencies in energy intake probably will become increasingly commonplace as genetic selection and exogenous hormonal treatments (bST) act to increase aver-
473
474
LUCY ET AL.
TABLE 1. Composition (%DM) and formulated energy content of the six diets fed to cows from 0 to 77 d postpartum.
Dietary treatment Ingredient
11
Beamudagrass hay Alfalfa hay A l M i a cubes Ca fatty acid3
... ... ...
...
... ...
... ... ...
27.8 17.4 7.3 14.5 14.7 3.6 1.74
21.8 17.4 7.3 14.5 14.7 3.6 1.70
27.8 17.4 7.3 14.5 14.7 3.6 1.70
Corn QLage
Ground corn Soybean meal Distillers grains whole cottonseed Minerals and vitamins NE1 Mcavkg
22
14.7
14.7
3 14.7
4
5
6
... ...
14.4
... 14.7 ...
2.2 272 17.1 7.1 14.2 14.4 3.4 1.75
...
... ... 11.4 ...
31.6 15.8 6.7 13.4 14.7 3.1 1.68
37.9 12.1 6.8 13.6 15.3 2.9 1.70
' U n W basal diet formulated to be deficient in nutrients supplied by 2.9 kg DM of long bermudagrass hay of€ered separately. 2~alancedbasal diet with long bennudagmss hay offered separately. 3ca1cium At of palm oil w e e c ) .
age lactational milk yield within dairy herds (23). During the period of energy deficit, the hypothalamic-pituitary-ovarian axis recovers from the influence of previous pregnancy and undergoes active changes leading to the reinitiation of estrous cycles (22). In most cattle, pituitary secretion of luteinizing h m o n e (LH) increases during the first 2 to 3 wk after pamuition (9, 10, 11). Luteinizing hormone acting on ovarian follicles induces waves of follicular growth, which lead to the selection and ovulation of a dominant follicle between 15 and 25 d after calving (15, 29). This ideal scenario apparently is affected by the postpartum EB of the animal. An extremely negative EB will dampen pulsatile secretion of LH (13, 17) and delay ovulation, which contributes to on-farm inefficiency through economic losses associated with postp m anestrus. Possible modulators of these events are opiate peptides acting at the level of the hypothalamus to decrease the release of GnRH and therefore LH (5). Alternatively, metabolically responsive hormones like insulin may act directly on the ovary to influence follicle growth or be transported to the cerebral spinal fluid to modulate hypothalamic function in accordance with energy status (2, 25). Isolation of insulin receptors within the hypothalamus (12) supports the concept of a possible role for insulin in the coordination of EB and reproductive function. Journal of Dairy Science Vol. 74, No. 2. 1991
Ultrasonic examination of the ovary has been shown to be an accurate and reliable method of quantitating ovarian follicle populations (24). The objective of this experiment was to evaluate the recrudescence to ovarian follicular growth using ultrasonography and to determine how EB and diet influence changes in ovarian follicular populations in postpartum dairy cattle whose reproductive cycles had been programmed for scientific study. MATERIALS AND METHODS
Animals and Diet
Fifty-twomultiparous Holstein cows calving between August 15 and November 17, 1988 at the University of Florida Dairy Research Unit (Hague, FL) were used. Animals were housed in open-air free stall barns with free access to unpaved exercise lots from 10 d prior to expected Calving until 77 d p~stpamUn.At C ~ V ing, cows were assigned randomly to one of six diets (treatment), which were fed for the entire experimental period. Diets included a TMR consisting of corn grain, corn silage, soybean meal, distillers grains, and whole cottonseed, supplemented with different species and forms of hay (Table 1). Bermudagrass was fed as long hay (diets 1 and 2) or chopped and mixed with the TMR (diets 3 and 4); alfalfa was cubed mechanically and mixed with the TMR (diet 6)
ENERGY BALANCE AND OVAFUAN FOLLICLES
475
or chopped and mixed with the TMR (diet 5). injected with 25 mg of prostaglandin Fz, (LutaDiet 4 was supplemented with Ca salts of long- lyse; Upjohn Co., Kalamazoo, h4l) and fitted chain fatty acid (2.2%of DM; Megalac, Church with an intravaginal progesterone releasing and Dwight Co., Inc., Princeton, NJ). Ration device (controlled internal drug release-bovine, ingredients were mixed daily and offered in CIDR-B; Eazi-breed, AHI Plastic Co., New two feedings at 0830 and 1430 h in amounts Zealand) containing a total of 1.9 g of progesthat allowed 5 to 10% refusals. The Calm terone for 15 d. This was done to standardize electronic door feeding system (American Ca- the ovarian follicular environment and prevent lan, Inc., Northwood, NJ3) was used to monitor ovulation during a period of reported negative feed intake and refusals of individual cows on a FB (i.e., 3 to 5 wk of lactation). On d 40 postpartum, the CIDR-B was removed, and daily basis. cows were observed for estrus for 14 d. Any cow not expressing estrus after 14 d was subCalculatlon of Energy Balance jected to ovarian ultrasound examination, and Feed intake was determined daily for each those cows not having a corpus luteum at that individual cow (feed offered minus orts). Dry time were diagnosed as having no ovulation (or matter content of the feed and orts were deter- estrus) after CIDR-B removal. The cows were mined by weekly analysis. Nutrient content of first inseminated at the first observed estrus ration components was determined by weekly following the CIDR-B-induced estrous cycle sampling and ration analysis (New York DHIA (about 60 d postpartum). Forage Testing Laboratory, Ithaca, NY). Calculation of the NE1 intake (NEI) was made by Ovarian Ultrasonic Examinatlon multiplying the DM consumption (TMR and Ovaries were examined using an Eguisonic hay) by the calculated AE, of the ration. Daily milk production (MP) was recorded model 300A linear array ultrasonic scanner electronically and B W measurements were equipped with a 7.5-MHz transducer CEQUisonmade weekly. Milk fat composition was deter- ics 300A, Keiki, Tokyo, Japan). Briefly, fecal mined by the DHLQ testing laboratory (Raleigh, material was removed from the rectum, and the NC) for weekly samples of four consecutive transducer was inserted into the rectum and (a.m. and p.m.) milkings. Net energy required held adjacent to the ovary. Serial secretions for body maintenance @ was I& calculated ,,) were examine size (diameter) and number of using the equation & = (kg BW.75) x .OS ovarian follicles (23 mm) were recorded onto (21). Daily BW of cows between measurements follicular maps. Ovarian follicles were orgawas estimated by h e a r interpolations. Net en- nized into four classes based on diameter: class ergy in the secreted milk was calculated using 1 , 3 to 5 mm; class 2 , 6 to 9 mm, class 3,lO to the equation NE1 = MF' x (.312 + [.0962 x fat 15 mm, and class 4, >15 mm. In addition, %]) (21). Net energy balance (NEB) was der- presence and diameter of corpora lutea (CL) ived on a daily basis using the equation NEB = were recorded. The relationship between calcuNE1 - NE, - NE1. Values then were subjected lated follicular volume (CFV, mm3; based on to a third-order regression analysis for each diameter determination by ultrasound) and aspicow to achieve a smoothed El3 curve. Predicted rated follicular fluid volume ( A t cm3) for EB (PEB) values h m these regressions for follicles greater than 5 mm in diameter was each day (0 to 77) then were used in statistical CFV = -73 + 1482 x AV (R2 = .73; K. L. Macmillan, W. W. Thatcher, unpublished obanalyses. servations). Ultrasound examinations occurred before inReproductive Management sertion of the CIDR-B (d 7, 16, 18,20,22,25), For the first 25 d postpartum, cows were during the period of CIDR-B insertion (d 30, treated for uterine or metabolic disorders. Ovar- 35, and 40),and after removal of the CIDR-B ian disorders (i.e., cystic structures), if diag- (every other day until estrus, and then on d 6, nosed, were not treated. On d 25, all cows were 12, and 18 of the subsequent estrous CycIe). Journal of Dairy Science Vol. 74, No. 2, 1991
476
LUCY ET AL.
Blood Collection and Analysis
Ten milliliters of blood were collected by coccygeal venipuncture on each day that a cow was examined by ultrasonography. Blood was collected into evacuated heparinized (143 units) tubes (Vacutainer, Becton Dickinson, East Rutherford, NJ)and put on ice and centrifuged at 11,OOO x g for 20 min within 30 min of collection. Plasma NEFA, glucose, and insulin were measured on d 7, 16, 20, 25, 30, 35, and 40 postpartum. Plasma NEFA were measured by an enzymatic colorimeaic method using a kit (Wako Pure Chemical Industries, Ltd., Osaka, Japan). Plasma glucose was measured in duplicate samples of (25 pl) using an oxidaseperoxidase colorimetric method (Glucose Number 510; Sigma Chemical Co., St. Louis, MO). Plasma insulin was m e a d in duplicate samples (150 pl) by radioimmunoassay described previously (6). Intraassay and interassay CV for insulin analysis were 2 and 12%, respectively. Statlstlcai Analyses
Total number of follicles within each size class was analyzed using the general linear models procedure of SAS (26). Models used included treatment (diet), animal nested within treatment, day postpartum, follicle size class, second- and thirdader interactions, and residual. Predicted EB for the specific day in which follicular data were recorded was included as a covariate in the analyses. Interactions of PEB with class variables also were tested. During the CIDR-B and post-CIDR-B periods, main effect of day was not significant, so models included treatment, animal within treatment, class, treatment by class interaction, and residual. In addition, the relationship between NEFA and EB, as well as interval to first ovulation and EB were tested by linear regression. Plasma insulin and glucose were analyzed using a model that included treatment, animal within treatment, day postpartum, the treatment by day interaction, and residual. Tests of significance for the effects of treatment were performed using animal within treatment as the error term. Other interactions and main effects were tested using the pooled residual. Signifcance was declared at Pc.10 unless otherwise noted. Journal of Dairy Scieace Vol. 74, No. 2, 1991
6
m
15
20
25
DAY POSTPARTUM
Pigum 1. Average number of follicles within different size classes (class 1.3 to 5 mm; class 2.6 to 9 mm; class 3, 10 to I5 mm; class 4, >15 mm) in cattle (n = 52) from d 7 to 25 after calving.
RESULTS
Follicular Development Before Day 25
Average number of follicles for each size class on successive days postpartum is presented in Figure 1. Before d 25, the average number of small (class 1) follicles decreased, whereas the number of large (classes 3 and 4) follicles increased with increasing days postpartum (follicular class by day interaction, Pc.01). Dietary treatment did not influence the total number of follicles recorded. However, a treatment by class interaction was found (Pc.06), suggesting that the number of follicles within each follicular size class was not similar among diets (Table 2). Including the PEB by class interaction in the statistical model increased the probability associated with the treatment by class interaction to 0.46 (day by class interaction remained significant, P
ular and ovarian function in postpartum lactating dairy cows. (Key words: follicle, energy balance, cows)
Multiparous Holstein cows (n = 52) were fed one of six diets consisting of a totally mixed ration (corn silage, corn grain, soybean meal, dried distillers grains, and whole cottonseed) plus either alfalfa hay, alfalfa cubes, or bermudagrass hay fed chopped as a component in the mixed ration or separate as long hay. predicted energy balance was calculated from DM intake, milk yield and composition, and BW. On d 25 postpartum, ovarian status was programmed by injecting 25 mg of prostaglandin F% and treating cows for 15 d with an intravaginal device containing 1.9 g progesterone. Before d 25, number of class 1 follicles (3 to 5 mm; detected by ultrasonography) decreased with increasing days postpartum, and number of class 3 (10 to 15 mm) and class 4 (>15 mm) follicles increased The number of class 1 and 2 follicles (6 to 9 mm) decreased with increasing energy balance, and number of class 3 follicles increased with energy balance. Before d 25, predicted energy balance explained treatment differences in the number of follicles within each size class. After d 25, energy balance did not affect the average number of follicles per cow, but diet affected the number of follicles within each class. F'redicted energy balance and dietary treatments influenced number of follicles at different times after calving. These results identify the importance of diet and energy balance to follic-
Abbreviation key: CFV = calculated follicular volume, CIDR-B = controlled internal drug releass-bovine, CL = corpus luteum, EB = energy balance, GnRH = gonadotropin-releasing hormone, LH = luteinizing hormone, MP = milk production, NE = net energy, NEB = net energy balance, NE1 = net energy intake, NE, = net energy required for body maintenance, PEB = predicted energy balance. INTRODUCTION
Received February 19, 1990. Accepted September 4, 1990. 'Florida Agricultural Experiment Station Journal Series Number R-00517. +o whom reprint requests should be made. 1991 J Dairy Sci 7 4 4 7 3 4 8 2
The importance of postpartum energy balance (EB) on the recrudescence of normal ovarian cycles in high producing dairy cattle has been recognized in several studies (5). Energy balance has been defined as the difference between the net energy (NJ3) intake of the animal minus the NE required for maintenance and the milk secreted. Dairy cattle undergo an energy deficit in early lactation because maximum milk production is attained prior to maximum feed consumption (3,7). This situation leads to a compensatory response (homeorhesis; 3) involving adipose tissue (increased lipolysis), liver (increased gluconeogenesis and glycogenolysis), muscle (mobilization of protein reserves), and bone (mineral mobilization). Eventually, increased gastrointestinal capacity and activity (turnover) lead to increased energy intake and a positive EB. Calculated EB may reach -15 Mcal/d in some postpartum cows, and an excess of 50 d of lactation may be required to achieve an energy intake that allows for positive EB (4, 31). Deficiencies in energy intake probably will become increasingly commonplace as genetic selection and exogenous hormonal treatments (bST) act to increase aver-
473
474
LUCY ET AL.
TABLE 1. Composition (%DM) and formulated energy content of the six diets fed to cows from 0 to 77 d postpartum.
Dietary treatment Ingredient
11
Beamudagrass hay Alfalfa hay A l M i a cubes Ca fatty acid3
... ... ...
...
... ...
... ... ...
27.8 17.4 7.3 14.5 14.7 3.6 1.74
21.8 17.4 7.3 14.5 14.7 3.6 1.70
27.8 17.4 7.3 14.5 14.7 3.6 1.70
Corn QLage
Ground corn Soybean meal Distillers grains whole cottonseed Minerals and vitamins NE1 Mcavkg
22
14.7
14.7
3 14.7
4
5
6
... ...
14.4
... 14.7 ...
2.2 272 17.1 7.1 14.2 14.4 3.4 1.75
...
... ... 11.4 ...
31.6 15.8 6.7 13.4 14.7 3.1 1.68
37.9 12.1 6.8 13.6 15.3 2.9 1.70
' U n W basal diet formulated to be deficient in nutrients supplied by 2.9 kg DM of long bermudagrass hay of€ered separately. 2~alancedbasal diet with long bennudagmss hay offered separately. 3ca1cium At of palm oil w e e c ) .
age lactational milk yield within dairy herds (23). During the period of energy deficit, the hypothalamic-pituitary-ovarian axis recovers from the influence of previous pregnancy and undergoes active changes leading to the reinitiation of estrous cycles (22). In most cattle, pituitary secretion of luteinizing h m o n e (LH) increases during the first 2 to 3 wk after pamuition (9, 10, 11). Luteinizing hormone acting on ovarian follicles induces waves of follicular growth, which lead to the selection and ovulation of a dominant follicle between 15 and 25 d after calving (15, 29). This ideal scenario apparently is affected by the postpartum EB of the animal. An extremely negative EB will dampen pulsatile secretion of LH (13, 17) and delay ovulation, which contributes to on-farm inefficiency through economic losses associated with postp m anestrus. Possible modulators of these events are opiate peptides acting at the level of the hypothalamus to decrease the release of GnRH and therefore LH (5). Alternatively, metabolically responsive hormones like insulin may act directly on the ovary to influence follicle growth or be transported to the cerebral spinal fluid to modulate hypothalamic function in accordance with energy status (2, 25). Isolation of insulin receptors within the hypothalamus (12) supports the concept of a possible role for insulin in the coordination of EB and reproductive function. Journal of Dairy Science Vol. 74, No. 2. 1991
Ultrasonic examination of the ovary has been shown to be an accurate and reliable method of quantitating ovarian follicle populations (24). The objective of this experiment was to evaluate the recrudescence to ovarian follicular growth using ultrasonography and to determine how EB and diet influence changes in ovarian follicular populations in postpartum dairy cattle whose reproductive cycles had been programmed for scientific study. MATERIALS AND METHODS
Animals and Diet
Fifty-twomultiparous Holstein cows calving between August 15 and November 17, 1988 at the University of Florida Dairy Research Unit (Hague, FL) were used. Animals were housed in open-air free stall barns with free access to unpaved exercise lots from 10 d prior to expected Calving until 77 d p~stpamUn.At C ~ V ing, cows were assigned randomly to one of six diets (treatment), which were fed for the entire experimental period. Diets included a TMR consisting of corn grain, corn silage, soybean meal, distillers grains, and whole cottonseed, supplemented with different species and forms of hay (Table 1). Bermudagrass was fed as long hay (diets 1 and 2) or chopped and mixed with the TMR (diets 3 and 4); alfalfa was cubed mechanically and mixed with the TMR (diet 6)
ENERGY BALANCE AND OVAFUAN FOLLICLES
475
or chopped and mixed with the TMR (diet 5). injected with 25 mg of prostaglandin Fz, (LutaDiet 4 was supplemented with Ca salts of long- lyse; Upjohn Co., Kalamazoo, h4l) and fitted chain fatty acid (2.2%of DM; Megalac, Church with an intravaginal progesterone releasing and Dwight Co., Inc., Princeton, NJ). Ration device (controlled internal drug release-bovine, ingredients were mixed daily and offered in CIDR-B; Eazi-breed, AHI Plastic Co., New two feedings at 0830 and 1430 h in amounts Zealand) containing a total of 1.9 g of progesthat allowed 5 to 10% refusals. The Calm terone for 15 d. This was done to standardize electronic door feeding system (American Ca- the ovarian follicular environment and prevent lan, Inc., Northwood, NJ3) was used to monitor ovulation during a period of reported negative feed intake and refusals of individual cows on a FB (i.e., 3 to 5 wk of lactation). On d 40 postpartum, the CIDR-B was removed, and daily basis. cows were observed for estrus for 14 d. Any cow not expressing estrus after 14 d was subCalculatlon of Energy Balance jected to ovarian ultrasound examination, and Feed intake was determined daily for each those cows not having a corpus luteum at that individual cow (feed offered minus orts). Dry time were diagnosed as having no ovulation (or matter content of the feed and orts were deter- estrus) after CIDR-B removal. The cows were mined by weekly analysis. Nutrient content of first inseminated at the first observed estrus ration components was determined by weekly following the CIDR-B-induced estrous cycle sampling and ration analysis (New York DHIA (about 60 d postpartum). Forage Testing Laboratory, Ithaca, NY). Calculation of the NE1 intake (NEI) was made by Ovarian Ultrasonic Examinatlon multiplying the DM consumption (TMR and Ovaries were examined using an Eguisonic hay) by the calculated AE, of the ration. Daily milk production (MP) was recorded model 300A linear array ultrasonic scanner electronically and B W measurements were equipped with a 7.5-MHz transducer CEQUisonmade weekly. Milk fat composition was deter- ics 300A, Keiki, Tokyo, Japan). Briefly, fecal mined by the DHLQ testing laboratory (Raleigh, material was removed from the rectum, and the NC) for weekly samples of four consecutive transducer was inserted into the rectum and (a.m. and p.m.) milkings. Net energy required held adjacent to the ovary. Serial secretions for body maintenance @ was I& calculated ,,) were examine size (diameter) and number of using the equation & = (kg BW.75) x .OS ovarian follicles (23 mm) were recorded onto (21). Daily BW of cows between measurements follicular maps. Ovarian follicles were orgawas estimated by h e a r interpolations. Net en- nized into four classes based on diameter: class ergy in the secreted milk was calculated using 1 , 3 to 5 mm; class 2 , 6 to 9 mm, class 3,lO to the equation NE1 = MF' x (.312 + [.0962 x fat 15 mm, and class 4, >15 mm. In addition, %]) (21). Net energy balance (NEB) was der- presence and diameter of corpora lutea (CL) ived on a daily basis using the equation NEB = were recorded. The relationship between calcuNE1 - NE, - NE1. Values then were subjected lated follicular volume (CFV, mm3; based on to a third-order regression analysis for each diameter determination by ultrasound) and aspicow to achieve a smoothed El3 curve. Predicted rated follicular fluid volume ( A t cm3) for EB (PEB) values h m these regressions for follicles greater than 5 mm in diameter was each day (0 to 77) then were used in statistical CFV = -73 + 1482 x AV (R2 = .73; K. L. Macmillan, W. W. Thatcher, unpublished obanalyses. servations). Ultrasound examinations occurred before inReproductive Management sertion of the CIDR-B (d 7, 16, 18,20,22,25), For the first 25 d postpartum, cows were during the period of CIDR-B insertion (d 30, treated for uterine or metabolic disorders. Ovar- 35, and 40),and after removal of the CIDR-B ian disorders (i.e., cystic structures), if diag- (every other day until estrus, and then on d 6, nosed, were not treated. On d 25, all cows were 12, and 18 of the subsequent estrous CycIe). Journal of Dairy Science Vol. 74, No. 2, 1991
476
LUCY ET AL.
Blood Collection and Analysis
Ten milliliters of blood were collected by coccygeal venipuncture on each day that a cow was examined by ultrasonography. Blood was collected into evacuated heparinized (143 units) tubes (Vacutainer, Becton Dickinson, East Rutherford, NJ)and put on ice and centrifuged at 11,OOO x g for 20 min within 30 min of collection. Plasma NEFA, glucose, and insulin were measured on d 7, 16, 20, 25, 30, 35, and 40 postpartum. Plasma NEFA were measured by an enzymatic colorimeaic method using a kit (Wako Pure Chemical Industries, Ltd., Osaka, Japan). Plasma glucose was measured in duplicate samples of (25 pl) using an oxidaseperoxidase colorimetric method (Glucose Number 510; Sigma Chemical Co., St. Louis, MO). Plasma insulin was m e a d in duplicate samples (150 pl) by radioimmunoassay described previously (6). Intraassay and interassay CV for insulin analysis were 2 and 12%, respectively. Statlstlcai Analyses
Total number of follicles within each size class was analyzed using the general linear models procedure of SAS (26). Models used included treatment (diet), animal nested within treatment, day postpartum, follicle size class, second- and thirdader interactions, and residual. Predicted EB for the specific day in which follicular data were recorded was included as a covariate in the analyses. Interactions of PEB with class variables also were tested. During the CIDR-B and post-CIDR-B periods, main effect of day was not significant, so models included treatment, animal within treatment, class, treatment by class interaction, and residual. In addition, the relationship between NEFA and EB, as well as interval to first ovulation and EB were tested by linear regression. Plasma insulin and glucose were analyzed using a model that included treatment, animal within treatment, day postpartum, the treatment by day interaction, and residual. Tests of significance for the effects of treatment were performed using animal within treatment as the error term. Other interactions and main effects were tested using the pooled residual. Signifcance was declared at Pc.10 unless otherwise noted. Journal of Dairy Scieace Vol. 74, No. 2, 1991
6
m
15
20
25
DAY POSTPARTUM
Pigum 1. Average number of follicles within different size classes (class 1.3 to 5 mm; class 2.6 to 9 mm; class 3, 10 to I5 mm; class 4, >15 mm) in cattle (n = 52) from d 7 to 25 after calving.
RESULTS
Follicular Development Before Day 25
Average number of follicles for each size class on successive days postpartum is presented in Figure 1. Before d 25, the average number of small (class 1) follicles decreased, whereas the number of large (classes 3 and 4) follicles increased with increasing days postpartum (follicular class by day interaction, Pc.01). Dietary treatment did not influence the total number of follicles recorded. However, a treatment by class interaction was found (Pc.06), suggesting that the number of follicles within each follicular size class was not similar among diets (Table 2). Including the PEB by class interaction in the statistical model increased the probability associated with the treatment by class interaction to 0.46 (day by class interaction remained significant, P
