GLUCONEOGENIC SUPPLEMENTS DO NOT AFFECT PRODUCTION , REPRODUCTIVE TRAITS AND BLOOD METABOLITE OF HOLSTEIN COWS DURING THE TRANSITION PERIOD

The use of gluconeogenic supplements for dairy cows during the transition period has produced contradictory responses in the literature, making it difficult to recommend them. The objective of this trial was to evaluate supplementation with propylene glycol (PG), calcium propionate (CP), and “Dairy Power Drench” (DR) on the transitional period of Holstein cows. Parameters studied were variation of body condition score (BCS), body weight (BW), milk production (MP), reproductive efficiency and nonesterified fatty acid concentration (NEFA). One hundred and sixty five animals from two commercial herds were used. Treatments consisted of: CControl; DRadministration of “Dairy Power Drench” (3 applications) during postpartum; CPdaily administration of calcium propionate (500 g); PGdaily administration of propylene glycol (500 mL). Variation sources studied were the effect of treatments and blocks as function of farm and parity. No interaction between time (weeks) and treatments, or treatment effects, were found for BCS. However, there was a quadratic effect of time. Body weight and its variation were not affected by treatment nor by the time × treatment interaction. However, a quadratic effect of time was observed. An interaction time × treatment was observed on MP, but possible differences within each week were not detected. Treatments also did not affect reproductive efficiency parameters. On average, the number of days to first detected postpartum estrus was 69.5, with 2.23 services per conception and 172.6 days open. The mean concentration of nonesterified fatty acids was 376.6 mmol L and no effect of the interaction time × treatment, or of treatment, was observed. However, a linear effect was observed with time, with a decrease of 48.2 mmol L per week.


INTRODUCTION
Milk production in recent decades has experienced marked increases as a result of intense genetic selection, advances in nutritional management, improvements in disease control, and other management practices (Miyoshi et al., 2001).The globalization of the economy has facilitated the commercialization of semen, embryos, and animals, and this has contributed to increase the world milk production, especially regarding genetic materials of Canadian and American origin (Philipson, 1990).A few basic points to observe to achieve high efficiency in intensive production systems are the use of specialized animals, and sanitary, nutritional and reproductive management practices as near as possible to the ideal, thus offering adequate comfort conditions, so the animals can respond with high yields.Management, in addition to the environment are the main causes of variation in reproduction and milk yield (Nebel & McGillard, 1993); milk production and reproductive performance play the most relevant roles in determining the profitability of dairy herds (Arbel et al., 2001).
Cow's efficiency in the subsequent lactation is determined during the transition period, and problems during this period result in losses from 4.5 to 9.0 kg day -1 in the peak of lactation, representing economic losses of U$300 to U$600 per lactation under consideration (Drackley, 1999).The transition period, according to Schlatter (1997), varies from three weeks before calving to different postpartum periods.This postpartum period may vary from three weeks or until the cow reaches lactation peak, with variations according to the management adopted for the animal.This period is characterized by great metabolic stress, and manifested by a decrease in voluntary feed intake, intense mobilization of body reserves, and increase in nutritional requirements, with consequences in yield and reproduction.
Nonesterified fatty acids (NEFA) are the main source of energy during the end of gestation and the beginning of lactation.Its concentrations in the plasma obey a growing pattern starting on the third prepartum week, with a peak near calving, decreasing through lactation (Drackley, 1998).
On the other hand, serum insulin concentrations follow an inverse pattern, decreasing considerably as calving approaches (Studer et al., 1993).Therefore, feeding the cow during this period is a challenge, since this stage is marked by endocrine changes, responsible for preparing the cow for the final growth of the fetus, calving, and lactogenesis (Grummer, 1995).
Even though a well-balanced diet may reduce problems in the transition period, the program can be compromised if a decrease in intake occurs, which leads to dietary deficiencies in energy and other nutrients.In this context, forced oral administration of nutrients could be a useful tool to increase blood concentrations of calcium, minimize energy deficiency, rehydrate the animal and stimulate ruminal fermentation (Stokes & Goff, 2001).Studer et al. (1993) demonstrated that the administration of energy supplementation in the transition period reduced the plasma concentrations of ketone bodies and FFA, and increased the concentrations of insulin, during the supplying period.
Researches have been conducted to find energy alternatives that would diminish the energy deficiency in the transition period, allowing the animals to show yields and reproduction compatible with their potential.In this sense, the use of plant and animal fats, propylene glycol, sodium propionate, calcium propionate, and ionophores have been evaluated.Another alternative has been the use of "Dairy Power Drench ® ", even though no literature data is available about this product.
Therefore, the objective of this study was to evaluate organic responses to the use of neoglycogenic substances on physical condition, yield, reproduction, and blood parameter, when administered daily and orally to dairy cows, from two weeks before the estimated calving date until seven weeks postpartum (calcium propionate and propylene glycol), and three postpartum applications, on consecutive days, beginning on the calving day (Dairy Power Drench ® ).

MATERIAL AND METHODS
Trials were carried out at two commercial doing farms (Holstein), located Araras (Farm 1) and Engenheiro Coelho (Farm 2), State of São Paulo, Brazil.One hundred and sixty five purebred and crossbred Holstein animals, 132 cows and 33 heifers, with calvings between March 15 and June 15, 1999 were used.Data collection extended from March 15 to August 12, 1999.Animals were managed under "free-stall" feedlot systems and total mixed ration (TMR) programs.Animals were not separated from the others and the management system adopted by each property was obeyed.

Experimental treatments consisted of a control and three application protocols:
Treatment 1 (control) -this group consisted of 31 cows and ten heifers submitted to the same management conditions as the animals that received the treatments.
Treatment 2 -oral administration of "Dairy Power Drench ®1 " to 42 animals (31 cows and 11 heifers), which received the treatments according to the manufacturer's recommendations: three applications, the first on the day of calving and the other two on the two Sci.Agric.(Piracicaba, Braz.), v.61, n.4, p.376-385, July/August 2004 consecutive days.The product, sold in 500 g packs, was diluted in 25 to 30 liters of warm water and administered orally with a feeding catheter.Product composition, according to the manufacturer, is: yeast (Yea-Sacc), calcium carbonate, sodium bicarbonate, potassium chloride, cobalt sulfate, magnesium sulfate, chelated chromium, Bcomplex vitamins, propylene glycol, dextrose, and flavorants.
Treatment 3 -daily oral administration of propylene glycol (500 ml day -1 ) to 43 animals (33 cows and ten heifers), from thirteen days before the estimated calving date until 49 days postpartum.Administration was done with the help of an extra person, always in the morning.
Treatment 4 -daily oral administration of 500 g calcium propionate to 39 animals (37 cows and two heifers), from 11 days before the estimated calving date until 51 days postpartum.The 500 g of product were placed in 1-L glass bottles and the volume was completed to 1 L, agitated to homogeneize the mix and provided to the animals at the same times as in the previous group.Obs.: in this group two heifers died and one cow had health problem due to product aspiration.
Body condition score (BCS) was evaluated weekly, with a scale from 1 (thin) to 5 (obese) 0.25 point intervals were adopted (Wildman et al., 1982).Body score variation was obtained from the difference between BCS in week six and week one for each animal.
Body weight was determined with a weight tape, on the same day as the body condition evaluation, always by the same person, to avoid possible assessor interferences.Body weight variation was obtained from the difference between weight in the 6 th and 1 st weeks for each animal.The daily variation in body weight corresponds to the daily weight loss during the experimental period.Animals were milked three times a day and milk yield was measured individually once a week for all animals.
Blood was sampled once a week in "vacuntainer" tubes, containing potassium oxalate and sodium fluoride, for nonesterified fatty acid (NEFA) determination.Samples (n=2) were placed in an isothermal box until they arrived at the centrifugation site, which took approximately 3 hours, and were then centrifuged at 3000 × g for 20 minutes.Plasma was placed in Eppendorf tubes, (n=2) and stored in freezer at -20°C for later analysis.Chemical analyses of nonesterified fatty acids were performed with the NEFA C KIT (WAKO Chemicals, Code no.994-75409E).This test is based on the colorimetric enzymatic method (D.L. Palmquist, 1999; personal communication) adapted for use with microtitration plates, and later reading on an "Elisa Reader" device, with a 550 nm filter, results presented as micromoles per liter.
Number of days to first postpartum estrus, days open, and number of services per conception were also evaluated.Heat was detected twice a day by visual observation.
Samples of TMR were collected monthly, and at every two collections a sample was taken and stored at -18°C for later chemical analysis, which was performed in the chemical analysis laboratory of AGRIBRANDS; the DM, CP, EE, CF, ash, Ca and P analyses were performed according to AOAC (1985); the NDF and ADF analyses were performed according to Van Soest et al. (1991).Chemical composition of diets is presented in Table 1.
Data were submitted to analysis of variance, with separation, as causes of variation, of the treatment effects and effects of blocks formed as a function of farm (Farm 1 and 2) and number of calvings (primiparous or multiparous).Analyses corresponding to milk yield values, body condition scores, body weight, daily variation of body weight, and nonesterified fatty acids were performed as previously described, but added of split-plot in time factor (weeks), corresponding to the several weekly data collection moments.Probabilities of interactions with time were determined by the Greenhouse-Geisse test, using the "REPEATED" command generated by the GLM procedure (PROC GLM) (SAS Institute Inc., 1985).Analyses by time were only performed when interactions between time and treatments were significant.The time effect evaluation within each treatment was performed by regression, using polynomial contrasts (PROC GLM).In the presence of linear or quadratic effects, the corresponding regression equation and coefficient of determination (R 2 ) were obtained.In the presence of treatment effects, comparisons among treatment means were made by Tukey's test.A 5% significance level was used in all tests performed.

RESULTS AND DISCUSSION
No interaction was found between farm and treatment in any of the studied variables.No interaction was observed between treatment and time for the body con- dition score variable.However, there was a quadratic effect of time on BCS (P=0.0001)(Table 2 and Figure 1).The equation was: BCS=3.433-0.135 week + 0.0118 week 2 , and the point of inflexion of the curve was approximately the 40 th day, or 6 th week postpartum.
BCS for the group treated with propylene glycol was 3.0.This result is similar to the 3.11 value found by Fonseca (1997), who used only 15 doses of 300 mL propylene glycol, in daily administrations (from the fourth day prepartum until calving) and on alternate days (from the 10 th to the 6 th day prepartum and from calving to the 16 th day postpartum).Studer et al. (1993) also found similar results (BCS of 3.1) to those obtained in the present experiment for cows that received 1 L PG, daily, during the 7 days that preceded the estimated calving date.Likewise, Juchem (2000) provided a 300 mL dose daily and the animals showed a BCS of 3.13 in the sixth week of lactation, but showed higher milk yields (31.2 kg) than animals in this trial.Perhaps the use of a smaller amount of propylene glycol might result in the same gluconeogenic effect, according to results obtained by Grummer et al. (1994), or maybe higher PG doses could decrease dry matter intake, according to Christensen et al. (1997).
Although the BCS data are different in different experiments, Gearhart et al. (1990) define the ideal body condition for cows at each lactation stage as that which optimizes milk yield, minimizes health problems and reproductive disorders, and maximizes the economic return of the farm.The ideal management for the cow during the transition period is one where calving occurs under good body condition, the incidence of fatty liver is reduced, dry matter intake is maximized, negative energy balance is not marked, and functions and reproductive efficiency are not reduced (Grummer, 1995).Also, no difference (P=0.7774) was verified in body condition variation among treatments during the experimental period (Table 2).On average, the cows lost 0.28 body score unit.Therefore, considering that loss of one unit represents 84 kg on body weight reduction (NRC, 2001), cows lost approximately 23.5 kg in 49 days after calving.The BCS for cows treated with calcium propionate was 3.2 at the end of the trial; this represents a 0.30 unit decrease when compared with the initial week, while Aslup & Schlatter (1998) obtained a 0.22 decrease, despite the fact that those authors only used 114 g of the same product.
Weight loss did not differ (P=0.8698)among treatments (Table 3 and Figure 1), and no effect of the time and treatments interaction was observed (P=0.8223).However, body weight showed a quadratic effect of time (P=0.0001),as follows: BW=571.952-15.207 week + 1.3729 week 2 , the inflexion point of the curve occurring approximately on the 39 th day or in the 6 th and a half week after calving.
Mean cow's weight during the experimental period, considering all treatments, was 539.6 kg (558.2 in the beginning and 530.2 kg at the end).These results concerning weight loss are in agreement with Butler & Canfield (1989), who stated that in the beginning of lactation the energy from feed intake is smaller than maintenance and production requirements.This is why cows with a high milk yield use their body reserves to supply such deficiencies, resulting in weight and body condition losses.
Cows that are obese at calving mobilize more body reserves during postpartum, because they are submitted to a more severe negative energy balance (Rukkwamsuk et al., 1998).During this stage, the use of  body reserves is indispensable to maintain milk yield until the requirements are completely fulfilled by feed intake.Energy that comes from lipid mobilization can represent up to 33% of the energy used for milk synthesis (Moe, 1965).Komaragiri & Erdman (1997) observed an average mobilization of 54 kg fat and 21 kg protein in Holstein cows, from 2 weeks prepartum until the 5 th week of lactation.In another study, Holcomb et al. (2001) seved postpartum weight losses of 38 kg for animals that, during the prepartum period, consumed concentrate diets in a restricted manner; therefore, those losses were almost 100% above the mean values observed in this experiment, even though their results were based on different diets and without energy supplementation.
On average, cows lost 0.79 kg day -1 during the entire experimental period (Table 4 and Figure 1).For this variable, no interaction between time and treatment (P=0.6252),nor a treatment effect (P=0.6684), was observed.However, a quadratic effect of time (P=0.0375)was verified on this variable, with the following equation: weight variation = -2.180+ 0.6218 week -0.0515 week 2 .The inflexion point of this curve was verified approximately on the 42 nd day, or the 6 th week postpartum.Buckley et al. (2000) observed that cows with high (28.2kg) and medium (25.0 kg) genetic merits for milk yield experienced weight losses of 0.80 and 0.65 kg day -1 , from the first to the fourth lactation week, respectively, which corresponds to approximately 22.4 kg for the first group and 18.2 kg for the second; these values are quite near to those obtained in this experiment.
Cows produced, on average, 27.3 kg milk day -1 (Table 5 and Figure 1) during the experiment, starting at 23.0 kg and finishing in the sixth week at 29.7 kg.An interaction was verified between the main factors, time and treatment (P=0.0076), for the variable under consideration.However, within each week, the analysis of variance did not detect any differences (Table 5 and Figure 1).
As time passed, yield increases were 12.5% in the second week relative to the first, 6.1% in the third week relative to the second, and so, with values of 3.8%, 3.4%, and 0.8% (Table 5).The control, "Drench", and propylene glycol treatments showed a linear effect of time (P=0.0001),and yield increases of 1.355, 1.337 and 0.999 kg were observed at each week, respectively.In turn, the calcium propionate treatment showed a quadratic effect with time (P=0.0072).
Mean daily milk yield of animals supplemented with PG in this experiment was 26.5 kg, which is lower than the 31.2kg found by Miyoshi et al. (2001), who supplied the animals with the same PG dose (500 mL).Stokes & Goff (2001) obtained mean values of 39.7, 41.1, and 42.8, for control, calcium propionate, and propylene glycol groups, respectively, using only two 300 mL PG doses and two 680 g calcium propionate doses.Similar results as those obtained in the present experiment were found by Studer et al. (1993) and Juchem (2000); their results did not show a PG effect, when supplied in the prepartum period, on milk yield.However, Fonseca (1997) observed higher milk yields only in the fourth and fifth lactation weeks for cows that received PG pre and postpartum.Sauer et al. (1973) and Formigoni et al. (1996) showed that a potential benefit conferred by PG would be expressed in a more consistent way under risk situations, as is the case with cows having an extreme body condition, a fact that did not occur in this experiment.
With regard to plasma concentrations of nonesterified fatty acids, no interaction of effects time × treatment was observed (P=0.4791).Supplying of "drench", calcium propionate, and propylene glycol did not affect this variable (P=0.4298).However, there was a linear effect of time effect (P=0.0001) (Table 6 and Figure 1).
Mean NEFA concentration during the experimental period was 376.6 mmol L -1 .In the week prior to calving, cows showed, on average, 420 mmol L -1 NEFA.In the first week postpartum, however, mean NEFA concentration increased to 474.7 mmol L -1 .Mean NEFA value during the experiment decreased by 41% in the 5 th week postpartum, relatively to the first week, with a reduction of 48.24 mmol L -1 per week.
Peripartum period is marked by striking alterations, some of which are medium-term, as is the case with blood concentrations of NEFA (weeks).Nonesterified fatty acids is the main energy source during the end of gestation and beginning of lactation (Drackley, 1998), acting as a glucose-saving mechanism, i.e., the peripheral tissues use NEFA rather than glucose and amino acids.Plasma concentrations of NEFA follow a growing pattern starting on the third prepartum week, with a peak near calving, decreasing through lactation.Values obtained in this study are within the limits found by other researchers.Vasquez-Añon et al. (1994) found  300 mmol L -1 on day -10 from calving, and 1014, 700, 445, 320, and 300 mmol L -1 on days zero, five, 10, 20, and 25 postpartum, respectively.Staples et al. (1990) did not find interaction effect between time and treatment, and observed a decrease in NEFA concentrations from the first week in relation to the ninth, with values of 612 mmol L -1 and 237 mmol L -1 , respectively.In the same experiment, those authors related the effect of NEFA concentrations on ovarian activity and observed that NEFA concentrations were not a good indicator of energy status.Juchem (2000) did not find a PG effect (300 mL), nor a treatment × time interaction, on the NEFA concentrations in the pre and postpartum periods, when the product was supplied in the prepartum period.Then again, Fonseca (1997) did not observe a PG effect during the prepartum period and the beginning of lactation (300 mL), on the concentrations of NEFA and glucose, and on the incidence of subclinical ketosis.However, Studer et al. (1993) demonstrated a decrease in FFA during prepartum and a reduction tendency in the postpartum, by supplying 1 L PG per day during prepartum.Waldo & Schultz (1960) concluded that the hyperglycemic effect of PG was a consequence, for the most part, of its intact absorption and, to a lesser extent, of the production of propionic acid, and therefore it was an excellent alternative for treating clinical cases of ketosis.Goff et al. (1996), supplying calcium propionate at calving and after 12 hours, observed a smaller concentration of NEFA 24 hours after calving in Jersey cows, but not in Holstein cows.However, its application reduced the occurrence of milk fever and subclinical hypocalcemia in both breeds, in herds where those problems occurred.
The cows in the present experiment, in both farms, were submitted to good nutritional management in the prepartum and postpartum periods.This can be demonstrated from the lack of occurrence of animals showing clinical ketosis, and from the NEFA concentrations, Sci.Agric.(Piracicaba, Braz.), v.61, n.4, p.376-385, July/August 2004 which were relatively low when compared with other experiments.Similarly, no animals were diagnosed with milk fever.
There were no differences between treatments (P=0.7399) with regard to the first detectable postpartum estrus.On average, cows showed the first estrus at 69.5 days postpartum (Table 7 and Figure 2).This result was superior to that obtained by Thatcher & Wilcox (1973), of 43 days, and by Miyoshi (1995), who found 53.1 days for animals that received 500 mL propylene glycol, and 56.8 days for animals in the control group.Likewise, Jorritsma et al. (2000) observed a 54.1 day interval for the first estrus, without administration of energy supplements.
In Brazil, Fonseca (1997) recorded a mean value of 40.33 days for the first estrus in animals with normal concentrations, and of 51.6 days in animals with acetonemia; these are values still lower than the 69.5 days found in this trial.However, that author did not find a PG effect on that variable, either.More expressive results, but also inferior than those found in the present study, were recorded by Juchem (2000), with means of 22.2 days until the first ovulation for animals that received propylene glycol, and 18.6 days for animals in the control group, detected by ultrasound.Also, the observed percentages of anestrus cows were 7% and 24%, for PG and the control, respectively.The difference, however, was not significant, suggesting that under adequate management conditions the potential benefit of this additive could be small, therefore making its use unviable due to its high cost.Nevertheless, Formigoni et al. (1996) found smaller percentages of anestrus cows (30 and 58%) at 90 days of lactation for cows receiving PG at pre and postpartum.Mean number of services per pregnancy was 2.2, with no difference (P=0.9342)among treatments (Table 7 and Figure 2).Lower values were found by Miyoshi (1995), of 1.7 service per conception (S/C), in animals also supplemented with 500 mL of propylene glycol.Higher values than those obtained in this experiment were found by Smith & Wallace (1998), with 2.7 S/C for cows that cycled before postpartum day 21; by Heuwieser et al. (1994), with 3.3 services per conception in animals with less days for the first service and lower yields; and by Foote & Riek (1999), with 2.6 services per conception for cows showing some reproductive problem.
The number of services per conception is a variable that results from a well-done heat verification, a wellconducted and successful insemination, a good fertilization rate and a low early embryonic death rate (Ayalon, 1978).The number of services per conception obtained in this experiment was relatively high, since a S/C value of 2.0 is considered normal or acceptable.The higher value obtained (2.3) could be due to an association of the factors mentioned above, and to the fact that repeat breeder cows were not excluded in the statistical analysis.

Figure 1 -
Figure 1 -Effect of treatments on body condition score (BCS), body weight and daily weight variation, milk yield, and nonesterified fatty acid concentration during the experimental period.

Table 1 -
Nutritional composition of diets at Farms 1 and 2, based on dry matter.

Table 2 -
Effects of treatments on body condition score and its variation during the experimental period.

Table 5 -
Effects of treatments on milk yield during the experimental period.

Table 4 -
Effects of treatments on the daily variation in body weight during the experimental period.CP=calcium propionate; PG=propylene glycol; SEM=standard error of mean; P=statistic probability.

Table 3 -
Effects of treatments on body weight and its variation during the experimental period.Effects of treatment × time: P=0.8223; time: P=0.0001; linear: P=0.0001; quadratic: P=0.0001 C=control; DR="drench"; CP=calcium propionate; PG=propylene glycol; SEM=standard error of mean; P=statistic probability; Variation=weight difference between week 6 and week 1

Table 7 -
Effect of treatments on number of days to the first estrus, service per conception, and days open during the experimental period.P= statistic probability, 1 st estrus= number of days to the first estrus; S/C= number of services per conception; Days open= number of days open; n= number of samples for each variable.Figure 2 -Effects of treatments on the number of days to first estrus, number of services per conception, and number of days open.