Effect of dry period length on the subsequent production and reproduction in Holstein cows

Calving records of Holstein cows from January 1983 to December 2006 comprising 1,190 herds with 384,717 calving events were used to evaluate factors affecting the dry period length and effect of current dry days on the next productive and reproductive performances of Holstein cows. Statistical analyses of dry days, productive and reproductive traits in this study were performed as linear mixed models. Dry period length (DD) of cows was grouped into 14 classes from < 10 days through > 130 days. Average DD was 100.46 days in Holstein cows. Primiparous cows had the lowest DD and the mean of DD decreased over the years from 1983 to 2006 and summer calvers had the shortest DD (p < 0.001). Cows within the DD classes of 51-60 and 61-70 had the greatest actual milk yield, mature equivalent milk yield, adjusted milk yield, adjusted fat yield, mature equivalent milk yield, adjusted protein yield and mature equivalent protein yield (p < 0.05). Also, cows within the DD class of 0-10 had the greatest adjusted fat percentage, mature equivalent fat percentage and adjusted protein percentage (p < 0.05). Our current results showed a reduction in calving interval and age at calving after cows had shorter dry days than with other longer dry period lengths (p < 0.05). Given the recent advocacy for shortened dry periods, it is worthwhile to emphasize that 51 to 70 DD does provide maximal performance in Holstein cows. Additional key words: dairy cow; dry days; productive performance; reproductive performance. * Corresponding author: nhosseinzadeh@guilan.ac.ir; navid.hosseinzadeh@gmail.com Received: 02-06-12. Accepted: 24-01-13. Abbreviations used: AAC (age at calving); AFP (305-d adjusted fat percentage of milk); AFY (305-d adjusted fat yield); AMY (305-d adjusted milk yield); APP (305-d adjusted protein percentage of milk); APY (305-d adjusted protein yield); bST (bovine somatotropin); CI (calving interval); DD (dry period length); DIM (days in milk); ME (mature equivalent); MEFP (mature equivalent fat percentage); MEFY (mature equivalent protein yield); MEMY (mature equivalent milk yield); MEPY (mature equivalent protein yield); REML (restricted maximum likelihood); RMY (unadjusted milk yield). Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA) Spanish Journal of Agricultural Research 2013 11(1), 100-108 Available online at www.inia.es/sjar ISSN: 1695-971-X http://dx.doi.org/10.5424/sjar/2013111-3165 eISSN: 2171-9292


Introduction
Optimum length of the dry period has been a topic of interest for many years, with recorded debate beginning as early as 1805 (Grummer & Rastani, 2004).During recent years there has been a renewed interest in dry period length, perhaps partly because of an ever increasing need for dairy farmers to maximize their income on investment.During the production cycle before the expected calving a period of rest is needed and in that time milking is ceased and thus the production of milk in the udder stops.The dry period is required for the regeneration of the mammary gland and its preparation for lactation, during that time papillae of the rumen and the small intestine are regenerated, and the organism of the cow prepares for an increased nutrient requirement of the mammary gland during lactogenesis (Capuco et al., 1997;Annen et al., 2004).We also need to take into consideration the intensive growth of the fetus during that period.
Cows reach increasingly higher milk yields and this suggests the need to investigate the length of the dry period, since it is considerably connected with the profitability of production.The 60-day dry period, used to date, was adopted in the early 1900s and since then the principle "one calf a year" has been applied, which for a 305-day lactation gives 60 days of rest.It was found that its elimination or reduction may have an effect on reduced yield while improving fat and protein contents (Rémond et al., 1997;Andersen et al., 2005).
According to Stockdale (2006), shortening or eliminating the dry period may result in a lower incidence of metabolic problems postpartum, and a reduced negative energy balance in early lactation due to the maintenance of dietary intake while milk yields and body condition loss are reduced (Pytlewski et al., 2009).
Since 1936 many observational and experimental data have been generated to establish an optimal drying off time for cows (S ¸ükrü Gülay, 2005).Observational studies (such as Dias & Allaire, 1982;Kuhn et al., 2006) have much greater statistical power and accuracy of estimates but lack control over some unknown variables; designed trials generally lack power and accuracy but have more control over some extraneous variables and thereby provide confirmation for results found in observational studies (Kuhn et al., 2006).Observational data will be affected by many factors, in addition to dry period length, that are highly related to subsequent milk production.For example, data from existing records often will not include the reason why a specific cow was dried off earlier than other cows or why cows were dried off late (< 60 d).Some cows cease lactation spontaneously or the dairy producer will dry off cows early because of insufficient milk production.Thus, the reason why cows had shorter dry periods most often cannot be learned from the milk yield records.Cows with short dry periods also may include those cows that calved early due to physiological problems, sickness or exposure to heat stress, among others.This would bias the estimated effect of dry days on milk yield in the subsequent lactation because of potential or actual problems during early lactation associated with early calving; this would affect the lactational performance.As a result, flaws in record analysis may produce a bias in the milk production records and this may result in insufficient information to estimate the true effects of dry period length adequately (Gulay et al., 2003).Some experimental studies with randomized assignment of cows to different dry period length have been conducted (Gulay et al., 2003(Gulay et al., , 2005;;Annen et al., 2004;Andersen et al., 2005;Rastani et al., 2005;Pezeshki et al., 2007).
Increased potential for milk yield may have made cows more tolerant of shorter dry periods.Conversely, higher production may also result in a demand for a longer rest period in order to maintain production, health and fertility in the subsequent lactation.Considerable research has been done regarding the effect of dry days on subsequent lactation milk yield but far less research is available on the effects for other economi-cally important traits such as milk components or fertility.Several recent studies (e.g., Gulay et al., 2003;Annen et al., 2004;Rastani et al., 2005) considered effects of dry days on fat, protein and somatic cell score.However, all of these studies were based on small numbers of cows and, although collectively such studies can be informative if enough of them are conducted, individually they lack adequate power to be conclusive (Kuhn & Hutchison, 2005).Furthermore, of the studies that have examined dry days effects on fat and protein yield, all have reported results in terms of either yield for partial lactations or in terms of 305-d, matureequivalent lactational yield.Recent research has shown, however, that dry days has larger effects on actual lactational milk yield than on records standardized to a 305-d basis; the very standardization of records to a common lactation length (305 d) and mature-equivalent basis, in effect conceals variation in production caused by dry days partly because of its effects on days in milk (DIM) and culling in the subsequent lactation (Kuhn et al., 2005b).If a short dry period, for example, caused problems in the subsequent lactation that resulted in early culling or earlier dry off, then extending records to a 305 d basis would tend to reduce or eliminate that effect.If short subsequent lactations are unrelated to dry days, then they will occur randomly across dry periods and cause no bias in the analyses (Kuhn et al., 2005b).Given the high phenotypic correlation of milk yield with both fat and protein yield (Welper & Freeman, 1992), it is likely that actual lactational records would be more informative for fat and protein yield as well.Since dairy producers are paid for actual yield rather than standardized yields, effects of dry days on actual yields should be ascertained (Kuhn et al., 2006).The effect of variation in dry period length on subsequent lactation production and reproduction, for modern day dairy cattle, is largely unknown and warrants re-evaluation.Therefore, the aim of this study was to evaluate the effect of different lengths of dry period on the subsequent productive and reproductive performances of Holstein cows.

Material and methods
Calving records from the Animal Breeding Center (Karaj, Iran), collected from January 1983 to December 2006 and comprising 384,717 calving events of Holsteins from 1190 dairy herds were included in the data set.The characteristics of dairy herds used in this Effect of dry period length on performance study were described in previous studies (Ghavi Hossein-Zadeh et al., 2009;Ghavi Hossein-Zadeh & Ardalan, 2011).The data included animal registration number, herd, calving date, parity, calving age, raw or unadjusted milk yield (RMY), 305-d adjusted milk yield (AMY), mature equivalent milk yield (MEMY), 305-d adjusted fat yield (AFY), 305-d adjusted fat percentage of milk (AFP), mature equivalent fat yield (MEFY), mature equivalent fat percentage (MEFP), 305-d adjusted protein yield (APY), 305-d adjusted protein percentage of milk (APP), mature equivalent protein yield (MEPY), calving interval (CI) and age at calving (AAC).Adjusted 305-d yields were actual yields of dairy cows which were corrected based on days in milk and twice daily milking.RMY was actual lactation milk yield, not standardized to 305 d.Mature equivalent (ME) yields/percentages were milk, fat or protein yields/percentages which standardized for age at calving.Records were eliminated if no registration number was present for a given cow.Records were also deleted from the analyses if there was no information on the productive and or reproductive performances.Months of calving were grouped into four seasons: April through June (season 1 = spring), July through September (season 2 = summer), October through December (season 3 = fall), and January through March (season 4 = winter).In addition, calving years were grouped into four classes : 1983 through 1988, 1989 through 1994, 1995  Statistical analyses of dry period length, productive and reproductive traits in this study were performed as linear mixed models (Proc Mixed) with the best fitted covariance structure of SAS (SAS Inst., 2002).The least square means were estimated by Restricted Maximum Likelihood (REML) method.The models used to analyze RMY, AMY, MEMY, APY, MEPY, APP, AFY, MEFY and AAC included the fixed class effects of herd, calving year, calving season, parity of dam, dry period length, interaction effects of year by parity, year by season, year by dry period length, season by parity, season by dry period length and parity by dry period length and linear and quadratic covariate effects of age.Linear covariate effect of days in milk was included in the model of analysis for RMY.In addition, AFP was analyzed in a model in which the following variables were included: herd, calving year, calving season, dry period length, interaction effects of year by parity, year by season, year by dry period length, and parity by dry period length and linear covariate effect of age.MEFP was analyzed in a model in which the following variables were included: herd, calving year, dry period length, interaction effects of year by parity, year by season, year by dry period length, season by dry period length and parity by dry period length and linear and quadratic covariate effects of age.dered as a random variable in all models of analysis for productive and reproductive traits.Phenotypic correlations between DD and productive and reproductive traits were calculated by Corr procedure of SAS (SAS Inst., 2002), and regression coefficients of productive and reproductive traits per one day change in DD were estimated using Reg procedure of SAS (SAS Inst., 2002).

Results
The average dry period length was 100.46 days in this study.The productive and reproductive performances at different dry period length groups in Holstein cows are shown in Tables 1 and 2 3 shows the regression coefficients of productive and reproductive traits on dry period length and the phenotypic correlation between dry days and subsequent performance of dairy cows.There were significant and negative relationships between dry days and RMY, MEMY, AMY, AFY, MEFY, APY and MEPY in Holstein cows (p < 0.01); therefore, the yields of milk, fat and protein decreased along with the increase in dry days.On the other hand, there were significant and positive relationships between dry days and APP, AAC and CI (p < 0.05); protein percentage, calving interval and age at calving increased along with increase in dry days.Also, there were non-significant relationships between AFP and MEFP and dry days of dairy cows.There were significant interaction effects of dry period length class by parity on the subsequent production and reproduction of dairy cows which were depicted in Figs. 2 to 5. Effects of DD on yield or percentage traits were, for the most part, consistent across lactations and the dry period length to maximize subsequent RMY, AMY, MEMY, APY and MEPY was generally the same (61-70) regardless of parity.There were decreasing trends for RMY, AMY, MEMY, APY and MEPY at the dry days of < 60 or > 70 over the parities.
The dry period length to maximize subsequent second lactation AFY and MEFY was 51-60, but corresponding dry period length for third and greater lactation AFY and MEFY was 61-70.There were decreasing trends for AFP, MEFP and APP from dry days of < 10 to 61-70 over the lactations, but an increasing trends were observed for AFP, MEFP and APP at dry days of 70 onwards (p < 0.001).Therefore, second, third and fourth and greater lactation AFP, MEFP and APP minimized at the dry days of 61-70.Also, there were generally consistent phenotypic trends for AAC and CI over the parities.Second and third parity cows within the dry days of < 10 had the lowest AAC, and fourth and greater parity cows within the dry days of < 10 and 31-70 had the lowest AAC.On the other hand, there were increasing trends for second, third and fourth and greater lactation CI along with increase in dry days (p < 0.001).Effect of variables affecting the DD of Holstein cows is shown in Table 4  period of 2001-2006 (p < 0.001).Summer-calved cows had the lowest DD but spring-and winter-calved cows had the greatest DD (p < 0.001).On the other hand, primiparous cows had the lowest DD and cows in their parity 4 and greater had the greatest DD (p < 0.001; Table 4).There was significant interaction effect of calving year by season of calving on the DD and cows which calved during the spring season of calving and years 1983-1988 had the lowest DD (p < 0.001).Also, primiparous cows which calved during 1983-1988 or summer season had the lowest DD (p < 0.001).Also, there were significant effects of age at calving and previous milk yield on DD (p < 0.001).The regression coeff icient of dry days on milk yield was -0.008 ± 0.00005 (p < 0.001); this indicated the reduction in dry days per kilogram of increase in milk yield of dairy cows.On the other hand, the regression coefficient of dry days on age at calving was 0.68 ± 0.03 ( < 0.001); Effect of dry period length on performance 105 this indicated the increase in dry days along with one month increase in calving age.

Discussion
The average of dry period length in the current study (100.46days) was lower than the report of Amasaib et al. (2011) who stated to be 133 days in crossbred dairy cows of Sudan.However, Musa et al. (2005) reported 112 days for Sudanese cattle, while Ishag (2000) found that the dry period for crossbred dairy cattle was 90.5 days.It is commonly estimated that a twomonth dry period provides a complete regeneration of udder glandular tissue and is favorable for the high production in the forthcoming lactation (Annen et al., 2004;Andersen et al., 2005).Capuco et al. (1997) showed that a dry period was necessary to replace mammary epithelial cells, thus providing one biological basis for the lower milk yield that has been observed with shortened dry periods.The results of analyses indicated that too short as well as too long dry periods have a negative effect on milk, fat and protein yield in forthcoming standard lactation.This conf irms the earlier results by Borkowska et al. (2006), Winnicki et al. (2008), Pytlewski et al. (2009) and We ¸glarzy (2009).Similar to our results, Kuhn et al. (2007) reported dry periods of 30 d or fewer resulted in large reductions in subsequent lactation production and short dry period was beneficial for fat and protein percentages in the subsequent lactation of Jersey cows.Given the negative correlation between percentages and milk yield (Welper & Freeman, 1992), the highest percentages associated with shorter DD would be expected.We ¸glarzy ( 2009) observed the highest mean milk yield for 305-day lactation was obtained in the dry days of 61-90.Bachman (2002) and Gulay et al. (2003) have reported no loss with dry periods < 60 d.Sample sizes in both of those studies were small, however, and it has been shown (Kuhn & Hutchison, 2005) that even estimates in the wrong direction are not unlikely with such small sample sizes.On the other hand, an increase in yields of milk and protein and a decrease in milk fat content with an elongation of the dry period were recorded by Degaris et al. (2008).Also consistent with us, Kuhn et al. (2005aKuhn et al. ( , 2007) ) observed the greatest number of dry periods was between 46 and 65 d in Jersey and Holstein  Standard errors are within the parenthesis.Means within a column that do not have a common superscript (a-d) are significantly different (p < 0.05).
cows.Rémond et al. (1992), Madsen et al. (2004), andRastani et al. (2005) also found higher protein percentages for cows with no dry period.Several studies using Holsteins (Funk et al., 1987;Kuhn et al., 2005b) and Jerseys (Kuhn et al., 2007) have investigated whether the effect of dry period length in the subsequent lactation depends on parity and have generally found either no or only small interactions with parity.Furthermore, even when DD effects were found to differ slightly across parities, the dry period length to maximize subsequent lactation performance was generally the same regardless of parity.Kuhn et al. (2005b), for example, found that dry periods of less than 20 d decreased subsequent lactation milk yield for Holsteins more in the second lactation than in later lactations, but 60 DD maximized yield in the following lactation regardless of parity.Inconsistent with the current results, Annen et al. (2004) found higher fat and protein percentages with fewer DD, but only for second lactation cows; there was no clear pattern for higher parity cows.It should be noted, however, that Annen et al. (2004) utilized bovine somatotropin (bST) in all treatment groups, which may have affected results for all traits since milk yield loss, generally associated with fewer DD, might have been mitigated by the use of bST in their study.Similar to the result of this study, Kuhn et al. (2007) observed the most notable effect of parity on dry period length was for first-parity cows to average about 3 to 5 fewer DD than later parity cows.This is likely due to the greater persistency of firstlactation cows (Stanton et al., 1992).Thus, although cows produce more total milk over the entire lactation in second and later lactations, first-lactation cows actually have higher yields at the end of lactation, which in turn leads to slightly longer lactations, on average, and fewer DD.Also, consistent with us, Kuhn et al. (2007) reported recent years had lower mean DD than previous years and summer months had lower DD than other months in Jersey cows.
There is limited published research on the effect of dry period length on reproduction and fertility.Our current results showed a reduction in calving interval and age at calving after cows had shorter dry days than with other longer dry period lengths.Using shortened dry periods to improve fertility in either breed is not likely to be of much overall benefit because the "benefit" in fertility results only from lowered milk yield.It is, first of all, questionable whether improving fertility by lowering milk yield is a prudent economic choice (Kuhn et al., 2007).However, even if improved fertility through lowered milk yield was shown to be of merit, there are almost certainly less expensive or more efficient ways to lower milk yield than by reducing dry period length, a practice that necessarily increases labor, time, and maintenance in the milking parlor, unless herd size is also reduced.Lower cost rations, for example, might lower milk yield and would simultaneously reduce costs rather than increase them, again if improved fertility through lowered milk yield was desired or found to be in some way a favorable alternative (Kuhn et al., 2007).Similar to the results of this study, Kuhn et al. (2007) and Watters et al. (2009) indicated an improvement in reproductive performance of Holstein dairy cattle when dry period length was reduced.Contrary to the current results, Pezeshki et al. (2007) did not f ind a consistent improvement in reproductive measures with decreased dry period length.On the other hand, Grummer (2007) reported shortening or eliminating the dry period may be a more successful approach to improving reproductive efficiency than diet manipulation because shortening or eliminating the dry period may enhance dry matter intake during the transition period, decrease milk energy output, or both.
As conclusions, average DD was 100.46 days in Holstein cows.Calving year, season of calving and parity were identif ied as effective factors and had significant effects on the DD of dairy cows.Primiparous cows had the lowest DD and cows in their parity 4 and greater had the greatest DD.The mean of DD decreased over the years from 1983 to 2006 and summer calvers had the shortest DD.As with milk yield, fat and protein yields are maximized in the subsequent lactation with a 51-70 dry days period.Also, fat and protein percentages for Holstein cows were actually favored by shortened dry periods.Our current results showed a reduction in calving interval and age at calving after cows had shorter dry days than with other longer dry period lengths.This research is one of the few studies to examine DD effects on subsequent lactation, using a large dataset, for traits other than milk yield, and in particular percentage traits and reproduction.

Figure 2 .
Figure 2. Distribution of raw (a), adjusted (b) and mature equivalent (c) milk yield over different dry period length classes and different parities in Holstein cows.Dry period length classes are described in Fig. 1.

Figure 3 .Figure 4 .
Figure 3. Distribution of adjusted (a), mature equivalent (b) fat yield, adjusted (c) and mature equivalent (d) fat percentage of milk over different dry period length classes and different parities in Holstein cows.a)

Figure 5 .
Figure 5. Distribution of age at calving (a) and calving interval (b) over different dry period length classes and different parities in Holstein cows.
The model used to analyze CI included the fixed class effects of herd, calving year, parity of dam, dry period length, interaction effects of year by parity, year by season, year by dry period length, season by parity, season by dry period length and parity by dry period length and linear and quadratic covariate effects of age.The model used to analyze dry period length included the class effects of herd, calving year, calving season and parity and interaction effects of year by season, year by parity and season by parity and linear and quadratic covariate effects of age at calving and covariate effect of milk.Animal effect was consi-

Table 1 .
. Cows within the DD classes of 51-60 and 61-70 had the greatest RMY, MEMY, AMY, AFY, MEFY, APY and MEPY, and cows Milk yield, fat yield and percentage at different dry period lengths in Holstein cows 1 RMY: raw milk yield (kg); AMY: adjusted milk yield (kg); AFY: adjusted fat yield (kg); AFP: adjusted fat percentage; MEMY: mature equivalent milk yield (kg); MEFY: mature equivalent fat yield (kg); MEFP: mature equivalent fat percentage.Standard errors are within the parenthesis.a-l Means within a column that do not have a common superscript are significantly different (p < 0.05).

Table 2 .
Protein yield and percentage and reproductive performances at different dry period lengths in Holstein cows a within the DD class of 0-10 had the lowest values of RMY, MEMY, AMY, APY and MEPY, and cows within the DD class of 11-20 had the lowest AFY and MEFY (p < 0.05).Also, cows within the DD class of 0-10 had the greatest AFP, MEFP and APP (p < 0.05).Table

Table 3 .
. DD was the lowest for the Regression coefficients of productive and reproductive traits on dry period length and correlations between dry period length and performance traits in Holstein cows ** p < 0.01.ns: non-significant.

Table 4 .
Effect of different variables on the dry period length of Holstein cows