Effect of breed, season, year and parity on reproductive performance of dairy cattle under smallholder production system in Bukoba district, Tanzania

L Asimwe and G C Kifaro

Department of Animal Science and Production, Sokoine University of Agriculture, P.O.Box 3004, Morogoro, Tanzania
kifaro@suanet.ac.tz

Abstract

A study was conducted to evaluate the effects of breed level and non-genetic factors on reproductive performance of dairy cattle under smallholder production system. The study focused on assessing the reproductive efficiency of dairy cattle under smallholder farms in Bukoba district, Tanzania. Data on age at first calving (AFC), calving to first service interval (CFSI), number of services per conception (NSC), days open (DO) and calving interval (CI) were collected from Kagera Dairy Development Trust (KADADET) in Bukoba district. General Linear models procedure of Statistical Analysis System (SAS) computer software was used to analyse the data. The fixed effects considered in the analyses were genetic group, parity, season and period of birth or of calving.

The overall mean for AFC was 35.1+9.7 months with a coefficient of variation of 18.6%, for CFSI it was 196.95+1.8 days with a coefficient of variation of 20% while the mean for NSC was 1.66+0.0 with a coefficient of variation of 21%. Further, the mean DO was 205.2+2.6 days with coefficient of variation of 19%, and CI averaged 480.4+2.4 days with a coefficient of variation of 22%. AFC was significantly affected by period of birth (P<0.001), level of exotic blood (P<0.05) and season of birth (P<0.05). Genetic group significantly influenced CFSI, NSC, CI (P<0.05) and DO (P<0.001). F₁ crosses performed better than high-grades in all the traits with 34.6 months of AFC, 171 days of CFSI, 182 days of DO, and 455 days of CI. Season of calving significantly influenced CFSI (P<0.01), CI (P<0.001), NSC and DO (P<0.05). Cows calving in the long rain season were superior with 17 days of CFSI, 21 days open and had shorter (by 27 days) CI than those calving in the long dry season. Parity and period of calving significantly affected CFSI (P<0.05 and P<0.001), NSC (P<0.05), DO (P<0.01 and P<0.001), CI (P<0.05 and P<0.001) respectively. For these traits, performance was improving by advancement in age of cows. Milk yield in the first 100 days of lactation had no significant influence on post-calving reproductive traits.

It was concluded that reproductive performance was best in the long rain season and that with respect to reproductive traits F₁ crosses were better than high grades in Bukoba district.

Key words: age at first calving, calving interval, calving to first service interval, days open, genetic and non-genetic factors, number of services per conception

Introduction

In developing countries the demand for dairy products is increasing as a result of fast growing populations and a rise in the per capita income (ILCA 1993). Improvement of production in dairy cattle is frequently considered possible by either improving the genetic merit and/or animal husbandry practices but often reproduction is neglected. Failure to maintain a high degree of reproductive efficiency has been associated with major economic losses in the dairy industry mainly through increase in the length of unproductive life of cows (Plaizier et al 1998).

Improvement in smallholder dairy production contributes to the availability of milk for public consumption and also improves the economic status of smallholder farmers. In order to maintain a long economically productive life of a cow, high reproductive efficiency and milk production are important (Das et al 1986). Reproduction is crucial for production of the necessary replacement stock, for reducing unproductive periods like dry periods and for increasing lifetime milk production and income (Das et al 1986; Nebel and McGilliard 1993). When the reproductive efficiency of a farm is poor the following losses can occur; less milk is sold as a result of longer days open, fewer calves are sold due to longer calving intervals, increased veterinary costs caused by greater number of problem cows, increased number of matings per conception as a result of poor heat detection and increased number of replacement heifers needed for non-voluntary reproductive culls.

While appreciating the importance of reproduction in the dairy enterprise, traits related to reproduction are mainly influenced by environmental factors such as feeding (Msanga et al 1999), tracking of reproductive cycles through record keeping (Balikowa 1997), access to bulls or efficiency with which AI is performed (Mulangila 1997), ability to detect heat (Ulmek and Patel 1992), reproductive health problems and other non-genetic factors (El-Keraby and Aboul-ela 1982). The aim of this study was therefore to assess the influence genetic and non-genetic factors on key reproductive traits in dairy cattle kept under smallholder production system in a high altitude humid area of the tropics. 
 

Materials and methods

The study area

This study was carried out in Bukoba district, which is one of the six districts of Kagera region in the northwest zone of Tanzania. The district lies between longitude 30⁰ 45^’ and 32⁰ 00^’ east and between latitude 1⁰ 00^’ and 3⁰ 00^’ south of the equator. It borders the Republic of Uganda on the northern side and lake Victoria in the east. The area receives bimodal rainfall that varies between 1000 and 2000 mm. The average temperature is 20⁰ with variations between 15⁰ and 28⁰C.

Data collection and derivation of study variables

Data were collected from Kagera Dairy Development Trust (KADADET) office, which had been monitoring on-farm performance of cows between 1990 and 2000. Collected information was on birth dates, calving dates, dates of first and subsequent heats and insemination dates. Pregnancy diagnoses dates and their results were noted. Records of milk yield produced during the first 100 days of lactation were also transcribed from individual cow records.

From the collected information, the following dependent variables of interest were derived; age at first calving (AFC) as the number of months from birth date to first calving date, calving to first service interval (CFSI) was calculated as the interval in days from the calving date to the first service date, days open (DO) as the interval in days between calving and conception and calving interval (CI) was the interval in days between two consecutive calvings. The number of services per conception (NSC) was obtained by counting the number of services the cow was mated or inseminated until she conceived.

Data handling and analyses

Data were classified into several categories for statistical analyses. Two genetic groups F₁  (with 50% Friesian blood) and high grades (with >50% but <100% Friesian blood) were identified. Four seasons were established based on weather and climatic conditions of the area (January to February as short dry period, March to May as long rain season, June to mid October as long dry season and mid October to end of December as short rain season). Further, all parturition numbers were classified into four groups; 1, 2, 3 and 4. All parities above 4 were pooled together in parity 4 because only a few cows that had completed more than five lactations. For age at first calving, the years of birth were grouped into the following periods of birth, those born in the years before 1985, between 1985 and 1990 and from 1991 to 1995. For CI, DO, CFSI and NSC, there was an additional period for animals that calved after 1995.

Data on NSC were tested for normality using UNIVARIATE procedure of Statistical Analysis System (SAS 2002) and found to conform to the normal distribution. The General Linear Models procedure of SAS (2002) was used to assess the effects mentioned above on the dependent variables. Data on AFC were analysed using model 1, while that of CFSI, NSC, DO and CI were analysed using model 2. Number of records available for analysis were; 436 for AFC, 2131 for CFSI, 2090 for NSC, 1985 for DO and 2009 for CI. Milk yield produced during the first 100 days of lactation was used as a covariate. The intention was to see if there was any significant association between level of milk production in early lactation and reproduction parameters considered. Least squares means from significant factors were compared by PDIFF procedure of SAS (2002).

Model 1

Y_ijkn = u + B_i + S_j + Y_k + e_ijkn

Where:

Y_ijkn = Age at first calving of n^th  cow born in k^th period of birth, j^th season of birth and of i^th genotype

u = population mean

B_i = effect of i^th genotype

S_j = effect of j^th season of birth

Y_k = effect of k^th period of birth

e_ijkn = random residual error distributed as N(O,Iδ² e)

Model 2

Y_ijklnm = u + B_i + S_j + Y_k + P_l + b(x_ijklnm – x) + e_ijklnm

Where:

Y_ijklnm = the m^th record of n^th cow in l^th parity, k^th period of calving, j^th season of calving and i^th genotype

 u = population mean

B_i = effect of i^th genotype

S_j = effect of j^th season of calving

Y_k = effect of k^th period of calving
P_l = effect of l^th parity

X_ijklnm = milk yield during the first 100 days of lactation of the respective individual animal

X = overall mean milk yield during the first 100 days of lactation

b = linear regression coefficient of Y_ijklnm on X _ijklnm

e_ijklnm = random residual error distributed as N (O,Iδ² e)
 

Results and discussion

Age at first calving

Analysis of variance for AFC and other variables is given in Table 1.

The mean age at first calving was 35.1± 9.7 months with a coefficient of variation of 18.6% (Table 2).

This mean is similar to values reported by Kiwuwa et al (1983) of 34.5 months and Balikowa (1997) of 36.7 months. It is however, higher than mean (32.4 months) reported by Syrstad (1996) under similar tropical environments. The period in which cows were born was a highly significant (P<0.001) source of variation. Heifers born in the period before 1985 had the highest mean (37.3+16.4 months) while heifers born from 1985 to 1990 had the lowest mean (33.3+12.9 months). Poor feeding and management could have been the reason for highest age at first calving during this period since a majority of farmers at that time had no experience in dairy cattle management.

Level of exotic blood among heifers significantly (P<0.05) affected AFC. The F₁ crosses were found to calve earlier than the higher-grade counterpart heifers (Table 2). Such findings conform to results found by Sabota and Gill (1992), Kiwuwa et al (1983) and Syrstad (1996). F₁ crosses were distributed to farmers when they were in-calf heifers but high-grade animals were born on-farm. The prolonged age at first calving of high-grade heifers could be attributed to factors such as poor nutrition and management practices including poor heat detection at the time of mating the heifers. With good nutrition it is expected that high-grade heifers would exhibit fast growth and attain higher weights at relatively younger ages.

AFC was also significantly (P<0.05) influenced by the seasons in which the heifers were born (Table 1). This conforms to the results of El-Keraby and Aboul-ela (1982) and Kifaro (1984). Differences in age at first calving between seasons might have been attributed to seasonal fluctuations in forage availability. Kifaro (1984) associated seasonal differences to variation in pasture availability and other supplementary feeds. Calves born during the dry season tended to have highest age at first calving of 36.6+16.5 months while those born during the long rain season had the lowest age at first calving (34.1+18.9 months). The variation could be attributed to differences in quality and quantity of forage, a common observation in the tropics.

Calving to first service interval

The overall mean for the interval between calving and first service was 169.9±1.8 days (Table 3), which is almost similar to results found by Estévez et al (1995) of 164.1 days.

However it is higher than the results reported by Dawuda et al (1988) and Mujuni et al (1990), which ranged from 93.2 to 151.7 days in various crosses. Such differences could have been caused by failure of farmers to detect heat signs after calving thus prolonging the interval, but could also be associated with low nutritional status of the cows, which did not allow them to recuperate fast enough after calving. Analysis of variance on calving to first service interval (Table 1) showed that period of calving was a highly significant (P<0.001) source of variation. Cows that calved in the period between 1985 and 1990 had the longest interval of 225.9+9.9 days from calving to first service followed by those that calved from 1995 onwards. The shortest interval of 145.3±4.9 days was observed in cows calving earlier than 1985 (Table 3). Season of calving was a significant (P<0.05) source of variation on calving to first service interval. Cows calving in the long dry season had the longest intervals 183.0+3.0 days, while those in the long rain season had the shortest interval 166.3+3.5 days. This could be caused by good pasture and its adequate availability during the rain season. The significant effects of period and season of calving on interval between calving and first service concurs with results reported by Dawuda et al (1988), Lozano et al (1992) and Estévez et al (1995). For example, Dawuda et al (1988) reported the interval to be 152.8 days in the rain season and 122.9 days in the dry season; here dry season calvers  had shorter CFSI by approximately one month. The effect of period of calving could be attributed to the improvement in management skills gained by farmers as years went by and possibly to random environmental effects.

Level of exotic blood and parity were other important sources of variation on interval from calving to first service (P<0.05). High-grade heifers had a mean interval of approximately 9 days longer compared to F₁ crosses. It was observed that first calvers had the longest intervals (of 184.7+3.1 days) from calving to first service, while those that calved in fourth plus parities had the shortest intervals (170.8+3.6 days; Table 3). Physiological stress which first calvers do experience in early lactation could partly explain the observed longer CFSI. The second explanation is the fact that after the first parity, animals continue to grow whereby the dietary energy intake is partitioned to meet the requirements for maintenance, continuation of growth, lactation and reproduction.

Number of services per conception

The overall mean number of services per conception was 1.66±0.02 (Table 3) with a coefficient of variation of 21%, is close to the findings by Maniack et al (1978) of 1.8 in Jersey x Red Sindhi, Carmona Solano and Sato Vargas (1987) of 1.81 in dual purpose cows of Costa Rica and by Bekele et al (1992) of 1.66 in Holstein x zebu cows. The present finding is lower than the results found by Chaudry et al (1989) of 3.0 among Jersey x Sahiwal cows and by Kumar et al (1991) which averaged 2.05. Such differences could be ascribed to differences in management practices and mating systems used.

The analysis of variance (Table 1) shows that level of exotic blood, parity, season and period of calving had significant (P<0.05) effect on the number of services per conception. Heifers/cows that calved during the long dry season required more services per conception, while those that calved during the long rain season had fewer services per conception. Poor nutrition has often been a limiting factor to dairy cattle performance particularly in the long dry season when nutritive value of pasture is very low (Msechu and Mkonyi 1984). Hafez (1974) reported that insufficiency or an imbalance of protein, energy, roughage, vitamins and minerals do result in repeat breeders as well as low ovarian activity. High-grade cows under this study were also observed to require more services per conception compared to F₁ crosses. Cows with higher milk yields are known not to breed quickly, have longer service periods and take a long time to conceive. Invariably, they require more services per conception than do cows with low milk yields (Berger et al 1981; Chauhan et al 1994).

Days open

The overall mean length of days open in this study was 205.2+2.6 days (Table 4) with coefficient of variation of 19%. This length of days open is similar to the results by Chaudry et al (1989) of 195.3 and Estévez et al (1995) of 197.7 days.