Livestock Research for Rural Development 13 (5) 2001

Citation of this paper

Reproductive performance of Tswana and Simmental x Tswana crosses in smallholder farms in Botswana

O R Madibela, W S Boitumelo and B Kiflewahid 

Sebele Station, Department of Agricultural Research, 
P/Bag 0033, Gaborone. Botswana.


Data from the Small-scale Dairy Project were analyzed to compare the reproductive performance of Simmental Tswana crosses and purebred Tswana milking cows from smallholder farms. The effects of season (dry and wet), breed group (Tswana and Simmental crosses) and their interaction on calving interval and days open were determined. The effects of breed and year and their interaction on calving rate were determined. Correlation of monthly calving and monthly rainfall was computed. 

There was no difference in calving interval and days open between Tswana (57611.3 and 296 11.3 days) and Simmental crosses (584 22.3 and 303 22.3 days) cows. There were no effects of season on the calving interval and days open. There was also no breed x season interaction. Peak calving occurred between October and January (52.3%) while between May and June 15.3% of cows calved. The mean calving rate of Tswana cows (68.2%) tended to be higher than that of. Simmental crosses (62.5%) cows. There was a significant effect of year on calving rate. The implications of the results are discussed in relation to the effects of breed and season on the reproductive performance of these breeds. 

Key words: Calving interval, calving rate, days open, Simmental  crosses, and Tswana cows.


The Dairy Research Program of the Department of Agricultural Research has investigated smallholder milk production around peri-urban areas of Gaborone since 1985. When the project started in 1985 it was envisaged that the adoption of improved dairy husbandry techniques by the traditional farmers would increase milk production and consequently diversify rural economic activities, improve human nutrition at household level and reduce the countrys dependence on imported fluid milk. It was found consistently that Simmental crosses produce more milk than the Tswana cows on the same management regime. During the 1989/90 lactation periods, the Simmental Tswana crosses produced twice the amount of milk as Tswana cows (806 vs. 462 kg/lactation) (APRU 1990). The difference in milk yield was attributed to the difference in the genetic potential of the genotypes. 

The National Development Plan 6 (1985-91) recognized the need to increase milk production by peri-urban small-scale farmers for home consumption and surpluses for sale to the rapidly growing urban population. Crop-livestock systems are such that cattle provide draft, meat and milk products. In Botswana cattle depend on natural pasture for their nutritional requirements and often experience nutritional deficiencies especially in the dry season. These nutritional stresses often lead to loss in weight and delayed reproductive activities, which further contributes to fluctuations in milk supply. 

Due to shortage of feed supply and reduction in the area available for grazing, nutrition has become a major constraint to animal productivity in many parts of Africa (Mukasa-Mugerwa et al1992).  This limitation can manifests itself in both beef and milking herds in terms of longer calving interval and reduced milk off take. According to Voh and Otchere (1989) the percentage calf crop is crucial for herd replacement while milk off take for calf and human is dependent on reproduction. 

The importance of regular calving interval in a milking herd is emphasized by the fact that it costs money in lost income for each day the calving interval is prolonged beyond 365 days. A significant increase to productivity could therefore be feasible by paying attention to problems of reproductive inefficiency (Mukasa-Mugerwa et al 1992). Hence an important starting point in any animal improvement package is to assess the reproductive performance of the herd (Voh and Otchere 1989). 

Though the major aim of the smallscale dairy project was to increase milk production through improved nutrition and crossbreeding, it was hoped that other important parameters like reproductive wastage due to nutritional stress would be reduced. It was thus deemed necessary to investigate the effects of improved feeding and cow genotype on the reproduction of the participating herds.


The objective of this study was to assess breed differences under improved feeding during the dry season on the reproductive performance of cows kept by smallholder milk producers around the peri-urban areas of Gaborone.  

Materials and methods 

Characteristics of the farming system

Farmers participating in the Smallscale Dairy Project on average cultivated 6 ha of land and practiced a mixed crop and livestock system within the communal grazing lands (APRU 1986). In Botswana grazing pasture is not improved and its seasonal variation in quantity and quality follows the annual rainfall pattern. The most widely reared cattle are the Tswana, which belong to the humped Sanga type. The rearing system operates on a communal use of grazing land and thus natural mating is  not controlled. However, the participating farmers observed cows that were on heat and confined them with a Simmental or crossbred Simmental bull in a kraal.  The wet season covered the months of November to April and the dry season covered the months of May to October. Generally cows calve during the wet season to take advantage of abundant feed supply which promotes better milk production hence a better chance of survival of the calf. However, some animals tend to calveduring the dry season.

Data Collection

The data were obtained from pooled records kept by the small-scale dairy research project participants located within a radius of about 20 km from Gaborone. The records cover a seven-year period (1985 to the 1992).  Forty-five farmers participating in the project, each with an average of 16 cattle were selected. They fed cereal crop residues, Lablab (Lablab purpureus) hay, sorghum bran, bone meal and salt during the dry season . However, the number of animals per farm per year varied as farmers could sell or slaughter their animals at will. During the wet season, feed resources were natural grazing, sorghum bran, bone meal and salt. To date most of the original farmers who participated in the on-farm research are keeping and milking their crossbred cows (mainly Simmental x Tswana crosses) and Tswana cows. 

The cows were milked either twice or once a day depending on either the managerial capacity of the farmer, the needs of the calves or the availability of feeds. The calves were allowed to suckle their dams before milking for a few minutes to stimulate milk letdown. After milking, the calves were allowed to suckle again to draw residual milk for 30 minutes.

Data analysis

Calving interval was taken as the period between two consecutive calvings and was not adjusted. The total number of records included in the analysis when calving rate was the dependent variable was 502, whilst 387 records were included in the analysis when calving interval and days open were dependent variables. Observations that had less than 325 days (gestation and uterus involution) between two consecutive calving were excluded. The effects of season, (dry and  wet), breed group (Tswana and crosses) and their interaction on calving interval and days open were determined using General Linear Model procedures (SAS ).Differences between means in interaction subclasses were tested by least significant difference (LSD). Cows that had three or more consecutive calvings were included in the analysis for the calving interval. Monthly calving frequencies were computed as percentages of the total number of calves born. The relationship between rainfall and monthly calving frequencies after a delay of 9, 10 and 11 months were established by means of calculating correlation coefficients (r). The calving rates between the two genotypes were compared using Chi-Square test. Calving rates were calculated for each year starting from 1986 to 1991. The effects of year on calving rates were tested using GLM procedure and means separation was achieved using Duncan multiple range test.  Results are reported as least square means standard error.   


Calving Interval

Within the period 1985 to 1992 the mean calving interval for the cows (n =387) was 577 days. Tswana cows tended to have a shorter calving interval and days open, compared to Simmental crosses (Table 1). This difference was not significant (p> 0.05). 

Table 1. Mean calving interval and days open of Simmental Tswana crosses and Tswana cows.


No of cows

 Calving interval (days)

Days open (days)

Simmental Cross








Level of significance




There were no effects (p> 0.05) of season on the calving interval and days open (Table 2).

Table 2. The effects of season on the calving interval and days open


No of cows

Calving interval (days)

Days open (days)









Level of significance




Simmental Tswana crosses that calved during the dry season tended to have a shorter mean calving interval than Tswana cows. Days open followed the same trend as calving interval (Table 3). Among the cows that calved during the wet season Simmental crosses tended to have a longer calving interval compared with Tswana cows. Days open had a similar trend as calving interval.   

Table 3. Interaction of season and breed on calving interval and days open 


Dry season


Wet  season

          P rob.  

Simmental cross



Simmental cross


Calving interval







Days open







Calving Rates

Monthly calving frequencies as percentage of the total number of calves born showed a bimodal calving pattern with peak calving occurring during October to January representing 52.3% of the calvings. The small peak occurred in May to June representing 15.3% of the calving (Figure 1).

Figure 1: Calving (% of total annual calving) and rainfall (mm) 10 months previously

The calving rates for the years 1986, 1988, 1989, 1990, and 1991 tended to be higher for Tswana cows than Simmental Tswana crosses. However, the differences were not significant except for the year 1987 (Table 4). The mean calving rate of Tswana cows was slightly higher (p=0.330) than that of Simmental Tswana crosses. The highly significant difference between breeds in calving rates during the year 1987 cannot be explained.

Table 4. Mean calving rates (%) of Simmental crosses and Tswana cows



Simmental x Tswana Tswana





































There was a significant (p<0.001) effect of year on calving rate (Table 5). Strong correlation was observed between calving and the previous monthly rainfall at 9 (r= 0.58; p<0.05), 10 (r=0.84; p<0.001) and 11 (r=0.72; p<0.01) months.    

Table 5: The effect of year on the calving rates



Calving rate  



















abc Figures with different superscript are different: p<0.05



Butterfield and Lishman (1990) attributed low fertility to deficiencies in management programs including feeding regimes. In the present study calving interval between Simmental Tswana crosses and Tswana cows was similar though Tswana cows tended to have a shorter calving interval. However in a typical traditional setup where no supplementary feeding is practiced, difference in performance between the indigenous and temperate breeds would be expected. It has been observed that under nutritional stress the temperate breeds succumb more than indigenous breeds. Peters (1984) attributed prolonged anoestrus when Bos taurus animals are introduced in the tropics to malnutrition. However, in the present study, cows were offered supplements, which perhaps explains the lack of difference between the genotypes. It is also possible that crossing Simmental with Tswana introduced traits into Simmental crosses, which enables them to withstand stress conditions. van Zyl et al (1992) found that cows with more than 50% Afrikaner breeding (Sanga type cattle of South Africa)  had longer calving intervals than Hereford and Bonsmara but not Simmentaler. Ideally the calving interval should be as near to 365 days as possible. However the values reported here are higher than 365 days for both genotypes, indicating that regardless of the supplementary feeding, reproduction was not optimal. Assuming that the gestation period is 280 days and then adding 45 days for uterus involution, the long calving interval is explained by the long period to conception (days open), which had a mean of 297 days. According to Peters (1984), to achieve a 365-day calving interval, calving-to-conception interval should not be more than 85 days. Reasons for a long re-conception period are nutrition, reproductive diseases and heat stress. In on-station studies, APRU (1991, 1992) reported calving interval for Simmental Tswana crosses of 385 and 400 days, respectively. These figures are closer to the 365 days than the values reported in the present study. This difference is indicating variation in management levels between on-station trials and on-farm situation. The implication therefore is that smallholder farmers face difficulties in attaining managerial levels found in the research station.  

The lack of seasonal effect on calving interval may imply that supplementary feeding obviated the effects of season. Often, during the wet season there is sufficient dry matter from grazing and nutrient supply is adequate for reproductive activities.  During a prolonged dry season, nutritional deficiencies cause cows to lose weight resulting in post-partum anoestrus and delayed re-conception (Rennie et al 1977).  

Voh and Otchere (1989) estimated a calving interval of 730 days for Zebu cattle under traditional agro pastoral management in Northern Nigeria. This is higher than the value found in the present study (563 days). Citing other workers (Pullan 1979; Wheat et al 1972; Oyedipe 1982), Voh and Otchere (1989) indicated that under improved nutritional conditions calving interval is shorter than 546 days. In the present study supplementary feeding probably played an important role in shortening the calving interval. The insignificant effects of interaction between season and genotype on calving interval suggest that both breeds may be responding the same to the effects of season. 

The slight difference in the calving rates indicate that slightly more Tswana cows were able to calve than Simmental crosses regardless of the same feeding management. These values are higher than those reported by Rennie et al (1977) of 46%  in Botswana, or by Gubbins and Prankerd (1983)  in Zimbabwe (40-52%) for cows kept as communally managed herds. The difference between these reports and the present study may be attributed to the supplements fed in our study. Rainfall data indicates a high amount of rain was recorded in the rainy season 1987/88 and 1988/89. However, high calving rates were observed in the years 1987 and 1989 followed by the years 1988 and 1990. The trends in calving follow the trends in rainfall that fell in the study area. However, the bimodal pattern reflects the benefits of supplementary feeding. Butterworth (1983) attributed the bimodal calving distribution in the Lowveld of Swaziland to the similarities in the nutritional levels of the forage during the dry and wet months. According to Martin et al  (1992), a seasonal pattern of food availability has resulted in an evolution of the reproductive mechanism of herbivore animals in which plentiful supply of nutrients is a cue to initiating reproductive function. In the present study the initiation of breeding activities was started after the rain due to abundance of feed but was only reflected through calving 10 months later (Figure 1). This observation was confirmed by Butterworth (1983) in which monthly calving of Nguni cattle was highly correlated with rainfall that fell 10 months earlier. The rainy season 1985/86 and 1989/1990 were drought years (Boitumelo, personal communication) and thus their effect on calving rate was reflected 10 months later in 1986 and 1991 respectively. Thorpe and Cruickshank (1980) attributed the effect of year on calving rate of Zebu and Sanga cattle of Zambia to the variation in quantity and quality of forage between years.

According to van Niekerk (1982) low calving percentages can often be ascribed to poor nutrition at critical stages of the cow’s reproduction cycle.  


Inefficiency in reproductive performance is caused by many factors and often it is difficult to identify which one contributes more to inefficient reproduction. However, in the tropics, under-nutrition plays a major role in depressing reproduction. In the present study it was postulated that dry season supplementary feeding of cereal by-products and home grown fodder to Simmental Tswana crosses and Pure Tswana will improve their reproductive performance. This appears to have been the case, since animal performance between the dry season and the wet season were similar, indicating that nutrient supply for reproduction during the dry season may have equaled nutrient supply during the wet season. However, reproduction was not optimal, probably indicating other problems such as diseases, which were not investigated in this study.  


The authors wish to thank B M Mosimanyana for his comments and the Biometrics staff for helping with statistical analysis.  B Mabechu and S Tshenyo provided technical assistance. The International Development Research Centre (IDRC) and Botswana Government sponsored the Small-scale Dairy Project. 


APRU 1990 Livestock and Range Research in Botswana. Government Printers, Gaborone, 23pp.

APRU 1991 Livestock and Range Research in Botswana. Government Printers, Gaborone, 21pp.

APRU 1992 Livestock and Range Research in Botswana. Government Printers, Gaborone, 28pp.

Butterfield W A and Lishman A W 1990 Progesterone profiles of postpartum dairy cows as an aid to diagnosis and treatment of reproductive disorders. South African Journal of Animal Science 20:155-160.

Butterworth M H 1983 Seasonality of calving amongst unimproved cattle in Swaziland. Tropical Agriculture (Trinidad) 60:218-220

Gubbins K E and Prankerd H A 1983 Communal Agriculture. A study from Mashonaland West. Zimbabwe Agricultural Journal 80:151-158.

Martin G O, Boukhliq R, Tjondronegoro S, Hotzel M J and Fisher J S 1992 The effects of nutrition on reproductive endocrinology. Proceedings of Nutrition Society of Australia. 17:77-185

Mukasa-Mugerwa E, Mutiga E R and Girma A 1992 Studies on the reproductive performance of Ethiopian sheep by means of Enzyme Immunoassay Technique; Review. Reproduction, Fertility and Development 4:523-32.

National Development Plan Six, 1985-1991 (NDB 6 1985-1991) Ministry of Finance and Development Planning, Botswana Government.

Peters A R 1984 Reproductive activity of the cow in postpartum period. 1. Factors affecting the length of the postpartum acyclic period. British Veterinary journal 140:76-82

Rennie R, Light D, Rutherford A, Miller M, Fisher I, Pratchtt D, Capper B, Buck N and Trail J C M 1977 Beef cattle productivity under traditional and improved management in Botswana. Tropical Animal Health and Production 9:1-6.  

SAS 1990 Statistical Analysis Software. SAS/STAT Users’ Guide. SAS Inst. Inc. Cary, N.Y.

Thorpe W and Cruickshank D K R 1980 Genetic and environmental influence on beef cattle production in Zambia. 1. Factors affecting weaner production from Angoni, Barotse and Boran dams. Animal Production 30:217-234

van Niekerk A 1982 The effect of body condition as influenced by winter nutrition, on the reproductive performance of the beef cow. South African Journal of Animal Science 12:383-387.

van Zyl J G E, Schoeman S J and Coertze R J 1992 Sire breed and breed genotype of dam effects in crossingbreeding beef cattle in the subtropics. 2. Calving interval and cow productivity. South Africa Journal of Animal Science 22:166-169.

Voh A A and Otchere E O 1989 Reproductive performance of Zebu cattle under traditional Agropastoral management in Northern Nigeria. Animal Reproduction Science 19:191-203.

Received 29 July 2001

Go to top