Introduction

Malawi is in Southern Africa and lies between latitudes 9º22′ and 17º03′ S and longitudes 33º40′ E and 35º55′ E. It is characterized by tropical climate with two distinct seasons, which are rainy and dry seasons with annual rainfall ranging from 700 to 2,400 mm (FAO 2006). Mean annual temperatures vary between 10°C and 25°C depending on altitude, which ranges from 37 to 3,000 m above sea level (Reynolds 2000). Dairy production is an important enterprise in Malawi and comprises a few large-scale farms and smallholder farms. The difference between large and smallholder farms is mainly determined by herd size, breeds raised and management level (Chagunda et al. 2006). Large-scale farms keep large herds of pure dairy breeds and involve high inputs in terms of land, labour, housing, feed and health management. There are about five large farms owning about 12,000 animals (Imani Development Consultants 2004). The Holstein–Friesian is the major breed available in large-scale farms, with a few Ayrshires and Jerseys (Chagunda et al. 2006; Chintsanya et al. 2004). Animals are bred through artificial insemination (AI) using imported semen, with a few standby bulls. The farms grow forages for grazing and make silage and hay. Concentrates are generally based on groundnut, cottonseed and sunflower cakes, maize, soybean and mineral supplements (Chintsanya et al. 2004).

Smallholder farms keep a few animals that are mostly crossbreeds using a low input production system. There are currently, over 8,000 smallholder dairy farmers (Chimbaza 2011) owning about 40,494 dairy cattle (Department of Animal Health and Livestock Development (DAHLD) 2010). The farmers generally practice mixed crop–livestock production systems (Chagunda et al. 2006) with crops mostly being the primary production. The farmers are organized into milk bulking groups where they collectively sell milk to processors. They receive support from government and non-governmental organizations (NGOs) such as Land O’ Lakes and Small-Scale Livestock Promotion Program. The support from the institutions includes access to initial stock and improved dairy technologies. Technologies included are on feeding, housing, health and breeding. The support plays an important role as most farmers may not be able to generate sufficient initial capital on their own to profitably engage in dairy farming (Chindime 2008). In terms of stock, government mostly provides one half Holstein–Friesian × Malawi Zebu crosses, while most NGOs provide pure breeds of Holstein–Friesian or Jersey. The crosses from government are bred within government farms (Gondwe 2011) while pure breeds are imported. The subsector has registered a rapid increase in the past 6 years (DAHLD 2009). The rapid growth of the sub-sector could largely be attributed to support from NGOs and government projects that target dairy production among smallholder farmers.

Despite the rapid growth, the smallholder dairy sector is faced with several challenges. These include limited breeding stock, low productivity, poor management and inadequate access to extension, health and other support services (Goyder and Mang’anya 2009). These challenges have negative impacts on milk productivity and reproduction (Chagunda et al. 2010). With respect to reproduction, the main challenge is low fertility. Reports from one of the high milk-producing areas, Blantyre Agricultural Development Division (BLADD) indicate that calving percentage in 2007 was as low as 30% (BLADD 2008). Fertility is defined by Mukasa-Mugerwa (1989) as the ability of animals to produce healthy offspring in abundance. The trait has two components: firstly, the heifer or cow showing normal oestrous cycles and, secondly, establishing and maintaining pregnancy following artificial insemination or natural service (Lovendahl and Chagunda 2010). For the smallholder dairy cattle producer in Malawi where record keeping is poor (Chagunda et al. 2006), the most practical estimate of fertility is the percentage of mated or inseminated cows that become pregnant (pregnancy rate) or finally calve (calving rate). Pregnancy rate is mainly assessed through failure to return to oestrus.

It is important that the major factors associated with low fertility are identified and addressed as fertility determines herd growth and productivity in terms of milk sale per year. Herd growth may not be the target for most smallholder farmers as keeping more animals may not be practical and economically viable. However, successful reproduction is still important as the farmers benefit through continued milk production and sales of weaned calves. The current paper analyzes the situation and outlines future prospects of smallholder dairy cattle production system in Malawi in relation to cow fertility and outlines major challenges and possible solutions. It is anticipated that the information generated could be used as a basis for interventions to improve dairy cattle productivity among smallholder farmers.

Methodology

A review of the smallholder dairy cow management and fertility was conducted using annual reports from DAHLD and other literature. Most studies on dairy production in Malawi report on the general production system and related challenges with little or no information on cow fertility. Therefore, the review used any information that could directly or indirectly be related to cow fertility to analyze the current status of cow fertility. Reports from DAHLD also provided some raw data on some aspects of fertility, mainly, cow inseminations and births. These data were from three agricultural development divisions (ADDs), Lilongwe, Blantyre and Mzuzu (Fig. 1), which house the major milk-shed areas in Malawi. Descriptive statistics were generated from the data, and comparisons were made between ADDs over a period of 3 years. Challenges and opportunities reported were also compared, and inferences were made based on this information. The productivity status was further compared to study findings from similar production systems in the tropics.

Fig. 1
figure 1

Map of Malawi showing the agricultural development divisions

Full size image

The data from literature and the survey enabled the paper to provide an overview of cow management and fertility in smallholder farms. Feeding, housing and breeding management was analyzed and related to the nutrition, health and fertility status of the cows. Calving rate and calving interval were analyzed to represent the direct indicators of fertility. Calving rate or percentage is defined as ‘the percentage of breedable females which calve during a given year’ (Perera 1999), while calving interval is the period between two consecutive calvings (Gietema 2005). Other commonly used measures of fertility such as submission rates, conception rates and conception to calving interval could not be used as they require detailed recording, which does not currently occur among the smallholder farms surveyed. Data on the dairy herd structure of the three ADDs from 2007 to 2010 were entered in Microsoft Excel 2007 and exported to SAS 9.1 where calving rates were calculated and analyzed using the general linear model at 5% probability. Calving interval data were only reported in 2009 in Blantyre ADD. Means were calculated in Microsoft Excel 2007, and the data were not subjected to further analysis as the dataset was too small.

The literature was supplemented with primary data obtained from a baseline survey conducted in selected smallholder dairy farms between December 2009 and January 2010. The farms represented the Central (Lilongwe and Kasungu), Southern (Thyolo) and Northern (Mzimba) milk-shed areas. Data were collected using structured and semi-structured questionnaires that were administered to individual dairy farming households and key informants, respectively. A total of 163 dairy farmers were interviewed where 14, 40, 43 and 66 were from Kasungu, Lilongwe, Thyolo and Mzimba Districts, respectively. District and EPA extension staff as well as local leaders served as key informants. Management practices in animal breeding, record keeping, milk production, feeding, housing and health were recorded. Access to inputs and other services such as extension and health was also investigated along with challenges and perceived solutions to challenges outlined. Individual farms also reported on how long it took to inseminate the animals after calving. This information was used to give a rough indication of the calving to conception interval.

The baseline survey data were entered and analyzed in SPSS 15.0 where descriptive statistics that included means, percentages, frequencies and crosstabs of variables were generated. Chi-square and analysis of variance were used to determine differences between districts at 5% probability. Differences between means were assessed using the Student–Newman–Keuls multiple comparisons test.

Results and discussion

Feeding system

The baseline survey showed that the majority of smallholder farmers (94%) use zero grazing feeding system. Only a few farmers in Lilongwe (4%) and Mzimba (2%) used herded grazing. In the zero grazing system, cows are kept in pens throughout the year, and feed is always provided for them. Maize bran was the major type of concentrate that most farmers (96%) provided to the animals. Most farmers had established pastures (76%). Significantly fewer (p < 0.05) farmers (37%) had pastures in Thyolo than in Lilongwe (95%), Mzimba (87%) and Kasungu (86%). This could be attributed to land being a limiting factor in Thyolo. A considerable proportion of farmers (65%) conserved feeds, and most of these were from Mzimba (25%) and Lilongwe (21%). The types of pastures were mainly grasses as evidenced by Napier (Pennisetum purpureum, 89%) and Rhodes (Chloris gayana, 10%) grass from the survey. These results are similar to findings from the central and northern milk-shed areas by Chindime (2008) where 94% and 5% of the farmers utilized Napier and Rhodes grass, respectively. The prevalence of Napier grass could be related to the high dry matter (DM) yields, which are more than most tropical grasses (Orodho 2006).

Legumes, which are important forages in supplementing protein, do not seem to be routine practice (Mtimuni 2011) although their use by smallholder farmers has been reported (Chakeredza et al. 2008; Chindime 2008). The reports, however, do not specify the quantities fed and the actual feeding practices. Fodder legume production has challenges such as inaccessibility of planting materials and farmers not appreciating their importance (Chindime 2008). Promotion of legume production coupled with proper demonstrations and diet formulation could be used to improve dairy cattle feeding. Access to financial resources has also been found to play an important role in dairy feeding technology adoption. Use of supplements such as dairy mash, molasses, legumes and minerals was positively associated with farmers that had access to credit (Chindime 2008). This implies that proper dairy management is dependent on several factors, and it is important that basic services such as extension and finance be made available so that other specific factors that need to be addressed are not masked.

The sizes of the pasture plots are often inadequate to provide feed throughout the year. From the baseline survey, plot sizes averaged 0.26 ± 0.19 ha with a median of 0.2 ha. Assuming an average DM yield of 16 t/ha/year (Orodho 2006), the Napier grass DM yield per year would be 4.2 t in these plots. For a lactating cow weighing 500 kg, required daily DM intake would be 15 kg (Orodho 2006), and this forage would last for about 9 months. Given that farmers tend to own more than one cow, the forage supply would last much less than 9 months. In addition, the feed presentation, frequency and amounts are often inappropriate. This implies that the dairy animals may not be getting adequate feed. Reports by Gibbons et al. (2010) and Munthali et al. (1992) allude to inadequate feeding of dairy animals by smallholder farmers. Munthali et al. (1992) reported that the quality of the feed available was insufficient to meet the nutritional requirements of dairy cows. Even where feeds were available, another challenge cited is the knowledge of extension workers in formulation of suitable rations.

The feeding of mostly maize bran and grasses may also mean that important nutrients such as protein and some minerals may be inadequate in the diets. Crude protein (CP) content in maize bran is 11% (Mtimuni 1995), while in grasses, it is mostly below 7% during the dry season (Mtimuni 2011). Moran (2009) reported that CP requirements range from 10% to 18% depending on stage of lactation. This shows that maize bran barely meets the requirement for dry cows while grasses are very deficient in CP. The inadequate feeding coupled with some nutrient deficiencies results in negative energy balance, which in turn contributes to low fertility. Friggens (2003) reported that animals with constrained energy reserves either reduce their ‘reproductive burden’ or delay onset of the next reproductive cycle. For a dairy cow carrying one foetus at a time, reduction of the ‘reproductive burden’ implies no pregnancy and delayed onset of reproductive cycles means long calving intervals. Lanyasunya et al. (2005) reported that poor nutrition is the major cause of infertility and low productivity of dairy cattle under smallholder systems and recommends more attention to be given to nutrition for productivity to improve.

The findings on the feeding system have implications on cow productivity including fertility. Zero grazing has advantages in that animals are protected from between herd disease transmission for diseases such as East Cost Fever. In addition, it is relatively easier for farmers to observe animals on heat and call for timely service. However, the system has a disadvantage where AI services are inefficient, and this has been the case in Malawi for some time (Chagunda et al. 1998). There are high chances of animals not being served even when oestrus is correctly detected. At the same time, the cows miss out on the chances of timely insemination that could occur with natural mating if they grazed with other herds running with bulls.

Housing system

The cattle housing structures vary in terms of construction materials used. Roofs are made of thatch grass, iron sheets or a combination of plastic sheets and thatch grass while the floors are from mud, concrete or sand with or without beddings. The baseline survey results showed that irrespective of construction material, most houses (83%) were adequately roofed with a few that were not roofed (11%) or inadequately roofed (6%). A few (23%) farmers used beddings on the floors with Mzimba District having about 51% of these. About 54% of the pens had good drainage. Feed stores, milking parlours, feeding and water troughs were present in 57%, 79%, 77% and 80% of the farms, respectively.

The results suggest that some farmers have not fully embraced the importance of proper housing of dairy animals. Inadequate roofing and poor drainage can result in accumulation of slurry during the rainy season, which would be difficult to clean given the many tasks farmers need to perform particularly in the rainy season when crop production is of prime importance. Hence, accumulation of slurry is inevitable and is observed in most dairy farms. Accumulation of slurry would be detrimental to the animals and provide a medium for pathogens as well as high chances of milk contamination. The use of iron sheet roofs also needs to be critically evaluated because they conduct heat and could be associated with heat stress mainly in the hot-dry and hot-wet seasons.

Dairy breeds and population

The Malawi Government has recommended Holstein–Friesian, Ayrshire and the Jersey as dairy breeds (DAHLD 2006). Results from the baseline survey showed that most farmers (54%) owned the Holstein–Friesians with no significant differences (p > 0.05) between districts. Table 1 shows that the number of Holstein–Friesians was significantly higher (p < 0.05) than that of other breeds while the numbers of the rest of the breeds were not significantly different (p > 0.05) from each other. Other farmers kept Holstein–Friesian crosses with the Malawi Zebu (16%) and Jerseys (13%). Respondents in Thyolo District kept the highest percentage (80%) of pure Holstein–Friesians and their crosses with Malawi Zebu. The rest of the districts were also dominated by farmers owning Holstein–Friesians but with noticeable percentages of farmers owning Jerseys and Jersey × Malawi Zebu crosses. Lilongwe District had the highest percentage (32%) of farmers keeping Jerseys, followed by Mzimba (25%). The differences in breeds kept could be attributed to availability of the animals as provided by NGOs and other supporting agents.

Table 1 Breeds of dairy cattle owned by smallholder farmers in Mzimba, Kasungu, Lilongwe and Thyolo Districts
Full size table

The national dairy population database does not specify the dairy breeds. Instead, the cows are only classified as either pure or crossbreds. The data show that there are more crossbreds than purebreds with crossbreds accounting for 79% of the dairy population in 2010. The trend is the same throughout all the ADD as shown in Table 2. The table shows that the crossbreds account for more than 55% of the dairy herd in each ADD with Salima ADD having a dairy population comprised of crosses only. The ADDs in the northern milk-shed area (Karonga and Mzuzu) have relatively higher proportions of purebreds compared to the central and southern milk-shed areas. The differences in the distribution of the breeds could be due to differences in availability and pricing of the cattle. Crossbreds are relatively more available than purebreds. Their availability is mostly through AI births and sales from Government farms. As such, they are relatively cheaper than the purebreds.

Table 2 Distribution of dairy cattle population by Agricultural Development Division in Malawi in 2007/2008
Full size table

The presence of more purebreds in the north could reflect that there are more farmers that can afford the purebreds or there are more NGOs supporting purebreds supply to farmers in the northern milk-shed area than the other milk-shed areas. The national census data indicate that there is a high proportion of crossbreds compared to the trend in the survey findings where farmers reported more purebreds. The reason for this discrepancy could be farmers not being able to distinguish between purebreds and high-grade crosses. It may also be a reflection of lack of proper record keeping by the farmers as well as lack of a proper national inventory.

Table 2 also shows that the highest proportion (73%) of the dairy population is in Blantyre ADD, while the populations in Mzuzu (11%) and Lilongwe (8%) come second and third, respectively. This shows that dairy production is concentrated around the cities with most production in the southern region. Imani Development Consultants (2004) reported that dairy production is most developed in the southern region and attributed this to small land holding size, which favours dairy production. Blantyre ADD is also within the Shire Highlands, which is characterized as having favourable climate with relatively low disease challenges (Imani Development Consultants 2004). Shire Valley ADD has the lowest (0.2%) proportion of dairy animals followed by Karonga (0.5%) and Salima (0.6%) ADDs, respectively. The reason for this could be that the respective agro-ecological zones (AEZ) do not favour dairy production and that dairy production support in these ADDs is relatively lower than in the other ADDs. Shire Valley ADD is within an AEZ characterized by low altitude and low annual rainfall (<600 mm) while Karonga and Salima ADDs are in the lakeshore low altitude rain shadow area with 600–800 mm annual rainfall (Jere 2007).

Overall, the dairy population increased by 65% from 2004 to 2010. Figure 2 shows the trend in the population of dairy cattle from 2005/2006 growing season. The figure shows that the total dairy population has been increasing over the years. The rate of increase of pure breeds decreased sharply from 61% to 20% between 2006/2007 and 2007/2008 growing seasons and gradually increased to 30% in 2009/2010. The increase rate of crosses has been declining throughout the years. The rate declined from 24% in 2005/2006 to 10% in 2009/2010 growing season. The results may be a reflection of the importation and reproductive management of dairy cattle in the country. The high rate of purebreds increase in 2005/2006 could be due to a high number of animals that was imported with slightly fewer animals imported in 2009 and 2010. The perpetual decline on the increase rate of crosses could be an indicator of low replacement rate of the dairy herd, which could be related to cow fertility. Low replacement rate is also evidenced in the dairy herd structure obtained from both the survey and data from the ADDs as discussed below.

Fig. 2
figure 2

Dairy cattle population trend in Malawi from 2005 to 2010 (Data Source: DAHLD 2010)

Full size image

Herd structure

In the smallholder dairy farms, the average herd size is generally two to four cows (Kasulo et al. 2010; Chagunda et al. 2006; Chintsanya et al. 2004). Chagunda et al. (2006) and Kasulo et al. (2010) reported average herd sizes of 3.5 and 2.2 for Lilongwe and Mzuzu milk-shed areas, respectively. Results from the survey showed an average herd size of 2.2 ± 1.6 (SD) with no significant difference between districts (Table 3). The herds are mostly comprised of cows. A very small percentage of farmers (9%) own bulls. The average number of cows per farm was 1.2 ± 0.78. The range of the number of cows was 0–4 while Chintsanya et al. (2004) reported a range of 2–4 cows for dairy farmers in Malawi. The differences in the minimum could be attributed to Chintsanya et al. (2004) reporting on well-established dairy farmers while the survey included even new entrants into dairy farming.

Table 3 The structure of dairy herds among smallholder farmers in Mzimba, Kasungu, Lilongwe and Thyolo Districts
Full size table

The herd structure shows that cows make up 36% of the herd while heifers and heifer calves make 19% and 15%, respectively. Based on data from ADD annual reports, a similar herd structure is observed for Lilongwe, Mzuzu and Blantyre ADD herds (Fig. 3). The percentage of the different classes of the herds remained relatively stable and similar between the ADDs since 2007. The average proportion of cows, heifers and heifer calves between 2007 and 2010 within the ADD herds was 45%, 21% and 12%, respectively.

Fig. 3
figure 3

The dairy herd structure in Lilongwe, Blantyre and Mzuzu ADDs between 2007 and 2010 (Data source: BLADD 2010; LADD 2010; MZADD 2010)

Full size image

Heifers and heifer calves are necessary as replacements for cows in the dairy herd, and surplus heifers are an important component of the income from dairy production. Therefore, proper reproductive management is important for successful dairying (Stewart 2005). Table 3 shows that the percentage of heifers is slightly below what may be considered sufficient as replacement stock while Fig. 3 shows that the ADDs could be within the ideal proportion if it is assumed that cows remain in a herd until five lactations (Gietema 2005). Due to few numbers of cows kept, it is more likely that farmers in Malawi may keep cows for more than five lactations as similar farms in Tanzania reported to eight lactations (Bee et al. 2006). For cows remaining in the herd for four to five lactations, 20% to 25% should be replaced every year. However, if it is taken into account that farmers own 1.2 cows on average, then the current proportions of heifers may not enable stable herds to be kept. This is evident when calculations are made according to Bebe (2008a).

Bebe (2008a) reported that heifers for stock replacement are described as the number of heifers surviving to breeding age and are expressed as the number of heifers per cow leaving the herd through death or sales. To maintain herd size, the number of heifers per cow leaving the herd has to be equal to or more than one. Table 3 shows that the average number of cows and heifers per farmer is 1.16 and 0.85, respectively. Assuming that one cow in the herd dies, then the replacement heifers per cow is 0.73 suggesting that the current herd structure cannot be maintained. Cow mortality is assumed based on mortality rates reported in the baseline survey (14%) and other similar production systems (19% and 12%) in Zimbabwe (Masama et al. 2006) and Kenya (Bebe et al. 2003), respectively.

It would be reasonable for farmers to target having at least 25% heifers for their herds to be maintained. Moran (2009) illustrated that a system with 36% heifers, for example, would have sufficient replacement stock as well as enable sale of breeding stock or allow higher culling rate to improve herd genetics. Most smallholder farmers may not necessarily aim for expanding herds but would benefit from sales of breeding stock.

Bebe (2008a) reported similar results for smallholder dairy farmers in semi-zero and zero grazing systems of Kenya. The reason for low replacement stock could be a reflection of low fertility, poor reproductive management and/or deliberate decisions by farmers. Bebe (2008b) reported that some Kenyan farmers deliberately have few replacement animals to minimize labour for feeding animals. Bebe (2008b) showed that the proportions of heifers to cows among smallholder farmers in Kenya were related to feeding resources. Farmers that practiced zero grazing kept the proportions low through early weaning and sales, and it was concluded that availability of sufficient pastures determined the ability to rear adequate stock replacement heifers. In Malawi, it is not apparent that farmers deliberately keep the replacement stock low. The low numbers of replacement stock could be more as a result of poor reproductive management and hence low fertility. Herds that cannot be maintained internally often rely on external replacement and this is the situation in Malawi. The implication is that the farmers can hardly produce surplus stock to be sold to other farmers. Hence, if dairy production is to expand, importation of dairy stock has to be in place.

Breeding

Cows and heifers are mated using AI (55%) or bulls that are hired (36%) or borrowed (1%) from other farmers. Most of the semen used is from the National AI Centre (NAIC) based in Blantyre ADD. NAIC currently has four Holstein bulls from which semen is collected, processed and distributed throughout the country (Mayuni 2003). However, the performance of the bulls has not been evaluated (Gondwe 2011). The farmers target to inseminate the animals immediately after the recommended 60 days voluntary period. However, semen availability is often a challenge due to storage and transportation problems (Masangano et al. 2009).

There were significant differences between districts (χ 2; p < 0.05) between use of AI and bulls. AI was mostly used in Thyolo (74%) and least in Kasungu District (36%). Hired bulls were mostly used in Kasungu District (64%) and least in Thyolo (19%). The reason for these differences could be associated with proximity to the source of semen straws. Thyolo District is closest to NAIC; hence, farmers from Thyolo could access semen straws relatively easier than those in the Central and Northern Regions. Mayuni (2003) reported low conception rates as distance from NAIC increased. It is possible that farmers in the north milk-shed area rely less on AI after experiencing low conception rates with use of semen from NAIC.

Another reason for more AI users in Thyolo could be the availability of well-trained and experienced AI technicians as dairy farming is more organized and intensive in BLADD (CYE Consult 2009). In other areas such as Kasungu and Mzimba, AI technicians have challenges such as lack of transport, shortage of AI equipment and high AI technician farmer ratios (Masangano et al. 2009). These challenges lead to delayed or missed inseminations thereby making AI seem unreliable to farmers.

Herd health

Farmers mostly access health services through Government assistant veterinary officers, private veterinary services and drug revolving funds. Munthali et al. (1992) reported that the health services were not adequate, while Mwale et al. (1999) reported that animal health problems were ranked first among problems that dairy farmers encountered. Survey results showed that 11% of the farmers did not treat sick animals due to either unavailable or unaffordable drugs.

Farmers report various health challenges in dairy production. A study by Chindime (2008) showed percentages of farms that reported particular diseases as follows: mastitis (32%), East Coast Fever (ECF, 28%), fever (16%), coughing (8%), pink eye (7%), diarrhoea (6%) and sore feet (5%). These challenges are similar to those found in the baseline survey where mastitis, ECF and diarrhoea were reported by 35%, 19% and 16% of the farms, respectively. Other diseases reported in the survey were black quarter, tuberculosis, brucellosis and foot and mouth disease. Munthali et al. (1992) reported that health services were biased towards control of ECF with important health problems such as gastro-intestinal parasites and pneumonia not receiving adequate attention. This is in agreement with Huttner et al. (2001) who reported high incidences of helminthiasis, fascioliasis besides tick borne diseases. In addition, Trypanosomiasis is reported in cattle close to national parks and game reserves while black quarter is prevalent in some areas where it is considered endemic. Some zoonotic diseases such as tuberculosis and rabies are also reported.

The results show that dairy cow health is a challenge among smallholder farmers. Disease control is further limited by not treating sick animals or waiting too long before diseases are reported. Absence of detailed record keeping and disease surveillance makes it difficult to quantify the actual prevalence and severity of diseases affecting dairy animals. Huttner (2000) reported malnutrition and poor animal husbandry as the major predisposing factors to poor animal health among smallholder farms. The combination of health and nutrition challenge in dairy production implies further suppression of animal productivity in terms of milk yield and fertility. It is therefore important to have in place programmes that constantly and systematically monitor and support disease control while providing timely treatments and vaccinations as need arises. This requires the collaboration and participation of farmers, service providers and other relevant stakeholders.

Herd productivity

Milk yield varies depending on the level of management and genotype used. Malawi Zebu produces 2–4 kg/day (Mwale et al. 1999) while crossbreds produce daily averages ranging from 3 to 11 kg (Chindime 2008; Mwale et al. 1999; Agyemang and Nkhonjera 1990; Mussa et al. 1986). The yields are generally low when the potential milk yield of the breeds used is considered. A wide variation in milk yield is observed between farms within individual studies, which might be a reflection of differences in feeding levels and health of the animals. Farmers with access to resources are associated with high yields (Chindime 2008). The wide variation between farms indicates that the potential to improve yield by adjusting management in terms of feeding and health exists. Another challenge related to cow productivity is the apparent lack of understanding among farmers of the link between management input and productivity. Most farmers record milk yield in order to track their payments from milk sales (Chindime 2008). However, the farmers seem not to evaluate milk yields to reflect on inputs used or level of management. It would be worthwhile to help farmers appreciate this linkage.

In addition to milk yield, body condition score (BCS) is an important trait which forms an integral part of dairy records in many countries and informs cow productivity. BCS is generally recognized to provide a measure of the body energy reserves (Roche et al. 2009). Low BCS during late gestation and early lactation is associated with delayed resumption of ovarian cyclicity which could reduce chances of subsequent reproductive success (Chagas et al. 2007) in terms of parturition and rebreeding interval. Where resources are limited, there may be an advantage in a greater interval between pregnancies, but the target is to improve resource availability to increase dairy productivity. Currently, smallholder farmers in Malawi do not monitor BCS. There is a need to demonstrate the relationship between BCS, milk yield and management input and link this to profitability and fertility so that farmers understand their relevance and hopefully adopt record keeping, evaluation and recommended husbandry techniques.

Status of cow fertility

Calving rates

Calving rate is one of the simplest means to estimate fertility. It is a retrospective measure, which is useful for evaluating overall herd performance in a given period. Ramsey et al. (2009) reported that calving rate gives an indication of production management skills and is negatively related to production costs. The overall calving rates in the ADDs ranged from 22% to 61% with the trend fluctuating between breeds and over time (Table 4) but with no significant differences (p > 0.05). Calving rates based on artificial insemination records for the ADDs are even lower with Mzuzu ADD having the lowest rate of 16% in 2008. Lilongwe and Blantyre ADDs had relatively higher calving rates than Mzuzu ADD. However, the number of inseminations reported for Lilongwe ADD is much smaller than the population of dairy animals in the ADD. This may mean that recording of AI in Lilongwe ADD may not be up to date or that more farmers rely on natural service than AI.

Table 4 Calving rates of dairy cattle in Blantyre, Lilongwe and Mzuzu ADDs in 2007, 2008 and 2009
Full size table

The differences in calving rates between ADDs could be attributed to differences in recording skills and frequency as well as availability of AI services. MZADD is furthest from the major source of semen. Mayuni (2003) reported low conception rates as distance from the source of semen increased. The low conception rates were associated with poor semen quality. Reports from respective ADDs indicate lack of AI equipment and inadequate AI technicians as the major challenge to AI service delivery (LADD 2010; BLADD 2009; MZADD 2009). In addition, Mzuzu ADD cites irregular supply of semen and liquid nitrogen (MZADD 2009) while Lilongwe ADD also cites lack of incentives for farmer AI technicians as a challenge (LADD 2010).

There could also be other factors such as experience and transport availability for AI technicians and the ability of farmers to detect and report heat on time as well as poor health, inadequate nutrition and farm management skills. For instance, in 2009, BLADD reported an overall pregnancy rate of 71%, but calving rate came down to 47%. This may be reflection of challenges in animal health and farm management, which result in mortality, abortions and still births. The low calving rates could also be related to inadequate nutrition as discussed above. Calving rate is influenced by factors related to animals, the farm management as well as insemination services (Nordin et al. 2007), and these may be difficult to isolate unless detailed studies are carried out.

Generally, the AI records from the ADDs are not clear on whether the births reported are only from AI or include births from natural service. As it is established that a considerable proportion of farms depend on natural service, it would be worthwhile to present records on natural service and subsequent births separately. Despite the shortfalls, the data provide an indication of the existing calving rates, which are generally low but improving in LADD and BLADD. According to Perera (1999), acceptable calving rates for smallholder farms should be more than 70%. The calving rates in the ADDs are below 70% and similar to those reported for Zimbabwean smallholder dairy farms where calving rates were 30% (Masama et al. 2003) and 53.2% (Masama et al. 2006). This may be depicting poor reproductive management coupled with inadequate nutrition as discussed above. This implies that the farmers need further capacity building in dairy cattle management besides the requirement for improved AI and extension service delivery.

Calving interval

Calving interval (CI) is a trait that shows overall fertility over time, and Mukasa-Mugerwa (1989) reported that it is the best measure of reproductive efficiency. As with calving rate, it is a retrospective measure that is used to evaluate farm reproductive efficiency. BLADD reported overall CI of 457 ± 59 days (with a range of 405 to 540 days) in 2009, while this information was not available from the other ADDs. The CI is lower than average of 485 days reported by Agyemang and Nkhonjera (1990) for smallholder dairy herds in Southern Malawi. The difference could be attributed to the current herds being a mixture of purebreds and crossbreds while (Agyemang and Nkhonjera 1990) reported CI for crossbreds only. The CI is also shorter than the average CI reported in smallholder farming systems in Tanzania (Swai et al. 2005), Ethiopia (Lobago et al. 2007; Shiferaw et al. 2003) and Zimbabwe (Masama et al. 2003) which were all greater than 500 days. Perera (1999) suggested CI between 13 and 14 months (390–420 days) as acceptable under improved smallholder conditions. Other studies among smallholder farms in Tanzania and Ethiopia have shown CI less than 420 days (Chenyambuga and Mseleko 2009; Yifat et al. 2009) indicating that achieving lower CI is possible. This suggests that there is room for improvement on the current CI in Malawi.

Calving to conception interval

Calving to conception interval is an integral part of the calving interval and is directly affected by the voluntary waiting period before service. Generally, the voluntary waiting period is 60 days (Perera 1999). Among farmers that rely on AI services, this period may be prolonged by availability of AI services, inaccurate heat detection, absence of observable heat and/or prolonged post-partum anoestrus. The survey provided an indication of days to service based on verbal reports by farmers interviewed. The days ranged from 60 to 270 with an average of 107 ± 38 days and a median of 90 days. However, only a small proportion of farmers (4%) from Thyolo and Kasungu indicated that they served their animals more than 150 days from calving. If this proportion of farmers is excluded from the data, the average is 97 ± 24 days (Table 5).

Table 5 Number of days from calving to service for dairy cattle in Mzimba, Lilongwe, Kasungu and Thyolo Districts
Full size table

There were significant differences between districts (p < 0.05) with 79% and 73% of farmers in Mzimba and Lilongwe Districts serving their animals within 90 days after calving compared to 49% and 29% of the farmers from Thyolo and Kasungu Districts, respectively. This might be an indication of differences in access to animal breeding information or technology adoption rates between districts. Generally, the days from calving to service are higher than the ideal of 60 to 65 days. Suggested acceptable days to service for smallholder dairy farms in the tropics are less than 90 days (Perera 1999), while Moran (2005) reported acceptable calving to conception interval of less than 115 days. Some farmers in Malawi could be within these suggested intervals, while others could be above the intervals. There is need for further improvement on breeding practices in order to shorten the CCI and subsequently the CI.

For some farmers, a CCI of less than 115 days is not possible due to inefficient AI service delivery. In other smallholder farms in the tropics, CCI above 115 days are also reported. Yifat et al. (2009) reported CCI of 135 days in Ethiopia. The long CCI could also be exacerbated by prolonged post-partum anoestrus as a result of inadequate feeding and stress as discussed above. The management-related issues such as accurate heat detection, nutrition and health status may also largely contribute to the current status of CCI and CI.

Pregnancy diagnosis

A few farmers (23%) indicated that pregnancy diagnosis (PD) was done through rectal palpation by either the government (65%) or farmer AI technicians (35%) following an insemination. The average timing of the PD was 91 ± 33 days. Other farmers indicated that they diagnosed pregnancy on their own using non-return to oestrus and physical appearance as indicators. The PD by the technicians is done too late for timely repeat inseminations. The rectal palpation method that is used can be used as early as 30 to 40 days from insemination (Fricke 2010; Sheldon and Noakes 2002). It is worthwhile to encourage farmers to conduct PD using the simplest method, non-return to oestrus. Although the non-return to oestrus is not reliable (Fricke 2010) due to missed oestrous behaviours, it would be much better in this case (with appropriate heat detection skills) than waiting for a technician to come for rectal palpation after 3 months.

Conclusion

Management and fertility of dairy cows among smallholder farms is faced with both challenges and opportunities to improved productivity that are related to feeding, housing, health and breeding systems. Both the challenges and opportunities are influenced by the extent to which farmers have access to important services such as extension, health, breeding and finance. Feeding systems used seem inadequate in terms of quantity and quality. However, opportunities for improvement are available in terms of availability of technologies and stakeholders to promote the technologies and build capacity in farmers and extension workers. Dairy housing also needs to improve with some aspects, such as the most suitable construction materials, needing further assessment. The efficiency of the health and breeding systems also needs improvement. There is need for systematic health surveillance and recording to be in place in order to monitor health and reproductive management in the farms. These would help to identify specific gaps and suitable approaches to be used to address the issues

The potential for improved productivity is evidenced by wide variation in the magnitude of various traits that were analyzed. Improvement in service delivery and further capacity building of both farmers and extension staff is required in order to improve dairy management skill and subsequent productivity. Further detailed research in specific aspects of feeding, housing, health and breeding systems would help to identify specific interventions that could be used to improve dairy cow productivity.