Livestock Research for Rural Development 18 (5) 2006 Guidelines to authors LRRD News

Citation of this paper

An analysis of smallholder farmers' willingness to adopt dairy performance recording in Malawi

M G G Chagunda, A C M Msiska*, C B A Wollny**, H Tchale*** and J W Banda*

Department of Animal Health, Welfare and Nutrition, Danish Institute of Agricultural Sciences, Research Centre Foulum,
P.O. Box 50, DK-8830 Tjele, Denmark
Mizeck.Chagunda@agrsci.dk
*Department of Animal Science, Bunda College of Agriculture, University of Malawi, Lilongwe, Malawi
**Animal Breeding and Husbandry in the Tropics and Subtropics, Georg-August University Goettingen, Goettingen, Germany
***Center for Development Research (ZEF), University of Bonn, Bonn, Germany


Abstract

Adoption of dairy recording schemes remains poor and there is no centralised dairy recording scheme in place in Malawi. To investigate some of the root causes, a study on farmer perception on record keeping and factors affecting its adoption was conducted. Eighty-six smallholder farmers from six dairy cooperative schemes of Lilongwe were randomly interviewed. A logistic regression approach was used to analyze the adoption decision.

The results indicated a positive relationship between participation in dairy recording and the individual assigned the recording task (P<0.01), milk recording using simple calibrated containers (P<0.05) and also herd size (P<0.05). A negative association was found between recording participation and recording using calibrated scales (P<0.01). Sale of milk at the informal market as opposed to the formal market, and use of natural service as opposed to artificial insemination for breeding slightly affected recording participation. Farmer education level, cattle genotype, and daily milk yield had no significant influence on the adoption of milk recording. Based on the experiences and lessons learnt, a farmer-participatory recording system using simple equipment like calibrated one-litre cups and based on daily recording was designed and tested.

Key words: Dairy recording, Malawi, smallholder, willingness


Introduction

Peri-urban smallholder dairy sector supplies about 60 % of the milk that is processed at the formal processing plants in Malawi every year (Banda 1996). Recent information (Malawi Government 1997) indicates that there are about 3600 smallholder farmers who use over 6000 Holstein Friesian x Malawi Zebu cows and about 1700 smallholder farmers who use an unknown number of Malawi Zebu cattle for commercial milk production in the peri-urban setting. In addition to the smallholder farmers, there are 15 private large-scale dairy farms accounting for about 2200 milking cows. The major differentiating features between smallholder and large-scale dairy farms are the holding size, the genotype of cattle raised and the level of management applied. The predominant genotype on the large-scale dairy farms is the Holstein Friesian although some of these farms also have few Aryshire and Jersey cattle while smallholder farmers utilize Holstein Friesian x Malawi Zebu crosses of different grades.. The total milk production from both the large scale and the smallholder sub-sectors as at the year 2001 was estimated at 35 000 metric tonnes per year (FAO 2005). The smallholder dairy farmers are organised in three milk shed areas around the three major cities of Malawi and operate under corporate approach where at local level farmers belong to milk bulking group. Farmers from within a radius of 8 kilometres km bulk their milk at a cooling centre from where milk processors collect it. Buying of the milk by the processors is in bulk and a bonus is paid for higher bulk quantities. Malawi consumes about 42 000 metric tonnes of milk per year (FAO 2005). With a population of about 11 million people, the estimated average milk consumption is 3.8 kg per capita. This average is very low even when compared with that for Sub-Saharan Africa, which is estimated at 30.8kg per capita (FAO 2005). Coupled with the fact that Malawians get 7.46 calories, 1.18 grams, and 0.27 grams per day from milk compared to 52 calories, 2.7 grams, and 2.9 grams of energy, protein and fat for Sub-Saharan Africa, the need for improving milk production and the consequent milk consumption in Malawi is heavily pronounced.

Although the smallholder farms play an important role in milk production in Malawi, holistic dairy development evaluations for herd monitoring and assessment based on large amounts of longitudinal data are currently difficult if not impossible to carry out because of the unavailability of systematic records (Chagunda et al 1998). As was indicated by Howard and Cranfield (1995), animal recording system is not only of importance for any breeding program, but can also be "a lead technology" for all actions, which aim at the improvement of the global productivity and efficiency of farm systems. Galal (1998) defines animal recording as an activity that involves the measurement of various indicators of animal performance and the use of that information in the decision making process. Flamant (1998) expands further on this and defines it as a process dedicated to the collection of information on animals, completed by its processing, interpretation, and, dissemination of the results in a perspective of decision making for choosing breeding animals for future generations. In modern dairy farming, successful management relies on good record keeping and on information that can be derived from it. Farm records are to be utilised routinely for daily management and to solve problems. A quantitative knowledge about a farm provides the basis for understanding where the dairy has been, where it is today and where it is going (Jelan and Dahan 1998). It is realized, however, that substantial improvements in milk production can be achieved through establishment of simple, accurate, understandable and easy to keep recording systems. Recording of animal performance is of enormous value for management decision-making for both individual farmer and for the industry or country as a whole. Animal recording for animal management involves monitoring of each animal's performance and use of that information in normal, day-to-day farm management (Galal 1998). This represents the integration of performance data into the farm management process and permits more effective decision-making at farm level on an on-going basis. Furthermore, milk records can be used as a diagnostic tool and for detecting different kinds of health and reproduction issues in the herd. In Malawi, a smallholder milk-recording scheme that was set up in 1974 by the Ministry of Agriculture did not pick up. In 1990 the Ministry tried to reactivated the recording system efforts but with no success (Zimba 1993). To date there is no coordinated animal performance-recording scheme in Malawi despite these efforts. In order to find out why past efforts failed, this study was aimed at determining the herd level factors associated with the willingness of smallholder farmers to uptake a recording scheme.


Materials and Methods

Farm selection

The study was conducted in Lilongwe milk shed area covering Lilongwe and Kasungu Agriculture Development Divisions (ADD) in the central region of Malawi. There are a total of 18 Milk Bulking Groups (MBGs) in this milk-shed area with approximately 257 smallholder dairy farmers. Farmers from six MBGs to the north, in the central, and to the south of Lilongwe city were involved in the study. The study was conducted in two phases namely, a survey and a monitoring study. Before the survey, reconnaissance group meetings with the smallholder dairy farmers, dairy extension officers, and veterinary field officers were held to introduce the general objectives and methodology of the study. Respondents in the survey were randomly selected from a list of all members of the MBG.

Interviews

Initial interviews utilized a structured questionnaire, which was administered to a sample 86 randomly selected smallholder dairy farmers. This was done in order to describe the social background, farm characteristics, livestock ownership, and self-evaluation of farmers. Interviews were conducted between February and March 2001. In addition, a concluding interview was conducted at the end of the monitoring period to get feedback from the farmers on the practicability and feasibility of the tested alternatives.

Monitoring study

The monitoring study commenced after the initial questionnaire and involved the development and testing of alternative recording systems. Ninety randomly selected farmers from the six MBGs were involved in the study. The farmers were randomly assigned to two multiple-trait recording systems within which were three recording intensities. This culminated into 6 system-interval combinations. These alternative recording systems were designed based on two hypothesized principles of labour intensity and level of sophistication as a direct lesson coming from the initial survey results. The two systems were a) a system that required farmers to record on milk yield, reproductive and mating activities, pedigree information, disease and their treatments, and b) a system where everything in system (a) was recorded with the exception of reproductive and mating activities. Within each system there were three recording intensities of daily, weekly, and monthly. No interventions were introduced in the livestock husbandry practices of the farms. Based on what was learnt in the survey, one-litre calibrated cups were distributed to every participating farmer in the monitoring study. Farmers were advised to measure and record cow information on individual basis. Plastic ear tags and names were used for animal identification. Literate farmers were encouraged to do the recording themselves while those who could neither read nor write were advised to identify one household member that could do the recording. A two-day training session on the record-keeping format was conducted before the commencement of the monitoring study. Recording sheets were provided in vernacular language. The initial duration of the monitoring study was 4 months. Local extension workers facilitated the activities of the study.

Data analysis

Data analysis included descriptive statistics and logistic regression. The logistic analysis was done to explain the relationship between the adoption of recording systems, which was the discrete dependent variable and the independent variables. The dependent variable in the empirical model is whether or not the farmer adopts performance recording. Independent variables included in the analysis were socioeconomic and animal-related ones and are described in Table 1.

Table 1.  Description of the value for each variable that were included in the analysis to determine factors affecting smallholder farmers’ willingness to record dairy performance

Variables

Description

Herdsize

Average 3.5 cows ranging from 2 to 6 cows

Recording task

Self-Recording

Spouse-Recording

Child- Recording

Measuring equipment

Weighing Scale

Calibrated. Containers

Recycled Bottles

Mating system

Artificial Insemination (AI)

Both AI and Natural Service

Natural Service

Milk marketing

Formal Market

Informal Market

Highest Education level

Illiterate

Lower primary schools (class 1 – 4)

Higher primary school (class 5 - 8

Secondary education

Gender of farmer

Female

Male

Genotype of cow

Malawi Zebu

Holstein – Friesian x Malawi Zebu cross

Malawi Zebu x Jersey crosses

Milk yield/day

Average 5.5 (SD =3.8) kg; ranging form 1.8 to 11.6kg

Age of farmers

Average 44.1 (SD = 14.3) years; ranging from 21 to 64 years

The effect of a change of an explanatory variable with respect to recording performance was predicted using marginal probability, which indicate marginal change in the adoption of a performance recording participation due to a one-unit change in the explanatory variable in accordance with Maddala (1998). The model applied to determine the effect of the different factors on the farmers' willingness to participate in milk recording was:

In the model:

HERD represents the total number of cattle owned by the farmer and were categorized into three groups of between one and two cows, between three and six, and more than six cows.
GENOT is the genotype of the cows, which could be crossbreed only, Malawi Zebu and Crossbreeds together, and Malawi Zebu.
BRM is the breeding method, which was artificial insemination (AI), natural service, or both AI and natural service.
EDUC is the education level of the farmer while
MMC is the milk-marketing channel. In Malawi, milk is being sold either through the formal or the informal market.
MYIELD represents the continuous variable of average quantity of milk per day while EQUIP is the milk recording equipment used. Milk in smallholder dairy farms has been known to be measured by a variation of equipment ranging from simple tools such as recycled bottles, to more conversional equipment like graduated cups or churns and weighting scales.
TASK represents who in the household did the recording.

When household size, age of farmer and gender were included in the model, model parameters were non-estimable. This was because these variables were confounded with the farmer. As a result household size, age of farmer and gender were not included in the final model. The adequacy of the logistic model to explain participation was evaluated using a Log-likelihood function with a Chi-square statistic (). All analyses were performed using SPSS 8.0.

In the monitoring study, farmer participation was used to assess the system that was of practical use and would be easily adopted by smallholder farmers. Following Rogers (1995), we define adoption as a decision to make full use of an innovation as the best course of action once the individual has known and assessed the attributes of the innovation. Participation was defined as an attempt to keep records at household level after being aware of the presence of the recording intervention in their extension environment. Drop out rate was defined as the proportion of farmers that failed to complete the recording exercise during the course of implementation for other reasons other than drying their cows. Farmers who had kept records up to the time their cows had dried off were considered to have participated fully. The drop out rate was calculated as percentage of the number of farmers from the start minus number of farmers at the end of the monitoring phase. Chi-square analysis was used to test the goodness-of-fit of the three recording intervals that were tested.


Results

Characteristics of the farmers

Of the interviewees in the study 79.1% were male and 20.9% were female farmers. On previous participation in performance recording, 58.1% and 17.4% of the male and female farmers, respectively, have participated in recording at one point or another (Table 2). The highest percentage of farmers who had participated in recording (39.4%) was within the category of those who have undergone senior primary education while 25.6% attained basic primary education.

Table 2.   Characteristics of the smallholder dairy farmers in Lilongwe milkshed area, Malawi

Status

Age range

Education level

Sex

21-40

41-60

>60

Illiterate

Literate

Male

Female

n

%

n

%

n

%

n

%

n

%

n

%

n

%

Non-users

5

5.8

11

12.8

5

5.8

6

7.0

16

18.6

18

20.9

3

3.5

Users

18

20.9

30

34.9

17

19.8

7

8.1

57

66.3

50

58.1

15

17.4

Number of farmers (n) = 86

Reasons for not participating in dairy performance recording

Among the farmers interviewed in this study, at least 75.6 % had participated in recording exercises at one point or the other. From the interviews, farmers indicated various main reasons why they do not participate in dairy recording (Table 3).

Table 3.  Main reasons for non-participation in previous milk recording exercises by smallholder dairy farmers

Reasons

n

%

Busy with other activities

17

18.3

Lack of knowledge/Ignorance

15

16.1

Education qualification

12

12.9

Done by non-family members (extension worker / milk buyer)

11

11.8

Low milk yield

10

10.8

No recording materials

9

9.7

Lack of business orientation

8

8.6

Small herd size

7

7.5

No clear objectives/ lack of interest by other stakeholders

4

4.3

Total

93

100.0

From the interviews, the main reason for farmers not participating in recording activities was that farmers were busy with other activities. This is basically a reflection of the production system, which is a crop-livestock mixed system. Results from the logistic regression analysis are presented in Table 4. Factors with statistically significant influence on the decision to participate in the recording exercises were; recording task (P<0.01), recording equipment (P<0.01) and herd size (P<0.05).

Table 4.  Factors affecting participation in milk recording by smallholder dairy farmers

Variables

Regression coefficient

Standard error

T-statistic

Marginal probability

Constant

-4.03

1.33

-3.04***

_

Herdsize

 

 

 

 

One or two cows

0.12

0.37

0.33

0.12

Between three and six cows

0.52

0.35

1.46*

0.44

More than six cows

-0.12

0.37

-0.33

-0.11

Recording task

 

 

 

 

Self-Recording

1.53

0.64

2.38**

0.78

Spouse-Recording

0.84

0.87

0.96

0.72

Child- Recording

-0.95

0.38

-2.53**

-0.82

Measuring equipment

 

 

 

 

Weighing Scale

-0.18

0.56

-0.32

-0.15

Calibrated. Containers

0.47

0.24

1.91*

0.19

Recycled Bottles

0.30

0.25

1.17*

0.16

Mating system

 

 

 

 

Artificial Insemination (AI)

0.51

0.23

2.23**

0.21

Both AI and Natural Service

0.11

0.25

0.42

0.03

Natural Service

0.003

0.08

0.03

0.001

Milk marketing

 

 

 

 

Informal Market

-0.12

0.23

-0.53

-0.18

Pearson Goodness-of-Fit Chi Square = 84.62               DF = 69               P = 0.02;
*** P<0.01; ** P<0.05 and * P<0.10.

Type of breeding service provided and type of milk market vaguely influenced the adoption rate (P<0.10). Cow genotype, education level of farmer, and milk quantities produced per day were tested in the model but indicated no significant effect on performance recording. The effects of each of the factors on farmer participation in dairy recording are presented in the sections that follow.

The effect of herd size

There was a positive and significant relationship between herd size of 3 to 6 cattle and recording participation (P<0.05). For herd size of between one and two, the probability of participation in recording increased by 11%. Herd size of between three and six cows had the highest increase in the probability for participation of 44% (P<0.005) but then the probability of recording participation dropped with increase in herd size to more than six cows per farmer by 11%. The decreased participation in herd size of between one and two cows can be explained by the fact that, most farmers are not likely to record as they claim to keep the transactions in memory if the herd size is small. As such the farmers might easily recall, where as a large herd might require putting down notes.

The effect of member assigned to do the recording

Results indicated that the probability of performance recording participation increased by about 78% when the farmers did the recording themselves as compared to when their spouses did the recording. When spouses did the recording the probability for participation decreased by 72%. However, if children were assigned to do the recording the probability is reduced further by about 82%. This implies that as farmers shift the responsibility of recording to spouse, children or other household members, the level of recording declines. Expected level in milk recording participation was highest with the farmer doing the recording reflecting the seriousness of the farmer over the recording responsibility.

Effect of measuring equipment used

A positive and significant level of farmer participation was found when farmers used calibrated containers, P<0.05. Probability for participation in recording increased by about 19% when farmers used calibrated containers compared to when they used other containers. The probability of recording increased by about 16% when farmers use recycled bottles. However, the probability of participation reduced by 15% when farmers used a weighing scale.

The effect of mating system

With regard to mating system, results show a positive and significant relationship between the marginal changes in using artificial insemination (AI) and recording participation. By using AI, the probability for participation in recording increased by 21%. The probability of participation however only increased by about 3% when farmers used both AI and natural service and increased by only about 0.1% when farmers used natural service alone. The probability of participation in performance recording when using natural service is very minimal.

The effect of marketing channel

The results indicated that farmer participation in recording increased when farmers sold milk through the formal market. A shift from the formal market to a combination of formal and informal market resulted in a decline in the probability of participation in recording by about 18%. The majority of farmers selling milk on the informal market still maintained some ties with the formal market by supplying some milk to the formal market for membership in the milk-bulking group.

The effect of recording intensity or interval

Results on the level of farmer participation at different recording intervals are presented in Table 5. The rate of participation for weekly recording interval was 60%. This was numerically higher than those for daily and monthly recording intervals, which were 50 and 46.7%, respectively.

Table 5.  Participation levels in milk recording at different recording intervals by smallholder dairy farmers

Farmer Status

Daily Recording

Weekly Recording

Monthly Recording

n

% Farmers

n

% Farmers

n

% Farmers

At start

30

100

30

100

30

100

No attempt

13

43.3

9

30.0

15

50.0

Participated

15

50.0

18

60.0

14

46.7

Drop out

2

6.7

3

10.0

1

3.3

n = number of farmers

Although not statistically significantly different, numerically the highest drop out rate was observed in the weekly interval, (10.0%) followed by the daily recording, (6.7%) with the monthly interval recording having the least drop out (3.3%). Further, the results showed that weekly recording interval had the lowest number of farmers that did not attempt to record (30% vs. 43.3 and 50% for daily and monthly recording intervals , respectively).


Discussion

In this study we examined the herd level factors associated with the willingness of a smallholder dairy recording scheme. Several important factors that influence the level of participation in milk recording were determined. Major of these factors were herd size, member of the household performing the recording task and milk measuring equipment. Of less but notable importance were mating system and milk marketing channel.

Very small on the one hand, and relatively large herd sizes on the other hand, were associated with low recording while medium herd size of between 3 and 6 cows were associated with high recording participation. This agrees with results in a case study of smallholder animal recording in Sri Lanka where Amarasekera (1998) indicated that smallholders having one or two cows very rarely keep individual production records. Also, Bachman (1998) indicated that most smallholder farmers having small herds of 1 to 2 cows have significant difficulties in recording adoption (Bachman 1998). The argument being that, in small herds, transactions are few and infrequent such that farmers are familiar with their herds and might account these to memory. However, memory recall is never accurate. With medium size herds, recording participation increased. This might be due to the fact that as number of animals increased, farmers realized increased milk yield and eventually recording participation also increased. However, recording for individual animals might be a lot of work such that farmers with a large herd (more than 6 animals) might see the exercise as cumbersome, thereby reducing recording participation. This agrees with Nicholson et al (1999) who indicated that in larger herds, good intentions to keep milk production records disappear over a period of time hence the more animals kept, the less likely the intuitive approach to individual animal recording will be accurate. In addition to small herd size, Nicholson et al (1999) indicated that where extensive management system is practised no records are kept. The only record that the farmers keep is the receipt issued by the milk-collecting centre for the supply of milk.

The other factor that was found to be of importance is the member of the household assigned to do the recording. Since the majority of dairy farmers are male, the spouses who might be assigned the recording task are women. In a number of households, women find themselves centrally involved with day-to-day affairs of their homes. Apart from routine household work, there is also labour demand for off-farm activities, social obligations and domestic chores for women (Kapalamula 1993). The additional activity of milk performance recording might strain labour demand on the spouse, as it would be taken as a supplementary chore in addition to the other household activities. As for children, milk recording, which is mostly done at almost the same time that they are going to school, culminates into a conflict of interest. Howard and Cranfield (1995) reported that education, training, and experience have been shown to influence farmers' adoption and farm management behaviour among other factors. All these three characteristics are individual-based and in the current study, they seem to be in favour of the farmers who are the owners of the milking cows. To assess the future of animal recording, there is need to look at the lessons to draw from the past experiences. In Pakistan, low literacy rate of farmers, lack of awareness on the importance of records, lack of incentives for recording, and lack of breeders associations have been the major constraints in setting up a viable dairy buffalo recording system (Bachman 1998). Low literacy rate of the farmers is frequently mentioned as one of the major obstacles to record keeping in so far as the tool is on written basis.

Equipment used in measuring milk was also found to be a factor of importance. While the simplest way of keeping records, information on overall herd production cannot provide practical guide to pin-point individual cow problems, it was the simple equipment that tended to be favoured by farmers. Use of calibrated containers and recycled bottles in measuring milk showed to have an important role in smallholder dairy recording. Use of calibrated containers might be related to the level of complexity in the equipment. Recycled bottles are usually used when selling milk at household level, most farmers take it for granted to document for milk sold to other private markets and hence call for no extra knowledge acquisition in their use. This means that farmers were not presented with the challenging task of reading calibrations but rather counted the number of scoops made. While this method has limitations of giving less accurate results, the advantage to smallholder farmers is becoming familiar with the need for record keeping. In addition, provision of the right facilities, which are adequate to do the job, will stimulate the collection of comprehensive information in milk recording (Diwyanto et al 1998). This is because farmers with limited education background might have limitations in carrying out the recording using sophisticated equipment, thereby decreasing participation.

To a lesser extent, the mating system and the milk marketing channel also affected uptake to recording. The implication of these results are that the more the farmers use natural service on their cattle, the less likely they are to participate in performance recording. One reason why level of participation in recording for farmers using AI is the husbandry practices that are required with AI. With AI, farmers closely observe their cows in order not to miss any signs of heat and hence call the inseminator. There is also an inherent connection with the production system. The majority of farmers that use AI have their cows under intensive or semi intensive production system. As well, farmers using AI are given AI cards, which act as a source of encouragement to the farmer to monitor changes in performance better. In most cases, natural service is widely used for farmers keeping their animals under extensive system of production. In this situation, recording would be a difficult task since the animals are left to graze freely. Even in the low-input production systems like the smallholder dairy, use of AI would promote setting up of genetic improvement programmes. Use of pedigree recording becomes important when technologies like AI are used to permit widespread use of selected parents across many herds and several production environments (Van Vleck and Henderson 1961). The information provides life-cycle productivity, which explicitly recognizes the long-term importance of reproductive, health related, and maternal productivity traits to overall economic value. It is necessary to know pedigrees if attempts to avoid close inbreeding are made. Although milk prices in the formal market, which are represented by dairy processors, are low, they are stable prices and the farmers are assured of a less risky source of income. This encourages the farmers to keep records albeit an accumulated milk volume delivered to the milk-bulking centre in any particular month. Although this form of records would not suffice for individual cow monitoring, they offer a possibility for herd monitoring in relation with events occurring in the management, climate and health conditions. These results agree with Diwyanto et al (1998), who noted that efforts to create and maintain a conducive environment for record keeping should include provision of guaranteed milk markets and attractive and stable milk prices. Record keeping can be improved by applying an attractive guaranteed price for the milk produced in different markets (Diwyanto et al 1998). The milk price differences do not only have an effect on performance recording but also on economic values derived from profit analysis as reported by Chagunda and Wollny (2002). In their study, Chagunda and Wollny (2002) reported that the real market price differentials had an influence on the economic value and consequently the genetic progress that could be obtained from any breeding program that could be instituted in that production system.

Peters and Thorpe (1988) reviewed the animal recording systems that existed in Sub Saharan Africa during the years 70s and 80s but regret that the recording systems only gave priority to breeding purposes and underestimated the interest of this tool as support for improving management, production efficiency, and health status of in-field animal production systems. Animal recording in low-to-medium input animal production systems is vital multipurpose tool and a platform for development, which can be achieved into various forms related to various objectives in respect to the specific local situations of animal production systems (Flamant 1998). However, the challenge is what records should be used to achieve what objective. For example, recording once a month is accurate enough for individual selection and once every two weeks for progeny testing and this form of recording is mainly used as a tool for breeding in Europe and North America (McDaniel 1969). However, it is questionable if monthly records are optimal for nutrition, general management and herd level performance decisions (Svennersten-Sjaunja et al 1997). Feeding recommendations based on monthly recording can be misleading as the animals might either be underfed or overfed as opposed to daily recording, which gives a true picture of the animal's performance.

Although the current study did not investigate the actual improvement in farmers' income due to performing animal recording, the association farmers, Government extension workers, and researchers with each of them having a specific function in the process, provided a social network that would provide an appropriate starting point for establishing relations between institutions and farmers. As pointed out by ILCA (1993), inadequate organizational structure, inadequate resources, weak linkages between extension staff and farmers, limited data processing and feedback mechanisms have been cited as some of the constraints to milk recording in a lot of countries. With the exception of Zimbabwe, only a minor proportion of the national dairy cattle population within the southern African region is covered by organized milk recording schemes (ILCA 1993). Further studies to critically assess and quantify how dairy performance recording actually improves that farmers' income are needed.


Conclusion
s


Acknowledgements

The smallholder farmers of Lilongwe milk shed area are acknowledged for their willingness to participate in the study. The financial and general support given by GTZ/SACCAR and the ENRECA projects based at Bunda College of Agriculture is gratefully acknowledged.


References

Amarasekera S K R 1998 Animal recording in smallholder farming systems - The Sri Lankan experiences. Ministry of Livestock Development and Infrastructure. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 69-77. http://dad.fao.org/en/refer/library/reports/anand.pdf

Bachman F 1998 Animal recording for livestock development experiences of the Swiss Agency for Development and Co-Operation and of Interco-Operation in India. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 59-68. http://dad.fao.org/en/refer/library/reports/anand.pdf

Banda J W 1996 A general survey of the dairy and meat sector in Malawi. Internal Report. Bunda College of Agriculture, Lilongwe, Malawi.

Chagunda M G G and Wollny C B A 2002 Consequences of differences in pricing on economic values for milk yield of dairy cattle in peri-urban farms in Malawi. 7th World Congress on Genetics Applied to Livestock Production. Montpellier. Communication No 25-03. August 19-23, 2002.

Chagunda M G, Wollny C, Bruns E and Kamwanja LA 1998 Evaluation of the artificial insemination program for small-scale dairy farms in Malawi. Archiv für Tierzucht (Archives Animal Breeding) 41 (1/2), 45 -51.

Diwyanto K, Angraeni A and Djajanegara A 1998 Practical experiences in milk recording in Indonesia. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 89-102. http://dad.fao.org/en/refer/library/reports/anand.pdf

FAO 2005 http://faostat.fao.org (Accessed, July 2005).

Flamant J C 1998 The impact of socio-economic aspects on the development and outcome of animal recording systems. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 267-317. http://dad.fao.org/en/refer/library/reports/anand.pdf

Galal S 1998 Guidelines for development of performance recording in low-to-medium input production systems - A perspective. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 425-426. http://dad.fao.org/en/refer/library/reports/anand.pdf

Howard W H and Cranfield J 1995 Ontario beef producers' attitudes about artificial insemination. Canadian Journal of Agricultural Economics 43 (2), 305-314.

ILCA 1993 Milk recording scheme for smallholder producers in East And Southern Africa. Draft Review by Participating NARS. Bamako. Mali. March 1993.

Jelan Z A and Dahan M M 1998 Monitoring Livestock Productivity in Malaysia. In: Trivedi K R (editor), International Workshop on Animal Recording for Smallholders in Developing Countries. ICAR Tech. Series No. 1, pp. 79-88. http://dad.fao.org/en/refer/library/reports/anand.pdf

Kapalamula M 1993 Comparative study of household economics of integrated agriculture-aquaculture farming systems in Zomba district. M.Sc. Thesis, University of Malawi, Zomba, Malawi.

Maddala G S 1998 Limited Dependent and Qualitative Variables in Econometrics. Cambridge University Press, Cambridge.

Malawi Government 1997 Annual livestock census. Department of Animal Health and Industry, Ministry of Agriculture and Livestock Development. Lilongwe, Malawi.

McDaniel B T 1969 Accuracy of sampling procedures for estimating lactation yield: A Review. Journal of Dairy Science 52, 1742.

Nicholson C F, Thornton P K, Mohammed L, Muinga R W, Mwamachi D M, Elbasha E H, Staal S J and Thorpe W 1999 Smallholder dairy technology in coastal Kenya. An Adoption and Impact Study. International Livestock Research Institute (ILRI), Nairobi, Kenya. 68pp.

Peters K J and Thorpe W 1988 Current status and trends in on-farm performance testing of cattle and sheep in Africa. 3rd World Congress on Sheep and Beef Cattle Breeding, Paris, Volume 1, 275-301.

Rogers E M 1995 Diffusion of Innovations. The Free Press. New York.

Svennersten-Sjaunja K, Sjaunja L O, Bertilsson J and Wiktorsson H 1997 Use of regular milk records versus daily records for nutrition and other kinds of management. Livestock Production Science 48, 167-174.

Van Vleck L D and Henderson C R 1961 Utilising both part and complete daughter records in sire evaluation. Journal of Dairy Science 44, 2068-2076.

Zimba A W C 1993 A review of milk recording in Malawi. A report Submitted to ILCAs Cattle Milk and Meat Thrust. Chitedze Research Station, Lilongwe, Malawi.


Received 23 September 2005; Accepted 15 February 2006; Published 11 May 2006

Go to top