EFFECT OF SELECTED INSECTICIDE ON WHITEFLY (Bemisia tabaci) INFESTING BRINJAL PLANTS 128 Comparison of Technical Efficiency and Socio-economic Status in Animal-crop Mixed Farming Systems in Dry Lowland Sri Lanka Upul Yasantha Nanayakkara Vithanage Postgraduate Institute of Agriculture, University of Peradeniya, Peradeniya, Sri Lanka Lokugam H. P. Gunaratne Professor; Department of Agricultural Economics and Business Management, Faculty of Agriculture, University of Peradeniya, Sri Lanka Kumara Mahipala M. B. P. Senior Lecturer; Department of Animal Science, Faculty of Agriculture, University of Peradeniya, Sri Lanka Hewa Waduge Cyril Professor; Department of Animal Science, Faculty of Agriculture, University of Peradeniya, Sri Lanka Abstract Pre-tested, structured questionnaires covered management aspects, inputs, outputs, socio- economic situations and constraints in dairy farming among Semi-intensive (SIFS) and Extensive farming systems (EFS) in dry-lowland Sri Lanka. Parametric data were analyzed using two- tailed‘t’ and ‘Z’ tests, and non-parametric values were analyzed using Chi-square and Fisher’s extract tests. Cobb-Douglas model was used to calculate meta-frontier and system-specific frontiers. Returns in SIFS are lower than EFS. Labor costs are 91.72% and 87.26% in EFS and SIFS respectively. Counting family labor, SIFS has no comparative surplus. Excluding this, dairying is profitable even in SIFS. Dairying provides EFS family insurance where selling animals increases income. Discouragement of this in SIFS impacts negatively on sustainable income. Integration is comparatively minimal in EFS. Established with the best practices and technologies available, SIFS requires external resources to enhance efficiencies. If all EFS farmers achieved best farmer TE, output could increase by 45.09%. Similarly, SIFS output could increase by 57.08%. Farmer education and training programs contribute to improved production efficiency. Grassland scarcity and low productivity affect output adversely; poor veterinary and extension services are major constraints. Farmers consider dairying as profitable, which secures its future. Contrastingly, 35.19% of farmers believe it is low status, preferring professional jobs despite lower comparative incomes. Keywords: Buffalos, cattle, farming systems, technical efficiency Introduction 1 Agriculture and livestock in Sri Lanka account for 11.2% and 0.84% of total GDP Corresponding author’s Name: Upul Yasantha Nanayakkara Vithanage Email address: nvuyasantha@yahoo.co.uk respectively (CBSL, 2011). National milk production in 2011 was estimated as 256.78 million litres with approximately 201.17 million litres supplied from neat cattle. It is estimated that total milk production grew by 2.3% from 250.99 million litres in 2010 (CBSL, 2012). In 2011, the value of the total milk and milk products importation was 38,182 LKR millions, a 30.69% Asian Journal of Agriculture and Rural Development journal homepage: http://aessweb.com/journal-detail.php?id=5005 mailto:nvuyasantha@yahoo.co.uk Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 129 increment since 2010 (CBSL, 2011). In 2007, the total milk consumption of Sri Lanka was around 612,000 MT (32 kg per capita consumption). Compared to the per capita consumption of developed countries, Sri Lankan per capita milk consumption is among the lowest in the world (Ranawana, 2008). Obviously, the dependency on milk powder importation and low liquid milk consumption due to low supply will probably continue. In Sri Lanka there are four cattle based farming systems namely up and mid country, coconut triangle, wet lowland and dry lowland (Ibrahim et al., 1999). Dairy (cattle and buffalo) in the dry lowland zones is an important capital asset for the peasant farmers. Moreover, dairy farming is predominantly a smallholder mixed crop- livestock farming operation (Ibrahim et al., 1999; Bandara, 2000; Premarathne and Premalal, 2005). In 2011, the dairy population of dry lowland was 1,009,390 (63.21% of the total population) and the total average daily milk production of dry lowland was 320, 850 L (45.6% of total production) (CBSL, 2012). Total land extent of the dry lowland is around 37, 600 km 2 and limited land for fodder and forage production is a critical issue in dairy production (Ranaweera, 2007). Therefore, national milk self sufficiency could be achieved only by enhancing the production in dry lowland. The development of the dairy industry in the country has stagnated mainly for political, technical and socio-economic reasons (Perera and Jayasuriya, 2008). Further, a poor understanding of socio- economic and cultural factors was the main reason for unsatisfactory outcomes of the dairy development programs in the country (Zemmelink et al., 1999). Against this background, the objective of this study is to identify the farm-level socio-economic and technical efficiency in the above systems that would provide further support to increase national dairy production, which would thereby ensure long-term sustainability. This finding may have regional application for countries with similar tropical farming systems. The paper is organized as follows. The following section presents the methodology adopted which will be followed by results, and discussion sections. The conclusions and implications are presented in the last section. Methodology Data collection The study was conducted in dry-lowland dairy farming systems in Polonnaruwa and Baticaloa districts of Sri Lanka. This area accounts 23.41% of dairy population and 30.02% of milk production out of total in dry-lowland in 2011 (CBSL, 2012). Prior to the survey, fifteen in-depth discussions were held with selected small and large scale farmers, officers of major large-scale dairy farming companies (i.e. CIC Holdings), officers of formal milk collecting networks (i.e. Milk Industries of Lanka Company Ltd., Nestle Lanka Ltd.), small-scale milk collectors, veterinary surgeons, livestock development instructors and artificial insemination technicians. Approximate numbers of dairy farmers were found from the different 10 administrative divisions. Then, stratified random sample was drawn, probability proportion to size thus end up with a sample of 96 farms. This information supplemented the conceptual framework of the survey instrument. The structured questionnaire was pre-tested twice. The questionnaire covered dairy management aspects, inputs used, outputs, socio-economic situations, constraints and involvement of households in dairy farming. Analytical framework The stakeholder discussions as well as field observations reveled that there are two distinct dairy farming systems existed in the study area that could be named as Extensive farming system (EFS) and Semi-intensive Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 130 farming system (SIFS). The socio-economic and production details were analyzed for two systems separately and compared if needed. Dairy production economics - A cost benefit analysis was carried out considering the expenditure and revenue on dairy farm activities for a period of one month. The cost of labor was calculated based on the duration (hours/day) spent on dairy farming activities and the daily labor wage following Mahipala et al., 2006. All data were analysed using SAS (1998) statistical package. Parametric data were analysed using the two-tailed ‘t’ test and ‘Z’ test. Non parametric values were analysed using Chi-square and Fisher’s extract tests. Technical efficiency - TE shows the farmer’s ability to produce the maximum level of output by using an existing set of inputs (i.e. output oriented) or produce the same level of output using fewer inputs (i.e. input oriented) (Lawson et al., (2004); Pierani and Rizzi (2003). Furthermore, TE is the ability of a decision-making unit (e.g. farm) to produce maximum output by using a given set of inputs and technology (Thiam et al., 2001). The analysis based on the concept of TE in terms of deviation from best-practice frontier as described by Farrell (1957). Further, TE is defined as the capability of the farm to reach its maximum possible production level by using a given set of inputs and available technology. According to Farrell (1957) there are two ways to quantify technical efficiency either by estimating econometric frontiers such as production or cost frontier or by estimating a free disposal convex hull of the observed input-output ratios using mathematical programming techniques. Since most of the agriculture production systems analysis used the stochastic frontier analysis (Bauer, 1990), this study deals with the parametric approach. In the parametric approach, the stochastic frontier production function approach independently proposed by Aigner et al. (1977) and Meeusen and Van den Broeck (1977) is used where the error term has two components that are systemic random error and one sided technical inefficiency terms. The stochastic frontier model is described as below, Yi = f (Xi; β) exp (Vi - Ui) Where i = 1, …, n observations in the sample; Yi is the observed output level of i- th farm; f (Xi; β) is a production functional form with respect to input vector, Xi, and unknown parameter vector β; Ui is a non- negative random variable representing technical efficiency; and Vi is the random error term which is assumed to be distributed independently and identically as N (0, σv 2 ) and independent of Ui. The frontier output (Yi * ) is represented by f (Xi; β) exp (Vi) (Gunerathne and Leung, 1997). Maximum likelihood estimate of the parameters are measured in order to parameterization, σv 2 + σ 2 = σs 2 and γ = σ 2 / (σv 2 + σ 2 ) where σ and σv are indicated standard deviation of Ui and Vi respectively. Also, the ratio of the observed output to the corresponding frontier output refers the TE of i-th farm. The Cobb-Douglas model is used here due to its simplicity, absence of multicolinearity and homogeneity. Therefore, TEi = Yi/Yi * exp (-Ui) The milk production is affected by several variables (Singh et al., 2006). The number of cows and all other costs including labor were considered as independent variables, while milk production represented the dependent variable. The Cobb-Douglas stochastic frontier in this study is specified as; Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 131 Where Yi is the value of milk production for the i th farm in the sample [in Sri Lankan rupees (LKR)/farm]; X1 is the total number of cows of the herd; X2 is the total cost of the farm (LKR/farm/month); and β1, β2 and β3 are output elasticity of total cows of the herd, total cost of the farm and other factors respectively. Technical efficiency was calculated by two methods using Frontier 4.1. Meta-frontier and system specific frontiers were estimated for each farming system. When TE is measured among different groups of famers, TE should be measured with respect to a common frontier (meta-frontier) instead of depending on separate frontiers for each group (system specific frontier) (Victor et al., 2010). A major assumption in meta- production function is that all considered groups have similar access to technology (Gunaratne and Leung, 1997). System specific frontier was calculated relatively to all farms belonging to a particular farming system (either EFS or SIFS) whereas meta- frontier was calculated considering all farms as a pool. The outputs obtained from the two models were compared. Non- compatibility of the data across the system is one of the major limitations when using meta-production function. The surveyed data was gathered by using same questionnaire over the same period throughout different systems. Therefore, it greatly eliminates the non-compatibility among the data (Gunaratna and Leung, 1997). In the second level, six other factors (size of the farm family, farmer’s ethnicity, age, education, knowledge of dairy farming and farming systems) were analysed to determine their effect on technical efficiency of dairy farming. Constraints and other social issues - The three main production constraints as well as service constraints were identified using open-ended questions. Other than that, social issues related to dairy farming i.e. reasons for being engaged in dairy farming, recognition as a dairy farmer, children’s willingness of continuation and the future of dairy farming were also identified using open-ended questions. Results Dairy production economics The direct costs and revenues of dairy farming were considered in calculating the profit of the dairy enterprises. Costs account for expenditure on feeding, breeding, health care, labor and so on. Revenues were generated from selling milk and live animals for meat purposes. Expenditure - The cost of concentrate and mineral feeding was significantly higher (p<0.05) in SIFS than EFS (Table 1). Furthermore, the cost of family labor and hired labor was significantly greater (p<0.05) in EFS than SIFS. The cost of drugs and total cost including family labor for dairy farming was found to be significantly higher (p<0.05) in EFS than SIFS. Table 1: Monthly expenditure on dairy farming Expenditure LKR/farm/month P values EFS 1 SIFS 1 Dairy animal purchasing 416.67 (2165.08) 622.58 (1404.07) 0.5838 Concentrate and mineral supplement 31.67 (161.55) 1081.40 (1892.00) 0.0206 Drugs 209.25 (340.45) 47.84 (103.83) 0.0006 Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 132 1Mean, Values in the parentheses are standard deviations Means significantly different between systems (p<0.05) Revenue - Dairy farming was the major income generating enterprise in EFS and it accounted for 63.91% of total farm income. The revenue from milk and meat (selling culled animals) was significantly (p<0.05) higher in EFS than SIFS (Table 2). There was a significant difference (p<0.05) in total dairy farm income between the two farming systems, and it was greater in EFS than SIFS. Moreover, the profit of the dairy farm including or excluding family labor was significantly greater (p<0.05) in EFS than SIFS. Table 2: Monthly revenue from dairy farming 1Mean, Values in the parentheses are standard deviations Means significantly different between systems (p<0.05) Veterinary service 19.60 (81.11) 17.33 (43.53) 0.8597 Artificial insemination 0.00 (0.00) 23.59 (62.13) 0.0523 Natural breeding (stud bull) 0.00 (0.00) 0.60 (4.98) 0.5344 Family labor 20995.74 (8534.82) 12375.00 (5103.97) 0.0001 Hired labor 1203.7 (3442.55) 0.00 (0.00) 0.0043 Equipment 0.00 (0.00) 10.43 (54.91) 0.3280 Other 15.43 (80.18) 2.66 (15.53) 0.2061 Total cost excluding family labor 1896.33 (4026.13) 1806.44 (2797.59) 0.9013 Total cost including family labor 19003.97 (11301.48) 13662.05 (6835.27) 0.0057 Revenue LKR/farm/month P values EFS 1 SIFS 1 Milk 25888.89 (21967.52) 3913.04 (4132.46) 0.0001 Meat (sell cattle/buffalo) 23725.31 (23066.40) 1077.29 (2361.07) 0.0001 Total dairy farm income 49614.20 (43222.12) 4990.34 (5269.97) 0.0001 Profit excluding family labor cost 47717.86 (42727.55) 3183.90 (4701.47) 0.0001 Profit including family labor cost 30610.23 (40081.09) -8598.32 (5712.65) 0.0001 Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 133 Technical efficiency The technical efficiencies in EFS varied from 0.05 to 0.8847 with respect to system specific frontier whereas it ranged from 0.0211 to 0.9316 in SIFS. The system specific values are significantly different between two farming systems. The mean TE in EFS (system specific frontier) was greater than the mean TE in all farms (meta- frontier) (Table 3). The mean TE in system specific frontier was lower than the mean TE in meta-frontier in SIFS. There was an association between the level of TE in dairy farming and variables such as farmers’ age, education, ethnicity and knowledge (Table 4). When considering Sinhala/Tamil farmers in both farming systems, the majority belongs to the TE level of 51-75%, whereas, majority of Muslims belongs to the TE levels of 25-50% in EFS and <25% in SIFS. In EFS >75% TE level was achieved by the farmers between 31-55 years only and that age group was more efficient than other farmers in the system. The majority of farmers who achieved TE level of >75% belong to the age category of >55years in SIFS. Meanwhile, TE increased with farmers’ age. The majority of the farmers in EFS who were educated up to primary level show the highest TE level compared to others. However, the TE did not increase along with education. In SIFS, the highest TE level was achieved by the greater part of farmers who have a basic education level. In addition, TE increased along with farmers’ education. Only the farmers having a ‘satisfactory’ level of knowledge indicated >75% TE level in EFS. In SIFS, the majority of the farmers who reached the >75% TE level have ‘moderate’ knowledge of dairy farming. Furthermore, in both systems the level of TE increased with farmers’ knowledge of dairy farming. Table 3: Mean technical efficiencies in EFS and SIFS based on system specific models and meta-production frontier models Mean TE System specific Meta-frontier Extensive farming system 1 0.5491 (0.2045) 0.4129 (0.1907) semi-intensive farming system 1 0.4292 (0.2638) 0.4719 (0.2425) P values 0.0453 0.2804 1Mean, Values in the parentheses are standard deviations Means are significantly different between systems at p<0.05 level Table 4: Level of TE in dairy farming Variables Frequency (%) TE in Extensive farming system TE in Semi-intensive farming system <25% 25-50% 51-75% <75% <25% 25-50% 51-75% <75% Ethnicity* Sinhala/Tamil 30.00 20.00 40.00 10.00 16.33 26.53 40.82 16.33 Muslims 12.50 68.75 18.75 0.00 80.00 20.00 0.00 0.00 Age* 15-30 years 22.22 55.56 22.22 0.00 100.00 0.00 0.00 0.00 31-55 years 7.14 57.14 28.57 7.14 19.51 29.27 36.59 14.63 > 55 years 66.67 0.00 33.33 0.00 18.18 18.18 45.45 18.18 Education* No formal education 0.00 100.00 0.00 0.00 100.00 0.00 0.00 0.00 Basic education (<5 grade) 33.33 66.67 0.00 0.00 25.00 50.00 0.00 25.00 Primary education (5-9 grade) 7.69 61.54 30.77 0.00 29.03 16.13 38.71 16.13 Ordinary level (10-11 grade) 0.00 100.00 0.00 0.00 5.56 38.89 44.44 11.11 Knowledge* Very poor 0.00 50.00 50.00 0.00 100.00 0.00 0.00 0.00 Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 134 Poor 25.00 62.50 12.50 0.00 25.93 25.93 33.33 14.81 Moderate 20.00 20.00 60.00 0.00 16.67 38.89 27.78 16.67 Satisfactory 0.00 0.00 0.00 100.00 12.50 0.00 75.00 12.50 *There is an association among TE and variables (p<0.05) Constraints and other social issues In EFS, the major production constraint among the majority of the farmers was lack of grasslands (Table 5). Low productivity of existing animals and water scarcity were identified second as production constraints. Unavailability of well-structured sheds was the third production constraint in EFS. Lack of grasslands was the major production constraint identified among the majority of the farmers in SIFS as well. Low productivity of existing animals and unavailability of well-structured sheds were the second and third production constraints in SIFS respectively. Table 5: Constraints among different farming systems *There is an association among farming systems and variables Poor veterinary service, poor extension service and lack of training were identified as the first, second and third service constraints respectively of the majority of the farmers in EFS. In SIFS, poor veterinary service and poor extension service were identified as the first and second service constraints respectively. Poor veterinary and extension services were identified as the third service constraint among majority of the farmers in SIFS. In both farming systems, a majority of the farmers engaged in dairy farming with the purpose of increasing their income (Table 6). A negligible number of famers rear dairy animals, mainly for the purpose of collecting cow dung. The status of the dairy farmer was well recognized in both farming systems; however a considerable number of farmers stated that there was no satisfactory recognition (‘bad’) for dairy farmers in the society. Issues Frequency (%) First Production Constraint* Second Production Constraint* Third Production Constraint* EFS SIFS EFS SIFS EFS SIFS Lack of grasslands 81.48 68.12 3.70 17.39 7.41 13.04 Low productivity of existing animals 7.41 10.14 33.33 33.33 25.93 21.74 Unavailability of well-structured sheds 7.41 15.94 11.11 15.94 37.04 26.09 Water scarcity 3.70 5.80 33.33 23.19 3.70 15.94 Lack of stud bulls 0.00 0.00 3.70 4.35 0.00 5.80 Scarcity of high yielding animals 0.00 0.00 0.00 4.35 3.70 13.04 Rules and regulations of the Mahaweli authority 0.00 0.00 0.00 1.45 0.00 4.35 Diseases 0.00 0.00 14.81 0.00 22.22 0.00 Issues First Service Constraint* Second Service Constraint* Third Service Constraint* EFS SIFS EFS SIFS EFS SIFS Poor extension 18.52 27.54 62.96 36.23 3.70 23.19 Poor veterinary service 48.15 33.33 22.22 27.54 25.93 23.19 Lack of subsidies 0.00 2.90 3.70 4.35 3.70 7.25 Lack of market facilities 18.52 24.64 3.70 13.04 7.41 20.29 Lack of trainings 3.70 8.70 7.41 17.39 40.74 20.29 Poor security 11.11 2.90 0.00 1.45 18.52 5.80 Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 135 Table 6: Social issues among different farming systems Frequency (%) Frequency (%) Reasons for doing dairy farming* EFS SIFS Recognition for dairy farming* EFS SIFS Increased income 88.89 78.26 Good 51.85 57.97 Home consumption 3.70 7.25 Bad 37.04 33.33 Preference 3.70 10.14 Neutral 11.11 7.25 Tradition 3.70 2.90 No idea 0.00 1.45 In order to get cow dung 0.00 1.45 Frequency (%) Frequency (%) Children’s willingness of continuation* EFS SIFS Future of cattle farming EFS SIFS Yes 85.19 81.16 Good 92.59 94.20 No 14.81 15.94 Bad 3.70 2.90 No idea 0.00 2.90 No idea 3.70 2.90 *There is an association among farming systems and variables In both farming systems, the majority of children (next generation) were interested in carrying out dairy farming as a source of income in future. Also, majority of the farmers stated that ‘the future of dairy farming is secured’. Discussion Dairy production economics – The majority of the farmers in dry zone (50% in the EFS and 51.18% in the SIFS) are educated up to primary level (grades 5-9). And, they mainly depend on dairy farming or crop cultivation, especially paddy (88.37% in the EFS and 86.72% SIFS) which is passed on over generations. Moreover, a lack of alternative sources of income or employment opportunities in the dry zone restricts many people to the only available option – livestock keeping (Ratnayaka et al., 1992). In these mixed farming systems, dairy acts as insurance. De Silva and Sandika (2012) have reported this especially for poor farmers if there is a loss by crop cultivation, especially paddy, the farmers try to compensate for it by selling animals such as bulls, male calves and old cows. Further, diversification in agriculture can act as a mitigating factor against crop failure. Dairying is emerging as a better alternative in diversification especially under mixed farming systems (Bharadwaj et al., 2006). Generally, farmers in the dry zone do not invest more in purchasing dairy animals. For instance, they try to breed animals themselves via natural breeding or artificial insemination (AI). One the interesting observation is that the farmers in the EFS generally do not spend money on breeding their animals. Because of the large herd size and the free grazing system, it is difficult to practise AI. On the other hand, it is not required to hire stud bulls as the herd already contains stud bulls. Farmers do not have knowledge on breeding to select stud bulls therefore; this situation leads to inbreeding and low production performances of the cows in the EFS. Reproductive and productive performances of dairy cows could be improved by adapting proper management practices, improved nutrition and new reproductive technologies (Kollalpitiya et al., 2012). Farmers in the SIFS usually practise both natural breeding and AI. However, their cost of natural breeding is closer to zero as they borrow bulls from others without any payment and this social exchange mechanism could minimize the trend of inbreeding. Feeding rice straw is common in the dry zone. Generally, concentrates are not fed (Ibrahim et al., (1999). Compared to the EFS, farmers in the SIFS usually spend more on concentrates and mineral supplements, around 7.63% of the total cost. There is a common trend in foot and mouth disease, mastitis and worm problems in the Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 136 EFS as the animals are enclosed densely in a paddock during night time. Therefore, the drug cost is significantly higher in the EFS. The labor cost is the major component of the dairy expenditure at 91.72 % and 87.26 % in the EFS and SIFS respectively. Moreover, 89% of the total expenditure on dairy farming was accounted for by labor in the dry zone Ratnayaka et al. (1992) and those figures concur with these research findings. Farmers require more labor (13.61 hr/day) on dairy farming in the EFS as farmers have to look after their animals during the whole grazing period. Some farmers require hired labor only during cultivating and harvesting periods of paddy, when they are unable to dedicate their time to dairy farming. Revenue is generated via on-farm or off- farm activities in both farming systems. On- farm revenue included the income generated by animal husbandry and crop cultivation. The income generated by selling milk is 6.6 times higher in the EFS than the SIFS as the total cow unit (CU) (Lactating/Milking/Dry cow = 1CU; Heifer = 0.75 CU; Calf = 0.50 CU; Bull = 1.5CU) is 11.86 times higher in the EFS than the SIFS. Income generated via selling animals for meat purposes was 22 times higher in the EFS than the SIFS for two reasons. One is the large herd size (51.33±8.46) in the EFS, and the other is the socio-religious background where the majority of the EFS population is Muslim. There is no religious barrier to dairy animal slaughtering in Islamic culture. Comparatively, in the SIFS, the herd size is smaller (4.64±0.43) and the majority is Buddhist, and animal slaughtering is not encouraged. The SIFS has not a surplus if family labor is taken into account compared to the EFS. If family labor is not considered, dairy farming is a reasonably profitable enterprise even in the SIFS. Parallel studies have also reported that Sri Lankan small scale dairy farmers earn a marginal profit by their dairy farming activities (Navaratne and Buchenrieder, 2003; DAPH, 2009). Technical efficiency - Even though the EFS showed the highest performances with respect to system specific frontier, it dropped by 0.136 when farms were compared with meta-frontier. The results revealed that the EFS farms have more capability to increase their production by using the best resources and technologies available in the area. Comparatively, SIFS where majority of the farmers are doing their farming efficiently, perform better than EFS. On the other hand, greater meta- frontier value in the SIFS implies the system has already adopted the best practices and technologies available in the area. In the light of that, the system required external resources and technologies to enhance their efficiencies. Paris (2002) reported that farm productivity and household incomes could be increased by introducing specific technologies. However, poor technology intervention could be observed among Sri Lankan smallholding dairy farmers because of a lack of understanding of the complexity of the systems (Bandara, 2000). Farmers who are in both farming systems carry out dairy farming with crop farming (upland and lowland crops). Therefore, new interventions should be introduced after considering animal and crop farming and their interactions. Further, in small-scale mixed farming systems in South Asia, low productivity of the animals could be observed because introduced technologies have not been adopted or their update has not been sustainable. Also, crop-livestock systems were considered as single production units and new technologies applied accordingly. Nevertheless, before applying new technologies, complexity and diversity of the farming systems should be considered (Thomas et al., 2002). The mean TE of farmers in the EFS was found to be 54.91% which indicates that the output could be increased by 45.09% if all farmers achieved the TE level of the best farmer. In the same way, the output of SIFS farmers could be increased by 57.08%. Some farms performed well, whereas others did not, even with the same system. This interesting observation could be found in both farming systems and this situation was created due to other factors such as farmers’ culture, Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 137 age, education and knowledge, which correlated with farm business management. Some farms owned by Muslims are usually managed by herdsmen. At the end of each year the owner earns a large sum of money by selling animals for slaughtering. The herdsmen look after the herds throughout the year and practise milking in order to fulfil their daily necessities, but they do not try to manage the business in an efficient way. Therefore, those farms are relatively less efficient than the others. On the other hand, Sinhalese and Tamils manage their farms efficiently by themselves, by using available resources and technologies. Dairy farming has been carried out by farmers over generations. For instance, a particular farmer engaged in rearing dairy animals since his childhood would have his farming experiences enhanced over his life time. Yeamkong et al. (2010) showed that farm production and revenue is increased with farmers’ experiences. Therefore, there was a positive relationship between TE and the age of the farmers in the SIFS. Moreover, positive relationships were found between education and knowledge of the farmers with TE in the SIFS. However, such an association could not be observed in the EFS where the majority consists of Muslims, because those farms were managed by the herdsmen. Farm could be changed by farmer education Ondersteijn et al. (2003) because there is a stronger influence of education on farm productivity in a modernizing environment (Phillips, 1994). As well, there is a positive relationship between farmer education and farm production (Yeamkong et al., 2010) and farm efficiency (Latruffe et al., 2005). Therefore, well planned farmers’ training programs (Wubeneh and Ehui, 2006) and policy interventions such as farmer education and agricultural extension (Fuwa et al., 2007) contribute to improve farm production efficiency. Integrated farming is a method in which two or more enterprises (crops, livestock, fish) are combined to get maximum efficiency from resources used, considering their relationships and interactions provided that an output from one enterprise would be an input for another (Ibrahim and Schiere, 2002). In contrast, there is no proper integration among different enterprises in either farming system. Moreover, paddy straw and crop residues (by-products of crop cultivation) are used as dairy animals’ feed in the SIFS during drought periods when fresh roughage becomes limited. Cow dung is used in upland crop cultivation, but not for paddy in the SIFS. Poultry manure is used neither in upland crop cultivation nor paddy farming in the SIFS. Only paddy straw is used as dairy feed and crop residues are not fed to dairy animals in the EFS. Neither cow dung nor poultry manure is used for upland crop or paddy cultivation in the EFS. Therefore, the integration among different enterprises is minimal in the EFS compared to the SIFS. Proper livestock- crop integration ensures increasing farm TEs in both farming systems. Furthermore, farm productivity could be improved by integrating crop and livestock enterprises. In both farming systems, farmers are rearing dairy and poultry. Even though there is a high potential in rearing goats, farmers did not take the opportunity. Therefore, farm efficiency could be increased by introducing goats into both farming systems. Further to this, it has been shown that mixed livestock farming is a very important method to increase farm efficiency (Gaspar et al., 2009). Constraints and other socially-related issues - Major production constraints were almost similar in both farming systems. Grasslands became limited over time due to the population pressure. Moreover, the grassland of Sri Lanka has low productivity and has deteriorated due to mismanagement under the existing socio-economic position in the country (Premaratne et al., 2003). Therefore, the lack of grasslands was the major constraint to expanding dairy farming today and in future. The best solution for the scarcity of grasslands was having a lower number of highly productive animals instead of having a larger number of unproductive animals. Other than that, an intensive or semi-intensive management Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 138 system was more viable than extensive management practices. Feed scarcity become a severe problem during the dry season and farmers should use feeding technologies such as urea treated straw feeding, urea molasses mineral block feeding etc. especially, during drought periods. Similar recommendations are suggested for dairy farms in Kurunagala district of Sri Lanka and the study revealed that proper feeding technology is required for the dry season. Further, farmers should be trained in feed management (Kubota et al., 2010). Extension officers and veterinary surgeons were reluctant to work in these rural areas due to poor infrastructure and other facilities. Therefore, the lack of extension and veterinary services has become a major barrier to expanding the dairy sector. Premarathna and Premalal (2005) reported that access to extension services in dry lowland is very difficult. Most of the studies revealed that efficient and effective extension services for the dairy farmers are a paramount requirement in many parts of the country (Bandara et al., 2011). A majority of the farmers consider dairy farming as a profitable venture, which secured its future as their children were willing to carry out dairying in future. In contrast, there was another group of farmers (35.19%) who believe it is a low status they occupation in their society. They prefer professional jobs even though they may generate a lower income compared with dairying. Conclusion Although dairy farming maintenance cost per animal was relatively high in SIFS, system returns were lower in comparison with EFS, which used low external inputs. Labor cost is considerably high in both systems. If family labor inputs are considered, SIFS has no surplus compared to EFS. However, if family labor is not considered, dairy farming is reasonably profitable. Dairying acts as insurance for the family if farmers are willing to sell culled animals. Selling animals for meat purposes was strongly discouraged due to cultural and religious ethics and barriers which have had a direct negative impact on monthly income as well as the financial security of the farm family. Appropriate livestock-crop integration and mixed livestock farming ensure increasing farm production and efficiency in both systems. Established with the best practices and technologies available, SIFS required external resources and technologies to enhance its efficiencies. In both systems, improved farm production efficiency was achieved through well- planned farmer education and training programs. Lack of grasslands, low productivity, water scarcity and unavailability of well-structured sheds affected production adversely, whereas poor veterinary services, poor extension services and lack of training were major service constraints. The future of dairy farming seemed secure because most farmers considered it to be profitable, and their children were willing to continue that way of life. However, policy implementers and social workers should work cooperatively to enhance the social status of dairy farmers. Acknowledgement The authors gratefully acknowledge the invaluable support given by Dr. H.L.J. Weerahewa, professor, Department of Agricultural Economics and Business Management, Faculty of Agriculture, University of Peradeniya; Audrey Cornish, English language advisor; dairy farmers, officers from Milk Industries of Lanka Company Ltd., Nestle Lanka Ltd., CIC holdings, veterinary surgeons, livestock development instructors and artificial insemination technicians who have given their invaluable support to make this study a success. References Aigner, D. J., Lovell, C. A. K., & Schmidt, P. J. (1977). Foundation and estimation of stochastic frontier production function models. Journal of Econometrics, 6, 21-37. Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 139 Bandara, B. H. W. M. U. S. (2000). The current status of smallholder dairy systems in Sri Lanka. Department of Animal Production and Health, Gatambe, Peradeniya, Sri Lanka. Bandara, D. M. D. S., Premaratne, S., & Dematawewa, C. M. B. (2011). Production and Economic Characteristics of Intensive and Semi-Intensive Dairy Cattle Management Systems in Vegetable Based Farming System in Welimada, Sri Lanka. Tropical Agricultural Research, 22(3), 314-323. Bauer, P. W. (1990). Recent developments in econometric estimation of frontiers. J. of Econometrics, 46, 39- 56. Bharadwaj, A., Dixit, V. B., & Sethi, R. K. (2006). Economics of buffalo milk production in Hisar district of Haryana state. Indian J. of dairy science, 59(5): 322-327. Central Bank of Sri Lanka (CBSL) (2011). Annual report. Central Bank of Sri Lanka, Colombo, Sri Lanka. Central Bank of Sri Lanka (CBSL) (2012). Economic and Social Statistics of Sri Lanka 2012. Central Bank of Sri Lanka, Colombo, Sri Lanka. Department of Animal Production and Health (DAPH) (2009). Estimation of cost of production of milk in different agro climatic zones of Sri Lanka. Livestock planning and economics division, Department of animal production and health, Gatambe, Peradeniya, Sri Lanka. De Silva, P. H. G. J., & Sandika, A. L. (2012). The Impact of Agricultural Credit and Farmer Trainings on Small Holder Dairy Production in Southern Region in Sri Lanka. Iranian Journal of Applied Animal Science, 2(3), 265-269. Farrell, M. J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society, Series A, 120(3), 253-281. Fuwa, N., Edmonds, C., & Banik P. (2007). Are small-scale rice farmers in eastern India really inefficient? Examining the effects of microtopography on technical efficiency estimates. Agricultural Economics, 36, 335-346. Gaspar, P., Mesias, F. J., Escribano, M. , & Pulido, F. (2009). Assessing the technical efficiency of extensive livestock farming systems in Extremadura, Spain. Livestock Science, 121, 7–14. Gunaratne, H. P., & Leung, P. (1997). Asian black tiger shrimp industry: a meta- production frontier analysis. Economics and Management of Shrimp and Carp Farming in Asia pp. 55-68. Ibrahim, M. N. M., & Schiere, J. B. (2002). Introduction to Agricultural System Studies. Faculty of Agriculture, University of Peradeniya, Sri Lanka. p. 9. Ibrahim, M. N. M., Stael, S. J., Danniel, S. L. A., & Thorpe, W. (1999). Appraisal of the Sri Lanka dairy sector, Vol. 2. (Synthesis report). Department of animal Science, University of Peradeniya; ILRI (International Livestock Research Institute), Nairobi, Kenya; and Ministry of Livestock Development and Estate Infrastructure, Colombo, Sri Lanka. Kollalpitiya, K. M. P. M. B., Premaratne S., & Peiris, B. L. (2012). Reproductive and Productive Performance of Up- Country Exotic Dairy Cattle Breeds of Sri Lanka. Tropical Agricultural Research, 23(4), 319-326. Kubota, S., Kato, Y., Kono, H., & Karunagoda, K. (2010). Feed utilization and improving management on small scale dairy farms in the Kurunegala District, Sri Lanka, Research Bulletin, Obihiro University, 31: 13-17. Mahipala M. B. P., Gunaratne, L. H. P., Bandara, A. B., & Peries, B. L. (2006). Farm-level economics, services and constraints of cattle farming in smallholder cattle farming systems in Up-Country Wet- Zone, in Sri Lanka. In: Proceedings of International Conference on Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 140 Livestock Services. China. pp. 242- 253. Meeusen, W., & Van den Broeck, J. (1977). Efficiency estimation from Cobb- Douglas production functions with composite error. International Economic Review, 18, 435-444. Latruffe L., Balcombe, K., Davidova, S., & Zawalinska, K. (2005). Technical and scale efficiency of crop and livestock farms in Poland: does specialization matter? Agricultural Economics, 32(3), 281–296. Lawson, L. G., Agger, J. F., Lund, M., & Coelli. T. (2004). Lameness, metabolic and digestive disorders, and technical efficiency in Danish dairy herds: a stochastic frontier production function approach. Livestock Production Science, 91, 157–172. Navaratne, M. A. J. S., & Buchenrieder, G. (2003). Economics of Small-Scale Dairy Farming in Sri Lanka - A Case Study from Coconut Cattle Silvo Pastoral System. Technological and Institutional Innovations for Sustainable Rural Development: Deutscher Tropentag 2003; International Research on Food Security, Natural Resource Management and Rural Development. Ondersteijn, C. J. M., Giesen, G. W. J., & Huirne, R. B. M. (2003). Identification of farmer characteristics and farm strategies explaining changes in environmental management and environmental and economic performance of dairy farms. Agricultural Systems, 78, 31– 55. Paris, T. R. (2002). Crop-animal systems in Asia: socio-economic benefits and impacts on rural livelihoods. Agricultural Systems, 71, 147–168. Perera, B. M. A. O., & Jayasuriya, M. C. N. (2008). The dairy industry in Sri Lanka: current status and future directions for a greater role in national development. Journal of National Science Foundation Sri Lanka, 36, 115-126. Phillips, J. M. (1994). Farmer education and farmer efficiency: A Meta-Analysis. Economic Development and Cultural Change, 43(1), 149-165. Pierani, P., & Rizzi, P. L. (2003). Technology and efficiency in a panel of Italian dairy farms: an SGM restricted cost function approach. Agricultural Economics, 29, 195- 209. Premarathna, S., & Premalal. G. G. C. (2005). Country Pasture/Forage Resource Profile. Food and Agriculture Organization, Rome, Italy. Premaratne, S., Premalal G. G. C., & Jayawardena, V. P. (2003). Sustainable management of grassland resources for ruminant livestock production in Sri Lanka. Tropical agricultural research and extension, 6, 60-65. Ranawana, S. (2008). Dairy Industry in Sri Lanka: Problems and Prospects. Economic review, Aug/Sep. pp. 2-5. Ranaweera, N. F. C. (2007). Improved market access and smallholder dairy farmer participation for sustainable dairy development. Central Bank of Sri Lanka, Colombo, Sri Lanka. Ratnayak A, D., Abegunawardena, H., Jayatilaka, M. W. A. P., & Abewansa, W. D. (1992). Identification of distinctive cattle farming systems in Sri Lanka and estimation of their production costs. Tropical Agricultural Research, 4, 225-245. Singh N. R., Chauhan, A. K., & Sharma, S. P. (2006). Economic analysis of milk production in tribal Ara of Udaipur (Rajasthan). Indian Journal of Dairy Science, Sep/Oct. 59: 5. Thiam, A., Bravo-Ureta, B. E., & Rivas, T. E. (2001). Technical efficiency in developing country agriculture: A meta-analysis. Agricultural Economics, 25, 235-243. Thomas, D., Zerbini E., Rao, P. P., & Vaidyanathan, A. (2002). Increasing animal productivity on small mixed farms in South Asia: A systems http://onlinelibrary.wiley.com.virtual.anu.edu.au/doi/10.1111/agec.2005.32.issue-3/issuetoc Asian Journal of Agriculture and Rural Development, 4(2)2014: 128-141 141 perspective. Agricultural Systems, 71, 41–57. Victor H., Moreira, H., & Boris E. Bravo- Ureta (2010). Technical efficiency and metatechnology ratios for dairy farms in three southern cone countries: a stochastic meta-frontier model. Journal of Productivity Analysis, 33(1): 33-45. Wubeneh, N., & Ehui, S. (2006). Technical efficiency of smallholder dairy farmers in the central Ethiopian highlands. International Association of Agricultural Economists Conference, Gold Coast, Australia. Yeamkong, S., Koonawootrittriron S., Elzo, M. A., & Suwanasopee T. (2010). Effect of experience, education, record keeping, Labor and Decision Making on Monthly Milk Yield and Revenue of Dairy Farms Supported by a Private Organization in Central Thailand. Asian-Australasian Journal of Animal Science, 23(6), 814–824. Zemmelink, G., Premaratne, S., Ibrahim, M. N. M., & Leegwater, P. H. (1999). Feeding of dairy cattle in the Forest- Garden farms of Kandy. Sri Lanka. Tropical Animal Health and Production, 31(5), 307-319. http://link.springer.com.virtual.anu.edu.au/journal/11123 http://link.springer.com.virtual.anu.edu.au/journal/11123