Innovations for smallholder dairy producers in Nepal

Material Information

Innovations for smallholder dairy producers in Nepal a study on the adoption and dissemination of mobile app feeding support tool and mastitis control technologies
Trimono, Rio ( author )
Physical Description:
1 online resource (64 pages) : illustrations ;


Subjects / Keywords:
Sustainable Development Practice field practicum report, M.D.P
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


This report describes the research and the findings from a field practicum undertaken in Nepal during summer 2019. The study was conducted as part of the monitoring and evaluation (M&E) activities of USAID's Feed the Future Innovation Lab for Livestock Systems at the University of Florida, which aims to sustainably intensify smallholder livestock systems in order to improve human nutrition, health, and incomes.1 The purpose of the field practicum was to assess the dissemination process of two innovations developed by Heifer International Nepal, with support from the Innovation Lab. The two innovations, Livestock Feeding Support Tools (L-FST) and Mastitis Control Technology Package, were designed and piloted to help smallholder dairy producers reduce the production costs and improve the quality of milk derivatives (Pfluger and Varijaksha 2018). The study on the adoption of innovations helps address the Sustainable Development Goals (SDGs) 1 -- No Poverty, 2 -- Zero Hunger, and 9 -- Industry, Innovation, and Infrastructure (Nepal Planning Commission 2015). The initiatives to reduce poverty, improve food security, and promote industry, innovation, and infrastructure are in line with Nepalese dairy sector strategies (Nepal Planning Commission 2015). Dairy production plays an important role in providing opportunities for livelihood generation and improving the nutritional status of thousands of households in Nepal. In 2017, there were 2.3 million cattle farmers and 1.7 million buffalo farmers in Nepal, with a dairy animal population of more than 7.30 million cattle and 5.17 million buffaloes (MALD 2017). However, the milk yields remained low, averaging 879 liters/milking animal in 2017 (755 liters/milking cow and 964 liters/milking buffalo) (MALD 2017). In 2016, the country's production of 1.9 million liters of milk per day could not meet the national demand of
2.4 million liters/day (Pant 2017). 2 Low milk production can negatively impact household incomes and nutritional outcomes, especially for young children (FAO 2014). Sustainable food productions in livestock and dairy systems require proper animal feeding management and Good Husbandry Practices (GHP) that can prevent and control disease, as well as ensure good quality and quantity of milk. Feed contributes to more than 60% of production costs in dairy farming, meaning that inefficient feeding practices and inconsistent availability of feed inputs largely affect the sustainability of dairy production (MALD, 2017). Moreover, farmers also face the challenge of the prevalence of mastitis, a bacterial disease that causes economic losses due to lower milk yield, quality, shelf life, and fat content, and additional cost to care for sick livestock (Ng et. al. 2010). In order to address the issues of inadequate feed and mastitis disease, researchers at Heifer International Nepal developed the L-FST and Mastitis Control Technology Package, which consist, respectively, of an app that helps farmers calculate the right feed composition for their animals, and a kit plus training to help them detect and control mastitis. Despite the potential benefits of adopting innovations, there may be obstacles to adoption due to potential users' knowledge gaps (lack of awareness of the technology or of its benefits), and reduced effectiveness of dissemination efforts. This study aims to assess the extent to which dissemination efforts have produced the intended results in terms of increased awareness, willingness to adopt and changes in feeding and mastitis control practices. The field practicum was conducted mostly in the Terai and to some extent also the Hill regions of Nepal to identify the enabling factors and obstacles affecting the awareness and the adoption of innovations.
Research findings can be used by research and extension organization to derive lessons from past efforts as well as contribute to improve the design and implementation of future dissemination strategies, in this and similar intervention programs. This work was funded in whole or part by the United States Agency for International Development (USAID) Bureau for Food Security under Agreement # AID-OAA-L-15-00003 as part of Feed the Future Innovation Lab for Livestock Systems.
Includes bibliographical references.
General Note:
Major departments: Latin American Studies, African Studies.
General Note:
Major: Sustainable Development Practice.
General Note:
Committee member: Serra, Renata.
General Note:
Committee member: Galindo, Sebastian.
General Note:
The MDP Program is administered jointly by the Center for Latin American Studies and the Center for African Studies.
Statement of Responsibility:
by Rio Trimono.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
037835494 ( ALEPH )
LD1780.1 2020 ( lcc )


This item is only available as the following downloads:

Full Text


Feed the Future Innovation Lab for Livestock Systems Innovations for Smallholder Dairy Producers in Nepal A Study on the Adoption and Dissemination of Mobile App Feeding Support Tool and Mastitis Control Technologies April 2020 Field Practicum Report Prepared by Rio Trimono Committee: Dr. Renata Serra, Dr. Sebastian Galindo Master of Sustainable Development Practice University of Florida


ii Acknowledgment This field practicum report could not have been completed without the enormous support and guidance from Dr. Renata Serra and Dr. Sebastian Galindo as advisory committee members and supervisors of this project, and Dr. Glenn Galloway and Dr. Andrew Noss as the director and program coordinator of MDP program, respectively. During the preparation and implementation of the field practicum, valuable support also came from Andrea Bohn, Brigitte Pfluger, and Lacey Harris Coble from the USAID Feed the Future Innov ation Lab for Livestock Systems at University of Florida / Institute of Food and Agricultural Sciences (UF/IFAS), as well as from Dr. Bhola Shankar Shrestha and the team from Heifer International Nepal, my field assistant, Binod Pokhrel from Agriculture an d Forestry University in Chitwan Nepal, and the Senior Livestock Development Officer in Gandaki Province Dr. Dirganath Dunganga. Disclaimer This work was funded in whole or part by the United States Agency for International Development (USAID) Bureau for Food Security under Agreement # AID OAA L 15 00003 as part of Feed the Future Innovation Lab for Livestock Systems. Any opinions, findings, conclusions, or recommendations expressed here are those of the authors alone.


iii Table of Contents Acknowledgment ................................ ................................ ................................ ................................ ......................... ii 1. INTRODUCTION ................................ ................................ ................................ ................................ ............ 1 2. CONTEXT ................................ ................................ ................................ ................................ .......................... 3 2.1. Organizations ................................ ................................ ................................ ................................ ............. 3 2.2. Innovations ................................ ................................ ................................ ................................ ................. 4 2.3. Country Profiles ................................ ................................ ................................ ................................ ......... 6 2.3.1. Geography, Demographics, and Economy ................................ ................................ .................. 6 2.3.2. Livestock Production Systems ................................ ................................ ................................ ....... 7 2.3.3. Dairy Cooperatives ................................ ................................ ................................ ........................... 9 2.3.4. Extension Services ................................ ................................ ................................ ......................... 10 3. LITER ATURE REVIEWS ................................ ................................ ................................ ............................ 11 3.1. Diffusion of Innovation ................................ ................................ ................................ ......................... 11 3.2. Defining Adopter ................................ ................................ ................................ ................................ .... 12 3.3. Variables Determining Adoption Rate ................................ ................................ ................................ . 13 4. CONTEXTUAL/CONCEPTUAL FRAMEWORK ................................ ................................ ............... 14 5. STUDY SITES AND METHODS ................................ ................................ ................................ ............... 17 5.1. Study Sites ................................ ................................ ................................ ................................ ................. 17 5.2. Methods ................................ ................................ ................................ ................................ .................... 18 6. RESULTS AND DISCUSSION ................................ ................................ ................................ ................... 20 6.1. Description of Respondents ................................ ................................ ................................ .................. 20 6. 2. Main Concerns in Dairy Sector ................................ ................................ ................................ ............. 21 6.3. Changes in Farming Practices and Dairy Sector ................................ ................................ ................. 22 6.4. Awareness, Trial, and Adoption ................................ ................................ ................................ ............ 24 6.5. Important Characteristics of Innovations ................................ ................................ ........................... 30 6.6. Barriers to Knowledge and Adoption of Innovatio ns ................................ ................................ ....... 35 6.6.1. Results: Knowledge and Persuasion ................................ ................................ ................................ ...... 35 6.6.2. Results: Decision, Implementation, and Confirmation ................................ ................................ .......... 38 6.6.3. Discussion: Comparing L FST and Mastitis Control Technologies ................................ ...................... 41 7. IMPLICATIONS ................................ ................................ ................................ ................................ ............. 43 8. CONCLUSION ................................ ................................ ................................ ................................ ................ 47 9. LIMITATIONS OF THE STUDY AND SELF REFLECTION ................................ ......................... 49 Bibliography ................................ ................................ ................................ ................................ ................................ 51 Appendix 1 Demographics of dairy farmers participating in the interviews ................................ .................... 54 Appendix 2. Demographics of Members of Producers Organizations and Technicians ............................... 54 Appendix 3. Interview and FGDs Questions Guideline ................................ ................................ ..................... 55 Appendix 4 . Response Matrix on Perceived Characteristics of Innovations ................................ ................... 56


iv List of Figures and Figures Figure 1. The three step evaluation process of the Innovation Lab for Livestock Systems ........................... 3 Figure 2 . L ................................ ................................ ................................ .................. 5 Figure 3. Mastitis control technology package ................................ ................................ ................................ ........ 6 Figure 4. Innovation Development Process ................................ ................................ ................................ ......... 11 Figure 5 . Contextual/Conceptual Framework of Field Practicum ................................ ................................ .... 15 Figure 6 Map of interviews/FGDs locations ................................ ................................ ................................ ........ 17 Figure 8 . Main concerns in Nepalese dairy sector expressed by respondents ................................ ................ 22 Figure 9 . Common changes in Nepalese dairy sector in the past 3 years ................................ ......................... 23 Figure 10. Awareness, trial, and adoption of L FST and mastitis control technologies (Producers) ........... 25 Figure 11. Awareness, trial, and adoption of L FST and mastitis control technologies (Technicians) ....... 26 Figure 12. Awareness, trial, and adoption of L FST and mastitis control technologies (FGDs) ................. 27 Figure 13. Awareness, trial, and adoption of L FST by districts ................................ ................................ ........ 28 Figure 14. Awareness, trial, and adoption of mastitis control technologies by districts ................................ 29 Figure 15. A pp Demonstrations ................................ ................................ ................................ .............................. 33 Table 1 Characteristics of Major Cattle Buffaloes Production Systems in Nepal ................................ ............ 8 Table 2 Common Characteristics of Cattle Buffalo Production Systems between Subsistence and Commercial Farmers ................................ ................................ ................................ ................................ .................. 9 Table 3. District Profiles of Study Sites ................................ ................................ ................................ ................ 18 Table 4. FST ................................ .................... 31 Table 5. ......... 32 Table 6. ................................ ................ 37


1 1. INTRODUCTION This report describes the research and the findings from a field practicum undertaken in Nepal during summer 2019. Th e study was conducted as part of the monitoring and evaluation (M&E) University of Florida , which aim s to sustainably intensify smallholder livestock systems in order to improve human nutrition, health, and incomes. 1 The purpose of the field pract icum was to assess the dissemination process of two innovations developed by Heifer International Nepal , with support from the Innovation Lab . The two innovations, L ivestock Feeding Support Tools (L FST) and Mastitis Control Technology Package, were designed and piloted to help smallholder dairy producers reduce the production cost s and improve the quality of milk derivatives ( Pfluger and Varijaksha 2018) . The study on the adoption of innovations helps address the S ustainable D evelopment G oal s (SDGs) 1 No Poverty, 2 Zero Hunger, and 9 Industry, Innovation, and Infrastructure (Nepal Planning Commission 2015) . The initiatives to reduce poverty, improve food security, and promote industry, innovation, and infrastructure are in line with Nepa lese dairy sector strategies (Nepal Planning Commission 2015) . D airy production plays an important role in providing opportunit ies for livelihood generation and improving the nutritional status of thousand s of households in Nepal . In 2017, there were 2.3 million cattle farmers and 1.7 million buffalo farmers in Nepal, with a dairy animal population of more than 7.30 million cattle and 5.17 million buffaloes ( MALD 2017). However, the milk yield s remained low, averaging 879 liter s /milking animal in 2017 (755 liters/milking cow and 964 liters/milking buffalo) (MALD 2017). In 2016, the country production of 1.9 million liters of milk per day could not meet the national demand of 2.4 million liters /day ( Pant 2017 ) . 1 See


2 L ow milk production can negatively impact household incomes and nutritional outcomes, especially for young children (FAO 2014) . Sustainable food productions in livestock and dairy systems require proper animal feeding management and Good Husbandry Practices (GHP) t hat can prevent and control disease, as well as ensure good quality and quantity of milk. Feed contributes to more than 60% of production costs in dairy farming, meaning that inefficient feeding practices and inconsistent availability of feed inputs largely affect the sustainability of dairy production (M ALD , 2017). Moreover, farmers also face the challenge of the prevalence of mastitis, a bacterial disease that causes economic losses due to lower milk yield, quality, shelf life, and fat content, an d additional cost to care for sick livestock (Ng et. al. 2010). In order to address the issues of inadequate feed and mastitis disease, researchers at Heifer International Nepal developed the L FST and Mastitis Control Technology Package, which consist, re spectively, of an app that helps farmers calculate the right feed composition for their animals , and a kit plus training to help them detect and control mastitis. Despite the potential benefits of adopting innovations, there may be obstacles to adoption du benefits), and reduced effectiveness of dissemination efforts. This study aims to assess the extent to which dissemination efforts have produced the intended results in ter ms of increased awareness, willingness to ad o pt and changes in feeding and mastitis control practices. T he field practicum w as conducted mostly in the Terai and to some extent also the Hill r egions of Nepal to identify the enabling factors and obstacles affecting the awareness and the adoption of innovations. Research findings can be used by research and extension organization to derive lessons from past efforts as well as contribute to improv e the design and implementation of future dissemination strategies , in this and similar intervention program s .


3 1. CONTEXT 1.1. Organizations ` Feed the Future Innovation Lab for Livestock Systems (Lab for short) is one of the USAID Feed the Future Innovation Labs and is i tute of Food and Agricultural Sciences (IFAS). The Lab develops long term research and capacity development efforts, primarily through multi disciplinary, integrated, and competitively funded applied research . 2 In Nepal, one of the Feed the Future countries, two out of four projects funded by the Lab were led by Heifer Project International Nepal , and were conducted between October 2016 and March 2018 . Figure 1 . The three step evaluation process of the Innovation Lab for Livestock Systems To monitor and evaluate the se and other projects , the L ab developed an evaluation roadmap in three stages as shown in Figure 1 ( Pfluger and Varijaksha 2018) . For the projects in Nepal, Stage 1 was completed in May 2018 and provide d information on the characteristics of the innovations and their relevance to the local context in the pilot project location, as well as the dissemination plan for potential uptake by a larger audience ( Pfluger and Varijaksha 2018) . The activities referr ed to in this report conclude Stage 2 , which seeks to examine the awareness and application of the innovations and the barriers preventing adopt ion . Heifer International Nepal is part of a global development organization , Heifer International, whose missi on is to end hunger and poverty in a sustainable way by supporting 2 See us/


4 and investing alongside local farmers and their communities . 3 Heifer International Nepal (Heifer for short), which started its official operations in the country in September 1997, is the main implementing partner in the research sponsored by the Feed the Future Innovation Lab for Livestock Systems in Nepal. B etween September and November 2018, Heifer organized, together with National Dairy Development Board (NDDB) , four one day training on L FST that w as attended by 131 participants (dairy farmers, cooperative members, and dairy entrepreneurs) from 12 different Distr icts . Heifer International Nepal also introduced L FST in one of the training sessions organized by Agriculture Information and Training Centre o n January 28 th and February 15 th , 2019 that were attended by 32 Livestock Officers of Department of Livestock S ervices. By the time of the field practicum, training on Mastitis Control Technology Package ha d not been extended beyond the pilot project locations. However , some of the training for mastitis control strategies ha d incorporated the topic of mastitis and control techniques including post milking teat dipping, hygiene management, and sanitation. The mastitis control technologies were also disseminated through Community Animal Health Workers (CAHWs) in Heifer project a reas, radio broadcast, and outreach workshops at the district and national level. 1.2. Innovations L FST is a mobile application based tool to formulat e the least cost, nutritionally balanced feed rations for dairy cattle and buffalo utilizing locally availabl e f odders and to predict milk yield. The app is compatible with Android phone s and available for download in Google Playstore. Figure 2 interface and features . The f ormulation allows users to calculate feed rations by enter ing dairy animal related data (i.e. body weight, pregnancy stage, and fat content) in the Input Form . Th en, users can customiz e the type of f ee d s and add price in the Ration Form to generate Ration Summary (feed ration, nutri ents composition, and the feed cost) . Feed Store contains information on the locally available feed s and the tool to add new ingredients 3 See us/index.html


5 or edit the existing list . Solution List shows the history of saved feed ration while Setting allows users to choose language in English or Nepali. During the development of the app, the pilot research by Heifer International found that 95% of over 200 dairy producers participating in the project reported a reduction of feeding costs and an increase in milk production by 15% in the first month and an additional 7.7% in the next two months ( Management Entity 201 9 a). Figure 2 . L The technology package for prevention and control of mastitis comprises good husbandry practices (GHP) ; mastitis detection using California Mastitis Test (CMT) ; and milk conductivity test ; and mastitis control using post milking teat dipping (PMTD) and dry cow therapy (DCT) technologies. CMT is a four compar tment paddle and reagent that provides an indication of the number of somatic cells found in the milk (McFadden 2011). Image 1 (left) shows the use of CMT to detect mastitis infection based on the level of somatic cells found in the milk sample . Meanwhile, clinical mastitis by measuring the absolute electrical conductivity values from milk samples. PMTD is a technique of dipping animal teats with disinfectants to prevent udder infection by reducing the


6 number of bacteria on teats (Canadian Bovine Mastitis Research Network 2011) . Figure 3 (right) m mary antibiotic therapy after the last milking of lactation to prevent new ma mmary infections during the dry off period ( DFA 2006). During the pilot project involving 422 dairy farmers, the evaluation study showed reduced sub clinical mastitis cases from 55% to 28% in dairy cows and from 78% to 18% in buffaloes within the six months of adoption , and no reporting of clinical mastitis cases ( Management Entity 201 9b ). Figure 3 . Mastitis control technology package. (Left) Milk samples tested using CMT compartments and reagents. One quarter show CMT positive result (McFadden 2011); (Right) Applying teat disinfection using teat dip (Source: Canadian Bovine Mastitis Research Network 2011) 1.3. Country Profile 1.3.1. Geography , Demographics , and Economy Nepal is landlocked between two rapidly emerging economies, China in the north and India in the south. With a total size of 143,350 km 2 land area, t he country is divided into three agro ecological zones mountains, hills, and Terai with elevation from 60 m in the Terai flatland in the south to 8,848 m in the Himalayas in the north (Thapa 2010). Hills has an elevation between 610 to 4876 and comprises 42 percent of the land area w hile Moun tains and Terai comprises 35% and 23% of land area respectively ( MoPE 2017) . Annual rainfall ranges from 1500 mm to 2500 mm and 60 80 percent of the rainfall is concentrated during the monsoon season from June to September (Thapa 2010). Based on the nati onal census in 2011, N epal has a population of 26,494,504 people with a 1.35 annual growth rate (M oPE 2017). As of January 2020, the population has reached


7 28,891,147 people with an average population density of 203 people per k m 2 ( Worldmeter 2020 ) . The w orking age population between 15 to 59 years represents 57 percent of the population ; m ore than half (50.3%) of the total population live in the Terai region (M oPE 2017). Increased migration from Mountains and Hills regions to Terai is caused by unequal distribution of basic facilities and infrastructure, availability of productive land in Terai, difficult topography of Hills and Mountains, and the disparity of socio econom ic development ( MoPE 2017). In 2015, the maternal mortality ratio wa s 258 death per 100,000 live birth (UNICEF 2015). The sex ratio at birth was approximately 105 male births for every female birth in census 1981 but switch ed to more female s in census 2011 (94 males born per 100 female) ( MoPE 2017). Remittances sent by migrants to Nepal are a very important source of income for households and amounted to nearly Karki 2017). GDP growth wa s at 6.3% in the 2018 f iscal y ear (FY) , down from 7.9% in FY2017 due to floods disrupting paddy production (ADB 2019) , but still considerably high . Nepal suffered a decline equal to 0.6 GDP growth in FY2016 due to 2015 earthquakes (ADB 2019). In FY2018, the agriculture sector contributed 28.2% of GDP, down from 38.0% in FY1999 (ADB 2019). The m ajority of farmers hold an average farm size of 0.8 ha with 47 percent of land holdings measuring less than 0.5 ha (Thapa 2010). The l ivestock sec tor contributes almost one third of the GDP from agriculture with dairy production representing the most significant sub sector within the livestock production systems based on the number of dairy producing households and the share d output to the GDP (Sapkota 2010). 1.3.2. Livestock Production Systems The Food and Agriculture Organization (FAO) defines li vestock production systems as a subset of farming systems in which livestock contribute more than 10 percent to total farm output either in the form of livestock value or intermediate functions such as traction and manure (Sere and Steinfeld 1996). Livesto ck systems are differentiated based on agroecological


8 zones ( such as arid, semi arid, sub humid, humid, and tropical highlands ) and classified into landless livestock systems, grassland based livestock systems, mixed rainfed livestock systems, and mixed ir rigated livestock systems (Sere and Steinfeld 1996; Teufel et. al. 2010). Based on the total land area, livestock production systems in Nepal are mostly mixed rainfed humid/sub humid (16.2%) , mixed irrigated arid/semi arid (15.7%) , and mixed rainfed tropical highland (11.4%) (Teufel et. al. 2010). Table 1 describes the characteristics of main livestock production systems in Nepal. Table 1 Characteristics of Major Cattle Buffaloes Production Systems in Nepal Systems Mixed Rainfed Humid/ Subhumid Mixed Irrigated Arid/ Semiarid Mixed Rainfed Tropical Highland Length of Growing Period More than 180 days Less than 180 days Less than 110 days and daily mean temperature during the growing period in the range 5 20 o C Irrigated areas Less than 25% More than 25% Less than 25% Zone The Terai Region The Terai and Hills Region The Mountain (> 2,000m) and Partly Hills Region (300 2 ,000 m) Ruminants Feeding Practices Grazing in terraces, fallow land , and forest areas. Feeding of crop residues and agro industrial by products. Harvesting green forages from croplands and the buds in the Summer. Backyard grazing. Stall fed with cut and carry fodder and straw. Summer: stall fed with cut and carry green grass and field weeds; short hour grazing (buffalo). Winter: few hours grazing during the daytime, stall fed on crop residues in evening Cropping Systems Maize wheat fallo w and maize rice wheat. Year round crop production, typically rice wheat systems. Rotational, m ixed, diverse, and subsistence oriented crops (e.g. maize potato wheat finger millet ) . Adapted from Teufel et. al. 2010 For the most part, milk production in Nepal is a subsistence oriented activity carried out under traditional mixed farming practices with small non commercial holdings (Dhakal et. al. 2014). Approximately 85% of cattle and buffalo farmers in the country are smallholders owning le ss than five animals per household; only about 1% farm households own more than ten animals (MALD 2017). There is an increasing trend in the number of commercial farms in Nepal, reflected in the growing number of improved breeds in the dairy sector (Dhakal et. al.


9 2014). Table 2 shows the common characteristics of subsisten ce and commercial farmers , which are classified based on the degree of commercialization (the percentage of sales in relation to the gross return) and the degree of dependence on livestoc k (Teufel et. al. 2010) . In cattle buffalo production system, subsistence farmers still mainly rely on communal grazing and diverse source of income. On the other hand, commercial farmers use purchased feed and forages, and mainly depend on dairy and milk production as the source of income. Table 2 Common Characteristics of Cattle Buffalo Production Systems between Subsistence and Commercial Farmers Subsistence Commercial Specialized Multiple Products Specialized Multiple Products Feed resource Communal grazing Communal grazing, cut and carry Purchased feed resources Purchased feed resources + Purchased feed + Forages + Forages Product Milk Milk, draught, meat, dung Milk Milk, meat The m ain source of income Diverse; typically crop production Diverse Dairy production Dairy and meat production Adapted from Teufel et. al. 2010 1.3.3. Dairy Cooperatives Dairy cooperatives play an important role in organizing milk production and marketing of smallholder farmers in Nepal. There are approximately 1,875 dairy cooperatives spread across 62 of the 75 districts in Nepal ( MALD , 2017). The burgeoning of dairy cooperatives in the country began after the formation of Dairy Development Corporation (DDC) in July 1969 whose major obje ctives include developing organized milk collection and marketing systems in the country (Sapkota 2010). ma rked the first formalization of dai r y cooperatives in Nepal (Sapkota 2010 ). MPCS is the first tier of the cooperative system whose main function s are to collect milk from the farmers (both the members and non members), run quality control, sell and deliver the milk to nearest milk chilling centers or milk processing plants of DDC and/or private


10 companies, as well as receive and distribute payment to th e individual milk suppliers (Sapkota 2010). Some cooperatives run their own milk chilling vat s /containers and milk processing operation to produce various dairy products (Sapkota 2010). The second tier of the system is represented by District Milk Producer s' Cooperative Unions ( DMPCUs ) whose tasks are to deliver programs for improving dairy production and product diversifications and accelerate poverty reduction among rural milk producers (Sapkota 2010). The apex associa t ion of the dairy cooperatives in Nepal is the Central Dairy Cooperative Association Limited Nepal (CDCAN) which in 2015 counted a national network of 1,050 cooperatives in 53 districts in the Terai and mid hill region of Nepal (SAMARTH 2015). 1.3.4. Extension Services In t he Nep alese agricultural extension system , the public se ctor has the essential function to conduct research for generating the technologies , and to provide extension services to farmers . One of the limitations in the system has been that extension f ocus es on tec hnology transfer instead of engaging through a participatory process ; there is limited follow up in ensuring the adoption of the technologies at the farm level (Dhital 2017; Thapa 2010). More recently , the ve been undergoing a transition in the approaches and administration, following the ongoing decentralization reforms, and the devolution of agricultural extension function from the federal to the local ly elected bodies (District Development Committees) (Dhit al 2017). Based on the delivery institutions, the extension services are moving towards three categories : 1) Public level, including District Livestock Offices (DLSOs), Agriculture Service Centre, Nepal Agriculture Research Council (NARC), Department of Co operatives, DDC, nongovernmental organization (iNGO) ; 3) Community based organization (Dhital 2017; Sapkota 2010). The rationale for this new decentralized system is that , once put in place and f unctioning, it can better serve the needs of the local communities.


11 2. LITERATURE REVIEWS 2.1. Diffusion of Innovation Innovations in agriculture refer to new products or new ways of doing something in order to increase effectiveness, competitiveness, and resilience (FAO, 2018). As shown in Figure 4 , t he innovation process start s from recognizing a problem or need, conducting innovati ve research, developing the innovation, to commerciali zation. Then, the inno vation is diffus ed and adopt ed by users , and create imp ac ts /con s equences . Widespread adoption of an innovation can be achieved through dissemination (active communication and facilitation approach done by change agents to promote adoption) or diffusion (the organic process by which an innovation spread through certain communication channels to the members of the social system) (Rogers 2003; Greenhalgh et. al. 2004). Figure 4 . Innovation Development Process, Stages of Innovation Decision, and Factors Affecting the Knowledge and Persuasion of An Innovation. Adapted from Rogers (2003) T he diffusion and adoption stage, as shown in Figure 4 , comprises five stages : gaining knowledge of an innovation ( knowledge ), forming an attitude towards it ( persuasion ), making the decision to adopt or reject ( decision ), implementing the innovation ( imple mentation ), and continuing


12 in using it ( confirmation ) (Rogers 2003). The knowledge can be distinguished into awareness knowledge or knowing that innovation exists, how to knowledge or knowing how to use an innovation properly, and principle knowledge or kn owing the underlying principles that make an innovation works (Rogers 2003). In the persuasion stage , five characteristics of innovation can determine t he rate of adoption : 1) Relative advantage better than 2) Compatibility 3) Complexity 4) Trialability and 5 ) Observability 3, 30 31). In the d ecision stage, Rogers (2003) distinguishes two types of rejection as (1) active rejection, which involves consideration to adopt (including its trial) followed by decision not to adopt it, (2) passive rejection, which never really consider s the use of innovation. In the implementation and confirmation stages, users assess whether to continue adoption, modify or change the use of the innovation to fit with their circumstances and preferences ( reinvention ), or stop the adoption altoge ther ( discontinuance ) (Rogers, 2003) . 2.2. Defining the A dopter According to Rogers (2003, 21), adoption is s micro level adoption of agricultural innovations in Nepal have use d , for exampl e in adopting improved soil conservation technology (Tiwari 2008), organic farming (Kafle 2011; Karki, Schleenbecker, & Hamm 2011), agricultural intensification (Raut 2011; Alomia Hinojosa et. al. 2018), agroforestry (Dhakal, Cockfield, & Mar aseni 2015), and new/improved crop varieties (Ghimire & Huang 2015 & 2016; Barrueto


13 2018). A dopter of management practices may be more d ifficult to define because farmers can partially adopt or gradually change behaviors (e.g. a farmer may be classi fied as an adopter of improved rice but still grow local seed in some areas of his/her farm) (Doss 2006) . 2.3. Variables Determining Adoption Rate Besides attributes of the innovation, Rogers (2003) argue that the adoption rate is affected by the type of decision to adopt ( optional, collective, and authority ) , communication channels and the nature of the social system (Rogers 2003). Individual characteristics (i.e. gender, age, experience, an d socioeconomic status) and personalit y traits (e.g. being progressive and commercial oriented) also influence the technology acceptance and usage behavior (Venkatesh et. al. 2003 ; Rogers 2003). In Nepal, m embership in a group stimulates participation in extension program s and peer learning that create the facilitating condition for adopti ng new management practices (Tiwari, 2008; Suvedi, Ghimire, & Kaplowitz, 2017; Adhikari, Timsina, & Lamichhane, 2018). The a doption of agricultur al innovations is usually constrained by factors such as limited information, labor scarcity, limited access to credits, and limited availability of inputs (Doss 2006) . In Nepal, labor scarcity and limited access to saving and credits are major constraints for farmers to adopt improved seed varieties and post harvest safety measures (Floyd, 2003; Flock, 2015; Kumar et. al., 2017). Socio economic factors such as years of schooling, age, and farm/herd size also affect the likelihood to grow new crops o r improved varieties and apply new farm management practices (Floyd et. al., 2003; Tiwari, 2008; Karki, Schleenbecker, & Hamm, 2011; Banjara, 2016). Caste and ethnicity have been included as explanatory variables affecting the adoption of agricultura l technologies where higher caste and privileged ethnic groups are more likely to undertake adoption (Floyd, 2003; Tiwari, 2008; Barrueto et. al., 2018).


14 3. CONTEXTUAL/CONCEPTUAL FRAMEWORK The c ontextual/conceptual framework shown in Figure 5 illustrates the underlying theories for this study, the pathway leading up to improved income and nutritional status of dairy producers in Nepal , and the area of contribution of the field practicum . Starting from the bottom of the figure, the conceptual framework borrows from the theoretical concepts illustrated in the previous literature reviews. different people these become the factor that affect s the ra te of adoption. The other factors affecting the adoption are at the individual/household level (socio economic status, age, and community level (environment, insti tution, and social systems). The rate of adoption is the result of a cumulative process of three nested aspects: awareness, trial, and adoption . Awareness in this study refers to awareness knowledge discussed in the previous section , that is, whether a person has heard about the existence of innovation and its purpose. Trial refers to whether or not a person has tried the innovation ( or not ) and is measured as the proportion of those who have tr ied out of those who are aware . A doption refers to the decision, implementation, and confirmation regarding the innovation ( , dropped , and ) , and is calculated as the proportion of those who have tried. The contextual framework component of Figure 5 characterizes the contribution and role of the different actors and the phase they participate in. Researchers based at Heifer and its in country partner organizations (i.e. Department of Livestock Services, Himalayan College of Agri cultural Science and Technology, Nepal Agriculture Research Council, and National Dairy Development Board), with the support of the Innovation Lab at UF/IFAS, developed the research and the initial proof of concept for the two innovations, which was pilote d in a small setting.


15 Figure 5 . Contextual/Conceptual Framework of Field Practic u m


16 Heifer and in country partner organizations, as well as the province and district level governments, have been involved in the Adoption/Diffusion stage, by disseminating innovations. One of their roles has been providing extension services through livestoc k technicians and extension agents to dairy cooperatives and producers. Disseminating L FST and mastitis control technologies to wider communities in Nepal is expected to scale up the positive impact of these innovations on the income of smallholder farmer s and the overall nutritional status of the people through better availability of quality animal sourced food. Ultimately, such benefits will contribute to the development of the Nepalese dairy sector and the achievement of e global poverty and hunger. The field practicum M&E team undertaken during the Adoption/Diffusion phase through research on the adoption and dissemination of technologies among the producer s and related organization s in Nepal. In particular, the field practicum research focuses on assessing the level of awareness, trial, and adoption of L FST and mastitis control technologies and on examining the factors affecting the rates of adoption (perc eived innovation characteristics and barriers in raising awareness, facilitating trial, and promoting adoption). Based on th e theoretical framework, the research questions developed for this study are as follows: 1. What is the level of awareness , trial, and adoption of L FST and mastitis control technologies among dairy producers and livestock technicians? 2. What characteristics of innovations are perceived as the most important? 3. What are the barriers to knowledge dissemination and adopt ion of the innovations? The remaining of the report describes the sites, the methods, and the main findings from the research in Nepal during summer 2019 (hence


17 4. STUDY SITES AND METHODS 4.1. Study Sites The study was conducted in 7 sites (6 districts and 1 metropolitan area): Chitwan and Greater Kathmandu Area in Province 3; Kaski in Province 4 (Gandaki Province); Kapilbastu, Dang, Banke, and Bardia in Province 5, as shown in Figure 6 . The main reason for selecting Chitwan, Kapilbastu, Bardiya, and Kaski is because of the representation of farmers and dairy cooperative members who have participated in the training of L FST (nearly 70% of trainee came from these districts). Meanwhile, the majority of DLS of ficers who have received training in L FST were stationed in the Greater Kathmandu Area that include d Kathmandu and Patan, so data collection was also conducted there. Bardiya and Dang were selected mainly because of the proximity to the pilot project loca tion of the mastitis control technology package, so the study could also capture the experiences of those involved in the project. Figure 6 districts, number of people trained in L FST, and number of producers and livestock technicians participated in interviews/FGDs.


18 The s even sites represent t wo agro ecological zones and development regions (Terai and Hills of Central, Western, and Midwestern Regions of Nepal) which when combined comprise a large proportion of dairy animal population s and contribute to more than 25% of the total milk produced in the country . Table 3 shows the elevation, area size, human population, number of milking cows and buffaloes, and the milk volume (in liter s ) produced by district . Table 3 . District Profiles of Study Sites Districts Elevation (m) Size (km 2 ) Population (2011) Milking Cows Cow Milk Milking Buffaloes Buff Milk Chitwan 100 815 2,238 579,984 21930 13971 32473 29658 Bardiya 138 1279 2,025 382,649 15595 8892 27566 23487 Banke 129 1290 2,337 491,313 12562 7032 17766 16549 Dang 213 2058 1,764 552,583 15787 8848 22150 24219 Kapilbastu 450 3937 1,738 571,936 14812 7789 20507 19596 Kaski 450 8091 2,017 492,098 16021 9773 31566 29704 Kathmandu 1280 1340 395 1,081,845 5674 4493 10014 10212 Source: Ministry of Agricultural Development 2014 4.2. Methods The data collection method for this study include s in depth interviews and Focus Group Discussions (FGDs) ( Denzin and Lincoln 2017 ) . Semi structured interviews with producers and technicians were predominantly conducted face to face (five interviews were throu gh phone calls with selected respondents ) in each study site after appointments were made one to several days in advance. R espondents for both producers and technician groups were selected through purposive sampling (Palys 2008) from those listed as training participants or those having a strategic position in the corresponding organizations. This sampling method was used to gain rich information about the nd experiences related to L FST and mastitis control technology, and dairy sector development at various levels. The interviews aimed to collect information on four main areas : 1) awareness of the innovations and experience in trying and adop ting them; 2) perception o f the characteristics of the innovations; 3) perception about enabling factors and barriers to adoption; and 4) dissemination activities. Interview modules also include d questions on the demographic and socio economic data,


19 problems or concerns on the farm or dairy sector, and self perceived changes related to dairy production practices in the past three years. A t otal of 61 interviews were conducted with 36 producers and 29 technicians (some interviews had more than one active interviewee ). The main reasons for using FGDs are twofold : first, to examine the mechanism by which the innovation was being diffused in the community/organizations; and second, to identify the novation . The themes that were pursued through FGDs were similar to what w as asked in the interview but of course , the FGDs methodology emphasize s the . Purposeful random sampling was used to select FGDs respondents among small producers who showed up during milk collection time in the dairy cooperative ( which happens twice daily, around 6:00 9:00 am and 5:00 7:00 pm). In two FGDs, respondents were purposely selected be cause of their participation in L FST training. There was a total of eight FGDs conducted in this study with each group consist ing of five to ten p articipants . Figure 7 shows the setting during milk collection time and the FGD with dairy producers from a d airy cooperative in Dang district . Figure 7 . Focus Group Discussion . in Chitwan; (Right) FGD with dairy producers in Dang, Nepal. (Source: Personal Documentation) Most interviews and FGDs with dairy producers and cooperative members were held in the Nepali language with the assistance of senior year student s from the Agriculture and Forestry University of Nepal. Meanwhile, most interviews with livestock officers and technicians were


20 mainly conducted in English . During the interview and FGDs, responses from respondents in the Nepali language were translated in to Englis h and recorded on paper as direct quotes used for the analysis . S ome audio recordings were later transcribed in English . The guideline s for interviews and FGDs questions are shown in Appendix 3. The i nterview guide includes close and open ended questions. Informed consent to do the interview, record the audio, and take documentation was asked to each interviewee and to the members of discussion groups. This study uses thematic analysis to examine the c lusters of perceived innovation characteristics and the factors affecting the decision to adopt L FST and mastitis control technologies a method for identifying, analyzing , and reporting patterns within data. Such an approach is used to capture rich information about enabling factors and barriers to innovation uptake from those who have received training or may have the opportunity to learn about the innovations from others. Data obtained from all transcripts and notes were coded into themes that commonly appeared and were strongly expressed by respondents. Themes referring to factors affecting the adoption of L FST and mastitis control technologies were linked to the factors discu ssed in the literature reviews . 5. RESULTS AND DISCUSSION 5.1. Description of Respondents Most dairy farmers participating in the interviews are between 46 to 55 years old, have higher secondary level certificates (12 th grade), live in a family of five, and own less than four lactating animals ( s ee Appendix 1). tasks entail ing feedin g the animals, cleaning shed, and milking every morning and afternoon are typically shared among husband, wife, and other adult members in the family. Farmers owning a larger number of cows/buffaloes (more than four) tend to hire workers and focus on expanding milk production and trading . Many respondents are smallholder farmers relying on other income generating activities such as


21 farming cash crops, collecting milk from other farmers , working in the cooperative, and receiving remittance from relative s working abroad. Cooperative officers participating in the interviews are mostly between the age of 36 to 45 and work as a cooperative manager ( seven people ), a chairperson/leader ( three people), and a private dairy owner/director ( three people) (See Appendix 2). Cooperative managers usually work on many different aspects of the organization such as milk collection, milk processing, finance and accounting, human resource management, provision of livestock, and many others. Meanwhile , most livestock tec hnicians participating in the interviews are between the age of 26 to 35, have bachelor 's s , and have 2 to 5 year experience working as technicians (See Appendix 2 ). Technicians interviewed are either private technicians working for coop eratives , owners of an animal pharmacy or government officers who operate in various levels (i.e. municipality, district, province) and sub sectors (feed quality control, livestock policy, vet hospital, breeding development, extension services, etc.). 5.2. Main Issues in Dairy Sector All r espondents were asked , s on the dairy farm and dairy and on average they would describe two to four problems they faced on the farm or dairy production related activities. Figu re 8 illustrates the common dairy sector related issues perceived by respondents. According to producers , the main issues in the dairy sector were the high cost of production and the prevalence of mastitis , followed by low milk prices . Instead technicians and government officers perceived scarcity of green grasses and lack of manpower as the ir main concerns in the sector and were the least concerned for low milk price s . The problem that were mentioned the least by the two categories combined we re: unproductive and male cow, lack of youth involvement, and lack of skills . Other concerns that seem ed important for a number of respondents include infertility, difficulties to sell milk, ineffective


22 government subsidies, limited access to veterinarian and livestock technician, low quality of commercial feed, and foot and mouth disease (FMD). Figure 8 . Main concerns in Nepalese dairy sector expressed by respondents 5.3. Changes in Farming Practices and Dairy Sector Respondents were asked , . Figure 9 shows the common dairy sector related changes based on the frequency expressed by the respondents. Da iry producers mostly cited improved feed management practices , the increased role of cooperative and the reduced prevalence of mastitis. Most of these respondents were involved in the pilot projects and training program in L FST and mastitis control techno logy package led by Heifer. They believe d that changes on the farm and community level have been indirectly affecting the process of adoption and dissemination of innovations at the community level.


23 Figure 9 . Common changes in Nepalese dairy sector in the past 3 years The major change mentioned by l ivestock technician s and development officers was the decentralization reforms, which lead to difficulties in planning and implementing livestock programming . Due to the change in mandate and staff position across offices , there were lack of manpower and infrastructur e in the local office as described in Quote 1 . Technicians recognized that local office would take time to adapt to the new system so they would focus mor e on building internal capacity and providing basic services to dairy producers. T hey indicated decentralization as one of the reasons for their limited knowledge and program implementation related with L FST and mastitis control technology package. a program nor plans to specifically train dairy farmers to use new technology. This office is new and most staffing posts are still empty because no one really wants to work in the local office . Maybe because of in adequate laborator y and other related facilities . Previously, we use d some technologies [provided by the central government] to help dairy responsibility of local government which has no [adequate] budget, human resources, and monitoring syst ems. The l ocal government may not prioritize the livestock [and dairy] sector . T hey usually allocate most of the budget for road construction and other things. The transition to new systems can also create confusion among farmers about where to go to test milk sample, to ask for animal treatment, and to seek extension services. (Livestock technician, 46) ( Quote 1 )


24 5.4. Awareness , Trial, and Adoption 6.4.1. Results Respondents were asked if they were aware of L FST and mastitis control technologies and ha d tried, used/applied, or already dropped them. For L FST, about 70% of dairy producer s interviewed were already aware of L FST since most of respondents selected had participation in the L FST training. However, less than half of those ha d tr ied and only 27% (3 out of 11) who had tried ended up using the App ( Figure 10 ). Those who used L FST stated that they did formulation for one animal and used the ration formula to make feed for multiple animals that shared similar characteristics (e.g. body weight, pregnancy). One formulation was used until there were changes in the availability of local feed sources, usually when the season changed. Adopters of L FST also used different ways of weighing the feed materials (e.g. usi ng scale, basket, or simply estimation), and liberally adapted the recommendations on the proportion of each component in the ration s (e.g. use d a slightly different volume of concentrate and exclude some material s that were limited in stock ). Quote 2 conve difficulties in starting to use L FST, the process taken to utilize the app, and its modification. worked, so I experimented with 1 cow. In one month, the cow produced more milk from 20 to 22 liters in average. Then, I used the app for formulating used the app again and changed formulation when Farmer, ( Quote 2 ) Despite the low adoption of L FST among the participants in L FST training, several farmers s aid that they understood the basic s of feed ration balancing ( i.e . mixing different feed ingredients and using mineral blocks) and had tried to implement the principle knowledge without using the app. Th ese respondents also observed positive changes in productivity and animal health, so they felt less need to use the app .


25 Figure 10 . Awareness, trial, and adoption of L FST and mastitis control technologies (Producers) In the case of mastitis control technologies, m ore tha n half of respondents from the producer group (20 out of 36) were aware of at least one component of the technology package (Figure 10 ) . Out of 20 producers who were aware , 80% had experienced implementing some strategies in preventing and controlling mastitis, including post milking teat dipping (PMTD) and mastitis detection kit. However, almost half of those who adopted the technologies (11 out of 25) had discontinued due to various reasons. PMTD was the most used method to prevent infection. Only few dairy farmers knew and used California Mastitis Test ( CMT ) , while none of the re spondents was familiar with dry cow therapy or milk conductivity test. For detection, many dairy producers still relied on veterinarians or technicians to help diagnose the occurrence of mastitis. They were also more familiar with the rapid mastitis test k it than with CMT. In terms of the management, most respondents who were aware of Good Husbandry Practices (GHP) had been improving the


26 sanitation and hygiene of their animal shed. Quote 3 shows the typical responses from producers when they were asked abou t the awareness and use of mastitis control technologies. miliar ( Quote 3 ) Figure 11 . Awareness, trial, and adoption of L FST and mastitis control technologies (Technicians) In the technician group, nearly h alf of the respondents (14 out of 29) were aware of L FST App . Half of technicians who were aware (7 out of 14) had tried the app to give a recommendation to farmers and as a quick reference in formulating animal feed ration without the need to use other complicated software or reading some literature. Almost half of those who had tried L FST (3 out of 7) had not had the chance to use it in their work. This was not surprising c onsidering the length between when the training ended and when interviews were conduc t ed. On the other hand, all respondents in the technician group knew about mastitis control technologies from the training they received as livestock technic i an. Almost all these


27 individuals (25 out of 29) had tried implementing mastitis control technologies in their work and more than half of those who had tried (14 out of 25) were adopters. Figure 12 Number of FGDs with members who were aware, tried, and adopted L FST and mastit is control technologies among producers in FGDs In the FGDs, six groups had at least a few members in the group who were aware of L FST as shown in Figure 12 . However, only two sixth s of the groups had members who had tried L FST and one sixth had members who adopted L FST. Among the members of the groups, L FST was not a common topic of discussion and it was not adopted voluntarily without intervention . This was d ifferent from the case of mastitis control technologies since some cooperatives actively help provide the tools and technological input for its members. All groups have at least a few members knowing about mastitis control technologies with three out of eight FDG s having many membe rs who were aware of the innovation. Almost all the FGDs groups had some members who tried and adopted mastitis control strategies . This was expected especially for some g roups in Bardiya and Dang w here the cooperatives had received training in the mastiti s control technology package by Heifer International Nepal.


28 Figure 13 . Awareness, trial, and adoption of L FST by districts Comparison by districts as shown in Figure 1 3 indicates that Bardiya Banke Dang has the smallest proportion of respondents who have heard of, tried, and adopted L FST, despite the three districts combined. Meanwhile, Kapilbas t u and Kaski were the only two districts with L FST adopters among dairy producers . In Chitwan, despite t he highest number of technicians who were aware of L FST, the number of people who had tried and adopted w as not very different from those in Bardiya Banke Dang and Kaski. This was due to the fact that many technicians interviewed in Kathmandu were station ed in the central office where mostly work in policy and administration and were not directly involved in delivering extension services to farmers. In contrast, awareness o f mastitis control technologies was more common for technicians in all districts. Th ere were more people who had tried and adopted mastitis control practices almost in all districts studied. Figure 14 shows the number of adopters of mastitis control technologies among both technician and producer groups by district .


29 Figure 14 . Awareness, trial, and adoption of mastitis control technologies by districts 6.4.2. Discussion To summarize, r esults show that despite the high awareness of L FST and mastitis control technologies among producers, the rate of adoption (tried and adopt) was generally low. Majority of respondents who received training in L FST had not tried the app , although some of them had implemented the knowledge of feed ration balancing, they received from the training. Very few respondents had tried and adopted both the app and the feed ration balancing practices. The app functioned mainly as a support tool to formulate one ration for multiple dairy animals undertaken every time the new season came. In the case of masti tis control, the proportion of adopters ( who had tried and adopted) is higher than for the L FST ( 50% vs. 27% ) . Producers were more familiar with PMTD than other technologies. A few of them were aware of CMT and used it to detect sub clinical cases. Howeve r, the proportion of producers who dropped (i.e. initially used but then discontinued) the mastitis control technologies (mainly PMTD) was higher than L FST . The reasons for this discont in uance will be discussed in section 6.6.


30 For livestock technicians, not many had heard of, tried, and adopted L FST. But most technicians were aware of and had tried mastitis control technologies . Technicians used the app to give recommendation to farmers but most respondents in technician group did not work directly in fa rmer extension. For mastitis , some of them used the CMT in their work and recommend ed farmers to use PMTD . 5.5. Important Characteristics of Innovations 5.5.1. Results Respondents were asked about the advantages of adopting the innovations, the complexity of implementation, the compatibility with their needs and condition, the trialability, and the observability of results. The r esponse matrix o n whether the innovations could help improve production, improve animal health, save time, save physical labor, and save money, and if the results of using the innovation could be easily experienced and can outweigh the cost is summarized in Appendix 4. For L FST, a ll respondents who were aware of the app agreed that it can help improve production and animal health ( Quote 4 and 5 , Table 4 ) . For some respondents , the App c ould also help save time, save physical labor and save money ( Quote 6 ) . T hese benefits c ould be easily observed and outweigh the cost if farmers truly decided to adopt L FST . However, since not many respondents have observed and experienced the benefits of practicing L FST at its fullest, most respondents were not sure how much positive change th ey could get from using the app. Most respondents perceived the app as easy to learn, to try, to use, and compatible with the community but they also stated that the training was insufficient so they could not fully understand how to use it ( Quote 7 ). Such a contradictive response was often expressed by the same respondent, stating that L FST was simple to use but he/she would need more training or assistance in order to be able to use it. M was limited by financial and environmental factors ( Quote 8 ).


31 Table 4 . FST Theme Quotes District Reduce p roduction c ost (relative advantages) The app helps me make feed using locally available sources so I can save money from buying commercial feed. P reviously should give to my cows. It turned out that I gave too much [grasses]. Since that much [grasses] was not needed, it was waste d. A fter giving a nutritionally balanced feed, I can manage the amount proportionately. It helps me to be more efficient . Farmer, 30 ( Quote 4 ) Kapilbastu and Kaski Increase p roduction (relative advantages) The app helps increase production. In about ten to fifteen days after application, I observed an increase of 1.5 liter milk produced from four animals. So far, my cows are healthier and there is no longer an infertility problem. I also see an increase in fat content . Farmer, 47 (Quote 5) Kapilbastu and Kaski Reduce f eeding t ime (relative advantages) I can save time by making feed rations from the App. B efore using it, I used to feed my cows and buffaloes four times a day . B ut now only twice a day. I can also save the feed for days, day to cut and carry fodder . Farmer, 27 (Quote 6) Kapilbastu and Kaski Require m ore t rial and l earning (trial a bility and complexity) don' t have time to learn and experiment it. I will just buy a ready made feed. I might slowly make feed rations based on what I learn ed from training but not by using the Farmer, 3 5 aining was not enough s o I firstly tried with only one cow . In a half month, I saw a change. The cow has produced more milk . Farmer, 27 needs tutorial video or example so people can learn more . Cooperative manager, 38 (Quote 7) Chitwan, Banke, Bardiya, Kaski Require more resources (compatibility) Not many farmers have a smartphone and can use a mobile app . Farmers who only have one or two cows also need more money to procure the suggested is more suitable for commercial farmers Cooperative manager, 33 Some fodders and feed ingredients suggested by the app are not locally available nor affordable. I t would increase the cost of production if I decided to gather these materials. Farmer 43 (Quote 8) Chitwan, Kapilbastu, Bardiya R espondents perceived various benefits in using L FST such as reducing cost, increasing milk production, and saving time (Table 4 ) . These relative advantages were important characteristics of L FST that motivate dairy producers who had tried the App to fully adopt the practice . M ost respondents were aware of those benefits but did not feel compelled to adopt due to the complexity, trialability, and compatibility of the app . According to them, developing the skill to use and experience with the app we re important preconditions to adopt L FST . Hence, they emphasized the need for more and better training or access to extension agents that could assist


32 them in implementing the change s . Some respondents also believed that L FST was more suited to those versed in using mobile apps (usually younger people) and those with the means to make feed rations (commercial farmers) . In the case of mastitis control technologies, all respondents who were aware of the technologies agreed that the technologies could help improves animal health and that the benefit could outweigh the cost ( Quote 9 ) . Most respondents also perceived that PMTD was easy to be learn ed and could be taught to farmers ( Quote 10 ). However, they believed that the technologies did not save physical labor and time ( Quote 11 ) and were limited by financial and market conditions ( Quote 12 ). Based on Table 5 , producers were motivated to adopt mastitis control technologies due to their benefits of improving animal health and preventing mastitis. According to many respondents, ease to learn and implement (trialability) and availability of affordable technological inputs (e.g. equipment and chemicals) enable them to change practices and observe results. However, many also said that performing PMTD c an be time consuming and labor intensive, so some of them stopped using the technologies , and instead planned to wait for the first signs or symptoms of the disease. Table 5 . Characteristics of Mastitis Control Technologies Theme Quotes District Improves animal health (relative advantage) has been reduced. By applying PMTD, I no longer find a sign of mastitis on my cows and buffaloes. Milk production also improves. PMTD makes the teat softer than before, and there is no cracking in the teat Farmer, 45 (Quote 9) Bardiya and Dang Easy to learn and implement (complexity) n ot complicated. Easy to learn and to teach other s . It used to be hard for me to perform PMTD because I had to tie the legs of my cows, but now I can manage . Farmer, 23 Using CMT and PMTD is easier than L FST. Farmers can simply follow instruction s from be ing shown how to do it. With the app, they have to be literate and understand ration balancing Technician, 27 (Quote 10) Chitwan, Kathmandu, Bardiya, and Dang Consume more time and labor (disadvantage) A dd ing new steps in taking care of the animals i s time consuming . I have to spend more than half an hour to do teat dipping for ten cows every morning . What if I have more cows? I will need more labor to work on it . FGD 5 o n a regular basis. So, I stopped doing it and resume if I see signs of infection or a decrease in milk production Farmer 41 (Quote 11) Bardiya, Dang, Kaski


33 Require access to inputs (compatibility) glicerin for PMTD is also available in the local market. For CMT, I have have no problem Farmer, 30 W e have limited dis i nfectant. So, we save it until we see signs of infection . Buying from the pharmacy here is expensive, so we will wait for the cooperative to provide FGD 3 (Quote 12) Kapilbastu, Bardiya, Dang, Kaski Respondents often expressed different concerns across districts regarding the important characteristics that motivate or discourage them from adopting L FST and mastitis control technologies. For example, farmers in Kaski and Kapilbastu did emp has ize the benefits of using L FST because they ha d experience in trying the app, while farmers in other districts did not believe the potential benefits were compelling, despite recognizing them. Possibly, producers lacked incentives to uptake the innovation when the potential benefits of using the app were not directly observabl e and did not clearly outweigh the cost. In the case of mastitis, respondents in Bardiya and Dang gave more evaluations and feedback regarding the mastitis control technologies since they had experiences getting in volved in the pilot projects led by Heifer . These respondents indicated that the limited availability and access of technological inputs in the local market and cooperatives caused discontinuance in the practices , while respondents in Kapilbastu, Chitwan, and Kaski who were closer to urban areas expressed less concern regarding market access. Figure 15 . App Demonstrations (Left) Demonstrating L FST to cooperative members during FGD in Kaski; (Right) App demonstration to local livestock technicians in Bardiya, Nepal. (Source: Personal Documentation)


34 Regarding L FST, about 30% of interview respondents (11 out of 36) and five (out of 8) focus groups had very little opinion about the perceived characteristic s of the app due to their lack of awareness . In that case, re s pondents were asked if they were willing to have a brief app demonstrat ion . If they agreed , field assistant would teach them on how to use the app and the L FST would be s so they could try on their own during the demonstration . R espondents were then asked about their views on the complexity and compatibility of using the app, and if they consider ed using it after the meeting. M ost of these respondents expressed the need to know how to estimate the weigh t of cattle effectively and how the proportion of mixed ingredients worked (if the proportion of grasses was increased , would the volume of concentrates be reduced ). Some of them were interested to experiment in making feed using the ration from L FST at home . They would consider adopting it when they could prove the potential benefit s of using the app and the ease to formulate nutritionally balanced feed rations . 5.5.2. Discussion Most producers perceived that L FST had potential benefits to improv e animal health and milk production while reducing feed costs, but they did not believe in using the App as the best option available to achieve those benefits . Some producers had changed their feeding practices based on principles of ration balancing they learn ed from the training without using L FST because they thought that using it would be more complex and they could not afford to learn through trial and error ( Quote 7 ) . L FST was also perceived as a solution that was limited by the financial conditions and the farm environment by some producers not suitable for subsistence farmers and those living in area with less diverse sources and availability of fodders, grasses, and other feed materials ( Quote 8 ) . Th ese aspect s of the app that require d more learning / trial and resources becom e the most common reasons that prevented producers from adopt ing L FST. While such initial costs were expected in technology adoption, most producers


35 had not experienced and observed the relative advantages of adopting L FST. Hence, their cost and benefit analysis resulted in the perception of cost outweighing the benefits. In the case of mastitis, most respondents who adopted control technologies recognized mastitis as a problem in their dairy farm and believed that the y needed to avoid the risk of reduced milk production and revenue from the sick animal. The relative advant age of adopting the technologies became an important factor for those who ha d observed loss caused by mastitis and the benefit of the technologies to animal health ( Quote 9 ) . The benefits offered by adopting the technologies were supported by the ease to learn and implement the technologies and the access to technological inputs ( Quote 10 ) . However, the technologies were also perceived as time and labor consuming ( Quote 11 ) . This is understandable si nce recommended management practices require more time and resources. Thus , those who no longer perceived the existence of the risk of mastitis infection and faced limited access to technological inputs ( Quote 12 ), were more likely to stop using the techno logies. The perceived characteristics were differently expressed across district s and geographical locations . Distance to urban areas, to market s , and to government facility for livestock services appeared to influenc e the perceptions of the motivating and discouraging characteristics of the innovations. 5.6. Barriers to Knowledge and Adoption of Innovations 5.6.1. Results: Knowledge and Persuasion Respondents were asked whether they have shared information about the innovations (if they were aware), what might cause the lack of awareness and what can be improved to promote the dissemination of innovation s . Most respondents voluntarily use d word of mouth to recommend the innovations to some people in their social networks. Some others did not share th e information for various reasons. For L FST, t he most common reason was due to the lack of understanding o f how to use the app and the lack of experience in using it despite participating in the training ( Quote 13 ) . Given that only a handful of cooperative members were literate and


36 well educated, many respondents used it as a reason for not sharing the information about the app to other people and to explain why not many people adopted the app. M any also indicated tha t knowledge and adoption of L FST were limited by the access to extension services and local livestock technicians. Hence, most respondents highlighted the importance of improving the access and the quality of training programs as well as adding more lives tock technicians and extension agents at the village level. training was like watching a cinema only watch ing the instructor giving a presentation. It went very quick ly Farmer, 4 (Quote 13) Although training plays a major role in the level of knowledge and persuasion among producers , such concerns vary across districts. For example, producers in Kapilbastu did not express any training related issues . Proximity to the L FST pilot project location s m ight explain why more producers in Kapilbastu were found adopting the innovation than those in other districts. Since L FST was piloted in the same district , the feed ingredients listed on the App represented the type of locally available materials and producers might know someone involved in the pilot projects . Nevertheless, a well designed and well implemented training program would increase the rate of ado ption for both L FST and mastitis control technologies. Table 6 describes some constraints or roadblocks of knowledge dissemination and adoption of innovations. From the interviews and FGDs, most producer s who received L FST training believed that coopera tive officers were not suitable to use and disseminate the app to the members, especially when they did not have dairy animals themselves ( Quote 14 and 15 ). Involvement of local technicians and key players in the dairy sector was also perceived important to support the dissemination of innovations ( Quote 17 ). Apart from the knowledge gained from the training, re spondents specifically highlighted the lack of persuasion or provocation from L F ST training organizers and other corresponding agents to turn knowledge into practice ( Quote


37 16 ) . These findings will shape the recommendation to Heifer and partner organizations in Nepal in disseminating innovations to dairy producers. Table 6 . Improvement Quotes The training did not target the right agents. he learned from it. He might come to the training only because he was invited or maybe because (Farmers in FGD 10) tru st me about telling them how to do feed management, so I only hand them a copy of the (Coop manager, 50) (Quote 14) T he t rainee did not seem to value information sharing no such farmers in my (Farmer, 27) (Farmer, 42) and might not use any smartphone. Many of them are not (Coop manager, 49) (Quote 15) The training was not accompanied by follow up up after the training and no extension agents working in this area, so no one helps (Coop manager, 37) arise. So, I restrained myself from telling anyone about the App. I wished an interactive menu is (Farmer, 30) (Quote 16) The training lacks the involvement of key players. (Technicia n, 30) (Technician, 34) (FGD 1) (Quote 17) Besides training, some respondents specifically suggested the following strategies to improve the widespread promotion of L FST: Providing video t utorial or modules to help learning and teaching. Using mass media to introduce the app to the public. Broadcasting the training event so farmers who really need the training can access the information and participate in the program . Adding short messaging services (SMS) for consultation if there are any questions regarding innovation use.


38 5.6.2. Results: Decision, Implementation, and Confirmation L FST Mobile App Respondents were asked about the difficulties of trying L FST and implementing feed ration balancing as a routine if there was any modification ( reinvention ), and why they decided not to adopt or drop the practices . Many producers had technical problems, related with unsupportive systems to install and use the app (e.g. owning iPhone instead of Android, old Android version, full memory storage, and unknown error in the installation) and the download link given during training which was not working or difficult to use at that time . L FST became available on Google Playstore after the training for dairy producers ended. Meanwhile, when the training for technicians was conducted , the L FST app was already av ailable on Playstore. By design, the L FST app was developed to be tried easily. But, the unavailability of the app for easy download during the training for producers and the unsupportive phone systems seemed to hinder this trial process. The device related problem also became the main reason why one respondent in Kaski dropped the app ( as his phone was damaged), even though he continued using the same formulation from L FST to make feeds. In the broa der context of why producers were not likely to adopt feed ration balancing, r espondents explained that felt need and financial constraint s determine d the innovation decision ( Quote 18 ) . They believed that only commercial farmers would feel the need to invest in improving milk production while also having the means to fully adopt ration balancing practices. In a n FGD with dairy farmers in Kaski, participants explained that subsisten ce farm ers face greater financial challenges to adopting L FST so they would not think of putting efforts that will cost them more money . others are reluctant to adopt L FST? an office job as the one or two cattle; they do livestock farming just for home consumption and they rely on other farm


39 activities. Third, farmers who received training but not implementing the knowledge and utilizing the opportunity. Lastly, commercial farmers whose farm and livestock a re treated as the main source of livelihood and business opportunity; they will likely using a smartphone. Besides, subsiste not worth the efforts and it might increase their cost of production especiall y when certain feed ingredients time dairy farmers are not Regarding commercial farmers being more likely to adopt , most of them were not actually using L FST because they felt the potential hassle of doing separate formulation and making different feed ration for every animal across the many cows/buffaloes they had. For this reason, they preferred to buy already made feed or make their own feed formulation in a bulk amount. Both commercial and non commercial farmers also face d similar challenges of lack ing manpower and time as they were already busy with their daily activities ( Quote 7 ) . Across districts, the re were variations in the local availability of feed ingredients listed on the app . Th is did not seem to be an issue in Kapilbastu, probably due to its proximity to the pilot project sites, so they w ould likely have the similar types of fodders. R espondents in Kaski and Banke suggested that the L FST app should help differentiate and identify fodders that were available in each agro ecological zone. As indicated before, the availability of ration inputs was perceived among the most key obstacles of L FST ( Quote 8 ) . Several com munities in Kaski even believed that their villages were not suitable for livestock rearing due to scarce and seasonal availability of grasses and fodder trees. Despite the potential benefits of L FST in helping producers address such issues, those communi ties had already developed an unfavorable attitude toward feed ration balancing because of the perceived incompatibility of livestock farming with their environment.


40 Mastitis Control Technologies For mastitis control technologies, respondents were asked the same questions about the difficulties to try and use, modification ( reinvention ), and the reasons preventing adoption or causing discontinuance. T he common factor limiting the adoption and continuation of use of the technologies wa s the availability of technology input s such as test kits, dip chemicals, and medicine ( Quote 19 ) . Many respondents described that having access to affordable technological input is crucial for sustained implementation of mastitis control. Beside s access to tech nological inputs, the ability to purchase them, the intensity of time to implement practices o n a regular basis, and the limited access labor seemed to be the major barriers for adoption and sustained use ( Quote 11, 12 ) . Some producers perceived mastitis control practices as time consuming, so enough labor to work on it. Respondents in Dang expressed some of these factors as the roadblock s to adopting the technologies : Facilitator: What are the barriers you could identify that impede you or other people in your community from using mastitis control technologies? mers who previously didn't involve I think it is because it adds new steps in taking care of the animals. I got bored of doing it every day . I also to hire helpers . Besides, we also need to work for other things project. Now they them . Unfortunately, those materials are not available in the nearby market. Some stores may have them, but they are so expensive. I only when th is cooperative can provide .. . I think we are willing to pay f or the medicines and other materials we need if cooperative can provide. pay ing is okay but we also have to spend money on other things, so it adds the overall cost of production. Hopefully , they are still affordable. Besides access to the technological inputs, the change the perceived need for implementing PMTD when there was no sign of mastitis seem s to be the reason for discontinuance. Some respondents got bored of doing the management routine , and others


41 explained t hey would stop doing PMTD when the condition of their animals improved showing no sign of mastitis infection. Th is perceived lack of urgency may help explain why many producers in Chitwan and Kaski did not adopt mastitis control technologies if there was less prevalence of mastitis case s in their area . 5.6.3. Discussion: Comparing L FST and Mastitis Control Technologies The availability of information and access to extension services are crucial for dairy producers to acquire awareness and knowledge about innovation. Some of the responses seem to suggest that the lack of extension services and low quality training had constrained the dissemination of L FST and mastitis control technologie s ( Quote 7, 13, 19 ) . This is in line with other studies reporting that level and distribution of adoption were significantly determined by participation in training, extension input, and frequency of farm visits by extension workers (Floyd et. al. 2003, Ka fle 2011, Karki et. al. 2011). within relation to L FST, most respondents also mentioned the importance of literacy skills and education which again finds support in s reception and acceptance of agricultural innovation (Suvedi et. al. 2017, Tiwari et. al 2008). In some study sites, dairy farmers have limited access to government extension services, so peer learning and self learning were among the ways to gain knowled ge on innovation. Knowledge dissemination of L FST App and mastitis control technologies was also low due to a lack of active information sharing by those who received the training. O rganizers of L FST training invited cooperative managers to participate under the assumption that their strategic position in the organization would allow them to reach out to many potential adopters. However, most cooperative managers involved in the study explained that they were not suitable to teach farmers how to do feed ration balancing ( Quote 14 ) . Most of them also stated that they would share the knowledge about L FST App and mastitis control only if farmers c a me to them seeking the information ( Quote 15 ) . These responses indicate the importance of targeting the


42 right a gents in the training sessions and of explaining the value of information sharing so other people would learn about the innovations. Results also indicate that most respondents developed an unfavorable attitude toward innovation due to lack of follow up after the training. They said that reinforcement from training organizers or other agents would help to convince them to take up the innovations ( Quote 13, 16 ) . According to Rogers (2003), during the persuasion stage, individuals seek informatio n that reduce s FST, respondents were left on their own without further involvement from key stakeholders. When they were at the training, some producers had planned to use L FST , bu t eventually change d their mind when they got home . In contrast, respondents in the project areas of the mastitis control technology package said that active engagement with the project personnel had helped the adoption of new practices. Considering the im portance of this persuasion stage, it seems evident that training programs should be accompanied by follow up and should reinforce the involvement of key players. L FST appeared to have more constraints for uptake than mastitis control technologies , which ha s been tried and adopted by more actors in the dairy sector. The perceived higher benefit s, as previously discussed , may help explain this outcome . In agricultural production where yields and revenue are determined by numerous external factors, risk and uncertainty influence technology adoption (Menapace et. al. 2013, Rogers 2003). In this case, dairy producers appeared to be more averse to expected losses caused by mastitis (animal death) than from low milk productivity . Therefore, in places where ma stitis cases were prevalent, producers felt a greater more need to adopt the associated technology solution . M astitis control technologies can be considered under the lens of preventive innovation. According to Rogers (2003), this type of innovation has a particularly slow adoption rate because the desired outcome is not immediate. This explain s why some components of the technology


43 package (i.e. mastitis detection test CMT and milk conductivity test) were less known and less adopted than PMTD , which control s the disease. W hen there is felt need for addressing the prevalence of mastitis , producers will likely adopt PMTD and will observe immediate results from using the technologies. On the other hand, L FST (despite improving production) is a preventi ve innovation with delayed reward in which the relative advantages appear distant in time . Producers will likely observe positive changes when they are committed to try and integrate the use of the app o n a regular basis. Both L FST and mastitis control technologies face similar challenge of availability of technological inputs. This constrain t can be overcome by commercial farmers who have more financial resources and motivation to invest in improving and maintaining their dairy production through the adoption of innovations. Hence, being a subsistent farmer , with a small number of dairy animals c ontribute s to the factors limiting the adoption. However, in the case of L FST, commercial farmers with a la rge herd size appeared to be less interested on the app due to the reasons discussed before. Implications of these findings will be discussed in the following section. 6. IMPLICATIONS The findings of this study can contribute to improving the innovati ve design and the dissemination strategies to promote the widespread adoption of L FST and mastitis control technology in the Nepalese dairy sector. The lack of acce ss to extension services has been a major problem for dairy farming in Nepal. While government and extension providers continue working in expanding their services to reach more communities, the dissemination of innovations can focus on improving the training program s . This can be done in at least two ways: 1) improve the length and the number of practical session s during the training, as well as the follow up support to facilitate change; 2) better selection of training participants. Results


44 show that the training in L FST (and possibly mastitis training) was too short and lack ed practical sessions , which may have contributed to hinder the decision to adopt. This outcome is understandable due to the limited funding and human resources. For this reason , it is even more crucial that organizers ensure that trainees are appropriately recruited. For example, in L FST, the trainee s should be given early access to the app and to the supportive devices to try and use it. Furthermore, providing follow up support after the training ends could assist potential adopters to apply the knowledge learn ed in the training and develop a favorable attitude toward adopting the innovations. The finding shows that targeting cooperative managers may not be as effective in promoting dissemination. Many respondents suggested that training should mainly target farmers who are motivated and have the means to adopt. Identification of training participant s can also consider the number of livestock holding. In L FST, the app appea red to be most appropriate for commercial producers with medium herd size (5 15 animals). Targeting subsistent farmers to adopt L FST and mastitis control technology can be challenging due to their intrinsic and situational barriers to adopting new techn ology. Excluding them in the dissemination of innovations , on the ground since they are less likely to adopt may however widen the inequality gap between the rich and the poor. Therefore , policy related to promoting inclusive credit access and strengthenin g the role of dairy cooperatives for small scale farmers should be revitalized . In the pilot projects, Heifer had facilitated adoption of the two innovations inclusively among members of the community. In the case of L FST training, the participation of c ooperative members was intended to leverage the role of cooperative in disseminating the App. However, the intention was not clearly communicated and manifested in the actions taken by key stakeholders, so most trainees saw the adoption more at the individ ual level. One of the cooperatives in the study shared their plan to develop group forages and feed products based on the L FST training they received. However, they were struggling in executing the plan due to a


45 lack of financial capital and government su pport. On the other hand, some district level government offices have programs in feed creation by which individuals or groups of farmers can participate and compete to get grants to develop their feed. Unfortunately, such a program is still limited to cer tain livestock, mainly poultry, and was less popular among the majority of farmers. When the government can offer similar programs for cattle feed and the dairy cooperative can take the opportunity to participate and use L FST, it can promote group adoptio n. Future research in promoting and examining group adoption is needed. L FST mobile app and mastitis control technologies are useful innovations that can help improve milk yield and animal health and reduce the cost of production. Efforts made by various organizations such as NDDB, DLS, Heifer International Nepal, and local government in disseminating these innovations have created positive impacts in improving animal feed management practices and reducing the prevalence of mastitis among dairy producers. Decentralization has been the challenge in expanding and delivering extension services to dairy farmers. But this can also be an opportunity to better engage local technicians and farmers. Based on the implications discussed before, this report provides th e following rec ommendations for future training programs: 1) Training should target the right agents. A t horough selection of training participants based on the need assessment and likelihood to adopt and disseminate will help achieve the desired outcomes of the program. Promotion of the innovations will be more effective by targeting those who have the right instruments to try (e.g. android phone users) and the motivation to learn new things. Further research is needed to identify the characteristics of potential adopters who are also willing to disseminate the innovations to their communities. 2) Training should explain the value of sharing. As the number of training attendants will be limited, the trainee should be encouraged to share the information and knowledge they


46 learn from the training. In this way, the trainee will be reminded to use the opportunity of learning to gain a clear understanding of the innovations since they somehow have to s hare it with other people. 3) Training should be accompanied by incentives for local technicians to follow up. Persuasion phase after trainee learn about innovation is crucial in developing an attitude toward the innovation. Providing incentives for local tec hnicians to encourage and assist farmers in adopting new practices will help facilitate the uptake. 4) Training should reinforce the involvement of key players. Collaboration with key players at the local level will allow local stakeholders to have ownership of the innovations . This can help individuals or households in adopting new practices through better allocation of local resources. The other important aspect of promoting widespread adoption of L FST is improving the innovation design. It is possible tha t the lack of interactive features to help to learn may contribute to factors hindering the adoption of L tutorial or help button that explains the process and component used in the app to facilitate teaching and l earning among dairy producers. Adding a tool to multiply the rations with the number of animals owned can also help reduce confusion among producers. Many of them excused themselves from using L FST because the app can only do one formulation for one anima l. While users can always modify how they use the results of ration formulation, some producers may have undermined the potential benefit of the app for large scale production.


47 7. CONCLUSION This study supports previous theories and findings on factors influencing the adoption of agricultural innovations. While much of previous research ha s focused on the quantitative explanation about such factors, the qualitative approach allows this study t o gain contextualized information on aspects of innovation attributes and social systems that affect the innovation decision process . Th e findings are useful for programming an intervention t o meet the needs of the target audience by focusing more on raisi ng awareness and providing information or in facilitating the change through reducing constraints in the implementation stage. Identification of barriers impeding the dissemination and adoption of L FST and mastitis control technologies can also help organ izations develop specific measures to tackle the problems. Based on the results and discussion, this study concludes the following points to answer the three main research questions: 1) Awareness , trial, and adopti on of innovations. More than half of the respondents were aware of the innovations. About two fifth s of those aware of L FST tried the app but only one eighth adopted . Producers adopting L FST used the app once per season to formulate one ration for multiple cattle/buffaloe s and slightly modified the proportion of formulated rations. Among technicians, nearly half of the respondents (48%) were aware of the app, 50% of th ose who were aware had tried it, and 57% of those who had tried adopted the app. In the case of mastitis, four fifth tried the technologies but only half of them adopted. Most producers were familiar with PMTD and only a few knew about CMT. One fourth of those who tried PMTD dropped the practice. Among technicians, all respondents were aware of masti tis control technologies, almost all of them had tried, and about half of them adopted the technologies in their work. 2) Important perceived characteristics of innovations. The relative advantages, complexity, and compatibility of innovations appeared to be important aspects that determine adoption. In L FST, reducing production cost, increasing milk yield, and


48 reducing feeding time became the major motivation for producers to adopt. Meanwhile, the complexity of the app that requires more trial and training and the compatibility of resources available seemed to contribute to the low adoption of the app. In the case of mastitis, improving animal health by controlling and preventing mastitis was a major reason to adopt. The ease to learn and implement also support the adoption process. However, some respondents perceived the technologies as time and labor intensive and , in some cases, incompatible with the economic situation of the producers. These characteristics were perceived differently across districts. 3) Factors limiting knowledge gained and adoption of innovations. The lack of training and follow up support from extension agents were perceived as major factor s affecting the level of awareness and adoption of innovations . In L FST, m any producers were constrained by the inability to try the app due to device incompatibility and the familiarity with a mobile application. Limited availability of technological inputs , limited financial capability of subsistence farmers and the small number of livestock holding also impede d the adoption of both innovations. This study provides recommendations to improve t he training program and extension services in four ways: (i) targeting the right agents; (ii) promoting the value of knowledge sharing; (iii) providing incentives for local technicians to follow up; and (iv) reinforcing the involvement of key players . While identifying likely adopters as participants of the training is strategic to ensure diff us ion of the innovations, dissemination efforts should consider ways to include participation of marginal farmers in extension services and other government program s such as subsidy and access to credits. Recommendation to improve innovation design that en sure s trialability and allow independent learning is also included in this report.


49 8. LIMITATIONS OF THE STUDY AND SELF REFLECTION The field practicum has provided valuable experience s for me in learning about the field of international development and research. First, I learned about the necessary preparation to undertake international work and understand the limitation in conducting such research. In this study, it is difficult to compare the two innovations d ue to the fact that L FST was disseminated outside the pilot project areas through several training programs while mastitis control technology ha d not. This situation affected the selection of research participants who were recruited mainly among those rec eived the training in L FST. Comparison between producers who received L FST training in 2018 with technicians who received it in early 2019 was also difficult because both groups would have a different experience (e.g. whether the app has been available i n Google Playstore or not) and different time span to decide using it (6 months for producers and 3 months for technicians). Several respondents around the mastitis pilot project sites were included to accommodate the poor representation of actors who have received training outside the pilot project. Moreover, respondents are aware of some of the aspects of technology package (i.e. PMTD) from various sources beyond the USAID zone of influence. Therefore, the mastitis control technology that is previously se en as a package of mastitis control and detection technology could no longer be viewed as a package. Future research should develop better study designs that make different innovations more comparable, and the data collection more reliable. As a micro leve l adoption study researching specific communities using cross sectional data, there are limitations related to scalability and broader application level of policymaking. A doption behavior and factors influencing adoption in one community might differ from those in other communities and change over time, thus the use of panel data is more useful to study adoption and conduct meta analysis . Doss (2006) argues that cross sectional microstudies lack of dynamics (i.e. comparison between before and after farmers encounter a new technology) and lack of variation within samples which allow policymakers to get enough information about


50 access to information and the impact of credits or labor market failures. This study will not be able to address systemic factors affe cting adoption such as the provision and the quality of the education system, as well as market incentives in the dairy sector. Furthermore, I also learn ed about the complex challenges and complex systems in development work . Behavioral change takes time and resources, but many factors (e.g. funding, human resources, political systems, etc.) have constrained organizations in delivering successful intervention programs with long lasting impacts. As a student who is trying to learn and understand more about agricultural innovations for marginal farmers, I was not surprised to find that financial constraint played a major role in limiting the adoption. While many more innovations utilizing mobile te chnology are introduced through development projects for farmers, the effectiveness of apps beyond the project period is very susceptible to the quickly changing world and widen the inequality gap. When such an issue is not addressed properly, there seems to be endless stubbornness to change. Hence, it can be easy for a change agent to fall into a pro innovation bias and blame people for not changing. Finally, my field practicum was a very humbling experience for me. T o meet with respondents and partners fr om various professions that dedicate their time and skills to improving the condition of the people in the country , I kept thinking to myself what I could do to effectively contribute my skills and create positive changes . Knowing so little about the count ry I lived in for less than two months, I do admit my limitation in proposing recommendations in this report. I believe these experiences ha ve been very useful for me to help to navigate my future academic and professional carrier .


51 Bibliography Asian Development Bank Vol. 7, No. 1, Kathmandu, Nepal, 2019 assessing factors affecting the adoption of an agroforestry base d farming system in Dhanusha Agr Sys 89 : 2015: 645 661 DOI 10.1007/s10457 015 9802 1 Adoption of food safety measures on milk gate prices and profi Food Policy 70 : 2017: 13 26 DOI: Binod Kafle World Journal of Agricultural Sciences 7 (5): 604 606, ISSN 1817 3047, 2011. and Post 04/20/2011 th Los Angeles: SAGE Publications , 2017 Farmers America, 2006 Nepal Environment Protection Center (NEPC) NDDB, Kathmandu Nepal, 2014 Journal of Advances in A griculture 1071 1082, DOI: 10.24297/jaa.v7i3.6287, 2017 E.M. Rogers. Di usion of Innovations, 5th ed . New York: Free Press, 2003. Last modified October 20, 2014, Brigitte Pfluger and Padmakumar Varijaksha , Evaluation Mission Report 02: Nepal, Feed the Future Innovation Lab for Livestock Systems, May 2018. tion Lab for Livestock Systems. Accessed April 2019, keshav prasad sah/


52 hancing Dairy Animal Innovation Lab for Livestock Systems. Accessed April 2019, bhola shankar shrestha/ (Preliminary) Report. Government of Nepal, National Planning Commission, Kathmandu, Nepal, 2015 Floyd, Agricultural Systems 76 : 715 738, 2003 K.R. Tiwari Environmental Management 42:2, 2008: 210 222; DOI: 10.1007/s00267 008 9137 z L. Karki, R. Schleenbecker Journal of Agriculture and Rural Development in the Tropics and Subtropics Vol. 112 No. 2: 113 123 (2011) Linda Ng , Christine Jost , Misha Robyn , I.P. Dhakal , Bernard Bett , Pramod Dhakal , and Rupak Khadkha Preventive Veterinary Medicine 96 , 179 185. DOI : 10.1016/j.prevetmed.2010.06.012 American Journal of Agricultural Economics, 95:2, 384 389, Jan 2013 MALD (Ministry of Agriculture and Livestock Development), Statistical Information on Nepalese Agriculture 2073/74 (2016/17) , Kathmandu, Nepal. 72 80. Mi nistry of Population and Environment , Kathmandu, Nepal: 2017 Michigan Dairy Review Vol. 16 No.2 4/76581/cmt.pdf , April 2011


53 Ministry of Livestock Development, Livestock Statistics of Nepal, Planning, Monitoring, and Evaluation Division, Kathmandu, Nepal. July 2017. 3 10. Farmers participation in exte nsion J Ag Ed Ext 23:4, 2017: 351 371; DOI: 10.1080/1389224X.2017.1323653 National Dairy Development Board, Study on the Cost of Milk Production in Nepal , Hariharbhawan, Lalitpur, 2018 (12). The Sage Encyclopaedia of Qualitative Research Methods. (Vol. 2). Los Angeles: SAGE Publication, pp. 697 8, 2008 KR Da iry Global 14 Sep 2017 , trends/Articles/2017/9/Cattle farming in Nepal Why farmers cry 183891E/ accessed in February 2019 Samarth Nepal Market Development Programme, 2015 Food and Agriculture Organization, 2010 Seré FAO Animal Production and Health Paper No. 127, 1996 Sub Pro Poor Livestock Policy Initiative No. 48 , 2010 T B Thapa Agriculture Organization, UN Complex, Pulchowk, Nepal: 2010 15 USDA V. Braun & 101, 2006.


54 Appendix 1 Demographics of dairy farmers participating in the interviews Appendix 2. Demographics of Members of Producers Organizations and Technicians Producers Organizations Technicians


55 Appendix 3. Interview and FGDs Questions Guideline What are your concern s , if any, related to dairy production? What changes have occurred in the past three years related to dairy farming practices? Awareness Knowledge: Have you heard of x innovation? Yes How to and Principles Knowledge: How and what do you know about x innovation? No Why are you unaware of x innovation? Trial: Have you tried operating x innovation? Yes What motivates and helps you to try x innovation? No x innovation? Ut i lization: Are you using x innovation o n a regular basis? Yes What are your experiences in using x innovation? What benefits do you associate with it? No Never use. x innovation? Discontinue. Why did you stop using x innovation? Reinvention: Have you modified the use of x innovation? Yes How and why did you modify x innovation? No What can be improved from the design and use of x innovation? Perceived innovation characteristics: 1. Relative advantages 2. Compatibility 3. Complexity 4. Trialability 5. Observability 1 Does x innovation help improve production, save money, save time, and save the environment? 2 Is x innovation suitable for your needs and capability? Does it help address your problems? 3 How easy or difficult is it to use x innovation? 4 What is the trial process of x innovation? 5 Could you observe the results of using x innovation? Enabling factors and barriers to adoption What can be improved to help you adopt x innovation? What are the difficulties in using x innovation? Dissemination Have you shared x innovation to others? Yes How and to whom did you share it? No What prevents you from sharing the information about x innovation to others?


56 Personal Information Full name: Gender: Age: Main profession: Secondary profession: Highest level of formal education completed: # Members living your household: Approximate ages of household members: Livestock Information Please answer the following questions regarding your farm and farm activities # Livestock Total # Lactating: Who owns XXX animal? Who feeds XXX animal? Who cleans XXX Who manages XXX animal when sick? Approximate monthly income from XXX animal # Cattle owned (total): # Lactating cattle owned: # Buffalo owned (total): # Lactating buffalo owned: # Goats owned (total): # Lactating goats owned: # Chickens owned (total): # Producing chickens owned: List other livestock owned: # Other animals owned: Who owns the farmland? (i.e., In whose name is the land title?) Do you have access to saving and credit? If yes, in what form? What is your main concern or issue on your farm? What farm related activity or product generates the most income for your household? Have any household members participated in livestock related training? If yes, please briefly describe the training topics.


57 Appendix 4 Response Matrix on Perceived Characteristics of Innovations Benefit L FST Mastitis Control Technologies The innovation Farmers Organizations Technicians Farmers Organizations Technicians 1. Improve the quantity and quality of milk produced Yes In some cases Yes Adopters and nonadopters believe that nutritionally balanced feeding ration can improve animal lactating and its productivity Yes Many believe that controlling mastitis can help maintain milk production but are not sure if it can increase the quantity o f milk produced Improved animal health positively affects the reproductive systems and production of milk Adopters reported a varying increase of milk volume by 2 to 3 liters and fat content from 4 to 4.7 after 1 month of application No Some argue that increasing production require more measures than just controlling mastitis 2. Improve the body condition of dairy animals Yes Yes Many believe that improved nutrients in the feed will improve the health of dairy animals Many believe that mastitis control can prevent the animal from getting the infection and improve its metabolic systems Adopters report that dairy animals look healthier and less stressful 3. Save time In some cases Yes In some cases No Yes Animals being fed consistently twice a day in the proportionate volume can save time for feeding too often and too many grasses They compare using the App is faster than doing manual or excel formulation Yes It can save time from treating sick animals and going to vet They need to commit to the new routine of applying mastitis treatment and control No Many are not used to do feed ration balancing, so it takes time to operate the App and to collect, sort, and weigh all the feed materials No It requires extra time for each animal to get treated with PMTD 4. Save physical labor In some cases No No Yes Feeding management became more labor efficient since farmers are no longer feed animals too often and too much Making balanced ration feed requires more manpower, especially for those having much livestock Can be repetitive and tiring for some farmers, especially when they have many other things to do No Still need to collect, weigh, and mix different ingredients for every animal


58 5. Save money In some cases Yes In some cases Yes Yes Cheaper than buying commercial feed Can reduce the cost of buying concentrate and ready made product Yes Cheaper than the cost of treating a sick animal, paying for the vet, and loss of animal death Save money from treating sick animals, and prevent loss of animal death No Still need to procure many feed materials No Still need to buy the medicine and the kits The results of using the innovation... Farmers Organizations Technicians Farmers Organizations Technicians 1. Can be easily observed No Yes In some cases Users need to apply and conduct an experiment to see if the App can help improve production Results of the formulation can be immediately generated Yes Visible improvement on teat condition after some treatments of PMTD; Mastitis detection results can be seen using CMT No Sometimes require a vet to sample and diagnose diseases 2. Can outweigh the cost In some cases Yes Yes When the financial cost of making feed is lower than buying commercial feed The decline in production and the risk of animal death are more costly than controlling mastitis No When the feed sources suggested by the App are not locally available and farmers have only one or two cows


59 Table Perceived complexity, trialability, and compatibility of L FST and Mastitis Control Technologies among groups of respondents L FST Mastitis Control Technologies Farmers Organizations Technicians Farmers Organizations Technicians 1. Easy to learn No In some cases Yes The app looks complicated and requires proper training to understand how it operates and is used in practice Yes The App can be easily learned by technicians and farmers who are used to mobile phone technology Extension agents can simply demonstrate how to use PMTD and farmers can easily follow No Require certain knowledge on animal physiology and in operating mobile App 2. Easy to try No Yes In some cases Require a compatible smartphone; it was hard to download the App from the link given during the training User can simply download the App from Google Play and use it Yes Anyone can try applying PMTD No Trying PMTD depend on the availability of the kits 3. Simple to use/apply In some cases Yes Yes Yes The App can automatically formulate feed ration just by two to three clicking The app is simpler than excel software for animal feed formulation There are no technical difficulties in performing PMTD No Users need to measure the weight of the animal body and fodders, update market prices, and adjust the results with the number of livestock; difficulties to add new feed sources 4. Compatible in my area In some cases In some cases Yes Many use smartphones and have started practicing feed ration balancing Yes Farmers have been practicing treatment techniques to control mastitis e.g. cold/warm therapy No Some perceive that their area is not compatible for livestock rearing and that the ingredients listed on the App are not locally available No There is no prevalence of mastitis in some areas