Agricultural nonpoint source pollution is recognised as a major cause of water pollution. The characteristics of nonpoint source pollution suggest that an efficient approach should focus on a source control and hence land-use management. Recently, the concept of Payments for Ecosystem Services (PES) has been advanced as an efficient market-based approach to protect in-stream water quality, while simultaneously supporting agriculture. Farmers can be rewarded for the adoption of certain ‘Best Management Practices’ (BMPs) in farming systems. But little is known about the adoption of BMPs in the context of Thai agriculture. This thesis examines the adoption of twelve BMPs on citrus farms in the Ping river basin in northern Thailand. In the context of potential PES development, three studies were undertaken using frameworks from economics and psychology. The first study used a bottom-up engineering approach to estimate economic costs of twelve BMPs at the farm-scale. The total annualised costs, including installation, maintenance, and land opportunity costs of each BMP were compared. The results indicated that land opportunity cost was the largest proportion of total costs. These estimates provided a basis for discussion on how the farmers’ perception of cost may influence their stated adoption intention. The second study used the Theory of Planned Behaviour (TPB), to investigate farmers’ intentions on adoption of twelve BMPs. A survey of 218 citrus farmers was undertaken in the application. Descriptive statistics and frequency of BMP selection were presented. The results showed that soil analysis was the most preferred BMP. A graphical analysis of other preferred measures suggested that these were not always consistent with the known cost information derived (above) and thus farmers’ perceived costs might not be the most important factor influencing adoption decisions. Further analysis based on TPB investigated other factors thought to be significant in farmers’ decision-making. Other potential external and psychological factors influencing adoption were investigated using a multinomial logistic model. The results indicated that the probability of adopting BMPs was associated with other psychological factors and external factors, rather than perceived costs. The significant psychological factors were farmers’ attitudes towards consumers and perception about farm returns, while the significant external factors were, for example, access to information and contribution of family labour to farm workload. The third study was based on the application of Q-Methodology, and aimed to obtain a deeper understanding of farmers’ perception towards BMPs. Seventy two participants were purposively selected from the 218 TPB observations. The results revealed four distinctive farmer groups holding different perceptions towards BMPs. The four groups were conservationist, traditionalist, disinterested, and risk-averse. These provided a specific segmentation to guide policy towards influencing attitudes and behaviours. The results suggested that farmers were not motivated solely by a profit maximisation goal. Overall, key findings from these three studies revealed some fundamental requirements for developing a water-related PES programme. These were: i) factors affecting eligibility to participate; ii) factors affecting desire to participate; and iii) factors affecting ability to participate. This information provided the basis for a set of recommendations addressing the development of the water-related PES programme in the Ping river basin.

For over 30 years, federal and state governments have been engaged in a collective effort to improve the water quality and living resources in the Chesapeake Bay (CB), focusing particularly on reducing delivered nitrogen and phosphorus loads. However, achievement of water quality objectives remains elusive. In Virginia, agriculture represents the single largest source of nutrient loads to the Chesapeake Bay. Despite aggressive regulatory efforts in other nutrient source sectors, state authorities rely on educational programs and voluntary financial assistance programs to induce landowners to adopt best management practices (BMPs) that reduce agricultural nutrient loads. This study explores two economic aspects of efforts to reduce agricultural nonpoint source (NPS) pollution in the Virginia portion of the CB watershed. Firstly, current and possible future transactions costs associated with specific aspects of agricultural NPS participation in water quality trading (WQT) programs are examined in Chapter 1. A case study approach is used to consider the possible cost consequences of expanding the use of NPS credits from agricultural 'working lands' BMPs in Virginia. Findings indicate that overall transactions costs for nutrient trades involving agricultural NPS in Virginia are currently relatively low, due to the type of activities being credited: simple land conversions. Based on best available evidence, the administrative transactions costs of creating credits on agricultural 'working lands' using management and structural BMPs will be 2 to 5 times more costly on a per project basis than for credits generated from land conversions. Compliance monitoring protocols were found to be a significant driver of costs for credits generated from working agricultural lands. These results suggest an important cost/risk tradeoff between verification cost and compliance certainty for program designers to consider. The second study (Chapter 2) considers the economic cost of meeting pollution reduction targets for the Virginia portion of the CB Watershed. Existing cost models are based on simplifying behavioral assumptions about public transactions costs, conservation adoption rates, and implementation costs of agricultural BMPs. This study builds on the existing literature and uses the estimates of transactions costs from Chapter 1 together with information on producer BMP adoption rates to examine the implications of including transactions costs and differential BMP costs and adoption rates when estimating the minimum costs of achieving specified nutrient reduction goals in Virginia. The paper uses a cost-minimizing mathematical programming approach and models a number of different cost scenarios. Results indicate that inclusion of transactions costs substantially affects estimates of total costs of meeting nutrient reduction goals; on average total costs increased by 44 percent, but ranged between 19 and 81 percent depending on the scenario analyzed. Analysis of the modelled scenarios shows that those BMPs that account for the most implementation costs do not necessarily account for the most transactions costs (and vice versa). This suggests that transactions costs should be acknowledged to vary with the type of practices being implemented, rather than being approximated as either a fixed amount or a fixed proportion of implementation costs. In addition, the analysis highlights the disproportionate costs associated with achieving nutrient reductions via high-cost adopters, and suggests there may be a role for education or extension to assist landholders to lower opportunity costs of participating in conservation. / Master of Science