Agent based modeling of payments for ecosystem services in Costa Rica – Adoption, dynamics and sustainability of silvopastoral practices

在拉丁美洲，永久性牧场上的养牛生产长期引起生物多样性丧失、温室气体排放增加和碳固存减少等方面的担忧。在过去，牧场生产力的下降导致了大规模的森林砍伐。生态系统服务支付（PES）已被确定为一种以市场为基础的机制，以激励重新造林和森林地区的保护。在这里，服务提供者因产生正外部性而由服务用户支付报酬。但是，森林的专属土地使用限制了土地所有者，降低了他们从农业生产中获得收入的能力。森林放牧的做法已经被发现增加了这两者；现场和非现场福利。在牧地种植乔灌木物种减少了水土流失，增加了生物多样性和碳固存。现场效益包括改善牧场，由于遮荫、木材、饲料和收入多样化而提高牲畜生产力。哥斯达黎加、尼加拉瓜和哥伦比亚率先实施了第一个森林畜牧业PES项目。2002年至2007年间实施的区域综合森林生态系统管理项目（RISEMP）被认为是一个巨大的成功，因为它带来了超过12000公顷的牧场，改善了生物多样性和碳封存。然而，该计划的可持续性仍然存在疑问，因为它需要有限的支付，这被假设为“打破平衡”。研究建议的第一个目标是阐明哥斯达黎加为森林收养所支付的款项的可持续性。直到现在，在停止支付十年之后，缓慢生长的树木和灌木物种所产生的全部效益才得以观察到。生态系统服务提供商的长期采用状态也是如此。第一个目标将通过补充二手数据的住户调查来实现。在项目期间和结束后收集的数据将允许调查一段时间内的采用状态和保护强度。第二个目标是确定技术采用的过程。最近的研究强调，非金钱驱动因素和沟通途径对采用很重要。也就是说，规范、有限理性模型和在社会网络中的地位与金钱激励相互作用。基于代理的模型已经证明了它们复制后者的能力。因此，第二个目标将通过开发和分析基于agent的PES采用silveropastoral实践模型来实现。在此，将通过模拟模型研究支付水平、保护强度和长期可持续性之间的比率。这些比率将在农业结构和社会嵌入发生变化的情况下进行检验。此外，对不同社会网络拓扑结构的分析将有助于理论的发展。该模型将用经验数据进行参数化。

Cattle production on permanent pastures in Latin America causes prolonged concerns with respect to losses in biodiversity, increased greenhouse gas emissions and reduced carbon sequestration. Decreasing pasture productivity has led to massive deforestation in the past. Payments for ecosystem services (PES) have been identified as a market based mechanism to incentivize reforestation and conservation of forest areas. Here, service providers are paid by service users for producing positive externalities. However, the exclusive land use of forestry constraints land owners and reduces their ability to generate income from agricultural production. Silvopastoral practices have been found to increase both; on-site and off-site benefits. The inclusion of tree and shrub species in the rangelands reduces erosion, increases biodiversity and carbon sequestration. On-site benefits are e.g. improved pasture, increased livestock productivity due to shade, timber, fodder and income diversification. Costa Rica, Nicaracua and Colombia pioneered the first PES program for silvopastoral practices. The Regional Integrated Silvopastoral Ecosystem Management Project (RISEMP) was implemented between 2002 and 2007 and was considered a huge success as it resulted in over 12000 ha of pastures with improved biodiversity and carbon sequestration. However, doubts remain on the sustainability of the program as it entailed finite payments which were hypothesized to “tip the balance”. The first objective of the research proposal is to elucidate the sustainability of payments made for silvopastoral adoption in Costa Rica. Only now, ten years after payments have ceased, the full scope of benefits arising from slow growing tree and shrub species is observable. The same accounts for the long-term adoption state of ecosystem service providers. The first objective will be approached with a household survey complementing secondary data. Data gathered during, and right after the end of the project, will allow investigating adoption states and conservation intensity over time. The second objective is the identification of processes of technology adoption. Recent studies emphasize that non-monetary drivers and communication pathways matter for adoption. That is, norms, models of bounded rationality and the position in social networks interact with monetary incentives. Agent based models have proven their ability to replicate the latter. Thus, the second objective will be approached by the development and analysis of an agent based model of PES adoption of silvopastoral practices. Here, the ratios between payment levels, conservation intensity and long-term sustainability will be investigated by the simulation model. The ratios will be tested in scenarios of a shift in farm structure and social embededdness. Moreover, the analysis of different social network topologies will contribute to theory development. The model will be parameterized with empirical data.