Abstract :
[en] As the rapid development of intensive animal husbandry industry, vast amounts of wastes are generated, of which, the main contributor to cause huge potential impacts on environment is manure, along with sewage (including feces, urine and wastewater). However, a high diversity exists in manure and sewage management (MSM) and has rarely been classified at a scientific and statistical level, thereby leading to complexity and difficulty in policy and sustainable evaluation. Concerning this issue, this study attempted to establish a quantitative typology to simplify MSM diversity. Taking Chinese dairy MSM strategies for example, we explored a MSM typology using the nationwide survey of 306 scale dairy farms in China. Seven well-known categorical clustering algorithms were implemented. The validation of clustering performance, i.e., clustering tendency analysis, six internal cluster validity indices and ranking method, were also conducted to recognize meaningful clusters and determine a suitable clustering algorithm contributing to reflect the real data structure. The clustering tendency analysis indicated the data highly clusterable with significant clusters. Furthermore, the ranking results showed that COOLCAT algorithm obtained the local optimal clustering performance and cluster number was verified to be four. Correspondingly, Chinese dairy MSM strategies were classified into four representative types based on six MSM variables involving various technologies or practices in collection, storage, processing and application stages. The typology we established could rapidly select representative types, deliver the outcomes to stakeholders straight forwardly, and make the best use of all available information. More importantly, it could capture the diversity of MSM strategies at national level, and support the mathematical-modelling evaluation of policy and sustainability at a higher scale.