Spatial-social evaluations of ecosystem services of adaptive aquaculture models using SAR and multivariate analyses: a case in the Vietnamese Mekong Delta.

Abstract

The presented study is conducted to investigate the efficiency of two important aquaculture models of the Vietnamese Mekong Delta (VMD)'s Soc Trang province via quantifying and mapping the supporting ecosystem services (ES). The study targets the two most prevalent rearing practices, intensive and semi-intensive, covering four rural districts: My Xuyen, Tran De, Cu Lao Dung, and Vinh Chau. A mixed-method approach was applied, combining remote sensing, grass-root social survey, and multivariate statistical analyses. First, image analysis using Sentinel-1A time-series data was conducted to detect the aquaculture areas across the study area based on temporal changes of VV backscatter of different land use/land cover (LULC) types, in which aquaculture receives relatively low backscatter values compared to other LULC categories except river and deeper water surfaces. Our analysis yields an overall accuracy of 91% with a kappa coefficient of 0.82. Second, using semi-structured questionnaires, a total of 140 shrimp farming households across the four focused districts were interviewed for their rearing experience. Thereupon, the collected responses were analyzed using two multivariate analyses, including principal component analysis (PCA) and hierarchical cluster analysis (HCA). In general, the intensive model could have generated more economic values of ecosystem services than the artisanal/semi-intensive model. Our analyses also took note of the potential barriers hindering the semi-intensive farmers from achieving higher economic income. These include (i) geographical factors, i.e., locations; (ii) social factors, i.e., experience, farming calendar, education; (iii) financial factors, i.e., investments; and (iv) technical factors, i.e., farm areas, productivity, rearing concentration. Since semi-intensive food is more appropriate for small-scale farming households, it is recommendable that addressing these factors can enhance the efficiency of this model as a profitable livelihood option.