Original ArticlesAssessing the relationship between ecosystem functions and services: Importance of local ecological conditions

- We evaluated the importance of local ecological conditions in the relationship between EF and ES through integration matrices.
- Hotspots of ES supply varied according to the matrix used, with possible consequences in land-use decision making.
- Expert knowledge elicitation techniques were useful for summarizing knowledge in the growing field of ES.
- Future studies should take into account non linear relationships between EF and ES which is also possible with expert elicitation techniques.

The development of standard procedures for mapping ecosystem services (ES) hotspots is an active research issue, mainly because it is very important to provide spatially reliable information to decision makers on land-use planning. When ES supply is estimated and mapped from sets of contributing ecosystem functions (EF), it is important to identify which of them contribute the most to a particular ES. Also, it is crucial to determine whether that contribution varies across different ecological contexts. In this work, we used an expert-knowledge-elicitation approach to build “integration matrices” that resume the relationships between EF and ES in four different ecoregions of southern South America (Valdivian Forests, Campos, Chaco and Pampas). To verify the utility of creating different matrices, we compared the results of the Pampas ecoregion with an average score matrix from the four selected ecoregions. Results showed that the contribution of EF to ES differed slightly using the Pampas and the general matrix. We also mapped hotspots of three ES in a real landscape from the Pampas, using both matrices and the ECOSER protocol for ES supply mapping. The location of ES supply hotspots differed between matrices, but the amount of discrepancies was not proportional to the differences in matrices scores. This suggests that the developed matrix that takes into account local ecological conditions provides a more reliable source of spatially explicit information on ES supply for decision makers than a general matrix. The integration matrices generated in this study are useful for different protocols of ES supply mapping.