Search

In recent years, the assessment of ecosystem services (ES) supply has been based on the use of Land Use/Land Cover (LULC) data as proxies for spatial representation of ecosystems. Nevertheless, some shortcomings of this method, such as uncertainties derived from generalization of the ecosystem types and assumptions of invariance across spatial scales, indicate the need for new approaches. Such approaches could be aimed at improving knowledge of the relationships between ecosystem services and landscape structure and the spatial characteristics of ES patterns. In this study, we propose an integrative approach that involves the generation and analysis of continuous maps representing the supply of five ES potentially related to the amount of biomass. Five remote sensing images of the Northwestern Iberian Peninsula, obtained with Landsat-5 TM, were used to generate a proxy for net primary production by combining the normalized difference vegetation index (NDVI) of each image to calculate a ΣNDVI index that could act as a potential indicator of some ecosystem services. This information was combined with three variables – terrain slope, population density and occurrence of protected areas – to produce spatial models for the five ES and eventually a series of five supply maps. Food, materials and energy provision services showed a clustered pattern, with high supply values in flat zones and areas with high population densities. In contrast, mass flow and climate regulation services were more widely distributed throughout the study area. The five ecosystem service patterns were analyzed at different scales by two methods: lacunarity and four term local quadrat variance (4TLQV) analysis. These methods revealed differences in the spatial pattern: lacunarity analysis was useful for detection of scale thresholds at the local level, whereas 4TLQV was more sensitive to scale thresholds at larger spatial levels. Thus, the variance analysis yielded higher values for larger windows sizes, particularly for provisioning services. The results demonstrated the suitability of the proposed approach for the spatially explicit modeling of ecosystem services, avoiding the uncertainty of other assessments such as those based on LULC data, and for the exploratory analysis of ES supply from a spatial point of view.

Ecosystem services research has become important in fields such as ecology and land planning in recent years. Several authors have emphasized the need to evaluate these services from a spatial perspective. The spatial distribution of the resources and processes that provide ecosystems have also been considered in relation to landscape. Because ecosystem services are provided by different ecological processes, the spatial scales probably also differ; this aspect may be of interest for analyzing the provision flow. In this study, we analyzed the spatial pattern of six services that are important in the study area (in Galicia, NW Spain). We first identified the landscape elements associated with these services, defining them as ecosystem service providers (ESP). To represent the ecosystem, we initially used cartographic information based on land use/land cover (LULC) and then generated two different raster ESP data sets: (i) binary and (ii) greyscale. We then explored the spatial patterns of ESP by lacunarity analysis, which is often used to study fractal elements, and by selecting landscape metrics. The results suggest that the spatial patterns of ecosystem services occur at different scales. We observed a strong relationship between lacunarity values and the different distribution patterns of ESP. Multiscale effects were also associated with changes in lacunarity values. Application of different spatial analysis techniques to study the relationships between landscape structure and service providers should provide a better understanding of service provision and enable evaluation of the ecological integrity of landscapes.