Over the last decades, the interest in mapping Posidonia oceanica beds has increased along with the improvement of the equipment’s precision of data acquisition. In Calvi Bay (Corsica, France) the meadows cover an area of about 5 km² and are found at a depth ranging from 3 m to 37 m. The availability of three distinct datasets for 1997, 2002 and 2010 allowed to assess changes in the patchiness of the meadows in the bay and to investigate evolution of maps precision through a surface analysis via GIS software. Thus, three maps were elaborated combining aerial photographs and side scan sonar images. The meadows percentage of cover through time was assessed using four bathymetric sections: 0-10 m, 11-20 m, 21-30 m and 31-40 m. Differences in the patchiness of P. oceanica meadows between 1997 and 2010 appear to be moderate (less than 3 %) in the sections 0-10 m and 11-20 m and then greatly increase with depth: 24 % at 21-30 m and 39 % at 31-40 m. This amazing regression seems hardly natural and unlikely given the slight quantity of human activities that can cause damages on the P. oceanic a meadows of the Calvi Bay. These results are likely to be mainly due to the improvement of precision and resolution of the aerial photographs (5 m in 1997, 0.8 m in 2002 and 0.5 m in 2010) and sonar images (5 m in 1997, 3 m in 2002 and 0.5 m in 2010). An issue of habitat determination (human vs instrumental) linked with the method adopted for mapping can also cause differences in the percentage of cover. Given the different accuracy among the three maps, the real regression and fragmentation of P. oceanica meadows could be hardly assessed. However, in several areas where the human activities are important, a clear regression or even a disappearance of the meadows has been observed. It is obvious that the last maps are more accurate than the previous ones and, thus, the former can be used for management purpose as well as for study on the patchiness; however, they still keep uncertainty no matter which method is used to create them.
Abstract Effective ecosystem‐based management requires understanding ecosystem responses to multiple human threats, rather than focusing on single threats. To understand ecosystem responses to anthropogenic threats holistically, it is necessary to know how threats affect different components within ecosystems and ultimately alter ecosystem functioning. We used a case study of a Mediterranean seagrass (Posidonia oceanica) food web and expert knowledge elicitation in an application of the initial steps of a framework for assessment of cumulative human impacts on food webs. We produced a conceptual seagrass food web model, determined the main trophic relationships, identified the main threats to the food web components, and assessed the components’ vulnerability to those threats. Some threats had high (e.g., coastal infrastructure) or low impacts (e.g., agricultural runoff) on all food web components, whereas others (e.g., introduced carnivores) had very different impacts on each component. Partitioning the ecosystem into its components enabled us to identify threats previously overlooked and to reevaluate the importance of threats commonly perceived as major. By incorporating this understanding of system vulnerability with data on changes in the state of each threat (e.g., decreasing domestic pollution and increasing fishing) into a food web model, managers may be better able to estimate and predict cumulative human impacts on ecosystems and to prioritize conservation actions.
The meadows formed by the Mediterranean seagrass Posidonia oceanica are subjected to various natural (e.g., water movement, light availability, sedimentation) and anthropogenic (e.g., anchoring, trawling, fish farms, explosives) phenomena that erode them and create diverse types of patches. The assemblage of the P. oceanica matrix and these patches creates particular seascapes. On the basis of this assessment, we aimed to investigate the importance of the patch type in structuring P. oceanica seascapes and to offer new prospects in the large scale studies of seagrass meadows. Five sites encompassing large P. oceanica meadows ranging from 1.86 km² to 4.42 km² along the Corsican coast (France) were considered. Eleven patch types with different sizes, shapes and origins were identified using side scan sonar images (sonograms). Five were recognized as natural and five as anthropogenic. One can be of both origins. The resolution of the sonograms allowed to detect patches of various sizes ranging from 1 m² to 111 829 m². The relation between structural characteristics of patches and the whole seascape aspect was explored using seven landscape metrics relevant for the study of meadows patchiness (patch area, mean radius of gyration, area-weighted radius of gyration, coefficient of variation of the Euclidean nearest-neighbor distance, area-weighted perimeter-area ratio, landscape division index, number of patches). Only a small number of patch types appears to play the strongest role in the characterization of the P. oceanica seascapes. Furthermore, the use of seascape structures seems to be suitable for the development of new tools like indices for the assessment of human impacts on P. oceanica meadows. In this perspective we propose a new and simple index, the Patchiness Source Index (PaSI), to estimate the origin of the patchiness (natural or anthropogenic) for a given area. A landscape approach, as well as information on patch dynamic, should be integrated in the new indices that aim to assess the state of conservation of the whole P. oceanica ecosystem.
