IENE 2016

Land transport infrastructure (LTI) projects, such as HSR projects, produce numerous impacts on biodiversity at many scale-levels throughout their development (land-buying, construction, exploitation etc.), and both strategic-impact studies and the scientific literature already show impacts at fine scale: roadkill, pollutions, barrier-effects etc. Models have been developed through geomatics to evaluate the diffusion of direct, indirect, cumulative or induced LTI impacts: some of them conclude that global landscape change may impact biodiversity, others that the barrier-effect e.g. should be observed till more than ten kilometres right of the line –depending on what species is impacted. It is chosen to analyse land-use dynamics in such spatial areas around HSRs, in a suburban-rural context, from the 47 land-use-types digital map Mode d’occupation du sol d’Île-de-France (Mos-IdF) realised regularly for seven times between 1982 and 2012 in the Paris region (France).

Methodology consists, according to distances of impact diffusion observed in the scientific literature, in calculating the frequency-of-change for each landscape patch in 5100m both-sides buffer-areas (except a thousand meters-large central strip) around HSRs, among six potential changes (seven observations) given by the Mos-IdF within thirty years. Average landscape-change frequency is then calculated for groups of patches selected by their "type" of land-use. Landscape-change frequencies at such "land-use" scale are finally compared to the age of closest HSR. Land-use type at patch level is defined as the most-frequent-in-time type among the seven regularly interspaced mappings. With help of spatial-join and distance tools, the age of HSR, that every landscape patch is closest to, is integrated to data, aswell as the distance (meters) of every patch to its nearest HSR. Geographic information system (GIS) and statistic softwares (ArcGIS™, Excel™) are used in this study.

Results show very few variations for landscape-change frequency at near or far distance from HSR. They also show some variations between single project areas and areas impacted by two different projects. In detail, two results can be observed. 1/ Frequency tends to decrease for some land-use types (greenhouse, glade, vacant land or logistic warehouse) according to the seniority of the closest HSR. 2/ On the contrary, frequency tends to increase for some other types (orchard, home garden, suburbanization or store) according to the seniority of closest HSR. These results suggest that HSR projects would tend to "fix" some land-use types in their surrounding landscape, while they would "mobilize" other types within the landscape change, so that they may impact diffusely biodiversity.

The willing interest of our study is to show how watching landscape diffuse change (in frequency; on fixed geometry) and dominant-over-time land-uses may help mitigating HSR impacts at long-term on biodiversity.