Plant species distribution survey and its explanatory variables

<p>Ten common herbaceous species were selected based on a survey of a 6.4-km² upstream catchment named « Bourdic » in southern France. Approximately 74 % of the catchment is agricultural (mainly vineyards), and 26 % is semi-natural (mainly woodlands and shrubs). The catchment has a Mediterranean climate with heavy rainfalls causing significant Hortonian runoff. The mean annual temperature is 14°C, and precipitation ranges from 600 to 800 mm per year with a drier period from March to October. Annual potential evapotranspiration is about 1100 mm. The altitude ranges from 55 a.s.l. at the outlet at the northeast to 128 m a.s.l. at the northwest.</p> <p>The surveys were conducted in July-August 2013 according to a non-destructive sampling procedure using GPS with an Android self-developed application; this enabled a location accuracy of 2 m. Agricultural ditches, including roadside ditches, were part of the study. Thirty-five kilometres of the drainage network (46%) were surveyed for presence/absence of the species. The remaining ditches were excluded from the analysis because surveying them was impractical or because recent management practices impaired species identification. After the survey, the georeferenced data were exported in a shapefile data format with line features.</p> <p>Also, explanatory variables of the dataset were reported, such as the geomorphological variables at the landscape scale that included the distance to the outlet (<strong>Doutlet</strong>), the drained surface area (<strong>Drain</strong>), the Multiresolution Index of Valley Bottom Flatness (<strong>Mrvbf</strong>), and the sun exposure of the slopes (<strong>Northness</strong>). The geomorphological variables at the local (ditch) scale were the slope (<strong>Slope</strong>) and solar radiation (<strong>Solar</strong>). All these variables are derived from a Digital Elevation Model (MNT) and a Digital Surface Model (MNS) taken in 2001 using an aerial lidar.</p> <p>We added also the distance to natural lands (<strong>Dnat</strong>) and distance to roads (<strong>Droad</strong>) on the basis of the manual classification of an orthophoto of the area taken in 2012.