This is a long standing problem & there are a number of existing approaches:
- I've used landuse=wood_pasture once for Wood pasture is the more generic English language term for Dehesa-like areas, As such areas exist in other parts of Europe it might be better to be less specific. Wood pasture exists in Galicia, Austria, Baltic States, England and no doubt elsewhere. Dehesa would be an inappropriate tag for these more northern wood pastures. Oliver Rackham covers some of this distribution in Woodlands, but also in his chapter on Dehesa in The Nature of Mediterranean Europe: An Ecological History. Mansey Common an area of relict wood pasture in England.
- Also there are also tags meadow=orchard and orchard=meadow. See https://wiki.openstreetmap.org/wiki/Tag:landuse%3Dmeadow_orchard for a discussion. I've never been clear if the meaning of these tags is suitable for Dehesa.
- Also 250 or so uses of natural=grassland with grassland=dehesa. In general Dehesa is closer to woodland in many attributes than grassland, but this has been used to code some CORINE data.
Other potential terms include agroforestry, silvopasture, etc. As I say I prefer wood pasture because it is not only used in English but should be readily understood. SK53 (talk) 10:54, 30 August 2019 (UTC)
Wood pasture not adequate
Wood pasture does not include the use of a dehesa for agriculture, and the same is true for grassland and meadow=orchard. After discussing my proposal through mailing lists, I'm currently considering changing my proposal to landuse=agroforestry, since that term wouldn't exclude any of the three mixed usages of a dehesa and it would also include similar land types outside the Iberian Peninsula. I'm currently waiting for feedback from other users.
- All the dehesa I'm familiar with in Extremadura is predominantly used for pasture, and certainly not for relatively intensive crops such as cereals (which may be related to poor soils). Of course the land will be ploughed periodically to prevent build-up of a scrub layer if it is not grazed and most areas I've seen have a ploughed area as a fire break. Of course there are some areas of relict dehesa with only a few trees and thus low (<5% tree cover) which can be used for crops. A sample of my photos of various bits of dehesa I surveyed in November 2016 are here on Flickr. There is one example of this latter relict dehesa form, which was being ploughed when I visited. I think if agroforestry is used then we should subtag it to show that it is dehesa and/or wood pasture. There is certainly something in discriminating boreo-temperate wood pasture from its Mediterranean equivalents. Note I also purchased the Spanish National Forest Maps for parts of Extremadura as part of my effort to examine dehesa more closely, but I never did a final analysis because of events in the family. SK53 (talk) 18:46, 4 September 2019 (UTC)
- @Seichter, the problem is that Dehesa really is definitely not either a meadow or farmland in all the classic OSM uses. No one can crop the grass as in a meadow because of the trees and many Dehesas have a scrub layer not suitable for pasture at all). In practice Dehesa is a form of woodland, but an unusual one sustained by man as well as animals., although there have been arguments (Vrba) that before mankind arrived in Europe, original woodland was closer to Dehesa (effect of large ungulates: elephants, rhinos, auroche etc). Although Dehesa and other wood pasture is now a relatively small proportion of European landcover I don't think we should reject it: we could tag glaciers as natural=water water=glacier on the same basis. SK53 (talk) 13:00, 21 November 2020 (UTC)