As Ornithologist, in the study of bird ecology o for modelistic purposes, the association maps represent clearly too detailed effort in terms of useful data. And, considering the main bird species, are very, very, "most useful" the land cover traditional maps. Because for animals, several times, is more important the structure rather that detailed information about vegetal species and vegetal associations. Furthermore, the association concept is too imprinted of "human kind" for me.... But I'm Ornithologist, not a phytosociologist! And repeat: is my humble opinion, as not expert in vegetation sciences.

Response to Morelli: Land cover maps are sometimes similar to Formation maps without the foundation subdivisions (alliances and associations) that may be of importance to smaller or less mobile creatures than birds. Perhaps we could enrich the categories currently used (natural, semi-natural, and cultivated) by dividing semi-natural into two categories, degraded and severely degraded.

Yes, I agree.

Everything depends on the scale you wish to work at. The more detailed the scale of the working map is, the more easy is to delimit vegetation units (whatever the syntaxonomic rank is). My experience in mapping vegetation throughout the Iberian Peninsula is that if you are able to identify higher vegetation units (e.g. alliances) in a given territory, you should equally be able to identify easily association units in the same degree and with the same effort (because in most cases one alliance usually is represented by one or two different associations which correspond to different, not coincident mapping polygons). Therefore, mapping vegetation to association level (and even to subassociation levels, when these are based on geographical patterns, not on ecological ones that should be regarded as vegetation variants) is very useful and not too much difficult as mapping higher syntaxonomical units. The most important goal to map at those detailed units is that the floristic, biogeographical, successional, and even phenological information you join to associations is lost when your mapping is reduced to higher levels (alliances, classes). And also one should think that all that kind of 'to-detail' information can be absolutely useful for conservation and/or management purposes. So, why should be refuse to do the effort...?

Interesting question, Garry!

They ARE worth the effort imo. However, it takes years of training to become a skilled phytosociologist who's able to define the units accurately. This is the reason that many vegetation map, based on phytosociological data are mediocre, at least in the Netherlands: they are made by small ecological enterprises/bureaus which are never trained scientifically.

The knowledge and literature behind well defined phytosociological units, especially when combined with soil data, are so enormous that it can hardly be replaced by other data. However, I think one requirement must always be fulfilled in order to give a vegetation map at the association level its significance: the units must be based on relevés always. This is the only way to check whether the appointment to associations is done correctly. Without such data, the units on the map just could mean anything... ergo nothing. A well executed survey in which relevés are made and clustered into units which can be assigned to phytosociological units in literature is hardly replaceable. Then, but only then, phytosociological vegetation maps with associations are worth the effort!

Thank you Rense. I guess the realistic answer to my original is "yes". This might be one example of technology reversing progress. It is much easier to learn to recognize gross morphology than species. And more funds are available.

Yes, Garry, but you could argue also that the effort it too large and the costs too high for such well defined phytosociological maps.

We use such maps for management planning and policy making, for instance for the Natura 2000 network in the Netherlands.

Yes. For many purposes they are too costly.

It will depend on where and for what purpose. My experience is in the tropics.

Clearly in some places there are relatively robust and consistent associations that can usefully be characterized and located on a map (for example the zonation we see on east African mountains are consistent in this manner). In other places (mixed lowland tropical forests in Borneo for example) it can be hard to identify a useful label for an association (I suspect that they will often exist in a loose statistical sense, reflecting soil conditions and the like, but will require a major research project to clarify them).

As a map user I like such summary data and insight it gives (especially when biophysical annotations are included) but would also want to highlight to any other users that such described associations are abstractions: they are what has been observed in some places at some point in history, they are not necessarily stable or robust, or transferrable among landscapes. (Lots more can be said on the role of disturbance history and Gleason's individualistic ecology concept, but the main point would be that some systems are more consistent than others, and the finer association divisions are less consistent and robust).

I have been involved in vegetation mapping in a couple of national parks in Africa more than a decade ago, and I now doubt (having been back a few times) that the detailed information we generated was used much by managers. I suspect that more training (how to use the maps), and more attention to what managers need to know, would have helpful. Nonetheless, the extensive ground-truthing generated by those projects remain a valuable reference point for assessing change. (I would still use those associations if I was looking for specific species or was discussing fire impacts etc., but few local managers know them).