Yes they are! Thermoclimatical and ombrothermic belts describes very well what can happen to woodlands in situatuions of precipitation decrease and raising temperature...

Rivas-Martínez S., Rivas-Sáenz S. & Penas A. 2011. Worldwide bioclimatic classification system. Global Geobotany 1: 1-634 + 4 Maps.

No, they're not! The future viability of tree species in any given locality is dependent on the outcome of the arms race between the trees and their pests and pathogens. Any climate based forecast of future range that doesn't take into account the effects of climate on those pests and pathogens is only looking at half the story. Since the impact of pests and pathogens depends on direct and indirect interactions with climate (including hard-to-predict extreme events) and other species (including novel or imported species), the future viability of tree species is a LOT more complex than bioclimatic envelopes can illustrate.

Furthermore, bioclimatic envelope studies depend on projections of future climates. These are only generally available up till 2100 (with decreasing certainty beyond the immediate future). But for many species and objectives (especially urban, conservation, landscape/cultural and hydrological objs) we need to grow individual trees for 150-400 years - timescales well beyond any certainty for climate prediction. Any confidence that bioclimatic envelopes can give us in such long timescales seems to me to be illusory.

Of course pathogens are note quantified in these approachs! It's a pathogen-free estimation...Altought natural woodlands are very resistant to them, mainly introduced by alochtonous forestswhich are more prone to be attacked! It's not a crime to try to know which floristic and climatic envelopoes could occurre, once we have the present known forests already described by geobotanical sciences, that can have pragmatic measures of what happens, for instance, if we get more hotter, termophilous woodlands should migrate latitudinally for nothern directions and higher positions up-hill (or the reverse), or even dryer vegetation series should leave rupicoulous stations and ocuppy mid-slope soils (if we get more dryer)...We can not stand fixed to "Ignoramus et ignorabimus"just because we dont control all the factors!

That's fair enough Carlos - of course, modelling is invariably about limiting variables to those which we can actually use.

But...we need to be very careful about the use of model outputs which are so incomplete, and that is my chief concern. In particular, the traditional view that natural woodlands are very resistant to pathogens is not supported by our recent experience of tree health problems around the world - the Chalara fraxinea dieback of ash across Europe is only the most recent headline example. The loss of many millions of hectares of pines in western Canada/US to a native pest - mountain pine beetle - is a clear example of an impact with a literally global impact that would not be predicted by bioclimatic modelling.

And finally - in terms of planning future woodlands, how does it help us to know the location of a suitable climatic envelope in 2080 (as used by Forestry Commission in UK) of a tree species we hope to grow for 300 years? Any decision based on that is surely flawed?

I think there's several different questions here – if the question is "are bioclimatic envelopes ALONE an adequate tool..." then the answer has to be no. But they may still make an important contribution.

They are most likely to be useful where the current distribution of a species is demonstrably controlled by climatic factors. For example, the northern limit of lime (Tilia cordata) in Cumbria is at the far northern limit for seed production (Pigott & Huntley 1978 and 1980), and in most localities there has been no reproduction from seed for at least a century. So if the climate changes in a direction which favours seed production, we might expect lime to spread north.

However, in practice, the distribution of all trees in Britain is heavily affected by human management. Lime is pretty unlikely to spread north naturally, because the woodlands are fragmented within a semi-natural matrix. Bioclimatic envelope models may be useful in helping us to think about when and where to introduce species beyond their current ‘natural’ range.

In many cases though, the ‘natural’ distribution of tree species does not correspond to current climate, but to historic factors. So beech probably never naturally spread further north than Norfolk in Britain, but has since been introduced much further north and reproduces prolifically in the north of Scotland. Since its current distribution is not related to climate, predictions of how its future distribution might be affected by climate are not necessarily useful.

Apart from climate and management history and current management, the actual distribution of a species – across the country or within a site – relates to a whole host of other factors. As Duncan points out, these will include pests and disease – and the impacts these have on competition with other tree species. Carlos suggests that natural woodlands are very resistant to pathogens, but I don’t think we can make this assumption. Dutch elm disease wiped out elm trees from much of the British landscape (although elm continues as an understorey species in many woodlands). We are now facing the challenges of Dothistroma needle blight on pine and Chalara on ash, to name but two. These have the potential fundamentally to alter the composition of our woodlands over the next few decades. Climate change is not the cause of these pathogens, but it will affect their spread and the way they affect our woodlands, so we cannot avoid taking them into account.

So I suggest that bioclimatic models may be a useful tool – but they can only ever be part of the suite of tools that we need to plan for the woodlands of the future.

Pigott, C. D. and Huntley, J. P. 1978 Factors controlling the distribution of Tilia cordata at the northern limits of its geographical range. I. Distribution in North-West England. New Phytologist 81, pp. 429-441.

Pigott, C. D. and Huntley, J. P. 1980 Factors controlling the distribution of Tilia cordata at the northern limits of its geographical range. II. History in North-West England. New Phytologist 84, pp. 145-164.

Dear all followers and contributors. Thanks for your input! I appreciate the statement of Jeanette that BE models alone are critical. I also see the recent BE models as a valuable tool, but their output alone should not be used for forest and nature conservation planning in the current state. But my questions to the BE modelling experts are: Are there existing approaches e.g. to combine BE tree species models with BE/impact models of biotic agents like bark beetles etc.? Or are there the ideas of integrated dynamic species range models of Brian Huntley (Huntley et al. 2010) already realized with integrated modules on climate suitablity (cf. recent BE models), habitat availability, population dynamics and dispersal of species. Are there other progress visible?