Hi Edward,

Here is a ref that might be useful for you:

Körner (2007), The use of ‘altitude’ in ecological research, Trends in Ecology and Evolution, Vol 22, pp 569–574.

Although I agree we have to be cautious, I do think sufficient (at least 1000 m) gradients can serve as altitudinal proxies. In my Arizona study area (a mountain canyon), the gradient ranges from semi-arid to temperate. See Crimmins TM, Crimmins MA, Bertelsen CD (2009) Flowering range changes across an elevation gradient in response to warming summer temperatures. Global Change Biology 15:1141-1152.

Edward, that also depends on what aspects of climate change you want to study. For instance, is it the current climate change, or the effects of past climate changes across various geographical and climatic domains? For deglaciation chronosequences in the current climate change and impact on biodiversity, space-for-time substitution experiments are being used. This is not a perfect approach, and, yes, it needs to be used cautiously but if applied properly can provide very useful results. I am using this to study the impact of melting ice in the Andes. Here are a few references about the method by various authors.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1559953/

http://www.ist-world.org/ProjectDetails.aspx?ProjectId=33f50bb093bc42c3a210954369171859

http://leml.asu.edu/jingle/Web_Pages/Wu_Pubs/PDF_Files/Wickham_Wu_B_97.pdf

I do agree with Bertelsen that certain long altitudinal gradients can serve as altitudinal proxies. beyond certain length of the altitudinal gradient, the variable altitude (temperature-barometric presure) becomes the overriding factor, while soil moisture and slope aspect become secondary ¿how long the gradient needst o be? this "altitudinal gradient length treshold" will varies from mountain to mountain depending on the particular environmental heterogeneity. The higher the environmental heterogeneity the longer the altitudinal treshold for a particular mountain.

See Vázquez and Givnish (1998) Vegetation of the Cerro Grande massif, Sierra de Manantlán, México: Ordination of a long altitudinal gradient with high species turnover. Boletín del Instituto de Botánica de la Universidad de Guadalajara. 6(2-3):227-250.

https://www.researchgate.net/publication/256488240_Vegetation_of_the_Cerro_Grande_massif_Sierra_de_Manantln_Mxico_Ordination_of_a_long_altitudinal_gradient_with_high_species_turnover?ev=prf_pub

Article Vegetation of the Cerro Grande massif, Sierra de Manantlán, ...