Radiative heaters, which are in widespread use, do not simulate
convective (diffuse) warming, but exert a directional heat with vertical gradients,
unlike that of a warmer climate, even if mean temperatures may match. The same
applies to soil warming, which, for physical reasons, induces water diffusion away
from the heat source. In addition, step increases of temperature in soils represent a
major disturbance which may take years to lead to a new steady state, with initial
responses in essence documenting the disturbance of the rather delicate balance between
plant roots, fungi, microbes and the soil fauna associated with it. Given these
intrinsic constraints, it is far safer to build upon short-distance natural topographic
Whenever possible, the various techniques should be combined to capitalise on
the advantage of each. I emphasise the simpler, often overlooked tools for biological
temperature research offered in situ, because the lack of high-tech facilities is often
seen to preclude upfront research. Controlling life conditions in closed research
units is and will remain a key tool for understanding plant temperature responses.
However, such data are not necessarily more ‘accurate’ or relevant than those obtained
in the field, although this assumption is the tradition that I and many of my
age class grew up with. I want to encourage the next generation to be more open
to the alternative approaches with much greater ‘experimental noise’ incurred, but
this may become manageable with high replication and with the modern statistical
and computational tools.