Experiments versus real-world studies

How close are experiments to nature? The answer isn't clean cut, partly because nature is messy.

Experiments, with the aspiration to undertake investigations in a controlled way that permits analysis and interpretation, are one of the cornerstones of modern science. But how valid are they at showing us what goes on ‘out there’ in the real world?

Baby kissing mirror image
Photo: roseoftimothywoods/Flickr

I’m sure those of an experimental science persuasion will say they are great, and the only way to really understand the complex natural situation. OK, but really, how valid are they at showing us what goes on in nature? Arguably, the only way to find out is to do an … experiment, which is what Maria Grazia Annunziata et al. did.

Some background: Illumination from the sun in the natural environment varies throughout the day, often fluctuating irregularly e.g. due to cloud cover, and with gradual shifts in dark and light at dawn and dusk. Plants in controlled environments, on the other hand, are exposed to constant irradiance during the day and experience abrupt light–dark transitions. How does plant metabolism compare between the two illumination regimes? Analysing a variety of metabolite profiles of their experimental organism – Arabidopsis thaliana (and don’t get me started on how relevant is this weedy crucifier as a model for each and all of the estimated 369,400 species of flowering plants!) – Annuunziata et al. revealed that carbon and nitrogen metabolism differed significantly between sunlight and artificial light conditions.

Cautious in their conclusion – as befits true scientists – they suggest that the variability of sunlight within and between days could be a factor underlying these differences, and that results obtained from plants grown with artificial lighting* might not be representative of natural conditions.**

Case closed: artificial [aka experimental/controlled] is not natural. I’ve always been deeply suspicious of studies that use plants sown in sterilised soil. In the real world – outside the lab/controlled environment facility/growth room – it’s been long-established that approx. 80 – 90% of plants have mycorrhizal associations (although the fraction is probably closer to 82%). I.e. real-world plants do not grow in sterile rooting media (and aren’t found in nature as a single species divorced from contact and interaction with other organisms). How valid or relevant are those experimental studies? Answer: Not at all(!)

* Their data did however suggest that energy-efficient LED [light-emitting diode] lighting is an acceptable alternative to fluorescent lights, which must give some comfort to somebody (and not just those who sell LEDs!).

** It should be noted that in this investigation, ‘natural’ was considered to be plants growing in a greenhouse (albeit one that was naturally illuminated…).

Reference List

Annunziata, M. G., Apelt, F., Carillo, P., Krause, U., Feil, R., Mengin, V., … Raines, C. (2017). Getting back to nature: a reality check for experiments in controlled environments. Journal of Experimental Botany, 68(16), 4463–4477. https://doi.org/10.1093/jxb/erx220

Brundrett, M. C. (2002). Coevolution of roots and mycorrhizas of land plants. New Phytologist, 154(2), 275–304. https://doi.org/10.1046/j.1469-8137.2002.00397.x