October 18, 2021

Pot Size – A Trade-Off Between Utility and Validity

Written by Dominic Hill:  My PhD project – investigating the effects of drought on potato phenotypes – involves the use of a very large, very heavy, multispectral imaging sensor. This sensor weighs around 150 kg and is, for all intents and purposes, a totally immovable object.

Thus, I am unable to conduct my research on potato in situ – i.e., to grow my plants in a field, wheeling the sensor over the plants to scan them in their “natural” habitat. Instead, I am forced to bring my plants to the sensor.

At first, this doesn’t appear to be an insurmountable problem. People grow plants in pots all the time. Unfortunately, as in almost exclusively the case in science, when one delves into the literature – searching for some sort of precedent – a seemingly simple issue invariably transforms into a much larger one.

The problem of pot size has been addressed before, most notably by Poorter, B Hler, van Dusschoten, Climent, & Postma (2012), who – after conducting a meta-analysis of 65 pot-based studies – concluded that pot sizes for phenotyping experiments should provide at least one litre of soil for every gram of dry biomass that that plant species would be expected to produce in the field.

If you know anything about potatoes, you’ll already know why this is a problem. A field grown potato – assuming the absence of severe drought of disease – can reasonably be expected to produce up to c. 700 g of dry biomass, the vast majority of which is contributed by the tubers (Wheeler & Tibbitts, 1987).

The previous recommendation of 1 g l-1­­ is, therefore, laughably impractical in potato. 700 litre pots would weight at least a few hundred kilos and could cost as much as £100 to fill with soil. They would also be impossible to move to the sensor, which is the whole point of growing them in pots in the first place!

With little help from the literature, I decided to devise my own potted potato protocol. I designed an experiment – making use of a range of reasonable pot sizes – to help make an informed decision on which pot size to use going forward. I was looking for a pot size which minimised the confounding effects of restricting growth on the plants’ phenotypes while simultaneously maximising the number of replicates I could reasonably grow and minimising cost and weight.

The results of this experiment were fascinating. I knew – from that previous paper – that pot size would affect the plants’ phenotypes. However, I assumed that small pots would simply stunt the plants growth, regardless of species or cultivar. What I observed, however, was that the smallest pots affected the two varieties I was growing very differently.

In the largest pots – which contained 40 litres of soil – the two cultivars looked relatively similar and almost indistinguishable from plants that I had previously observed in the field. In the small pots, however, the two cultivars looked remarkable different: Maris Piper was tall and leggy, while Charlotte was short and bushy – regardless of is the plants were drought stressed or not.

Figure 1. Plants of two cultivars of potato in 5 litre pots. Plants of the cultivar Maris Piper were very tall, with long internodes, and few stems (left). Plants of the cultivar Charlotte were much shorter, bushier, with a higher stem count (right).

Clearly, pot size was affecting the growth of my plants, far more than the drought conditions to which half were subjected. Had I never run this experiment, and simply followed a previously published protocol, which often use comically small pots (often <5 litres), I would have had no idea that many of the effects I was witnessing had nothing to do with drought and everything to do with pot size.

As a result, I have been careful to modify my methodologies for the future of my project. My protocols now include growing my plants in long, deep, and narrow troughs, more akin to how potatoes are growing in the field. These can be mechanically lifted on to a handy little trailer and rolled down to the sensor for phenotyping.

As for the mechanism behind the bonsai effect in potatoes grown in small pots, that remains unknown. My next experiment will address this directly, hopefully providing new methods to practically grow many plants for phenotyping experiments, without the need for ridiculously large pots.

My hunch is that the small pots dried out so quickly – having a much higher surface area to volume ratio than the large pots – that even the well-watered spuds were actually drought stressed. Thus, because the two cultivars differ in terms of drought tolerance, this accounted for the differences seen in the phenotypes of the two varieties.

How I test this hypothesis, I’m not sure, but I’m sure it’ll involve going back to the literature and probably ending up down another rabbit hole…


Poorter, H., B Hler, J., van Dusschoten, D., Climent, J., & Postma, J. A. (2012). Pot size matters: a meta-analysis of the effects of rooting volume on plant growth. Functional Plant Biology: FPB, 39(11), 839–850.

Wheeler, R. M., & Tibbitts, T. W. (1987). Utilization of potatoes for life support systems in space: III. Productivity at successive harvest dates under 12-h and 24-h photoperiods. American Potato Journal, 64, 311–320.