Trying to gather insight from an unresolved discussion with a colleague here, and either the available literature surrounding this topic and/or my keywords in search of are lacking.
Context: When conducting plant research using multiple replicates within several treatments, how important do you rank watering uniformity? Another way to say this would be, how important is the notion of every plant within an experiment receiving the same volume of water (or nutrient/fertilizer solution)? The two schools of thought between a colleague and myself are as follows:
1. All plants need the same volume of water in order for the effects of the (non-moisture related) treatments within the experiment to be valid.
versus
2. Plant replicates within a treatment/experiment may be watered singly as needed.
Additional and specific context: As always, it is best to begin a treatment/experiment with uniform plant replicates. In our case at times we have some outlier replicates from asexual propagation that may allow for non-uniformity with a treatment, a few weeks after transplant (clones of Cannabis sativa). Or, if one larger plant replicate dries out more quickly than a smaller replicate (within the same treatment or experiment), the act of providing water to one-few replicates of an experiment will not de-value the effects of the non-moisture related treatments in downstream data analysis.
In the case of some plants being larger and drying out more quickly than others: The above thought #1 would provide all plants, whether in moist or dry soil, with smaller volumes of water in order to prevent wilting of the larger plant whilst maintaining watering-volume uniformity of all reps. within an experiment. On the contrary, thought #2 would thoroughly irrigate the larger and drying-out plant and repeat this methodology as individual plants within a treatment begin to dry out. So, water individual replicates as needed based on soil moisture.
***Of course, the water holding capacity and saturation point of the growing medium would be the best way to conduct this, and presumably the maintaining of all replicates somewhere between 70-85% soil saturation would be the most ideal, but due to the size/scale of our operations this is not feasible.***
Moving forward, the current thinking is eliminate outliers/over-plant replicates in a treatment and just rogue out the outliers/smaller plant replicates, while still maintaining enough replication to gather statistical difference data.
Any and all insight is highly appreciated!