QUESTION (from Ashley):

Mitch and I were talking and thought a drought trial on 6 different varieties of Beta vulgaris in the greenhouse would be cool. We would have one set that would get watered normally and the other would not. My question is what type of sampling system do you think would be best? Take 3 different samples from 5 different plants per box and maybe taking 3 samples per leaf to see how consistant the device is? I am just a little stuck on how I would formulate the questions in the project builder to allow the easiest way to analyze the data.

ANSWER (from Greg):

Well, one thing we've learned so far is sample early. Often, especially in drought experiments, if you want too long the effects become visibly obvious and then collecting fancy sensor data is kind of pointless. So sample early.

Generaly we try to get 25 - 45 samples on a given population if we don't have control of lighting, timing or device. So maybe choose 1 plant per variety (3 measurements per plant). Measure each variety once, and repeat 3 - 4 times. I'd consider doing this once in the morning, and once in the afternoon on your sampling days. Ideally choose a day with relatively consistent light lighting isn't changing every few seconds as happens with really patchy clouds. I would include the following questions.

• Variety (1 - 6)
• Water regime (1,2)
• Leaf location (for beta vulgaris, this may be top/middle/bottom, or perhaps location on the leaf itself since they are so big... not sure, the intent is to account for the effect of leaf development so however that is appropriate).
• Time of Day (you always get a time stamp anyway, but sometimes it's easier to analyze if you just includes 'morning' 'afternoon' and 'evening' anyway)
• Days since planting (nice for time series data)