Help analyzing data over time

#1
I'm conducting a biostatistics project in which myself and a DIY Biology group are trying to determine the efficacy of Spirulina platensis for carbon capture. Our null hypothesis being that Spirulina grown with kombucha does not significantly reduce the levels of CO2 in the air put out by kombucha, and our alternative hypothesis being that Spirulina does significantly reduce CO2 levels in the air when grown next to kombucha. We also want to correlate the concentration of Spriulina (mg/mL) with the amount of CO2 reduced each day.

What we did was we grew kombucha in a miniature greenhouse [kombucha produces CO2 via fermentation] with a CO2 meter that took measurements of carbon dioxide in PPM, dew point, and temperature every 30 minutes for two weeks. Then we grew Spirulina in the mini greenhouse, taking concentration measurements at the beginning and end of each week, for two weeks. We plotted the data, looked at the trends, and I've been trying to compare the means of the CO2 levels before and after growing Spirulina with kombucha; including the overall means, the mean CO2 levels of each day, and the mean of the means of each day. The CO2 output of kombucha grown without Spirulina has random variability and displays a positive secular trend, and the CO2 output of kombucha grown next to Spirulina has random variability and displays a negative secular trend.

The part where I'm banging my head against a brick wall is that I'm having difficulty trying to find the right statistical model to apply to this project, to ensure that my analysis is correct. Every statistical model I've learned, including before and after treatment models, involves taking several individuals from a population and comparing the measurements from individuals in those populations. This is different. This project involves one kombucha culture, measured over a period of four weeks in the absence and presence of Spriulina platensis. The data sets are too skewed, but then again I don't think I can even apply the kinds of population distributions I've been taught to these data sets, so I'm also not sure which test I'm supposed to be using [although I'm positive that ANOVA is out of the question]. Right now I can't be sure that either of the statistical tests I've run are valid [Z test and T test].

If you would like further details you can go to our experiment.com page where I have posted our lab notes.

Maybe I'm just missing/forgetting something obvious, and I just need a fresh brain to look at this and hit me with a "blinding flash" of what should be obvious. Any help/perspective would be appreciated.