A few months ago, we published a post describing seed storage on our farm. In particular, we described a design for a cost-effective, below-ground root cellar of our own design. After over month of temperature recordings, our data is in and we are pleased with the results.
Dry season represents the extreme of temperature fluctuation of Thailand. In January and February, ambient temperatures reach down to 10C, and below 0C in the mountains. Summer regularly sees temperatures above 50C during March and April. The basic idea of a root cellar operating during such extremes would present little challenge to improve seed storage settings. Even the slightest effort would make a huge difference. Instead, we chose to test it in wet season, when ambient temperatures are much more mild with less variation throughout the day.
As stated in our prior post, our root cellar has a simple construction. After digging a pit deep enough for the storage bin and wide enough for a recycled tire laid on its side, cover the bottom of the pit with 2-3 inches of sand. Place one tire in the pit with the bin sitting in the middle of the pit. Then, fill the empty space in the tire with sand. Next, place the second tire over and around the bin and fill this with sand. Backfill the space between the pit walls and the tires with excess dirt. Finally, pour water over the tires to fill until the sand in the tires is saturated at the top. The basic premise of the idea is that the wet sand will help to insulate the root cellar from ambient temperatures and also draw off heat.
Our test was aimed at testing the cooling effectiveness of the root cellar. Future tests could be aimed at testing the germination success rate of seeds stored in the cellar and our farm house. However, given that temperature variation and extreme temperatures have been attributed to decreased in seed viability in available literature, we felt this initial temperature test would be sufficient for the time being.
Data collected matched our expectations. The root cellar was warmer in the morning at 8am on average (27.8C) than the farm house (27.7C). At 1pm, the root cellar averaged a temperature far below that of the farm house (28.9C to 31.0C). In the afternoon, 5pm, the root cellar averaged a lower temperature again (28.3C to 29.5C). Simply by evaluating the average temperature readings, it is clear that the root cellar is successful in not only lowering the seed temperatures, but also in lowering temperature fluctuations (as seen in a higher average temperature in the mornings).
As a bonus, for our statistic nerd friends, our own and outside consultation statistical analysis shows that our values were statistically significant with midday and afternoon temperatures having significant difference using a t-test with p < 0.01.
Overall, we are very pleased with this first model of root cellars and look forward to both recommending it to farmers in the area as well as performing further testing and modifications.