Digitizing colors of soils in mesocosm wetlands using Nix sensor

Forty mesocosms located at the Ahn Mesocosm Compound at George Mason University have been a part of legacy studies for 12 growing seasons since 2012. In a previous study, each mesocosm wetland was planted with different amounts of species richness. After pre-COVID mesocosm studies halted, vegetation communities gradually changed. The Nix color sensor scanned mesocosm soils that were collected to produce 15 unique color variables for each mesocosm. Data was grouped by mesocosms that shared the original number of species planted. Preliminary results show that there are two out of five groups that have similar color variable values. Further analysis will continue to see if any differences arise and whether the soil conditions have been altered.

Hello! My name is Seung Han, and my project is “Digitizing color of soils in mesocosm wetlands using Nix sensor.”
When we look at soil, the color tells us about soil health and properties such as the presence of organic matter, mineral contents, and changes from fluctuating water levels between the different layers of soil. Here we can see the differences between the soil colors with the very grey wetland soil on the right versus the dark, fertile soil on the left. The standard method of analyzing soil color has been performed by using the munsell soil color chart. This method requires training, is costly, and analysis can be varied from person to person.
A new digital method of analyzing soil color is the Nix color sensor. Each Nix scan produces values for 15 color variables from 5 different color spaces. The color variables are l-a-b, c-h, r-g-b, x-y-z, c-m-y and k.
The study occurred at the Ahn Wetland Mesocosm Compound located in George Mason University’s West Campus. Back in 2012, forty mesocosms, or experimental ecosystems, were set up to study the effects of planting different amounts of species in each tub, ranging from no plants to 4 different wetland plants. The mesocosms were maintained by weeding out unwanted species and keeping the water levels above 5 cm. After pre-COVID studies were completed, the maintenance of these mesocosms stopped in 2019. Now for the big question. After studies and maintenance stopped, are the mesocosm wetland soils still wetland soils?
So now we begin. Each mesocosm had soil cores extracted from each of these five sections. The soil core was cut along the top to create a flat surface to scan, transferred on top of a white sheet, and the Nix was placed on top of a flat, smooth section of the soil for a scan to be performed. This was repeated at least 3 times for each core. Each scan data was saved in the Nix app on a smartphone, and the scans were exported into a csv file for analysis.
Data was organized by grouping mesocosms together with their original amount of planted species. Preliminary results analysis show that mesocosms originally planted with two and four species seem like they both have similar amounts in ten different color variables. A, b, and c, h in these graphs, and z, g, b, c, m, and y in these graphs.
Data analysis continues to see if the differences can be explained by differences in vegetation communities by identifying, counting species, and calculating the amount of percent cover.
I would like to thank my mentor, Dr. Changwoo Ahn, for guidance, encouragement, and believing in me. Thank you to Dr. Stephanie Schmidt for feedback and support this semester. Thanks to Rylee for coming out on hot days to help with some cores and note taking. Special thanks to Trinity Lavenhouse for being a great lab mate, giving support, and being a positive presence. And thanks to OSCAR for funding this URSP project. Take care!