One of the first jobs I had when I was still in school was working for the Agricultural Research Service (ARS) at the Northern Great Plains Research Laboratory in Mandan, ND. They were a USDA affiliated lab, and I worked with a soil scientist on a crop rotation program. The project had gone on for quite some time, and I spent most of my days there on test plots making sure everything was going smoothly.

We started in spring by drilling in seeds, created quadrants, fertilized and tended those quadrants, and then harvested them in the fall. Since everything was laid out in a grid and we had information going back for years on everything including: soil moisture, nutrient/fertilizer loads, plant seed chemical analysis, and pesticide and herbicide use. This information was essential in creating a guide to which crops to plant for successive years. The office even produced a computer program for farmers to plan their fields out for years in advance.

After each quadrant was harvested, we used a shaker machine to separate the chaff, and then a laser seed counter to count how many seeds were produced. Those seeds then went to a grinder and then shipped out for analysis. It was so long ago that I did this work, Don Tanaka or Justin Hartel may be a better resource for questions.

One of my primary functions was bagging, counting, and organizing the seeds after they came off the sieve/shaker machine. The laser counters had to be calibrated and tweaked to get the most accurate count. So I spent quite a bit of time counting out 100 seeds and then running them through the machine to see how accurate we could get it. Weight wasn’t a reliable number due to each plant having different sized seeds.