Last week was bittersweet, as we said goodbye to Lance (fourth from left) but also congratulated him on completing his MS in Plant Breeding and Plant Genetics. One focus of Lance’s research was to measure the genetic correlation between the Hancock Agricultural Research Station (where we conduct our preliminary and advanced yield trials) and two commercial farms, over a two year period. For every trait examined, including yield, shape, size distribution, and specific gravity, the research station seemed representative of the “population of environments” on the commercial farms. Lance used the uniplot visualization technique to reach these conclusions, as shown below for the % of marketable yield greater than 6 oz.
Each environment had a three digit code: the first letter indicating the location (H = Hancock, C = Coloma, N = Nekoosa) and the second two digits indicating the year. For environments that fall on the circle, the correlation equals the cosine of the angle defined by the points and center of the circle. This geometric interpretation becomes less exact as the points move toward the center of the circle (in which case the full correlation matrix should be consulted).
Lance is headed to Mount Vernon, Indiana, for a position in the Dow maize breeding program. He will be missed!