Climate change necessitates new crop breeding objectives to address temperature and precipitation shifts, including tolerances against climate extremes such as droughts, winter kills, and floods. Using digital technologies in plant breeding aims to aide in reducing climatic stresses to crops. Yet, Using digital technologies requires identifying improvements in yields and other crop variety evaluation criteria without endangering the underlying crop genetic resources. With a focus on winter wheat (Triticum aestivum), this paper evaluates whether new varieties shift the breeding frontier by adding improved traits. We base our analysis on a unique and detailed unbalanced panel dataset of the German variety admission of 229 winter wheat varieties for 2008-2016, where we observe 32 variety-specific characteristics related to cultivation, yield and quality. Using non-parametric FDH and skyline algorithms, we benchmark newly admitted wheat varieties against existing varieties. Preliminary results indicate that throughout the observation period, nearly all new varieties have pushed the frontier, although the observed frontier shift is small. To improve further our analysis, we plan to address concerns related to the high dimensionality of the wheat characteristics, considering dimension reduction tools such as principal component analysis. Our procedure may serve as a first step towards developing a decision support system selecting varieties during variety admission processes. 

Accelerating modern plant breeding and stirring it towards stewardship of agricultural systems whilst dealing with global policrises demands a better insight in how modern knowledge systems in plant breeding play coevolve with the crops they create. Modern plant breeding together with its tools for gene editing, global marketing strategies for seed, and climatic changes of the last decades have led to a multitude of selection, replication, and innovation mechanisms that need to be put in context on multiple levels. Drawing from Multilevel Selection Theory, I analyze how variation, selection, replication, and innovation operate across different levels in knowledge systems of plant breeding science, based on the cases of the German seed system in canola and winter wheat. Cultural evolution introduces additional mechanisms of replication, including the cumulative retention of information in technologies, which are embodied in crops and their underlying genetic diversity. I propose a framework for looking these with a coevolutionary lens that explicitly includes and explains the socio-cultural factors influencing path-dependencies in plant science. Goal is to develop a framework useful to analyze genetic diversity in crops to investigate patterns of change over time.