Over the past century of maize (Zea mays L.) breeding, grain yield progress has been the result of improvements in several other intrinsic physiological and morphological traits. In this study, we describe (i) the contribution of kernel weight (KW) to yield genetic gain across multiple agronomic settings and breeding programs, and (ii) the physiological bases for improvements in KW for US hybrids. A global-scale literature review concludes that rates of KW improvement in US hybrids were similar to those of other commercial breeding programs but extended over a longer period of time. There is room for a continued increase of kernel size in maize for most of the genetic materials analysed, but the trade-off between kernel number and KW poses a challenge for future yield progress. Through phenotypic characterization of Pioneer Hi-Bred ERA hybrids in the USA, we determine that improvements in KW have been predominantly related to an extended kernel-filling duration. Likewise, crop improvement has conferred on modern hybrids greater KW plasticity, expressed as a better ability to respond to changes in assimilate availability. Our analysis of past trends and current state of development helps to identify candidate targets for future improvements in maize.

The search landscape is a common metaphor to describe the structure of computational search spaces. Different landscape metrics can be computed and used to predict search difficulty. Yet, the metaphor falls short in visualisation terms because it is hard to represent complex landscapes, both in terms of size and dimensionality. This paper combines local optima networks, as a compact representation of the global structure of a search space, and dimensionality reduction, using the t-distributed stochastic neighbour embedding algorithm, in order to both bring the metaphor to life and convey new insight into the search process. As a case study, two benchmark programs, under a genetic improvement bug-fixing scenario, are analysed and visualised using the proposed method. Local optima networks for both iterated local search and a hybrid genetic algorithm, across different neighbourhoods, are compared, highlighting the differences in how the landscape is explored.