Key points are not available for this paper at this time.
A comprehensive understanding of leaf shape is important in many investigations in plant biology. Techniques to assess variation in leaf shape are often time-consuming, labour-intensive and prohibited by complex calculation of large data sets. We have developed LeafAnalyser, software that uses image-processing techniques to greatly simplify the measurement of leaf shape variation. LeafAnalyser places a large number of evenly distributed landmarks along leaf margins and records the position of each automatically. We used LeafAnalyser to analyse the variation in 3000 leaves from 400 plants of Antirrhinum majus. We were able to summarise the major trends in leaf shape variation using a principal components (PC) analysis and assess the changes in size, width and tip-to-base asymmetry within our leaf library. We demonstrate how this information can be used to develop a model that describes the range and variation of leaf shape within standard wild-type lines, and illustrate the shape transformations that occur between leaf nodes. We also show that information from LeafAnalyser can be used to identify novel trends in shape variation, as low-variance PCs that only affect a subset of position landmarks. These results provide a high-throughput method to calculate leaf shape variation that allows a large number of leaves to be visualised in higher-dimensional phenotypic space. To illustrate the applicability of LeafAnalyser we also calculated the leaf shape variation in 300 leaves from Arabidopsis thaliana.
Weight et al. (Fri,) studied this question.