Comparison of two methods for measurment of stomata in pedunculate and sessile oak

Abstract

Stomatal traits, such as density and size, significantly influence gas exchange; therefore, their analysis is essential for understanding plant physiology, their adaptation to environmental conditions, and their responses to climate change. The aim of this study was to determine whether manual measurements of stomata in tree seedlings could be replaced by automated measurements, which would substantially increase the number of samples that can be analysed due to time savings. Epidermal imprints of stomata were prepared from seedlings of pedunculate oak and sessile oak growing at the same site. Ten randomly selected micrographs of the preparations at 200 × and 400 × magnification per species were used to analyse the number of stomata as well as the length and width of guard cells and stomatal pore, using both manual measurements in ImageJ and automated measurements in StoManager1. The results were statistically evaluated. No statistically significant differences between the two methods were found for measurements of stomatal number. Bland–Altman analysis showed good agreement between manual and automated measurements. Although the automated method slightly overestimated the number of stomata, the error was small. For size measurements, statistically significant differences between the two methods were observed, and Bland–Altman analysis revealed deviations between the methods at larger size ranges of the measured structures. Regression analysis indicated that agreement between manual and automated measurements was reasonably reliable within the size range of 20–25 µm (larger structures were not measured), but decreased with decreasing structure size, most likely due to image segmentation in the automated method and the associated less precise edge detection. In small structures, even minor segmentation differences can lead to larger errors. On the other hand, manual measurements in microscopy images can also become increasingly subjective as the size of structures decreases. Our results suggest that automated measurements using StoManager1 represent a suitable alternative to manual measurements of stomatal number/ density and dimensions of larger structures within the stomatal complex for the selected tree species.