Summary: Embolism spreading in angiosperm xylem occurs via mesoporous pit membranes between vessels. Here, we investigate how the size of pore constrictions in pit membranes is related to pit membrane thickness and embolism resistance. Pit membranes were modelled as multiple layers to investigate how pit membrane thickness and the number of intervessel pits per vessel determine pore constriction sizes, the probability of encountering large pores, and embolism resistance. These estimations were complemented by measurements of pit membrane thickness, embolism resistance, and number of intervessel pits per vessel in stem xylem (n = 31, 31 and 20 species, respectively). The modelled constriction sizes in pit membranes decreased with increasing membrane thickness, explaining the measured relationship between pit membrane thickness and embolism resistance. The number of pits per vessel affected constriction size and embolism resistance much less than pit membrane thickness. Moreover, a strong relationship between modelled and measured embolism resistance was observed. Pore constrictions provide a mechanistic explanation for why pit membrane thickness determines embolism resistance, which suggests that hydraulic safety can be uncoupled from hydraulic efficiency. Although embolism spreading remains puzzling and encompasses more than pore constriction sizes, angiosperms are unlikely to have leaky pit membranes, which enables tensile transport of water.

In this paper, various kinds of applications are presented, in which tomographic image data depicting microstructures of materials are semantically segmented by combining machine learning methods and conventional image processing steps. The main focus of this paper is the grain-wise segmentation of time-resolved CT data of an AlCu specimen which was obtained in between several Ostwald ripening steps. The poorly visible grain boundaries in 3D CT data were enhanced using convolutional neural networks (CNNs). The CNN architectures considered in this paper are a 2D U-Net, a multichannel 2D U-Net and a 3D U-Net where the latter was trained at a lower resolution due to memory limitations. For training the CNNs, ground truth information was derived from 3D X-ray diffraction (3DXRD) measurements. The grain boundary images enhanced by the CNNs were then segmented using a marker-based watershed algorithm with an additional postprocessing step for reducing oversegmentation. The segmentation results obtained by this procedure were quantitatively compared to ground truth information derived by the 3DXRD measurements. A quantitative comparison between segmentation results indicates that the 3D U-Net performs best among the considered U-Net architectures. Additionally, a scenario, in which “ground truth” data is only available in one time step, is considered. Therefore, a CNN was trained only with CT and 3DXRD data from the last measured time step. The trained network and the image processing steps were then applied to the entire series of CT scans. The resulting segmentations exhibited a similar quality compared to those obtained by the network which was trained with the entire series of CT scans.

Pit membranes in bordered pits of tracheary elements of angiosperm xylem represent primary cell walls that undergo structural and chemical modifications, not only during cell death but also during and after their role as safety valves for water transport between conduits. Cellulose microfibrils, which are typically grouped in aggregates with a diameter between 20 to 30 nm, make up their main component. While it is clear that pectins and hemicellulose are removed from immature pit membranes during hydrolysis, recent observations of amphiphilic lipids and proteins associated with pit membranes raise important questions about drought-induced embolism formation and spread via air-seeding from gas-filled conduits. Indeed, mechanisms behind air-seeding remain poorly understood, which is due in part to little attention paid to the three-dimensional structure of pit membranes in earlier studies. Based on perfusion experiments and modelling, pore constrictions in fibrous pit membranes are estimated to be well below 50 nm, and typically smaller than 20 nm. Together with the low dynamic surface tensions of amphiphilic lipids at air-water interfaces in pit membranes, 5 to 20 nm pore constrictions are in line with the observed xylem water potentials values that generally induce spread of embolism. Moreover, pit membranes appear to show ideal porous medium properties for sap flow to promote hydraulic efficiency and safety due to their very high porosity (pore volume fraction), with highly interconnected, non-tortuous pore pathways, and the occurrence of multiple pore constrictions within a single pore. This three-dimensional view of pit membranes as mesoporous media may explain the relationship between pit membrane thickness and embolism resistance, but is largely incompatible with earlier, two-dimensional views on air-seeding. It is hypothesised that pit membranes enable water transport under negative pressure by producing stable, surfactant coated nanobubbles while preventing the entry of large bubbles that would cause embolism.

Pit membranes between xylem vessels play a major role in angiosperm water transport. Yet, their three‐dimensional (3D) structure as fibrous porous media remains unknown, largely due to technical challenges and sample preparation artefacts. Here, we applied a modelling approach based on thickness measurements of fresh and fully shrunken pit membranes of seven species. Pore constrictions were also investigated visually by perfusing fresh material with colloidal gold particles of known sizes. Based on a shrinkage model, fresh pit membranes showed tiny pore constrictions of ca. 20 nm, but a very high porosity (i.e. pore volume fraction) of on average 0.81. Perfusion experiments showed similar pore constrictions in fresh samples, well below 50 nm based on transmission electron microscopy. Drying caused a 50% shrinkage of pit membranes, resulting in much smaller pore constrictions. These findings suggest that pit membranes represent a mesoporous medium, with the pore space characterized by multiple constrictions. Constrictions are much smaller than previously assumed, but the pore volume is large and highly interconnected. Pores do not form highly tortuous, bent, or zigzagging pathways. These insights provide a novel view on pit membranes, which is essential to develop a mechanistic, 3D understanding of air‐seeding through this porous medium.

In this paper, three-dimensional (3D) image data of ore particle systems is investigated. By combining X-ray microtomography with scanning electron microscope (SEM)-based image analysis, additional information about the mineralogical composition from certain planar sections can be gained. For the analysis of tomographic images of particle systems the extraction of single particles is essential. This is performed with a marker-based watershed algorithm and a post-processing step utilizing a neural network to reduce oversegmentation. The results are validated by comparing the 3D particle-wise segmentation empirically with 2D SEM images, which have been obtained with a different imaging process and segmentation algorithm. Finally, a stereological application is shown, in which planar SEM images are embedded into the tomographic 3D image. This allows the estimation of local X-ray attenuation coefficients, which are material-specific quantities, in the entire tomographic image.