Detection and mapping of resin canals by image analysis in transverse sections of mechanically perturbed, young Pinus radiata trees

We describe a novel, semi-automatic method for the detection, visualisation and quantification of axially oriented resin canals in transverse sections of Pinus radiata D. Don (radiata pine) trees. Sections were imaged with a flatbed scanner using circularly polarised transmitted light, with the resin canals that contained only primary cell walls appearing dark against a bright background of highly-birefringent tracheids. These images were analysed using ImageJ software and allowed for a non-biased, automated detection of resin canals and their spatial distribution across the entire stem. We analysed 8-month-old trees that had been subjected to tilting to induce compression wood and rocking to simulate the effects of wind. These experiments showed that both rocking and tilting promoted the formation of wood and confirmed that resin canals were most common adjacent to the pith. Both the rocking and tilting treatments caused a decrease in the number of resin canals per unit area when compared to vertical controls, but this change was due to the increased formation of wood by these treatments. In tilted samples, however, analysis of resin canal distribution showed that canals were more common on the lower sides of stems but these canals were excluded from regions that formed compression wood.