Pit membrane ontogeny in radial walls of Ginkgo biloba tracheids was followed using transmission electron microscopy. Torus initiation occurs prior to initiation of the pit border and without benefit of a microtubule plexus. The developing pit membrane is associated with masses of wall material located within plasmalemma invaginations. Wall material is added in such a manner as to form a torus with highly irregular surfaces. Margo and torus are traversed by plasmodesmata, whose channels are connected by extcnsive median cavities. Matrix material is removed from both margo and torus shortly after hydrolysis of the adjacent cytoplasms. Matrix removal begins at the pit membrane surface and is not preferentially associated with the plasmodesmata. Tori in aspirated pit membranes have compacted fibrils, and their fibrillar compaction might reduce permeability to air embolisms.
Intervascular pit membranes of Cercocarpus possess torus thickenings. The thickenings, or pads, consist of lignified, secondary wall material. Torus pad deposition occurs late in cell ontogeny and is not associated with a microtubule plexus. Half-bordered pit pairs between tracheary elements and parenchyma cells often have a torus pad on the membrane surface facing the conducting cell. In contrast, a thick protective layer fills the pit cavity on the side of the parenchyma cell. Ontogeny of the torus thickenings in Cercocarpus represents a third mode of torus development in eudicots when compared to that occurring in Osmanthus/Daphne and Ulmus/Celtis.
Of 22 species of Daphne surveyed, 19 possessed tori in their intervessel pit membranes. The torus has a circular shape and is centrally-located on the pit membrane. The fibrils of the surrounding margo show a random arrangement. In some specimens, the fibrils are obscured by material that impregnates the margo, coats the torus, and lines the cell lumens. The margo has small pores. In those species without tori, the intervessel pit membranes, when intact, show randomlywoven fibrils and small pores. Air-dried membranes of these speeies tend to separate into two layers along the line of the middle lamella The presence of a torus is not correlated with evergreen or deciduous habit, but absence of a torus seems to be limited to species of the section Mezereum within the genus Daphne.
Torus thickenings of pit membranes are found not only in gymnosperms, but also in certain genera of dicotyledons. One such genus is Osmanthus. Wood from 17 species of Osmanthus was searched for tori. Fourteen species from three of the four sections investigated possessed these thickenings. Ten of the species represent new records. Only the three New Caledonian species of Section Notosmanthus lacked tori. This observation in combination with other factors serves to isolate this section from the remainder of the genus.
This study compares intertracheary pit membrane structure and ontogeny in Abies firma (Pinaceae) and Metasequoia glyptostroboides (Cupressaceae). Initial phases of pit membrane development are the same for both species. Branched plasmodesmata are present in the earliest stages of pit membrane development observed. Torus thickening of the pit membrane occurs early in pit development prior to pit border initiation. During pit border enlargement, plastids frequently occlude the apertures. Cell lysis is associated with complete wall matrix removal from pit membranes of Metasequoia. By contrast, cell lysis in Abies results in loss of matrix material from the margo, whereas the torus remains largely unaffected. Torus extensions in pit membranes of A. firma retain variable amounts of matrix material. Either a difference in chemical composition of the torus or a difference in autolytic enzymes is hypothesized to explain developmental differences between pit membranes of the two species.
Intervascular pit membranes were investigated in species of Daphne, Wikstroemia, and other allied genera of the Thymelaeaceae. Results confirmed a previous study showing that, except for section Mezereum, all sections of Daphne had pit membranes with tori. Taxonomically isolated species D. aurantiaca and D. genkwa had tori, but lacked a G-layer. Tori similar in structure to those of D. aurantiaca and D. genkwa were observed in three species from the subgenus Diplomorpha of Wikstroemia. Tori of a slightly different morphology were noted in W. kudoi (subg. Daphnimorpha). Tori appeared absent from species of the subgenus Wikstroemia (= Euwikstroemia of Domke), and from the genera Drapetes, Edgeworthia, and Eriosolena. These results suggest a close relationship between Daphne and Wikstroemia. The degree of torus development and the distinctiveness of helical thickenings suggest that smaller tracheary elements serve as a backup water-conducting system to larger vessel elements.
Botrychium dissectum Sprengel rhizomes were examined at monthly intervals from February 1993 through December 1994. Sampies taken ranged from those with an inactive cambium and only mature tracheids to those having an active cambium and immature tracheids. The vascular cambium became activated in the early fall prior to maturation of the leaf and fertile spike complex. Intertracheid pit membranes with tori were present in all sampies, although the morphology of the torus varied. The presence of tori was first observed in a tracheid that had just initiated its secondary wall formation. As the pit membrane matured, matrix material was hydrolyzed first from the margo area, then from the torus, and eventually the pit membrane was represented only by a very thin network of microfibrils. In addition, studies confirmed that tracheids bordering parenchyma cells developed a torus thickening, aIthough no thickening of the parenchyma cell wall occurred. Torus ontogeny in B. dissectum combined features previously described for angiosperms and gymnosperms.
Atomic force microscopy was used to compare the structures of dried, torus-bearing pit membranes from four woody species, three angiosperms and one gymnosperm. Tori of Osmanthus armatus are bipartite consisting of a pustular zone overlying parallel sets of microfibrils that form a peripheral corona. Microfibrils of the corona form radial spokes as they traverse the margo. Margo microfibrils are loosely packed thus facilitating passage of water molecules. The pustular layer is removed by acidified sodium chlorite. Tori of Cercocarpus montanus also have a pustular surface, but lack a corona. Tori of Pinus taeda have a finely granular to amorphous torus matrix. Ulmus alata tori have microfibrils traversing the surface. The atomic force microscope proves itself a useful tool for high resolution study of pit membranes with only minimal specimen preparation.
Microdistribution of non-cellulosic polysaccharides in pit membranes of bordered pits (intertracheid pits between adjacent tracheids), cross-field pits (half bordered pits between tracheids and ray parenchyma cells) and ray pits (simple pits in nodular end walls of ray parenchyma cells) was investigated in mature earlywood of juvenile Scots pine and Norway spruce seedlings using immunocytochemistry combined with monoclonal antibodies specific to (1→4)-β-galactan (LM5), (1→5)-α-arabinan (LM6), homogalacturonan (HG, LM19, LM20), xyloglucan (LM15), xylan (LM10, LM11) and mannan (LM21, LM22) epitopes. Using phloroglucinol-HCl and KMnO4 staining, lignin distribution in pit membranes was also examined. Apart from cross-field pit membranes in Scots pine, all pit membranes observed showed a positive reaction for lignin with differences in staining intensity. Ray pit membranes showed strongest reaction with lignin staining in both species. Intensity of lignin staining in bordered pit membranes was stronger in Norway spruce than in Scots pine. With localization of non-cellulosic polysaccharide epitopes, Scots pine showed differences in cross-field pit membranes (rhamnogalacturonan-I (RG-I), HG and xyloglucan epitopes) from bordered and ray pit membranes (RG-I and HG epitopes). In contrast, Norway spruce showed significant differences in ray pit membranes (RG-I, HG, xyloglucan, xylan and mannan epitopes) from bordered and cross-field pit membranes (HG and no/trace amount of RG-I epitopes). Distributional differences in HG epitopes depending on antibody type/ membrane regions were also observed in cross-field pit membranes between the two species. Together, the results suggest that distribution patterns of lignin and non-cellulosic polysaccharides in pit membranes differ significantly between pit types and between Scots pine and Norway spruce. Compared with the same types of pit membranes in hardwoods, the results for Scots pine and Norway spruce (softwoods) differed significantly.