Printer Friendly

Wood anatomy of Brassicales: new information, new evolutionary concepts.

15. Resedaceae (Carlquist, 1998b; Schweingruber, 2006) (Fig. 17).

The six genera of Resedaceae are small, with only Reseda containing more than 10 species (Bolle, 1936). The few Resedaceae that qualify as woody (Caylusea, Ochradenus) do not have the clearly woody texture that shrubs in Ericaceae or Rhamnaceae do. One could regard these Resedaceae as herbs the longevity of which exceeds two or three years.

Wood plan. Growth rings are present or are lacking, depending on seasonality of growth in particular species (Fig. 17a).

Storying. Wood non-storied in all species studied (e.g., Fig. 17b). This may be correlated with the relative juvenilism of resedaceous woods.

Vessels. Vessels are circular in outline. Mean number of vessels per group ranges from 1.8 (Reseda crystalline Webb & Berth.) to 3.1, and averages 2.3 for the family as a whole. Solitary vessels are about as common as groups of two or three vessels (Fig. 17/a). Vessels are narrow, ranging from 15 [micro]m mean diameter in Caylusea (Fig. 17a) to 31 [micro]m (R. luteola), and 25 pm for the family as a whole. Mean number of vessels per sq. mm. ranges from 33 to 301; the lower vessel density is in R. alba and reflects the wood mesomorphy in this cultivated species. Vessel element length for the family has a range from 95 to 197 pm (145 pm for the family as a whole).

Perforation plates are all simple. Vessel pitting is alternate (Fig. 17c and d). Pits are not vestured in my material (Fig. 17d), although Schweingruber (2006) thought there might be some vesturing in R. suffruticosa Loefl. On the basis of light microscopy. Grooves that widen some pit apertures horizontally and interconnect two or three pit apertures (Fig. 17c) are present in some species. Fine striations were seen with SEM on vessel walls of R. luteola (Fig. 17f).

Imperforate tracheary elements. These cells have been reported to be libriform fibers, but SEM study reveals small borders (Fig. 17e), so fiber-tracheids may be present at least in some species.

Axial parenchyma. Vasicentric scanty axial parenchyma is present. In Caylusea, however, vasicentric axial parenchyma is scarce, but banded and confluent patterns occur (Fig. 17a). Marginal (probably terminal) axial parenchyma occurs in R. alba, R. crystallina, and R. lutea L.

Rays. Rays are (narrow) multiseriate and uniseriate in about equal numbers (Fig. 17b). Ray cells are predominantly upright as seen in radial sections,

Crystals. No crystals have been observed in wood of Resedaceae, although they are present in some other plant portions (Bolle, 1936).

Comments. In comparison to Gyrostemonaceae, Resedaceae are relatively nonsucculent. In the perennial or shrubby species, wood xeromorphy is evident in the narrowness of vessels and in the short vessel element length. Resedaceae have more paedomorphic character state expressions than do Gyrostemonaceae.

16. Stixaceae (Carlquist et al., 2013) (Figs. 18 and 19).

The family Stixaceae Doweld is tentatively recognized here because as shown by Su et al. (2012) and Flail et al. (2004), the family Capparaceae becomes monophyletic only with the removal of Forchhammeria, Stixis, and other genera once included in it by such authors as Pax and Hoffmann (1936). We need wood data from Neothorelia, Tirania, and more species of Stixis, and such data are likely to yield more diversity in this family, judging from some preliminary work (Carlquist, unpublished data). The grouping in the family Stixaceae of three Asiatic genera together with the neotropical Forchhammeria may seem phytogeographically less than plausible, but equally long disjunctions separate the two genera of Akaniaceae; instances of marked disjunction in other families could also be cited.

Wood plan. Successive cambia (Fig. 18a) are present in all species of Stixis and Forchhammeria for which material is available and has been studied. The plan of successive cambia follows that seen in other angiosperms with successive cambia (Carlquist, 2007), with vascular increments separated by conjunctive tissue.

Storying. Storying is not present. Although larger stems of Forchhammeria and one species of Stixis were available, secondary xylem consists of bands each generated by a vascular cambium (which in turn is produced by a master cambium). Thus, each vascular increment can be regarded as a short segment of ontogenetic change, with more extensive ontogenetic change (needed for storying to be achieved) observable in angiospenn species with indefinite production of a single wood cylinder by a single vascular cambium.

Vessels. Vessels are solitary in species of Forchhammeria studied to date. Grouped vessels are advantageous in angiosperms at large unless (in accordance with the concept (Carlquist 1984) vessels are associated with tracheids. This phenomenon demonstrates that tracheid presence is more effective than vessel grouping in promoting conductive safety. It also incidentally shows that solitary vessels would be advantageous in any wood if selection for xeromorphy is not an issue. Figures for vessel density in Forchhammeria are provided by Carlquist et al. (2013). Some vessel grouping was observed in S. philippinensis Mert., and this may related to the fact that vessels can be embedded not in a mass of tracheids, but in some places, in axial parenchyma. Vessel density is lower (16 per square mm) than for angiosperms as a whole, in accordance with the figures for successive cambial species by Carlquist (1975).

Pit apertures as seen on the inner surfaces of vessel walls are slit-like (Figs. 18d and 19e), and pit apertures adjacent in a helical direction may be coalescent. Vesturing is maximally present in vessels of Forchhammeria trifoliata (Fig. 18d), Stixis parviflora (Fig. 19e), and S. philippinensis (unpublished data). Vestigial vesturing was observed in other species of Forchhammeria. Pits cavities are circular and pits are alternate. Perforation plates are all simple.

Imperforate tracheary elements. Tracheids with conspicuous bordered pits occur in Forchhammeria (Fig. 18c) and Stixis (Fig. 19d). Pits may be vestured (Fig. 19d) or non-vestured (Fig. 18c) in Forchhammeria, but vestures in Stixis parviflora and S. philippinensis. The length of tracheids in Forchhammeria as a whole (1279 [micro]m) is greater than the corresponding vessel element length (206 [micro]m), suggesting that there is considerable intrusive growth.

Axial parenchyma. Axial parenchyma in Forchhammeria is diffuse, with tendencies toward vasicentric in F. pallida and F watsonii (Fig. 18a) and diffuse-in-aggregates in F. trifoliata. Axial parenchyma is sparse in Stixis parviflora, but abundant, both paratracheal and in bands, in S. philippinensis. Axial parenchyma cells in S. parviflora have oval pits unlike the small slit-like pits one associates with librifonn fibers. Strands of axial parenchyma are two to five cells in Forchhammeria.

Rays. Rays are predominantly multiseriate, with occasional uniseriates (F. watsonii) or extremely few (F. trifoliata. Stixis pauciflora). Rays are composed mostly of procumbent cells. Thus, they are transitional between Heterogeneous II and Homogeneous II of Kribs (1935). Rays are thus similar to those in Gyrostemonaceae.

Crystals. There are no reports of crystals in secondary xylem per se.

Conjunctive tissue. Conjunctive tissue may consist wholly of parenchyma, as in F. watsonii (Fig. 18a) or Stixis (Fig. 18c, bottom), or it may contain a band of sclereids, as in F. pallida (Fig. 19a). The sclereids of F. pallida contain a large polyhedral crystal each (Fig. 19b). A band of cells, some walls of which are thicker, suggestive of incomplete sclereids, occur in the last-formed conjunctive tissue in Stixis pauciflora (Fig. 19c). The crystals in this layer of cells are varied in shape and size.

Secondary phloem. Each vascular cambium in a species with successive cambia is capable of producing more secondary phloem (1sp, Fig. 18b) as well as more secondary xylem (sx) overtime. Bands of crushed secondary phloem (Fig. 18b, csp) are evidence of continued secondary phloem production.

Starch. Starch grains are common in phloem rays (Fig. 18b, middle right).

Comments. The inherent taxonomic interest of so many distinctive features in which Stixis resembles Forchhammeria seems clear: successive cambia, ray histology, tracheid presence, vestured pits, and crystal-bearing cells in conjunctive tissue.

The vessel diameter, vessel density, and vessel element length of Forchhammeria and Stixis would not suggests wood xeromorphy. However, various botanists have noticed that in Mexico, shrubs of Forchhammeria have green leaves during the dry months when other vegetation in the dty scrub habitats of Forchhammeria show drought deciduousness in foliage. The only wood feature that would explain this is presence of tracheids and vestured pits in secondary xylem. Tracheids are present in such large numbers that one can hypothesize that they play a role not merely in preservation of water columns, but in active conduction also. Occasional portions of secondary xylem in Forchhammeria are free from vessels.

17. Capparaceae (Figs. 20, 21, 22 and 23).

No survey of wood anatomy of Capparaceae exists. This is understandable in view of the wide geographical distribution and varied degrees of woodiness of the family. There are not even any generic monographs of wood anatomy within the family. Descriptions of wood of assemblages of species based on floristics exist (e.g., Cozzo, 1946; Stem et al., 1963). Gregory (1994) lists 39 such references, to which can be added Fahn et al. (1986). The accounts of Solereder (1908) and Metcalfe & Chalk (1950) therefore are by default, the best overall sources of information on wood anatomy. SEM images in the present essay are all original (as are the light microscope photomicrographs).

Capparaceae are recognized here in the sense of Hall et al. (2002), excluding Cleomaceae and Brassicaceae. With that definition, each of these families becomes monophyletic. The earlier vision of the family (Pax & Hoffmann, 1936) was more inclusive, and could not be maintained with the advent of molecular data. For the history of the taxonomy and phylogeny, the account of Hall et al. (2002) is admirable and need not be repeated here.

The revised Capparaceae includes some true trees (e.g., Boscia, Cadaba) as well as shrubs branched from the base (e.g., Capparis). Even though Capparaceae s. s. is a woody group, there are no wood features that clearly separate the family from Cleomaceae and Brassicaceae.

Wood plan. Wood consists of an ordinary woody cylinder, with growth rings minimal in the majority of species (Figs. 20a-d and 21a and e). However, some degree of successive camial activity has been reported in Boscia (Adamson, 1936), Cadaba, and Maerua (Metcalfe & Chalk, 1950).

Storying. Storying is minimally present, perhaps because intrusive growth of imperforate tracheary elements is prominent. Some examples of storying may be seen in Figs. 20b and 21b.

Vessels. Vessels are in radial multiples or chains (Figs. 20a and 21a and e). In some species, vessels are dimorphic in diameter, markedly in Apophyllum anomalum (Fig. 21e and 1), and less clearly so in Cladostemon, Maerua, Niebuhria, Quadrella (Fig. 20d), Ritchiea, and Stuebelia. Mean number of vessels per group ranges upward from about 1.5 (Capparis, Cladostemon, Quadrella, Fig. 21e) but mostly lies below 2.5.

Perforation plates are simple. Lateral wall pits of vessels is circular and alternate. Pits appear to be non-vestured in Maerua (Fig. 22d), and minimally vestured in Capparis spinosa (Fig. 20f). Vestures that are prominent but confined to pit apertures were observed in Atamisquea (Fig. 22f) and Apophyllum (Fig. 22c). Boscia hildebrandtii has vestures limited to the central portion of the pit cavity; these can be seen from both the inside (Fig. 22e) and on the outer surface of vessels (Fig. 22f). More abundant vestures can be seen on vessels of Crataeva (Fig. 23a and b), Quadrella (Fig. 23c) and Steriphoma (Fig. 23d). Crataeva has very distinctive vestures as viewed from the insides of vessels (Fig. 23f). The vestures extend from one widened pit aperture to another in a kind of network; there is merging of the abundant vestures.

Prominent grooves interconnecting pit apertures were observed in narrower vessels of Apophvllum (Fig. 2If). Helical thickenings occur in vessels of Atamisquea (Fig. 22f).

Imperforate tracheary elements. Librifonn fibers or fiber-tracheids with minimal borders (Fig. 20d) are present. A number of Capparaceae exhibit fiber dimorphism, a phenomenon observed in Asteraceae (Carlquist 1958, 1961) and in a range of other angiosperm woods (Carlquist, 2014). This feature, which denotes alternating patches of wide (presumably living) fibers and narrow fibers, can be seen to advantage in Capparis flexuosa (Fig. 20a-c), Quadrella cynophallophora (Fig. 20d), Cladostemon (Fig. 21a-c), and Apophyllum (Fig. 21e). No fiber dimorphism was observed in Maerua, Morisonia, Niebuhria, Ritchiea, and Stuebelia. Fibers are notably thin-walled and wide in Ritchiea.

Axial parenchyma. Axial parenchyma is mostly vasicentric scanty, which is also the predominant type in Brassicales. More abundant paratracheal parenchyma (about two layers thick around vessels) can be seen in Quadrella cynophallophora (Fig. 20d). Aliform-confluent parenchyma characterizes Niebuhria linearis (Fig. 2Id). Axial parenchyma can be easily distinguished from instances of fiber dimorphism, and the two co-occur in species with fiber dimorphism. Axial parenchyma strands are composed of two to five cells (Fig. 21b).

Rays. Both multiseriate and uniseriate rays, corresponding to Heterogeneous Type IIB of Kribs (1935), occur in Atamisquea, Maeriia, and Niebuhria. Multiseriate plus uniseriate rays composed of procumbent cells only, the Homogeneous Type II of Kribs, occur in Capparis (Fig. 20b) and Quadrella. Rays that are multiseriate and consist wholly of procumbent cells, the Homogeneous Type 11 of Kribs, were observed in Capparis flexuosa, Cladostemon kirkii (Fig. 21b), and Stuebelia.

Crystals. Rhomboidal crystals were observed in ray cells in Atamisquea (Fig. 23e), Cadaba (Fig. 23f), Capparis, Morisonia, Niebuhria, Quadrella, and Steriphoma. The crystals of Atamisquea are noteworthy for their lamellate structure. The crystals in Cadaba are apparently octahedral.

Comments. Wood features of Capparaceae are quite diverse, and include most of the character states found in other Brassicales, excluding presence of tracheids and interxylary phloem. Successive cambia, although reported in a few genera, are not characteristic of all species of those genera, may be infrequent, and need further investigation. Fiber dimorphism certainly occurs in several genera of Capparaceae characteristically, and has not been reported in any other family of Brassicales with the exception of Cleomaceae. Vestured pits are certainly well developed in the majority of Capparaceae and may represent the prime component of wood xeromorphy in the family. Vessel dimorphism, prominently present in Apophyllum, also occurs in Tropaeolaceae, and may represent another type of wood xeromorphy. There is a notable absence of juvenilistic features in wood of Capparaceae, suggesting that despite the abundance of herbaceous Cleomaceae and Brassicaceae, the three crown group families of Brassicales may have a woody ancestry.

18. Cleomaceae (data original and from Metcalfe & Chalk, 1950) (Fig. 24).

Because of the recent separation of Cleomaceae from Capparaceae (litis et al., 2011), there is no thoroughgoing account of wood of the family. Descriptions of wood of individual genera may be found in Solereder (1908) and Metcalfe & Chalk (1950) under Capparaceae.

Wood plan. Stems with a single cambium except for Cleome droserifolia Delile, which has successive cambia, illustrated by Fahn et al. (1986). The conjunctive tissue in that species contains a band of sclerenchyma. The remaining Cleomaceae studied have relatively wide rays (2-5 cells) separating fascicular areas with one or two radial strips of vessels (Fig 24a). The vessels are separated from rays by several layers of fibers, creating a "vessel restriction pattern" (Carlquist, 2009b).

Storying. Libriform fibers as well as vessels are storied in Peritoma (= Isomeris), which is a shrub with sufficient secondary xylem accumulation to show storying. Most species of Cleomaceae have less wood than Peritoma, hence, storying is not present in them.

Vessels. Vessels are in groups ranging from a mean of 1.4 vessels per group in Cleome artomala Kunth to 4.0 in Peritoma arborea (Fig. 24a). Vessels are not dimorphic in size, and are rather narrow (Fig. 24a and d) in them species studied.

Perforation plates absent. Vessel-to-vessel pits alternate and circular (Fig. 24c). Pits are prominently vestured (Fig. 24b). A wall interface between two vessels (Fig. 24c) show vestures on the wall seen from the lumen side (left), but vesture absence on the facing wall (right. This discrepancy may be due to handling, but it illustrates that not all vessel walls are alike when one searches to see whether vesturing is present.

Imperforate tracheary elements. Monomorphic libriform fibers are present in most species. Dimorphic fibers were observed in Peritoma (Fig. 24a, center).

Axial parenchyma. Vasicentric scanty, in strands of two cells, or undivided in Peritoma.

Rays. Rays are Heterogeneous Type IIB of Kribs (1935). Stems of Cleomella oocarpa A. Gray are rayless in earlier-formed secondary xylem (original observation).

Crystals. No crystals were observed in species studied.

Starch. Starch is abundant in the rays of Peritoma arborea.

Comments. The vessel restriction patterns of Cleomaceae are distinctive, and we have only preliminary explanations for this type of structure (Carlquist, 2009a) or several other non-random types of vessel distributions in woods. The high degree of vessel grouping in Peritoma is indicative of wood xeromorphy. Presence of dimorphic fibers in P. arborea is perhaps a link to Capparaceae and Brassicaceae. Although not all genera of these three families have dimorphic fibers, they are the only families in Brassicales in which they occur.

19. Brassicaceae (Metcalfe and Chalk, 1950; Carlquist, 1971; Schweingruber, 2006) (Figs. 26, 27 and 28).

References that describe wood of one or only a few species are listed by Gregory (1994). As the largest family (3710 species) of Brassicales, Brassicaceae is also minimally known with respect to wood anatomy. We do not even have thorough descriptions of ontogenetic stages and mature structure of secondary xylem of the cultivated species. To no small degree, this is related to the bias against less woody species where studies of secondary xylem is concerned. Absence of secondary xylem in herbaceous eudicots is, in fact, rare, and secondary xylem in herbaceous eudicots is worthy of investigation.

Wood plan. Most species have a single cambium; growth rings are inconspicuous (Carlquist, 1971). Latewood (as well as adjacent earlywood) bands of axial parenchyma that contain vessels are shown here for Stanleya pinnata (Fig. 25a, b and d). Such bands, which do not appear to be modifications of the basic axial parenchyma types (e.g., Kribs, 1937) also occur in Alyssum spinosum L., Brassica fruticiilosa Cyrillo, and Vella spinosa Boiss. (Metcalfe and Chalk, 1950). The entire background tissue may consist of storage parenchyma, as in roots of Raphanus sativus L. or thick-walled parenchyma may surround vessels, with thinner-walled parenchyma more distal, as in Armoracia (Fig. 27b), or an entirely fibrous background consisting of libriform fibers may be present, as in most of the woody Brassicaceae studied by Carlquist (1971). Lateral meristem activity, which can add some vascular tissue as well as parenchyma, is evident in Armoracia lapathifolia (Fig. 27a-c) roots ("rhizomes" of some authors). Interxylary phloem strands are present in Hirschfeldia incana (= Brassica nigra and B. aspera of various authors), as shown in Fig. 26c. Such variations are probably more widespread in Brassicaceae, and would provide interesting material for further examination. The annual habit may be regarded as formation of a single growth ring. (Fig. 26a and b) and often features diminution in vessel diameter as secondary growth ceases. The 3710 estimated species of Brassicaceae are mostly annuals and thus this pattern is probably widely represented.

Storying. Various degrees of storying from inconspicuous (Fig. 25c) to prominent (Crambe strigosa L'Her.), but storying may be entirely lacking in many annual species.

Vessels. Compared to other families of Brassicales, Brassicaceae have relatively narrow vessels, with diameter means from 16 to 71 [micro]m. with most species ranging from 30 to 50 [micro]m in the species assemblage studied by Carlquist (1971). The widest vessels in that survey occurred in a laurel forest shrub from Madeira, Cheiranthus mutabilis L'Her. (54 (tin, species mean); this species also has the longest vessel elements recorded for the family thus far, 100 [micro]m. Number of vessels per group is highest in Stanleya pinnata (Fig. 25a and b) and lowest in Cheiranthus mutabilis. Vessel grouping most commonly takes the form of radial multiples or radial chains, or a mixture of those with non-radial multiples (Fig. 25a and b; Fig. 26a and b).

Perforation plates are simple. Lateral wall pitting of vessels consists mainly of alternate circular bordered pits. The pit apertures facing the lumen may be interconnected by grooves (Figs. 26f and 27d) or such grooving may be absent, as in Matthiola. Degrees of prominence of the grooving as seen with SEM are shown in Fig. 26e, f and Fig. 27d. Pits are vestured (Figs. 26d-f, 27d-f and 28a-f). Vestures are evident if vessels are viewed from outer surfaces (Figs. 26d, 27c and 28b-f) or inner surfaces (Figs. 26e-f, 27d, f and 28a). The appearances in these two views are different. As viewed from the lumen side of a vessel wall, the vestures appear as branching corailoid outgrowths (Figs. 27f and 28a), which may appear to be in contact with the pit membrane--probably an artifact produced by drying. As seen from the outer surface, the vestures are seen as tips of the outgrowths (compare Figs. 27e with 27f; Fig. 28a with 28b). Thus far, there is no report of any Brassicaceae in which vestured pits are absent from vessels.

Vestured pits have hitherto been figured only in secondary xylem, and are shown mostly in vessels, although tracheids may also have vesturing (e.g., Bataceae, Koeberliniaceae and Stixaceae). However, my material of Raphamis raphanistrum proved favorable for demonstrating vesturing in primary xylem. Vestures can occur on the helical thickenings of primary xylem of both stems (Fig. 28d) and roots (Fig. 28e-f). In vessel elements that are transitional between helical and reticulate, vestures may be restricted to or more common in the angles where adjacent helices fuse (Fig. 28f). Vestured pits of four species of Brassicaceae were illustrated by Carlquist & Miller (1999).

Imperforate tracheary elements. Metcalfe & Chalk (1950) indicate that pits of "fibers" of wood of Brassicaceae may have simple or minutely bordered pits, and thus in the terminology used here, be libriform fibers or fiber-tracheids, respectively. In a survey of fiber dimorphism (Carlquist, 2014), Brassicales were not included, but on the basis of the present study, fiber dimorphism occurs in some Capparaceae as well as some Brassicaceae. Fiber dimorphism is present in all of the Brassicaceae studied by Carlquist (1971) except for Lepidium fremontii S. Waston and Parolinia ornata. Is axial parenchyma different from the wide, thinner-walled fibers where fiber dimorphism is present? In Brassicaceae, axial parenchyma is scanty vasicentric, and only exceptionally are wide seasonal bands (e.g., Fig. 25b) present. Fiber dimorphism, on the contrary, tends to occur in irregular patches, as illustrated here for Capparaceae. In addition, axial parenchyma in Brassicaceae is often subdivided into strands of two cells (arrow in Fig. 25d), whereas subdivided wide fibers are unusual, a distinction evident in radial sections. In some species, the two cell categories appear to merge, as they do not in the instances described in Carlquist (2014), in which clear instances of fiber dimorphism were selected for illustration.

Axial parenchyma. In addition to the vasicentric scanty parenchyma mentioned in the preceding paragraph, paatracheal sheaths more than one cell in thickness have been observed in Cheiranthus, Crambe, Descurainia, Lepidium, Sinapidendron, and Stanleya. In the roots of Armoracia lapathifolia, vessels are sheathed in thick-walled axial parenchyma, whereas thin-walled axial parenchyma forms the background tissue of axial xylem (Fig. 27a-c).

Some Brassicaceae have parenchyma bands interpolated into fibrous secondary xylem, notably Stanleya pinnata (Fig. 25a and b). These bands occur as latewood events, but are much wider than the one or two layers common in terminal parenchyma, and can include some vessels from other portions of a growth ring. Schweingruber (2006) offers some additional examples: Arabis alpina L., A. ciliata Clairv., and Ptilotrichium spinosum Boiss.

Rays. Rays of Brassicaceae can mostly be classified as Heterogeneous Type IIB of Kribs (1935). Upright cells, however, are more common in rays of such species as Cheiranthus mutabilis and Stanleya pinnata (Fig. 25c), suggesting that in a number of species, ray histology is transitional to Paedomorphic Type I (Carlquist, 1988); this indicates some degree of protracted juvenilism in ray structure. In Lepidium serra H. Mann the rays are all multiseriate with upright cells very common, so that Paedomorphic Type II is present. In Parolinia ornata Webb, rays are Homogeneous Type I of Kribs (1935). Ray cells are notably thick walled in Descurainia millefolia and Parolinia ornata, and could be termed sclereids. Raylessness occurs in Stanleya pinnata (Fig. 25a), in which rayless secondary xylem is the first secondary xylem produced; rays originate soon. Raylessness is likely to be found in other Brassicaceae with habits similar to that of Stanleya.

Crystals. Small rhombic crystals are present in the libriform fibers of Parolinia ornata (Carlquist, 1971; Schweingruber, 2006). Rhomboidal crystals were reported for Descurainia briquetii.

Comments. Although wood anatomy of Brassicacease is very similar to that of some Capparaceae and Cleomaceae, we as yet know relatively little about Brassicaceae in comparison to the size of the family. Although a number of Capparaceae are clearly woody, woodiness is limited in Brassicaceae: the shrubby Parolinia ornata and woodier cultivars of Brassica oleracea L are among the woodiest representatives of the family. Although other less woody representatives of the family may appear less pertinent to study for those grounded in timber species, wood anatomy of less woody species very likely has much to offer.

Character Change and Adaptation in Wood of Brassicales

This section is designed to show how Brassicales exemplify ecophysiological principles in their wood character states. These concepts, however, should apply to woody angiosperms at large, differing most notably where ecology and growth form differ modally from those of the various brassicalean families. For each category of structure, an interpretive hypothesis is given, and evidence from histology of wood of Brassicales is cited in a series of statements that follow the hypothesis.

1. Imperforate tracheary elements. Bailey & Tupper (1918) offered a muchreproduced drawing schematizing the tracheid as a primitive type of tracheary element in angiosperms; from this, vessels (first with scalariform, ultimately simple perforation plates) were progressively derived on the one hand, whereas on the other hand, imperforate tracheary elements progressively lost pit borders and changed (by implication, gradually) from conductive to mechanical cells. Some data by Metcalfe & Chalk (1950, p. xlv) seem to support this as a generalization, but is it always true? Can, in fact, this "trend" run in the other direction in some instances, or even abruptly change from tracheid to libriform fiber with few if any intermediate stages? Brassicales are an ideal group for demonstrating phyletic events because we have good molecular data showing the probable clades of the order, and we have a number of instances of shift from one type of imperforate tracheary element to another.

Hypothesis: Although in angiosperms at large the tracheid may have been the ancestral imperforate tracheary element type, the phylogeny of Brassicales (Fig. 1, column ITE at right) suggests five instances in which tracheids have been evolved from fiber-tracheids (Tropaeolaceae, Koeberliniaceae, Emblingiaceae, Pentadiplandaceae, and Stixaceae). Almost all vascular plants have the genetic information for form bordered pits, as vessels demonstrate, so transferring this information to imperforate tracheary elements can result in formation of tracheids in a clade that ancestrally has fiber-tracheids. In a similar way, simple pits on imperforate tracheary elements (= libriform fibers) have developed in some clades (Salvadoraceae, Tovariaceae, Borthwickiaceae, Resedaceae, Capparaceae, Cleomaceae, Brassicaceae); angiosperms have the genetic information to form fibers with simple pits (e.g. phloem fibers), so this genetic information can be applied to formation of imperforate tracheary elements.

Evidence for the above can be found in the following:

a. The early-departing branches of the Brassicales clade lack tracheids (in Tropaeolaceae, the tracheids are vasicentric tracheids, probably derived by vessel dimorphism). In Setchellanthaceae, Koeberliniaceae, Emblingiaceae, and Stixaceae, evolution of tracheids correlates with highly arid habitats. This evolution of "secondary tracheids" or "neo tracheids" can be noted elsewhere in angiosperms, as in Krameriaceae, Fabiana of Solanaceae, Rosmarinus of Lamiaceae, etc. (Carlquist, 1985a, 1992a, 1992b, 2005; Carlquist and Hoekman, 1985).

b. Tracheids confer safety to a conductive system because under conditions of negative pressure in water columns, the pit membranes prevent entry of air, whereas in vessels, and if a tracheid does embolize, emboli do not spread into other tracheids as they do in vessel elements, in which the simple perforation plates allow the cavitation to spread into much of a vessel (Zimmermann, 1983).

c. The families most likely to be ancestral to Brassicales (Fig. 1) are Malvales and Sapindales (Soltis et al., 2011), which most commonly have fiber-tracheids as an imperforate tracheaiy element type.

d. Libriform fibers in Brassicales may be considered, in agreement with the Bailey and Tupper (1918) scheme, to be an imperforate tracheary element type in which borders have been lost from pits on fibriform cells. If libriform fibers are dead at maturity, they may be presumed to have a mechanical function rather than a storage function.

e. Libriform fibers may be living fibers (fibers with extended longevity); septate fibers may be considered the most frequently encountered type of living fiber. Living fibers (Akaniaceae, Moringaceae) are believed to be involved in storage of water and photosynthates, but may play, at the same time, a mechanical role (Wolkinger, 1969).

f. The parenchyma background of secondary xylem of Caricaceae, like roots in some Moringaceae, can be regarded as libriform fibers that have transitioned out of a mechanical function into a storage function entirely. The mechanical tissue of Caricaceae consists of abundant secondary phloem fibers, by way of compensation.

2. Fiber dimorphism has been often overlooked, but must be included in any understanding of wood physiology. Brassicales and related orders offer occurrences that are likely to offer such understanding.

Hypothesis: Fiber dimorphism is a mechanism for production of wide, thinner-walled libriform fibers together with narrower, thicker-walled libriform fibers, the two types co-occurring in patches that are not discretely defined as are parenchyma bands, and are usually not subdivided into strands (Carlquist 1958, 1961, 2014). This dimorphism offers enough differentiation between mechanical and storage cells to qualify the two types of resultant cells as functionally different. The occurrences of fiber dimorphism in Brassicales and related groups offer the following correlations:

a. In species with fiber dimorphism, there is not a proportionate decrease in axial parenchyma volume, so the wide living fibers offer a net addition of storage tissue.

b. As shown by Sauter (1966), the storage in wide fibers of Acer (formerly Aceraceae, now Sapindaceae) is in the form of starch which can be rapidly hydrolyzed into sugar that increases the sugar content of sap in vessels. This mechanism has not been studied in other species, but undoubtedly is operative, judging by the occurrence of wide fibers similar to those of Acer.

c. Fiber dimorphism can be identified in some species of each of the three families Capparaceae, Cleomaceae, and Brassicaceae, which form the "crown group" of Brassicales. Similar distributions occur in Sapindales (Carlquist 2015a, 2015b). In all of these instances, fiber dimorphism is probably an apomorphy.

3. Vestured pits. Vestured pits occur in lateral wall pits of vessels of about half of the families of Brassicales that have been investigated with SEM (Fig. 1, column VES at right). Because all of the Brassicaceae investigated in this respect have proved to have vestured pits, the proportion of Brassicales with vestured pits may be much larger at the species level. Sampling the secondary xylem of thousands of species of Brassicaceae is not practicable, but sampling species selected according to ecological categories and systematic groupings would probably give good indicators of the presence of vestured pits in the family at large.

Hypothesis: Vestured pits offer a mechanism for prevention of, or repair of embolisms in vessels, or possibly both. Thus, vestured pits would be valuable in habitats where embolism risk is high: areas that are seasonally very dry, or where freezing of soil (which makes moisture unavailable to the plant). Potentially, vestured pits, vesturing on vessel walls, and similar structures (warts, which may not be separable from vesturing) are the most effective form of xeromorphy, but they may be complicated structures to evolve and therefore are not widespread. A very plausible functional explanation, that they are hydrophilic, has not been offered yet, but may prove to be operative.

a. The work of McCully et al. (2014) shows that in Zea, walls of vessels are mostly hydrophilic, but with hydrophobic patches. These authors find that pit borders are hydrophobic, but turn hydrophilic when water touches them. We do not have any data yet on whether or not vesturing is hydrophilic or hydrophobic, because no species with vestured pits have been studied by the methods used by McCully et al. (2014). The results of such studies are likely to help us refine our ideas about the function of vesturing. Work on surface topography effects of vessel walls (Kohonen, 2006; Kohonen & Helland 2009) are very likely relevant in this respect. Angle of structures related to pits has been mentioned (e.g., Jansen et al., 2003), but this does not explain instances of vestured pits that occur as mere irregularities along the margins of pit apertures (e.g., Batis, Setchellanthus). The ideas offered to date on the mechanisms for functioning of vestures are unsatisfying or vague (see Jansen et al. 2003).

b. Circumstantial evidence from geographic distribution of vestured pits has been offered by Jansen et al. (2004), who claim that vestured pits are most common in both mesic and dry tropical lowlands. Only a small number of eudicot families have vestured pits: 48 families in 11 orders (Jansen et al., 2001). These authors claim that the percentage of vestured pits drops to zero in boreal regions, but the large family Brassicaceae (3710 species) is largely boreal and cold temperate, and no species of that family were included in the survey by Jansen et al. (2004). To date, however, no species of Brassicaceae has been reported to lack vestured pits. The radiation of Eucalyptus and other Myrtaceae in Australia as well as the occurrence of vesturing in a number of Fabaceae, account for the idea that tropical and subtropical lowlands host the majority of examples reported to have vestured pits, so systematics and speciation of particular groups is certainly involved here. The most speciose families of Brassicales, Brassicaceae and Capparaceae, may owe their success to having vestured pits.

c. Vesturing does not evolve readily (present in 48 eudicot families according to Jansen et al., 2001) whereas other forms of topographic relief on vessels (helical thickening, grooves) are apparently easily achieved morphogenetically (note that they probably parallel cyclosis in the last stages of wall formation). In fact, we may be amazed that vestured pits have evolved as frequently as they have, considering their complexity. That they are as widespread as they are despite the precision and intricacy of their structure evidences the strong selective value that they have. Vestures may be lost, or suppressed, within particular families (e.g., Moringaceae). Vesturing is uncommon in plants with succulence (e.g., Caricaceae, Gyrostemonaceae and Moringaceae in Brassicales).

d. The examples of vesturing in Brassicales show that imaging vestures from the outer surfaces of vessels is insufficient. As one example, the rich assemblage of fused vestures on the insides of vessels of Crataeva (Capparaceae) would have been overlooked had only outer walls of vessels been studied. Vesturing in pits is three-dimensional and deserves study that can explore that. Sections through pits and TEM studies may also prove useful.

e. The imperforate tracheary elements of some Brassicales have vestured pits, notably those of Batis (Bataceae), Forchhammeria and Stixis (Stixaceae), Koeberlinia (Koeberliniaceae), as well as Pentadiplandra (Pentadiplandraceae). Occurrence of vestured pits on imperforate tracheary elements is always associated with fully bordered pits, and together, these two features are sufficient to declare such elements to be conductive cells, and therefore tracheids.

f. The demonstration of vestures on helices of primary xylem of Raphanus (Brassicaceae), a first report for angiosperms, is interesting in that there is not a clear connection between this occurrence and functioning of vestures, and it may represent a morphogenetic or developmental phenomenon in which complete exclusion of vestures from primary xylem in a species with vestured pits in secondary xylem is not always achieved. The occurrence of the vesture-like papules on flanges paralleling perforation plates in Azima is also an unprecedented occurrence.

4. Helical sculpture in vessels. Brassicales contain all of the main kinds of helical sculpture in secondary xylem vessels: helical thickenings, grooves widening pit apertures or connecting helical series of pit apertures (coalesced pit apertures), and thickening bands adjacent to grooves. The varied contexts of these occurrences suggest several possible interpretations.

Hypothesis: The hydraulic significance of these structures is related to change in surface topography that controls wettability of surfaces. Kohonen (2006) suggests "engineering the roughness of the capillary [inner surface of a vessel or tracheid] walls to achieve complete wettability." Enhanced wettability would permit recovery from cavitation, whether caused by drought or freezing.

a. McCully et al. (2014) and Brodersen and McElrone (2013) have found patchy distribution of hydrophilic wall surfaces, although their work does not specifically address helical sculpture. Further studies in this regard are much needed.

b. The occurrence of helical thickenings in the wet forest Akaniaceae may relate to cold conditions rather than to drought. Aridity seems related to helical thickenings in Setchellanthus.

c. Grooves interconnecting pit apertures in vessels are almost universally present in the woody Brassicaceae studied by Carlquist (1971). These species experience wet winters but dry summers.

d. Helical sculpture in vessels is highly correlated with xeric habitats in Asteraceae (Carlquist, 1966) and in the woody flora of southern California (Carlquist and Hoekman, 1985).

5. Vessel grouping, Although vessel grouping or lack of vessel grouping in woods has been figured beginning in the earliest works on plant anatomy, such as that of Grew (1682), Its physiological significance has been appreciated only recently (Carlquist, 1984).

Hypothesis: Vessel grouping is a form of xeromorphy that is present in angiosperm woods that do not have tracheids as am imperforate tracheary element type. Tracheids (sensu Bailey, 1936; Carlquist, 1988; IAWA Committee on Nomenclature, 1964; Sano et al., 2011) in angiosperm woods are more effective than vessel grouping as a xeromorphic feature, because vessels are mostly solitary in woods that have tracheids (or abundant vasicentric tracheids) as an imperforate tracheary element type. In woods with libriform fibers or fiber-tracheids, degree of grouping is proportional to habitat aridity. Vessel grouping represents a nonrandom form of vessel placement (Carlquist, 2009a). Solitary vessels randomly placed would be expected in angiosperm woods unless there is some factor promoting non-random placement. The effect of vessel grouping is one of redundancy, not of embolism reduction per se.

a. The vessel grouping of some species of Brassicaceae is relatively elevated: 5.7 vessels per group in Cheiranthiis scopariits Brouss., 5.9 in Stanleya pinnata\ most Brassicaceae have values between 1.5 and 2.2 (Carlquist, 1971).

b. Radial chains of vessels potentially represent not merely redundancy, but commissioning of new vessel elements so as to maintain the pattern of water columns in xylem as older vessels cease to conduct actively. This tendency is shown conspicuously in Cleomaceae.

c. Minimal vessel grouping is shown in the brassicalean families Emblingiaceae, Koeberliniaceae, Pentadiplandraceae, and Stixaceae. These families all have tracheids as an imperforate tracheary element type. Tropaeolaceae, which have abundant vasicentric tracheids, also exemplify this phenomenon.

6. Vessel diameter. Vessel diameter was early recognized as exemplifying xeromorphy (Carlquist, 1966, 1975). The resistance of latewood vessels to cavitation was demonstrated by Hargrave et al. (1994). Wide vessels have conductive efficiency, as cited by Zimmermann (1983).

Hypothesis: there is a trade-off between the conductive ability of wide vessels and the conductive safety of narrow vessels. Both can be accommodated in a single wood, by means of growth rings in which latewood features narrowing of vessel diameter. Vessel diameter may increase in diameter in a stem if a plant is able to tap deeper and therefore moister levels of water availability, but in many groups, decrease in vessel diameter accompanies senescence. The more shallow the roots of a plant at maturity, the more likely it is to show narrowing of vessels with age. Particular growth forms (vines, lianas) show wider vessels than one would expect from shrubs or trees of the same stem diameter, but vessel dimorphism (few wide vessels, more numerous narrow vessels) is common in scandent species.

a. Mesomorphy in woods of Brassicales is unusual, but is represented in Akaniaceae, as well as in species of Capparaceae which are native to moist forest areas. The wide vessels of Caricaceae and Moringaceae may relate to succulence more than to steady soil moisture (Olson & Carlquist, 2001).

b. Scandent species are few in Brassicales, but the wide vessels of Tropaeolaceae exemplify the tendency for vining species to have wider vessels and vessel dimorphism.

c. Brassicales that represent particular types of xeromorphy and grow in dry habitats include Koeberliniaceae (deserts with summer rainfall), perennial Resedaceae (arid scrub), and Setchellanthus (subtropical desert scrub). Bataceae grow in moist maritime habitats that are saline, which is a type of physiological drought. This is also true of Cleome droserifolia (Fahn et al., 1986).

d. Annuals such as Hirschfeldia incana represent a single growth ring, with narrow vessels formed as the plant goes into flowering and fruiting. Vessel dimorphism occurs in non-scandent Brassicales. It is particularly common in Capparaceae (Apophyllum, Atamisquea, Capparis, Maema, Niebuhria, Quadrella, Ritchiea, and Stuebelia). These are characteristically dry land shrubs in which narrower vessels could maintain conductive pathways even if wider vessels in a group cavitated.

7. Vessel density. Vessel density has been considered, along with vessel length and vessel diameter, a chief indicator of conductive characteristics. It should be roughly inverse to vessel diameter, based on packing considerations.

Hypothesis: deviations from the expected value (inverse of vessel diameter) do occur (Carlquist, 1975, p. 183), and may tell us some unexpected information about the wood plans of particular species and their function. In particular, vining species have fewer vessels per sp. mm than expected because their conduction is related to the fourth power, not the square or cube, of the vessel diameter (the Hagen-Poiseuille equation). Succulents probably have less than peak flow because of lowered transpiration rate, and have lower vessel density accordingly. Plants with successive cambia have lower density because older vessels are still active in conduction, so the conductive area of a stem is greater than in species that retire vessels sooner.

a. Tropaeolum does not have fewer vessels per sq. mm than the vessel diameter would dictate, but it does have fewer wide vessels. The narrower vessels confer conductive safety, countered by the high flow capacity of the larger vessels.

b. More succulent portions of Moringa (Olson & Carlquist, 2001) have fewer vessels per sq. mm. than expected, probably because peak transpiration is lower during the warmest months, this in turn related to drought deciduousness. Caricaceae (Carlquist, 1998a) has only slightly less vessel density than expected based on vessel diameter, probably because its foliage has greater transpiration than does the foliage of a succulent with few or no leaves during the warmest months.

c. The imperforate tracheary elements of Forchhammeria (Stixaceae) is consist wholly of tracheids, which have great conductive safety (Carlquist et al., 2013). The foliage of Forchhammeria consists of narrow leaves with recurved margins and thick cuticle, likely to have transpiration characteristics lower than those of broad, thin leaves. More importantly, each of the successive vascular cambia continues to produce secondary phloem for long periods of time, suggesting that vessels function for longer, and thus a greater actual conductive area is achieved in a given stem without adding very many vessels in newer vascular increments. Some portions of Forchhammeria secondary xylem are devoid of any vessels.

8. Vessel element length follows trends in vessel diameter, vessel density, and vessel wall sculpture by being shorter in plants of xeric regions. The reasons for this may be multiple.

Hypothesis: Because air emboli tend to stop at perforation plates, regardless of whether perforation plates are scalariform or simple (Slatyer, 1967), shorter vessel length may be disabled in a plant with shorter vessel elements. In addition, shorter vessel elements, characteristic of plants of drier regions (Carlquist, 1975; Carlquist & Hoekman, 1985), are not likely to be concomitantly wide: deviations from a straight-line length/width ratio do occur, but are not extensive (Carlquist, 1975, page 183), for reasons of optimal cylindrical strength design.

a. Vessel element length is notably short in Brassicaceae (Carlquist, 1971), which are mostly short-lived annuals or perennials.

b. The only family of Brassicales with relatively long vessel elements is Akaniaceae, in which both monotypic genera are native to moist forests.

c. Within Brassicaceae, the longest vessel element lengths occur in the species with the most mesic habitat, Cheiranthus mutabilis, from Madeiran laurel forest openings (Carlquist, 1971).

9. Axial parenchyma. Scanty vasicentric is the prevailing type of axial parenchyma present in Brassicales. The presence of other types of parenchyma in particular families invites interpretation based on ecophysiological and growth form characteristics primarily.

Hypothesis: A few cells adjacent to a vessel or a vessel group (vasicentric scanty parenchyma) suffice to control the conductive process in vessels. Additional thickness of parenchyma sheaths (paratracheal abundant) around vessels may serve for intermediate steps in the conductive process, such as storage. Although some of this process remains hypothetical (Zwieniecki and Holbrook, 2009), the role of axial parenchyma and its exchange of solutes with vessel sap seem clear in terms of histological association (Carlquist, 2015a). Although mechanical tissue tends to be prevalent in first-formed wood of a stem in a sufffuticose shrub, bands of parenchyma (which are especially common in latewood positions) may be added and have functions related to such functions as prevention of freezing damage. Conversion of part or all of the axial secondary xylem to parenchyma relates to succulence and storage.

a. The parenchyma bands seen in Arabis, Ptilotrichium, Stanleya and other genera of Brassicaceae occur in genera and species that experience considerable drought and/ or freezing exposure (Carlquist, 1971; Schweingiuber, 2006).

b. Instances of aliform-confluent axial parenchyma in Brassicales are scarce (Niebuhria, Fig. 2Id; Sinapidendrori), and may relate to the balance between storage and mechanical strength and woodiness of these species.

c. The water, starch, and sugar storage by parenchyma of widely known vegetables (e.g., Brassica oleracea) and condiments (Armoracia lapathifolia), although increased in extent in cultivars, is basic to storage for flower and seed production or for perennation.

d. Succulence in Caricaceae and Moringaceae, achieved by pervasive axial parenchyma in secondary xylem, relates to survival of their arborescent and sarcorhizal growth forms in habitats that provide a seasonal water supply (Olson & Carlquist 2001).

e. Bordered pits on cross-walls of axial parenchyma strands (Akania and Bretschneidera) are indicative of active axial flow of photosynthates in axial parenchyma,

10. Rays. The function of rays relates to histological feature: cell shape; types of pitting; and contents of ray cells. The predominant ray type in Brassicales is Heterogeneous Type II (which is also the most common type in angiosperm woods).

Hypothesis: Heterogeneous Type II rays feature multiseriates with a central group of procumbent cells, plus some upright sheathing cells, and few or no wings at tips of multiseriate rays. More abundant upright cells in such a ray qualifies it as Paedomorphic Type I, and relates to axial flow patterns in rays, whereas predominance of procumbent cells (Homogeneous Type I or II) indicates radial flow in rays. Active flow of photosynthates is indicated by presence of bordered pits, which are especially common in tangential walls of procumbent cells, indicating radial flow in procumbent cells. Raylessness is often exhibited in the form of first-formed secondary xylem that has only mechanical elements, and rays are initiated later; it is indicative of a trade-off between mechanical strength and radial flow of photosynthates in ray parenchyma in "woody herbs" that exhibit other features of secondary woodiness (Carlquist, 2015b).

a. Raylessness occurs in a number of Brassicaceae in which there is branching of stems from the base of a shrub or subshrub, and in which each of these branches bears an appreciable weight in terms of flowers and fruits. Raylessness at the beginning of secondary growth, yielding to development of rays thereafter, was observed in the brassicaceous species Cheiranlhus mutabilis, Lepidium fremontii, Matthiola maderensis Lowe, Parolinia ornata, Sinapidendron sp. (Carlquist 2760, RSA), and Stanleya pinnata (Fig. 25a) (new reports)

b. Procumbent cells are most abundant, even present exclusively in some species, in Akaniaceae, Capparaceae, Koeberliniaceae, and Salvadoraceae. This is especially true in the larger wood samples.

c. Presence of upright ray cells predominantly or exclusively is seen in Tropaeolum, some Resedaceae, and numerous insular Brassicaceae (Carlquist, 1971; Lens et al. 2013). This ray feature is prominent in instances of secondary woodiness (Carlquist, 1962, 1969; Lens et al., 2013). This feature is present in species with protracted juvenilism (also termed paedomorphosis, which implies sexual reproduction while in a juvenile state of development).

Brassicales and Beyond: Basic Ideas in Wood Evolutionary Theory

Reversibility in Wood Characters: How Prevalent Is It?. In a thoughtful review of this topic, Baas and Wheeler introduced us to some elements of the question. Bailey (1944) assert4ed that trends associated with vessel element specialization are irreversible. He used vessel element length as a kind of key to advancement in wood features. Leaving aside fluctuation in character expression, was he right? Baas and Wheeler (1996) think that reversion can take place. However, reversibility/irreversibility is not a single concept, but several.

Hennigian character state methodology and its problems. The level of irreversibility that has received attention with respect to cladistics methodology is character state change as seen from morphological studies. For example, Koeberliniaceae, Pentadiplandraceae, and Stixaceae have tracheids, whereas fiber-tracheids with nearly simple pits appear basic to Brassicales. Are these families reverting to an ancient type imperforate tracheary element? Bailey & Tupper's (1918) diagram of evolution of progressively more specialized vessels on the one hand, and progressively more specialized imperforate tracheary elements on the other, beginning with a tracheid with fully bordered pits as the ancestral cell type. So have the three families of Brassicales just named reverted to an ancient cell type? Not really. The genetic information for formation of bordered pits has not been lost, as can be seen from bordered pits on vessels in all Brassicales, so it's merely a matter of applying that information to the formation of imperforate tracheary elements. Hence, we have "secondary tracheids" or "neotracheids" in Koeberliniaceae, Pentadiplandraceae, and Stixaceae. Baas and Wheeler might see this as a character reversion. We have been much influenced by the 0 and + designation of character states in the Hennigian mode. But character states are rated by human observation, whereas the plant may arrive at a character state in any of various ways. Similar instances could be cited elsewhere in angiosperms. Fabiana is exceptional in Solanaceae in having tracheids, Rosmarinus similarly exceptional in having vasicentric tracheids in Lamiaceae (Carlquist 1992a, 1992b). In fact, Baas & Wheeler (1996) do present a table citing reversions for the prevalent sequence from bordered to simple pits on imperforate tracheary elements in angiosperms. If Baas and Wheeler could have used the concept of "secondary tracheids," their table would be different.

Unfortunately, a table of Baas & Wheeler (1996) combines both the pit evolution and evolution of simple perforation plates from scalariform perforation plates, so we cannot separate their reversion estimates for these two characters. In another table, they list families in which scalariform and simple plates are both present. However, the families cited have different stories. Dilleniaceae has long scalariform perforation plates in Dillenia front New Caledonian rain forest, but simple perforation plates in Hibbertia from dry Western Australian scrub. Araliaceae and Styracaccae, on the other hand, have scalariform and simple perforation plates in the same wood--the scalariform plates more common in latewood. And one can cite other modes of occurrence, showing that a simple 0 and + designation may be used in a data matrix, but may be misleading.

Ratchet Theory. Olson (2014) and Jansen & Nardini (2014) have attributed to me the idea of an evolutionary ratchet, which denotes evolutionary progression in a character and associated characters to the extent that reversion is not possible. Thus, change to a simple perforation plate in a clade may involve loss or silencing of information leading to the formation of bars on a perforation plate, and inevitably, other features in a wood (e.g., ray type) also change. This might explain the instances of perforation plates in Akaniaceae and Tropaeolaceae. In these families, perforation plates are simple, except for a very few plates that are imperfect versions of scalaforn patterns. A scalaforn perforation plate pattern will likely not return to such clades. We are not sure why the occasional malformed plates occur, although one can furnish ideas.

If scalaforn perforation plates arc associated with mesic conditions, simple plates in any given clade are not disadvantaged in mesic habitats, should there be a shift from more arid habitats where simple plates evolved in a clade, into some new mesic conditions. Virtually all of the plants on the summit of Mt. Waialeale, supposedly the wettest place on earth, have simple perforation plates.

Gene theory. If one looks at evolution at the gene level, one has a different perspective. Evolution is a forward progression, Even if a few genes (like those for variegation) can switch expression readily, most genes cannot, and a series of processes--silencing, modification, multiplication, loss, inversion of segments, pleiotropy--may be involved. Obtaining a picture of gene changes in wood anatomy for any given clade is so complex and perhaps impossible that we cannot readily think in these terms. But if we do use this perspective, we see that successive gene and gene combination changes are ongoing, they never return us to the ancestral DNA sequences.

Xeromorphy in Wood: Tiers of Effectiveness and Types of Action in Characters. In Asteraceae, number of vessels per group as seen in transverse section increases markedly with habitat aridity, as in Olearia (Carlquist 1960, 1966). Yet Krameria, Prinus, and some other desert genera never group vessels. Noting this, I reviewed the data on vessel grouping and for imperforate tracheary element type for each family in Metcalfe & Chalk (1950), and found a hitherto unappreciated correlation. Vessels also do not group in genera with abundant vasicentric tracheids (Eucalyptus, Quercus). Thus, the conductive safety provided by vessel grouping is subordinate to the safety provided by tracheids (Carlquist, 1984). The ultimate conclusion of this line of thinking is that no two xeromorphic adaptations have the same value, and the occurrence of one precludes the occurrence of others. Most characters, other than the pair just mentioned, are additive in nature.

Tier 1. Tracheids or abundant vasicentric tracheids; vestured pits.

Tier 2. Vessel diameter; vessel density (number of vessels per sq. mm); vessel length; growth rings.

Tier 3. Vessel grouping; vessel diameter dimorphism; helical thickenings in vessels.

Tier 4. Vessel element length; water storage parenchyma in wood.

The assignment of relative values to each of these characters is tentative. Rankings may change if a family or genus lacks a particular character. For example, numerous genera of Brassicales have vestured pits; Asteraceae lack vestured pits and tracheids (except for Loricaria), so Asteraceae show more reliance on Tier 3 characters. Interestingly, Brassicaceae with wood or leaf succulence (Bataceae, Caricaceae, Gyrostemonaceae, Moringaceae, Resedaceae) have minimal vesturing or none at all.

Particular families show quite different priorities for characters that tend to insure conductive safety. In cacti, Mauseth (1993) cites wide-band tracheids (characteristic of no other family except for Anacampserotaceae) and water storage parenchyma, which may be placed variously with respect to mechanical tissue. The succulent nature of stems as a whole in cacti permit the volumetric change that extends to wide-band tracheids, which would be non-adaptive in a family with no volumetric change of stems in accordance with changing water content.

The co-occurrence of two Tier 1 features, vestured pits and tracheids or vasicentric tracheids, in Myrtaceae can be correlated with the amazing radiation of Eucalyptus and other myrtalean genera in Australia. In this respect, we should note that vestured pits are varied, and that they do not lend themselves to experimental work in wood physiology because of their minute size.

Heterochrony: An Extensive and Nuanced Source of Diversity in Angiosperm Woods. Heterochrony is the umbrella term for protracted juvenilism and accelerated adulthood. The terms paedomorphosis and progenesis have been used for those two phenomena, respectively, but the terms are zoological ones and are doubtfully applicable to organisms with an open system of growth. The wood of conifers shows no protracted juvenilism (except perhaps in rays of Welwitschia), However, angiosperms have a rich assemblage of juvenile features the expression of which may extend for various periods of time--perhaps for the entire vegetative history of a particular plant (Carlquist 1962, 2009a). Once can say that all of these features relate in some way to cambial activity:

Juvenile

Cells upright only or predominantly in multiscriatc rays

Rays little changed during ontogeny

Raylesss

Length-on-age curve for vessel clement length descending

Storying absent or onset delayed

Cambium in bundles absent (monocots)

Pseudoscalariform pitting on vessels

Adult

Multiscriatc ray cells procumbent, upright ray cells on sides and tips of rays

Rays subdividing, widening, or initiated more frequently as a stem or root grows.

Rays present

Length-on-age curve for vessel elements

Once can say that all of these features relate ascending, then leveling

Storying onset earlier

Cambium present

Rapid progression to circular or oval vessel pits.

The inclusion of absence of cambium in monocots and a few non-monocots (Nymphaeales), may seem unexpected, but absence of a cambium is the ultimate juvenile condition. One can see minimally active cambia in some Saururaceae, leading to cambial absence; Saururaceae are not ancestral to monocots (although they are not far from monocot origins), but the reduction of cambial activity in their stems very likely simulates the stage preceding monocot stem origin. Could one assume that short duration of cambial activity is equivalent to juvenilism, and that prolonged cambial activity equates to woodiness? As a generalization, this seems permissible. Annual herbs can certainly be considered a juvenile growth form, albeit with some special features, and wood of annuals contains some or most of the features listed above.

There is no one kind of descending length-on-age curve for vessel elements (Carlquist 1962, 2013). These remain to be explored.

Juvenilistic wood characters can and do occur independently of each other, and one may find only one, or several in a given wood.

Do Brassicales show protracted juvenilism? Certainly the many annuals in Brassicaceae qualify, as do the perennials (Lens et al., 2013). Olson (2007) shows how Moringa stem and root anatomy can be interprete4d in terms of heterochrony, and similar considerations can be applied to Caricaceae. Tropaeolaceae can be considered a relatively non-woody derivative from ancestors like the arboreal family Akaniaceae. Tropaeolaceae have upright ray cells exclusively, as do some Resedaceae and some Gyrostemonaceae. The non-woody Cleomaceae have juvenilistic secondary xylem. With the exception of Brassicaceae, Brassicales are not exceptional among angiosperm orders in their degree of heterochrony.

Heterochrony in wood and in other respects of plant structure is basic to large territories of angiosperm evolution. The orders Asterales and Lamiales may be the richest assemblages of species representing juvenilistic structure modes, both in xylem and in other plant portions. A case has been made for considering basal angiosperms as stemming from less woody ancestors (Donoghue & Doyle, 1989; Taylor & Hickey 1996; Carlquist, 2009b), and we can no longer believe that trees are the basic growth form from which other angiosperm growth forms were derived. The full implications of heterochrony in angiosperms remain to be explored, but we can safely say that heterochrony, in addition to the simultaneity of embryo and endosperm development, was a major factor in the rise of angiosperms.

In Wood Anatomy, Which Characters are "Functional," Which are "Taxonomic"?. Solereder's thesis (1885), Uber den systematischen Wert der Holzstrukture bei den Dicotyledonen," set a precedent followed by his (1908) compilation "Systematic anatomy of the Dicotyledons," and Metcalfe & Chalk's (1950) "Anatomy of the Dicotyledons." The arrangement of the data by families is not just for the convenience of accessing data on particular plants: it was developed from the conviction that wood features could be used as systematic tools for wood identification. The use of such data for wood identification continues to the present. However, any thoughtful viewer of these books will say, "why do particular woods have the assemblages of features that they do?"

One must remember that the great majority of woody species is in tropical latitudes, whereas academic and especially forest institutes where wood anatomy was studied were in cold temperate regions. Thus, students of wood anatomy did not know the growth form and ecology of the woody species they studied. More importantly, with many different woods to compare, evolution of particular characters with respect to ecological and climatic factors was not highlighted. A situation of great simplicity was offered by the family Asteraceae, a family of about 26,000 species in which woods are essentially identical (owing to the recentness of radiation of the family) except for vessel characters that relate to ecology. By studying a large number of asteraceous woods from known habits, the patterns of vessel evolution were dramatically evident (Carlquist, 1966). The application of the knowledge gained in Asteraceae to angiosperms at large proceeded (Carlquist, 1975; Carlquist & Hoekman, 1985). Now, each family for which wood anatomy is reviewed, such as those of Brassicales, can be compared to our knowledge of ecological wood anatomy. Of course, physiological principles underlie the correlations, and students of wood physiology have been eager to discover those principles (e.g., Jansen & Nardini, 2014).

With the advent of molecular phylogeny, we no longer look to wood anatomy as a source of materials from which to build phylogenetic systems. More importantly, we realize that wood anatomy represents a series of character combinations, unlike from clade to clade, that offers various ways of satisfying the water economy needs of particular plants. We have now attributed ecophysiological functions to many wood characters. Not all wood characters relate to water economy. Mechanical functions of wood; carbohydrate and water storage; and herbivore deterrence are chief among the non-hydraulic features wood offers. If we are willing to interpret features like starch storage; crystals, silica bodies, gum or resin-like compounds; and various wood "fiber" configurations in terms of function, we have no residue of "purely taxonomic features." Assemblages of functional wood character states in one genus or family can, however, be compared to those in another. Thus, in Brassicales, we see that rays of two distinct sizes, septate or living fibers, helical thickenings in vessels, and curious rare malfonned scalariform perforation plates link Akania and Bretschneidera, previously widely separated in phylogenetic systems, into the family Akaniaceae. We also find resemblance between wood of Tropaeolaceae and that of Akaniaceae. Resemblance in functional systems between taxa that are shown to be closely related on the basis of molecular phylogeny is an expected outcome. There are some instances in which habit, wood anatomy, and habitat of one clade have veered markedly away from close relatives (e.g., Limnanthaceae, Koeberliniaceae, Stixaceae). Wood and other features of the family of annuals Limnanthaceae show few resemblances to the families closest to them in the system of Brassicales, Caricaceae, Moringaceae and Setchellanthaceae (Fig. 1).

Are there "relictual" features in angiosperm woods that persist even though they have little function? The scalariform perforation plate is an obvious example, because such perforation plates are common in basal woody angiosperms. The available evidence, inadequate though it is (see Jansen & Nardini, 2014), suggests that scalariform perforation plates may serve more than one function, including ones those authors do not list (e.g., Slatyer's 1967 idea that embolisms tend to terminate at perforation plates and thus stop disabling of an entire vessel). Those who do experimental work are understandably uncertain about how to proceed in such situations. Baas & Wheeler (1996) cite Zimmermann's (1978) suggestion that bars on perforation plates sieve out bubbles that form after frozen water in vessels thaws, but only half of boreal angiospenns have scalarifonn perforation plates. Moreover, the majority of scalariform perforation plates are not located in areas that freeze, but rather in frost-free cloud forests of Malaysia, the Andes, equatorial Africa, and Indonesia. When confronted with multiple and uncertain scenarios like these, those interested in wood hydraulics understandably opt for some experimental procedure that does not involve scalariform perforation plates. Synthesis is the ultimate goal, but wood anatomy is such a varied terrain in angiosperms that both separating out individual functions and then synthesizing them across phylogenetic lines requires a level of knowledge that may not be possible, even in a group of well-chosen collaborators. Ironically, when we look at anatomical preparations of woods of the world, we can see the results of natural experiments, but the complexity of angiosperms at large is so great, and the limitations of any individual's expertise are so real that while individual wood features and individual species can be studied, the needed synthesis of information all too often remains elusive.

DOI 10.1007/s12229-016-9161-2

Acknowledgments Dr, Mark Olson kindly provided material of Forchhammeria and Moringa; his contribution to study of Brassicales extends well beyond those samples, however. Dr. David Boufford kindly provided materials of Borthwickia and Stixis. Wood samples from the xylarium of the Forest Products Laboratory of several families of Brassicalcs were provided through the kindness of Dr. Regis Miller and other staff members of that institution. Wood samples of Capparaccac were provided through the kindness of the curator of the wood collection of the U. S. National Museum of Natural History (Smithsonian Institution). Most of the SEM micrographs were obtained with the SEM at Santa Barbara Botanic Gardens, and the directors of that institution (Dr. Edward L. Schneider, Dr. Steven Windhager) deserve thanks for giving me access to that machine and keeping it in repair. Dr. Jocelyn Hall and the University of Wisconsin Herbarium furnished a twig of Penladiplandra arborea. John Garvey applied scales, lettering, and symbols to the photographic figures.

Literature Cited

Adamson, R. S. 1936. Notes on the stem structure of Boscia rehmanniana. Proceedings of the Royal Society of South Africa 23: 297-301.

Arnold, G. H. & L. G. M. Baas Becking. 1949. Notes on stem structure of Carica papaya. Annals of the Botanical Garden, Buitenzorg 51: 199-230.

Baas, P. & E. A. Wheeler. 1996. Parallelism and reversibility in xylem evolution. A review. IAWA Journal 17: 351-364.

Bailey, I. W. 1936. The problem of differentiation and classification of tracheids, fiber-tracheids, and libriform fibers. Tropical Woods 45: 18-23.

-1944. The development of vessels in angiosperms in morphological research. American Journal of Botany 31: 421-428.

-& W. W. Tupper. 1918. Size variation in tracheary cells. 1. A comparison between the secondary xylems of vascular cryptogams, gymnosperms, and angiosperms. Proceedings of the American Academy of Arts and Sciences 54: 149-204.

Bolle, F. 1936. Rcscdacoac. In: A. Engler & K. Prantl (eds). Die naturlichen Pflanzenfamilien 17b: 659-682. Verlag Wilhelm Engelmann, Leipzig.

Brodersen, C. R. & A. J. McElrone. 2013. Maintenance of xylem network transport capacity: a review of embolism repair in vascular plants. Trends in Plant Science 4: 1-11.

Carlquist, S. 1958. Wood anatomy of Heliantheae (Compositae). Tropical Woods 108: 1-30.

--1960. Wood anatomy of Astereae (Compositae). Tropical Woods 113: 54-84.

--1961. Comparative plant anatomy. Holt, Rinehart & Winston, New York.

--1962. A theory of pacdomorphosis in dicotyledonous woods. Phytomorphology 12: 30-45.

--1966. Wood anatomy of Compositae: a summary, with factors controlling wood evolution. Aliso 6(2): 25-44.

--1969. Wood anatomy of Lobelioideae (Campanulaceae). Biotropica 1: 47-72.

--1971. Wood anatomy of Macaroncsian and other Brassicaccac. Aliso 7: 365-384.

--1975. Ecological strategies of xylem evolution. University of California Press, Berkeley.

--1978. Wood anatomy of Bataceae, Gyrostemonaceae, and Stylobasiaceae. Allertonia 5: 297-330.

--1982. The use of ethylene diamine for softening hard plant structures for paraffin sectioning. Stain Technology 57: 311-317.

--1984. Vessel grouping in dicotyledon woods: significance and relationship to imperforate trachcary elements. Aliso 10: 505-525.

--1985a. Observations on functional wood histology of vines and lianas: vessel dimorphism, tracheids, narrow vessels, and parenchyma. Aliso 11: 139-157.

--1985b. Vegetative anatomy and familial placement of Tovaria. Aliso 11: 69-76.

--1988. Comparative wood anatomy. Springer Verlag, Berlin.

--1992a. Wood anatomy of Lamiaceae: a sur vey, with comments on vascular and vasicentric tracheids. Aliso 13: 309-338.

--1992b. Wood anatomy of Solanaceae. A survey Allertonia 6: 279-326.

--1996. Wood anatomy of Akaniaceae and Bretschneidcraceae. A case of near identity and it systematic implications. Systematic Botany 21: 607-616.

--1998a. Wood and bark anatomy of Caricaceae: correlations with systematics and habit. IAWA Journal 19: 191-206.

--1998b. Wood anatomy of Resedaceae. Aliso 16: 127-135.

--2001. Comparative wood anatomy 2nd ed. Springer Verlag, Berlin.

--2002. Wood and bark anatomy of Salvadoraceae: ecology, relationships, histology of Interxylary phloem. Journal of the Torrey Botanical Society 129: 10-20.

--2005. Wood anatomy of Krameriaceae with comparisons with Zygophyllaceae: phylesis, ecology, and systematics. Botanical Journal of the Linnean Society 149: 257-270.

--2007. Successive cambia revisited: histology. Diversity, and functional significance. Journal of the Torrey Botanical Society 134: 301-332.

--2009a. Xylem heterochrony: an unappreciated key to angiosperm origin and diversification. Botanical Journal of the Linnean Society 161: 26-65.

--2009b. Non-random vessel distribution in wood: patterns, modes, diversity, correlations. Aliso 27: 39-58.

--2010. Caryophyllales: a key group for understanding wood anatomy character states and their evolution. Botanical Journal of the Linnean Society 164: 342-393.

--2012. How wood evolves: a new synthesis. Botany 90: 901-940.

--2013. More woodiness/less woodiness: evolutionary avenues, ontogenetic mechanisms. International Journal of Plant Sciences 174: 964-991.

--2014. Fiber dimorphism: cell type diversification as an evolutionary strategy in angiosperm woods. Botanical journal of the Linnean Society 174: 44-67.

--2015a. Living cells in wood, 1. Absence, scarcity and histology of axial parenchyma. Botanical Journal of the Linnean Society 177: 291-321.

--2015b. Living cells in wood. 2. Raylessness: histology and evolutionary significance. Botanical Journal of the Linnean Society 178: 529-555.

--& C. J. Donald. 1996. Wood anatomy of Limnanthaceae and Tropacolaceae in relation to habit and phylogeny. Sida 17: 333-342.

-& D. A. Hoekman. 1985. Ecological wood anatomy of the woody southern California flora. IAWA Bulletin, new series 6: 319-347.

--& R. B. Miller. 1999. Vegetative anatomy and relationships of Setchellanthus caeruleus (Setchellanthaceae). Taxon 48: 289-302.

--, B. Hansen, H. H. litis, M. E. Olson & D. L. Geiger. 2013. Forchhammeria and Stixis (Brassicales): stem and wood anatomical diversity, ecological and phylogenetic significance. Aliso 31: 59-75.

Cozzo, D. 1946. Relacion anatomica entre la estructura del leno dc las especies argentinas dc Capparis y Atamisquea. Lilloa 12: 29-37.

Donoghue, M. & J. A. Doyle. 1989. Phylogenetic analysis of angiosperms and the relationships of Hamamelidae. In: P. R. Crane & S. Blackmore (eds). Evolution, systematics, and fossil history of the Hamamelidae, 17-45. Clarendon, Oxford.

Erdtman, G., P. Leins, R. Melville & C. R. Metcalfe. 1969. On the relationships of Emblingia. Botanical Journal of the Linnean Society 62: 169-186.

Fahn, A., E. Werker & P. Baas. 1986. Wood anatomy and identification of trees and shrubs from Israel and adjacent regions. Israel Academy of Sciences and Humanities, Jerusalem.

Fisher, J. B. 1980. The vegetative and reproductive structure of papaya (Carica papaya). Lyonia 1: 191-208.

Gadek, P. A., C. J. Quinn, J. E. Rodman, K. 1. G. Karol, E. Conti, R. A. Price & E. S. Fernando. 1992. Affinities of the Austrlian endemic Akaniaceae: new evidence from rbcL sequences. Australian Systematic Botany 5: 717-724.

Gandolfo, M. A., K. C. Nixon & W. L. Crepet. 1998. A new fossil flower from the Turanian of New Jersey, Dressiantha bicarpellata gen. et sp. nov. (Capparaceae). American Journal of Botany 84: 964-974.

Gibson, A. C. 1979. Anatomy of Koeberlinia and Canolia revisited. Madrono 26: 1-12.

Gregory, M. 1994. Bibliography of systematic wood anatomy of dicotyledons. IAWA Journal, Supplement 1: 1-265.

Grew, N. 1682. The anatomy of plants. Rawlings, London (reprinted 1965 by Johnson Reprints, New York).

Hall, J. C., K. J. Sytsma & H. H. Iltis. 2002. Phytogeny of Capparaceae and Brassicaceae based on chloroplast sequence daa. American Journal of Botany 89:1826-1842.

Hall, J. C., H. H. litis & K. J. Sytsma. 2004. Molecular phylogenetics of core Brassicales, placement of orphan genera Emblingia, Forchhammeria, Tirania, and character evolution. Systematic Botany 29: 654669.

Hargrave, K. R., K. J. Kolb, F. W. Ewers, & S. D. Davis. 1994. Conduit diameter and drought induced embolism in Salvia mellifera Greene (Labiatae). New Phytologist 126: 695-705.

Heimsch, C. 1942. Comparative anatomy of the secondary xylem in the "Gruinales" and "Terebinthales" of Wettstein with reference to taxonomic grouing. Lilloa 8: 83-198.

IAWA Committee on Nomenelature. 1964. Multilingual glossary of terms used in wood anatomy. Verlagsanstalt Buchdruckerei Konkordia, Winterthur.

Iltis, H. H., J. C. Hall, T. S. Cochrane & K. J. Sytsma. 2011. Studies in the Cleomaccae. 1. On the separate recognition of Capparaceae, Cleomaceae, and Brassicaccae. Annals of the Missouri Botanical Garden 98: 28-36.

Jansen, S., P. Baas & E. Smets. 2001. Vestured pits: their occurrence and systematic importance in eudicots. Taxon 50: 135-167.

--, --, P. Gasson & E. Smets. 2003. Vestured pits: do they promote safer water transport? International Journal of Plant Sciences 164: 405-413.

--, --, --, F. Lens & E. Smets. 2004. Variation in xylem structure from tropics to tundra: evidence from vestured pits. Proceedings of the National Academy of Sciences 101: 8833-8837.

--, & A. Nardini. 2014. From systematic to ecological wood anatomy and finally plant hydraulics: are we making progress in understanding xylem evolution? New Phytologist 203: 12-15.

Kohonen, M. M. 2006. Engineered wettability in tree capillaries. Langmuir 22: 3148-153.

--& A. Holland. 2009. On the function of wall sculpturing in xylem conduits. Journal of Bionic Engineering 6: 324-329.

Kribs, D. A. 1935. Structural lines of specialization in the wood rays of dicotyledons. Botanical Gazette 96: 547-557.

--1937. Salient lines of structural specialization in the wood parenchyma of dicotyledons. Bulletin of the Torrey Botanical Club 64: 177-186.

Lens, F., N. Devin, E. Smets & M. del Arco. 2013. Insular woodiness in the Canary Islands: a remarkable case of convergent evolution. International Journal of Plant Sciences 174: 992-1013.

Mauseth, J. D. 1993. Water-storing and cavitation-preventing adaptations in wood of cacti. Annals of Botany 72: 81-89.

McCully, M., M. Canny, A. Baker & C. Miller. 2014. Some properties of the walls of metaxylem vessels of maize roots, including tests of the wettability of their luminal wall surfaces. Annals of Botany 116: 1-13.

McLaughlin, J. 1959. The woods and flora of the Florida Keys. Wood anatomy and phytogeny of Batidaceae. Tropical Woods 110: 1-15.

Metcalfe, C. R. & L. Chalk. 1950. Anatomy of the dicotyledons. Clarendon, Oxford.

Olson, M. E. 2001. Stem and root anatomy in Moringa (Moringaceae). Haseltonia 8: 56-96.

--2002. Combining data from DNA sequences and morphology for a phytogeny of Moringaceae. Systematic Botany 27: 55-73.

--2007. Wood ontogeny as a model for studying heterochrony, with an example of pacdomorphosis in Moringa (Moringaceae). Systematics and Biodiversity 5: 145-158.

--2014. Xylem hydraulic evolution, I. W. Bailey, and Nardini & Jansen (2013): pattern and process. New Phytologist 203: 7-11.

--& S. Carlquist. 2001. Stem and root anatomical correlations with life form diversity, ecology, and systematics in Moringa. Botanical Journal of the Linnean Society 135: 315-348.

den Outer, R. W. & W. L. H. van Veenendal. 1981. Wood and bark anatomy of Azima tetracantha Lam. (Salvadoraceae) with description of its included phloem. Acta Botanica Neerlandica 30: 199-207.

Pax, F. & K. Hoffmann. 1936. Capparidaceae. Pp 146-233. In: A. Engler & K. Prantl (cds). Die naturlichen Pflansenfamilien 17b. Verlag Wilhelm Engelmann, Leipzig.

Rodman, J. E., P. S. Soltis, D. E. Soltis, K. J. Sytsma & K. G. Karol. 1998. Parallel evolution of glucosinolate biosynthesis inferred from congruent nuclear and plastid gene phylogenies. American Journal of Botany 80: 443-459.

Sano, Y., H. Morris, H. Shimada, L. Ronse de Craene & S. Jansen. 2011. Anatomical features associated with water transport in imperforate tracheary elements of vessel-bearing angiosperms. Annals of Botany 107: 953-964.

Sauter, J. J. 1966. Untersuchungen zur Physiologic der Pappelholzstrahlen. I. Jahresperiodischer Verlauf der Starkespeicherung im Holzstrahlenparenchym. Zeitschrift der Pflanzenphysiologie 55: 246-258.

Schweingruber, F. H. 2006. Anatomical characteristics and ecological trends in the xylem and phloem of Brassicaceae and Resedaceae. IAWA Journal.

Slatyer, R. O. 1967. Plant-water relationships. Academic, New York.

Solereder, H. 1885 Uber den systematischen Wert dcr Holzstruktur. T. Oldenbourg, Munchen.

Solereder, H. 1908. Systematic anatomy of the dicotyledons (trans. Boodle & Fritsch). Clarendon, Oxford.

Soltis, D. E., et al. 2011. Angiosperm phylogeny: 17 genes, 640 taxa. American Journal of Botany 98: 704-730.

Stern, W. L., G. K. Brizicky & F. N. Tamolang. 1963. The woods and flora of the Florida Keys: Capparaccac. Contributions from the U. S. National Herbarium 34: 25--43.

Stevens, P. F. 2015. Angiosperm Phylogeny Website, Version 13 (accessed September, 2015).

Su, J.-X., W. Wange, L.-B. Zhang & Z.-D. Chen. 2012. Phylogenetic placement of two enigmatic genera, Borthwickia and Stixis, based on molecular and pollen data, and the description of a new family of Brassicales, Borthwickiaceae. Taxon 61: 601-611.

Taylor, D. W., L. J. Hickey. 1996. Evidence for and implications of an herbaceous origin for angiosperms. In, D. W. Taylor & L. J. Hickey, eds.. Flowering plant origin, evolution, and phylogeny, 232-266. Chapman & Hall, New York.

Wolkinger, F. 1969. Morphologic und systematisehe Verbreitung lebender Holzfasem bei Straucher und Baumen. Phyton (Austria) 14: 55-67.

Zimmermann, M. H. 1978. Structural requirements for optimal water conduction in tree stems. In: P. B. Tomlinson & M. H. Zimmerman (eds). Tropical trees as living systems, 517-532. Cambridge University Press, London.

--1983. Xylem structure and the ascent of sap. Springer Veriag, Berlin.

Zwieniecki, M., & N. M. Holbrook. 2009. Confirming Maxwell's demon: biophysics of xylem embolism repair. Trends in Plant Science 14: 530-534.

Sherwin Carlquist (1,2)

(1) Santa Barbara Botanic Garden, 1212 Mission Canyon Road, Santa Barbara, CA 93105, USA

(2) Author for Correspondence; e-mail: s.carlquist@vcrizon.net

Published online: 18 March 2016
COPYRIGHT 2016 New York Botanical Garden
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:p. 56-90
Author:Carlquist, Sherwin
Publication:The Botanical Review
Article Type:Report
Date:Mar 1, 2016
Words:12855
Previous Article:Wood anatomy of Brassicales: new information, new evolutionary concepts.
Next Article:A Phytogeographical Metaanalysis of the Semiarid Caatinga Domain in Brazil.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters