When looking at the intricate and delicate network of leaf veins, it is not necessary to be a botanist to feel amazed at such a wonderful example of evolutionary engineering (Fig. 1). This complex plumbing system is finely designed to assure capillary water delivery to photosynthetic mesophyll cells, thus replacing the huge amount of water they lose to the atmosphere during the transpiration process. In fact, water loss is unavoidably coupled to stomatal opening and CO2 diffusion from the atmosphere into the leaf interior, which is crucial to fuel photosynthetic processes. The structure–function relationships of this marvellous ‘irrigation’ system have intrigued researchers for many years. Only during the last decade, however, have we finally started to understand the important connections between the overall patterns and length of veins fitting in a unit of leaf surface area, generally termed ‘vein density’ (and expressed in mm mm−2), and the leaf efficiency in terms of photosynthetic capacity and resistance to environmental stresses (Sack & Scoffoni, 2013). In a new study by Feild & Brodribb, in this issue of New Phytologist (pp. 720–726), novel evidence suggests that high leaf vein density is driven by the evolution of vessels and a decrease of vein thickness.
‘The structure–function relationships of this marvellous “irrigation” system have intrigued researchers for many years.’
From a physiological point of view, the achievement of high vein densities is crucial to shorten the distance that water has to travel outside leaf vasculature in a high-resistance pathway across cell walls or membranes, thus improving water delivery rates and assuring the sustainability of transpiration volumes coupled to high CO2 uptake rates (Brodribb et al., 2007). Apparently, this surge in photosynthetic productivity guaranteed by high vein density was key to the burst of angiosperm diversification during the Cretaceous (Brodribb & Feild, 2010). High vein density has also been proven to be a key innovation allowing plants to better resist both biotic and abiotic stress factors, thus permitting angiosperm radiation in unfavourable habitats. In fact, highly reticulated and redundant venation patterns guarantee several alternative water pathways in case of partial leaf damage leading to isolation or blockage of some xylem conduits in veins, which can be caused for instance by herbivores (Sack et al., 2008; Nardini et al., 2010) or due to drought-induced embolism formation (Nardini et al., 2001; Scoffoni et al., 2011).
A major, still unanswered question is why only angiosperms, and more specifically eudicots, were capable of gathering this enormous competitive advantage by developing vein densities up to 25 mm mm−2, while non and basal angiosperms produced at best vein systems with densities c. 5 mm mm−2 (Boyce et al., 2009). Feild & Brodribb provide novel, helpful insights to answer this intriguing question, suggesting that the evolution of xylem vessels was a fundamental pre-requisite for the evolution of high vein densities. According to the authors, derived angiosperm clades with vein densities > 12 mm mm−2 are characterized by the presence of vessels with simple perforation plates, while veins of species with lower vein densities contain tracheids or vessels with scalariform perforation plates. Feild & Brodribb suggest that the appearance of simple perforation plates represented a key innovation improving the hydraulic efficiency of the leaf veins and may have assured adequate efficiency of water transport (Christman & Sperry, 2010). Overall, this finding is in line with the traditional ‘Baileyan’ trends in tracheary element evolution, which imply that vessel elements arose from species bearing tracheids with scalariform pitting, and that specialization was associated with evolutionary changes in the perforation plate morphology from scalariform and oblique to simple and horizontal, with intertracheary pitting evolving from scalariform to alternate (Bailey & Tupper, 1918; Bailey, 1954; Fig. 2). This irreversible trend in perforation plate evolution, which was considered to be one of the strongest phylogenetic trends in plant anatomy for many years, is generally supported by fossil data and remains largely accepted despite suggestions of reversibility in some clades (Feild et al., 2002).
The second major finding of the study by Feild & Brodribb is that the increase of leaf vein density and origin of vessels was strongly associated with a decrease in leaf vein thickness. Although the authors did not measure the tracheary element diameter in leaf veins and conduit density, it can be postulated that an increase of vein density would require an increase in vein diameter and conduit number if tracheids were used as conductive cell types only (see fig. 1 in Feild & Brodribb). The authors further speculate that an excessive increase of vein thickness would have reduced spacing between veins below critical values potentially obstructing adequate CO2 diffusion into the leaf mesophyll cells. It can also be hypothesized that highly efficient conducting units significantly decreased the carbon costs associated with the construction of vein tissues (Blonder et al., 2011; Nardini et al., 2012), thus making the construction of very dense vasculature patterns in eudicot leaves metabolically affordable.
Although the study by Feild & Brodribb sheds light on an important evolutionary trend in leaf hydraulics, with consequences for leaf construction and competitive success of angiosperms, a number of critical issues still await further experimental testing. First of all, to what extent is the distribution of vessels and tracheids associated with conduit size and vein order? Although conduit size is not reported in this study, major veins (first to third order) are more likely to show wide vessels with simple perforation plates, whereas narrow vessel elements with scalariform perforation plates or tracheids are more common in minor veins (Coomes et al., 2008). Indeed, elimination of scalariform perforation plates as a result of selection for promotion of higher conductive rate and volume may explain why narrow vessels retain scalariform perforation plates, while wider ones show simple perforation plates (Oskolski & Jansen, 2009). Such a relationship between conduit size and morphology may also explain the differences in conduit morphology between protoxylem and metaxylem as reported by Feild & Brodribb (supporting information table S1). Furthermore, it seems unlikely that vein thickness is highly representative for the conduit size or conduit density, because veins may also include considerable variation in the amount of sclerenchyma cells, which might play important roles in drought-adapted plants (Jacobsen et al., 2005).
It is interesting that some of the anatomical observations by Feild & Brodribb rely on earlier xylem observations from the primary stem. Based on maceration of leaf veins, the authors suggest a high similarity in tracheary element morphology between leaf veins and primary xylem of the stem. This analogy is certainly not surprising given that the xylem tissue forms a continuous system throughout the plant body, although the origin and specialization of vessels in eudicots is generally considered to occur in secondary xylem first, with some evidence that there is an evolutionary lag of specialization in (primary) xylem of leaves and inflorescence appendages (Bailey, 1954). Detailed observation of xylem anatomy and in particular the first appearance of vessels across plant organs at different developmental stages would be desired. Moreover, results based on the maceration technique should be interpreted with care because of possible preparation artefacts (Jansen et al., 2008).
In conclusion, this interesting paper illustrates that xylem anatomy deserves more attention and could play an important role in understanding functional traits allowing the development of high vein density in angiosperms. Besides xylem, one may wonder if similar evolutionary fine-tuning of vascular tissue could be found in the phloem sieve elements in angiosperms, which are traditionally classified into sieve elements of seedless vascular plants, ‘primitive’ sieve cells in gymnosperms, and the more efficient sieve-tube elements with various types of accompanying cells in angiosperms. Structure–functional relationships between phloem transport and the micromorphology of sieve elements would be most welcome to stimulate such research (Jensen et al., 2012).