(guest post on a new Nature Plants paper by lead author/past postdoc Dr. Ilaine Matos)
Over the past 400 million years, the leaves of vascular plants have diversified to generate an astonishing variety of forms and venation architectures (Fig.1). But exactly when, why, and in which plant clades the new vein architectural combinations arose and how these innovations influenced leaf functioning remains a mystery.
Figure 1. Time-calibrated phylogeny of 1,000 vascular plant taxa. Red branches indicate fossil taxa (N = 120), and black branches indicate extant taxa (N = 880). The other coloured branches indicate the taxa depicted in the outer images. Parenthetical numbers indicate the number of extant (black) and fossil (red) taxa sampled in each clade. Internal nodes for each clade are indicated by the labelled, coloured rectangles. The colours of the outer circles represent the clades depicted in the caption.
Previous work has focused on investigating how vein density, particularly on the smallest vein sizes, has increased over time, as leaves evolved to achieve higher photosynthetic capacity. However, the complex network of veins can be described and differentiated by many other important architectural traits besides vein density (Fig.2) and can be involved in several functions (e.g. mechanical support, resistance to damage) besides photosynthesis. For example, whether the veins only branch or form loops, as well as the number, shape, and size of such loops, all have an impact in the leaf capacity to perform its different functions under distinct environmental conditions. Additionally, these architecture traits can vary independently across vein sizes (i.e. small, medium, large veins), resulting in the diversity of venation networks observed in both extant and extinct leaves.
Figure 2. Examples of leaf networks with low and high values for four architecture traits. (a) VD (mm mm−2) quantifies the length of total vein segments per unit of leaf area, that is, if leaf networks have fewer (low VD) or more (high VD) veins per area. (b) MST (dimensionless) describes the degree of reticulation or branching, that is, whether leaf networks have more loops (low MST) or fewer loops (high MST). (c) Loop elongation ratio (ER) (dimensionless) describes the shape of the loops, that is, if loops are more circular (low ER) or more elongated (high ER). (d) Loop CR (dimensionless) describes the degree of loop smoothness, that is, if loops are more (high CR) or less (low CR) smooth.
In this Nature Plants paper, we assembled venation networks from 1,000 extant (N = 880) and extinct (N = 120) taxa across all major plant vascular clades (i.e. Ferns, Gymnosperms and Angiosperms). This unprecedented dataset was then used to investigate (1) How key venation architectural traits evolve across different vein sizes and major plant clades; and (2) Which abiotic (e.g. air temperature and CO2 atmospheric concentration) and biotic (e.g. insect herbivory pressure) drivers may have contributed to leaf venation evolution over the geological time.
Using state-of-art phylogenetic comparative methods we found that (1) venation networks evolved from having fewer veins and less smooth loops to having more veins and smoother loops, but these changes only occurred in small and medium vein sizes (Fig.3a). The diversity of architectural designs increased biphasically, first peaking in the Paleozoic, then decreasing during the Cretaceous, then increasing again in the Cenozoic, when recent angiosperm lineages initiated a second and ongoing phase of diversification (Fig.3b). Vein evolution was not associated with temperature and CO2 fluctuations but was associated with insect diversification. Our results highlight the complexity of the evolutionary trajectory and potential drivers of venation network architecture.
Figure 3. (a) Summary of the overall evolutionary trends of venation architectural traits. (b) Temporal changes in the venation architectural space occupation under a gradual model of evolution. Black and red lines indicate the median disparity metric value (median centroids – describes the position of an species in the morphospace of venation architecture compared to the centroid of this space; sum of variances – characterizes the size of morphospace occupied by each species) for each approximately 10 million year time slices and the corresponding 50% and 95% confidence intervals. The coloured vertical lines indicate the approximate age of origin of each vascular plant clade. The dashed black line indicates the different phases (phases 1–4) of disparity variation over time.





Comments
3 responses to “New paper: Leaf venation network evolution across clades and scales”
Great work, Benjamin! The artistic graphs are perfect.
Thanks for your support of our work – I hope you are doing well!
Thank you! Ilaine worked very hard on this study and the figure – I was lucky to hire her!