We have a publication on venation network architecture now out in Journal of Ecology. The study is focused on the drivers of network reticulation (looping) Some leaves have venation networks that do not form loops (e.g. Gingko biloba, the maidenhair tree), while others do (e.g. Citrus sinensis, the sweet orange).
Why does this variation exist? Venation networks are thought primarily to be hydraulic transport systems that move sugar and water throughout the leaf. In support of this hypothesis, a large number of studies have articulated a relationship between the density of minor veins and resource transport capacity. However, venation networks have also been hypothesized to have structural or defensive functions, e.g. by providing mechanical support and toughness, or by re-routing flows in the event of damage from water stress or herbivores. A smaller number of studies – primarily experimental, and restricted to a small set of species – have provided some support for this hypothesis.
Variation in reticulation may influence all of these hydraulic, structural, and defensive functions, but there had been limited empirical comparative datasets available to determine which functions were most important, or which aspects of network architecture were directly related to each function.
Our study measured these functions, as well as several aspects of network architecture, for more than 100 species growing in a range of environments spanning treeline in the Andes to the lowland Amazon basin of Peru. The study is an output of the CHAMBASA project, with Yadvinder Malhi, Brian Enquist, Sandra Díaz, and Greg Asner as PIs, and with Lisa Patrick Bentley, Norma Salinas, and Alexander Shenkin as core leaders (to name only a few of the team!). Collecting data at this scale was a multi-month undertaking in the field and a multi-year undertaking on the computer, and required a large international field team that included many Peruvian researchers and students. Six years after the project was planned and five years after the fieldwork was completed, the study is finally out.
We found that reticulation is not a single axis of variation from low to high – rather, we found that reticulation requires at least three axes, describing branching vs. reconnecting structures, compact vs. elongated loops, and low vs. high density veins. We also found that the first two axes – the most important ones – were most closely aligned with structural and defensive functions, while the third axis was most closely aligned with hydraulic functions. Thus the study indicates that venation network architecture reflects tradeoffs among multiple competing functions. Because not all functions can be simultaneously optimized (for example, a very mechanically strong network may have low per-mass photosynthesis capacity), the networks we observe in nature represent a balance between these costs and benefits in different contexts. My hope is that the study will stimulate more nuanced discussion of the ways in which we quantify network architecture, and will encourage more thinking on the tradeoffs underlying this architecture.
There is still much to learn about the significance of the intricate network of veins we see across leaves – our study was limited to descriptions of the loops at the finest spatial scales in the network. We are working now to better understand the functional significance of the larger loops, and hope to have new work to share soon. For now, you can read the article at the journal website.
