Figure 3.

Schematic of environmental and tree-internal drivers and regulators represented in the wood formation models discussed in this paper. Drivers reported are (top-left to bottom-right) water, temperature, either daylength (phenological) signal or carbon, hormonal/daylength (phenological) signal, carbon (note the absence of nutrients as growth rate modifiers in all models). The positioning along the axes within each box reflects (1) the number of cell types affected by an external/internal driver and (2) the level of detail driver-cell interactions are resolved. Left to right: First square: one cell type is affected only (this could be e.g., cambial cells or wall thickening cells only), last square: all three cell types are affected. Bottom to top: low level of complexity with which an environmental driver influences the model e.g., as single part of a physical equation (e.g., through a threshold parameter to promote an on-off switch environmental switch (e.g., assume metabolic activity occurs only above 5°C (Deleuze and Houllier, 1998))) next square: as part of a response-function e.g., increasing enlargement rate with temperature (Fritts et al., 1991; Friend, 2020). Complexity can increase even further to a detailed physiological level, for example through spatial interactions e.g., diffusion of carbon across the developing radial file, followed by arriving carbon being included into the thickening cell wall, as sort of done in Friend (2020). Note that for Schiestl-Aalto et al. (2015), we interpret the “ontogenetic development” used as one the modifiers of cell proliferation dynamics by the authors, as an internal signal, and therefore place it in the bottom-left square, as a hormonal/daylength (phenological) signal. If a model is absent from an environmental factor matrix, it does not resolve this particular environmental factor. Within the squares, no ranking is attempted. Note that the relative position along the y-axis is not comparable between the environmental /internal factors. Note that this is a rough scheme, derived from text-descriptions and equations in the publications, which sometimes may not represent the entirety or complexity of an operation in the model itself. Only the latest version of Treerings (Treerings3) is shown here.