Environmental factors have a greater influence on photosynthetic capacity in C4 plants than biochemical subtypes or growth forms

Abstract

Our understanding of how photosynthetic capacity varies among C₄ species and across growth and measurement conditions remains limited. We collated 1696 CO₂ response curves of net CO₂ assimilation rate (A/C_i curves) from C₄ species grown and measured at various environmental conditions and used these data to estimate the apparent maximum carboxylation activity of phosphoenolpyruvate carboxylase (V_pmaxA) and CO₂-saturated net photosynthetic rate (A_max), two key parameters describing photosynthetic capacity. We examined how V_pmaxA and A_max vary with species-specific traits, growth and measurement conditions. We found little systematic variation of V_pmaxA and A_max across the classical C₄ biochemical subtypes or growth forms, but showed that growth temperature and measurement conditions are major factors determining C₄ photosynthetic capacity. We found no evidence that common C₄ model species (e.g. maize, sorghum and Setaria viridis) differ in photosynthetic capacity from other C₄ species when grown in controlled environments. However, C₄ model species showed up to twice the photosynthetic capacity of other C₄ species when grown in the field. Our multivariate model accounts for 47-51% of the variation reported in V_pmaxA and A_max, and we argue that environmental conditions have a greater influence on C₄ photosynthetic capacity than biochemical subtypes or growth forms.

Keywords: A/Ci curve; Amax; C4 biochemical subtype; C4 photosynthesis; Vpmax; environmental response; photosynthesis modelling.

Fig. 1

Summary of data points (by groups) based on (a) growth forms, (b) locations, (c) CO₂ treatments, (d) fertilisation treatments and (e) watering status. Decreased and elevated CO₂ treatments consist of measurements made on plants with a growth CO₂ concentration of < 400 and > 400 ppm, respectively. See the Materials and Methods section for more information.

Fig. 2

Correlation between A _max and V _pmaxA estimated from data collected from a range of species measured at various growth and measuring conditions. Point shapes reflect important species groups, and colours denote measurement temperature (T _leaf). (a) All available A _max‐V _pmaxA data; (b–d) data subsets measured at 25–30°C, 30–35°C and 35–40°C, respectively. Equation, adjusted R ² values and P values of individual fits are shown in each panel. Shaded areas represent 95% confidence intervals around predicted means from the fitted curves.

Fig. 3

V _pmaxA and A _max plotted by C₄ biochemical subtypes (a and b, respectively), growth forms (c and d, respectively) and growth locations (e and f, respectively). Data presented here are a mixture of measurements done at various growth and measuring temperatures and irradiance. Raw data points are plotted as coloured symbols, with different symbol shapes reflecting important species groups. Model‐predicted marginal means and intervals are shown as a horizontal line. On this line, black circles indicate model‐predicted meta‐analytic means of V _pmaxA or A _max; thick bars are 95% confidence intervals, and thin bars are 95% prediction intervals. On the right‐hand side of each panel, the number of unique species per category (n.spp) and the number of individual studies (e.size) are indicated. P values of the multivariate linear mixed‐effects model are indicated in each panel (see Table 1). Significant codes: *, P < 0.05; **, P < 0.01.

Fig. 4

Comparison between V _pmaxA and A _max with growth location coloured by species group (C₄ non‐model species vs model species). A linear mixed‐effect model was conducted to examine V _pmaxA and A _max between two species groups at a location, or within a species group at both locations. Statistical results of comparisons are denoted with horizontal lines (i.e. the two bars at the beginning and the end of a horizontal line are compared) and asterisks indicate statistical significance (***, P < 0.001) from this lnear mixed‐effect model. The sample size (n) for each group is indicated on plots. Boxplots show the median (horizontal line within each box), interquartile range (IQR; box edges spanning the 25^th to 75^th percentiles), and whiskers (1.5 × IQR).

Fig. 5

Relationships between V _pmaxA and A _max with mean T _max (a and b, respectively) and growth CO₂ levels (c and d, respectively). Data points in (a) and (b) are coloured in a gradient by the measurement temperature (T _leaf). Solid lines represent model‐predicted values of V _pmaxA or A _max at a given mean T _max (equations are shown at the top of each panel), dashed lines indicate 95% confidence intervals, and dotted lines show 95% prediction intervals. P values of the multivariate linear mixed‐effects models are indicated. Significant codes: **, P < 0.01.

Fig. 6

Relationships between V _pmaxA and A _max with T _leaf (a and b, respectively) and photosynthetic photon flux density (PPFD) (c and d, respectively). Data points in (a) and (b) are coloured by the mean maximum growth temperature (mean T _max). Solid lines represent model‐predicted values of V _pmaxA or A _max at a given T _leaf or PPFD (equations are shown at the top of each panel), dashed lines indicate 95% confidence intervals, and dotted lines show 95% prediction intervals. P values of the multivariate linear mixed‐effects models are indicated. Significant codes: ***, P < 0.001.

Fig. 7

Contour plots illustrating the model‐predicted responses of (a) V _pmaxA and (b) A _max (μmol m⁻² s⁻¹) to T _leaf and mean T _max. P values of linear mixed‐effects model testing the interaction between T _leaf and mean T _max are indicated (see Table 1). Black circles represent comparisons of V _pmaxA and A _max at the same mean T _max and T _leaf of 20°C, 30°C and 38°C, with the colour gradient reflecting changes in V _pmaxA and A _max (see the colour legend on the right‐hand side of each panel). Significant codes: *, P < 0.05.