The sampling site was located at the campus of Nanjing Forestry University (118°48′35′′ E, 32°04′67′′ N), Nanjing, Jiangsu Province, China, which has a subtropical monsoon climate. The mean annual precipitation is 1,156 mm, and the mean annual temperature is 15.6°C based on the climate data collected between 1951 and 2014.¹ The rainy season is from June to August, approximately accounting for half of the annual accumulated precipitation.

A total of 1,736 leaves were collected from 40-50 Photinia × fraseri “Red Robin” plants in 2021. Newly developed leaves in early March 2021 were collected in the middle of March, April, May, June, and July 2021. These collections are labeled as N3, N4, N5, N6, and N7, respectively (to denote “new” in each month) (Figure 1 and Table 1).

We measured the fresh mass of each lamina using an electronic balance (Type: ML 204; Mettler Toledo Company, Greifensee, Switzerland; measurement accuracy 0.0001 g) and scanned the leaves to bitmap images at a 600-dpi resolution using a photo scanner (V550, Epson Indonesia, Batam, Indonesia). Adobe Photoshop (CS6, version: 13.0) was used to obtain a black and white image of each lamina. The M-file based on MatLab (version ≥ 2009a) developed by Shi et al. (2018) was used to extract the planar coordinates of each lamina. And then leaf A was calculated using an R script (version 4.0.3; R Core Team, 2020) developed by Shi et al. (2018) and Su et al. (2019). The fresh leaves were then dried to constant weight in an oven (Type: XMTD-8222; Jinghong Experimental Equipment Co., Ltd., Shanghai, China) at 80°C for 48 h to measure dry mass using the same electronic balance.

We used the analysis of variance (ANOVA) with a 0.05 significance level to test whether leaf age had a significant effect on the leaf lamina area, the ratio of lamina dry mass to lamina fresh mass (i.e., LDM/LFM), the lamina fresh mass per unit area (LFMA), and the lamina dry mass per unit area (LMA). Provided that a significant effect of leaf area on each of the measures was found, we used Tukey’s honestly significance difference test with a significance level of 0.05 (Hsu, 1996) to test the significance of the difference between any two groups in the leaf lamina area, LDM/LFM, LFMA, and LMA. Additionally, we used a power-law function to fit the scaling relationship between any two variables of interest:

where Y₁ and Y₂ represent the two interdependent variables, respectively, and α and β are the scaling exponent and the normalization constant, respectively. In order to stabilize the variance of leaf measures, both sides of the equation were log-transformed (Niklas, 1994; Niklas et al., 2007):

where y = ln Y₁, x = ln Y₂, and γ = ln β. The numerical values of α and γ were obtained using reduced major axis regression protocols (Niklas, 1994; Smith, 2009). The bootstrap percentile method (Efron and Tibshirani, 1993; Sandhu et al., 2011) was used to compare the significance of the differences of the numerical values of α between any two leaf-age groups. For any two leaf-age groups (denoted as groups A and B), we calculated 4,000 replicates of α for each group using the bootstrap method (Efron and Tibshirani, 1993). Denoting D as the differences in the replicates of α between groups A and B, we observed whether the 95% CI of D included 0. If the lower bound of the 95% CI of D is larger than 0, it indicates that the estimated α-value of group A is larger than that of group B; if the upper bound of the 95% CI of D is smaller than 0, it indicates that the estimated α-value of group A is smaller than that of group B; if the 95% CI of D includes 0, it indicates that there is no significant difference in the estimated α-values between groups A and B (see Sandhu et al., 2011 for details). To measure the goodness of fit, the root-mean-square error (RMSE = the square root of the quotient of residual sum of squares and sample size) was used. All statistical analyses were carried out using R (version 4.0.3; R Core Team, 2020).

Prior research has shown that the photosynthetic capacity of leaves decreases with increasing leaf age (Horsley and Gottschalk, 1993; Bauerle et al., 2020), and that this might be associated with the changes in chemical composition, leaf nitrogen content, and CO₂ diffusion limitation (Zhang et al., 2008). Photosynthetic rates of fully expanded leaves generally show a decline with increasing leaf age due to the reallocation of resources to young leaves to optimize the overall systemic photosynthetic returns of an individual plant, rather than resulting from leaf functional degradation (Field and Mooney, 1983; Hikosaka et al., 1994). Moreover, the photosynthetic capacity of young leaves has been shown to be significantly lower than mature leaves (Hikosaka and Shigeno, 2009; Liu et al., 2020; Ji et al., 2021). Although leaf functional traits have been regarded as important references for understanding plant ecological and growth strategies (Westoby et al., 2000, 2002), there are trade-offs among different leaf functional traits (e.g., lamina area, LMA) under the limitation of resources. If plants increase their investment to one functional trait, they are likely to reduce their investment to other functional traits (Falster et al., 2018).

The data presented here show that the leaf lamina area, the ratio of LDM to LFM, and LMA generally increase with increasing leaf age, which is consistent with previous reports in other species (Niklas, 1991; He and Yan, 2018; Liu et al., 2020; Guo et al., 2021). In addition, LFMA decreased at the leaf maturation stage. There are two interrelated explanations for this phenomenology. First, over the course of leaf development, maturation, and subsequent senescence, the dry matter investment disproportionately increases with increasing age, and, second, the symplastic volume fraction (the living contents of a leaf, i.e., the protoplasmic contents) systematically decreases with increasing age. Both of these trends are reflected in the trends of lamina area, LDM/LFM, LFMA, and LMA reported here with increasing leaf age, and both are clearly and intrinsically interrelated, although the proximate mechanism(s) underlying each of these trends differs. During leaf expansion, leaves generally have a low cost of construction with increasing leaf area because leaf expansion involves the volumetric increase in cell size (and increase in the symplastic volume fraction) rather than the addition of significant amounts of cell wall (apoplastic) materials. The reverse is generally true during leaf maturation during which the apoplastic volume fraction increases relative to the symplasic volume fraction as a consequence of the deposition of secondary cell wall materials including the lignification of vascular tissues. This general phenomenology is consistent with the trends in our data (see Figure 2).

The aforementioned growth dynamic has obvious physiological and biomechanical consequences. For example, species with long versus short leaf longevities must adopt specific and different strategies to maintain growth (Mediavilla et al., 2014). It is not surprising therefore that previous studies have shown that variations in LMA are correlated across plant functional groups (such as trees, shrubs, herbs) as well as abiotic variables such as light, temperature, water and nutrient availability, and atmospheric composition (Ackerly, 1992; Poorter and de Jong, 1999; Niinemets, 2001; Poorter et al., 2009). In our study site, ambient temperature and precipitation gradually increase during the rainy season, especially in June, July, and August, over the course of leaf initiation, expansion, and maturation (see the online Supplementary Material in Li et al., 2022). LMA has been demonstrated to have plastic responses to climate, e.g., in arid environments plants can reduce water requirements, which results in a higher LMA relative to those in humid environments (Poorter et al., 2009). Similarly, at the leaf maturation stage, the ratio of lamina dry mass to fresh mass reached a stable status, and the precipitation might play an important role in affecting the numerical value of LMA. For example, June and July receive greater amount of rain than May, which perhaps to a certain degree accounts for why the LMAs in June and July are significantly but slightly lower than that in May (Figure 2). Thus, changing but predictable environmental conditions might have exerted an influence to a certain degree on the expression of leaf functional traits (Poorter et al., 2009). Nevertheless, it requires additional controlled experiments for testing whether the amount of precipitation can change LMA, i.e., a plasticity of LMA to weather, in future studies.

The scaling exponents of LDM vs. LFM were found to be significantly different between March and April, likely because new leaves were growing expanding at this time. Figure 2 shows significant increases in both the lamina area and LDM/LFM between March and April. With the exception of March, the scaling exponents of LDM vs. LFM decreased toward 1 with increasing leaf age (Figure 6). We interpret these data to indicate that leaf water content and leaf dry mass tend to synchronously increase.

Liu et al. (2020) confirmed that for leaves of Alangium chinense (Lour.) Harm and Liquidambar formosana Hance, the scaling exponents of the lamina mass vs. area scaling relationship during the spring are significantly higher than during the summer. This is consistent with our results. In the case of N3 and N4, the numerical values of the lamina mass vs. area scaling exponents were significantly smaller compared to those for N5, N6, and N7 (Figure 6), indicating that the larger leaf area of mature leaves require greater apoplastic investments. Under any circumstances, it is obvious that the investment in leaf construction changed between the early rapid growth and leaf maturation stages. Changes in leaf investment may be related to photosynthetic returns (Day et al., 2001; Bielczynski et al., 2017), such that young leaves produce less photosynthates (Falster and Westoby, 2003) compared to more mature leaves (Pan et al., 2012).

The numerical values of the scaling exponents of the lamina mass vs. area scaling relationship for each of leaf-age groups were in excess of unity, and thus conform with the phenomenon called “diminishing returns” (see also Huang et al., 2019; Liu et al., 2020). In general, although larger leaf lamina area can yield a higher photosynthetic capacity, progressively larger leaves generally require a disproportionately greater investment in their mechanical support (Niklas, 1992, 1999; see also Gibert et al., 2016). This disproportionate investment in the apoplastic content relative to the symplastic content per unit leaf area can account for the numerical values of the lamina mass vs. area scaling relationship approaching and exceeding unity (i.e., α> 1.0).