2. Materials and Methods

2.1 Site description

The experiment was conducted at the Dongzhaigang National Nature Reserve (110°32´–110°37´E and 19°51´–20°01´N) in northeastern Hainan Province, China. The reserve is the earliest established mangrove reserve in China, covering 3337.6 ha. This area is characterized by a semi-enclosed estuary with a muddy bottom, nourished by four small rivers. It experiences semidiurnal tidal cycles, averaging a tidal range of 1.6 to 1.8 meters. The climate is characterized as a tropical maritime monsoon with an average annual rainfall of 1676.4 mm and a mean annual temperature ranging from 23.3 to 23.8°C (Li et al. , 2016). A total of thirty-five species of mangrove plants have been documented across 25 genera and 18 families. This included 24 species of true mangroves, which belong to 14 genera and 10 families, as well as 11 species of minor mangroves, categorized under 11 genera and 8 families (Jiang et al. , 2023). Deforestation ceased in 1986 when the bay was declared a national nature reserve. The dominant mangrove species are Avicennia marina , Aegiceras corniculatum ,Bruguiera sexangula , Ceriops tagal , and Rhizophora stylosa .

2.2 Field survey

The field survey was conducted during the peak of the rainy season. We selected five tree species and five shrub species for this study based on previous field investigations and literature research (Bai et al. , 2021; Yu et al., 2023). Among our sampled species, five were located in the low intertidal zone, and five were in the high intertidal zone (Table 1). Four plots (10 m×10 m, >1 km apart) were established for each species. The height and diameter at breast height (DBH) or basal diameter of each individual tree and shrub were recorded. For each species, we collected 30 current-season, fully expanded, light-exposed mature and healthy green leaves from three adult individuals per plot and mixed them as a composite sample. All leaves were placed in plastic bags and immediately stored in a cooler with ice. Subsequently, we transported the samples to the laboratory for the measurements of leaf structural traits.

2.3 Leaf traits

The fresh leaf chlorophyll content (LCC) was estimated with a portable optical chlorophyll meter (SPAD-502, Konika-Minolta Inc., Tokyo, Japan). The leaf area (LA) was determined with a leaf area meter (LI-3000c, Lincoln, Nebraska, USA). Additionally, leaf thickness (LT) was measured using a digital micrometer (Digimatic micrometer, Mitutoyo, Japan). This measurement was derived from the average of three randomly selected positions on each leaf, deliberately avoiding the prominent veins to ensure accuracy on flat leaf surfaces. Leaf fresh mass (LFM) was weighted using a balance (0.0001 g, Meilen, Meifu Electronics Co. Ltd., Shenzhen, China). Following the rehydration procedure, the leaves were carefully dabbed with tissue paper to eliminate any residual surface moisture prior to measuring the leaf saturated mass (LSM). Samples were subsequently oven-dried to a constant mass at 65°C for at least 48 h and then weighed to obtain the leaf dry mass (LDM). Leaf volume (LV) was estimated using LA multiplied by LT. The leaf mass per area (LMA), the reciprocal of the specific leaf area (SLA), was calculated using the LDM divided by the LA. Leaf density (LD) was calculated by dividing LMA by LT. The leaf dry matter content (LDMC) was calculated as the ratio of LDM to LSM. Finally, water saturation deficit (WSD), a critical parameter widely utilized for assessing plant tolerance to temporary water shortages, was calculated as follows (Lalet al. , 2009):
\begin{equation} \text{WSD}\left(\%\right)=\frac{\left(LSM-LFM\right)}{\left(LSM-LDM\right)}\times 100\%\nonumber \\ \end{equation}

2.4 Statistical analyses

All the statistical analyses were conducted using R (version 4.3.0, R Core Team 2023). Normality, homoscedasticity, and model fit were assessed using residual plots, Shapiro-Wilk test, and Levene’s test. First, we conducted two-way analysis of variance (ANOVA) using general linear model procedures to test for the main effects of intertidal gradients and growth forms and their interactions on leaf traits. When the effects of treatments were significant, mean comparisons were performed using the ‘emmeans ’ package. Second, phylogenetic signals of all traits were calculated with Blomberg’s K statistic (Blomberg et al. , 2003) using the ‘picante ’ package. This test compares the variance of the phylogenetically independent contrast of the study trait against those obtained with data randomly reshuffled in the phylogeny. A K value close to 1 indicates a significant phylogenetic effect, while a value close to 0 suggests no phylogenetic signal. In this study, the K values were less than 1, and the corresponding p values were greater than 0.05 for all traits, suggesting a lack of phylogenetic conservatism (Appendix Table S1). To investigate multivariate trait relationships, we performed principal component analysis (PCA) on all 11 leaf traits and plant sizes using the ‘vegan ’ package. Finally, we used simple regression analyses to examine the effects of plant height and diameter on LCC, LD, LDMC, WSD, and LMA and used general linear models to test the difference in regression slopes between intertidal zones and growth forms.