Divergence of vessel diameter explains interspecific variation in hydraulic safety to salinity in the Sundarbans mangrove ecosystem

Sundarbans (~10000 sq km), the world’s largest single-block mangrove ecosystem, provides important ecosystem goods and services to > 7.5 million coastal people in two neighboring countries, Bangladesh (~60%) and India (~40%). It has extensive variability in environmental variables, including siltation and salinity, which are influenced by freshwater damming in the north and changing seawater levels from the south. Based on spatial salinity gradients, this ecosystem is divided into three contrasting salinity zones: the low salinity zone (LSZ, eastern and northeastern part), the medium salinity zone (MSZ, central and southern part), and the high salinity zone (HSZ, western and southwestern part). Forest growth, such as tree height, increases from the HSZ to the LSZ and varies between salinity zones. However, human disruptions and rapid environmental changes, such as sea level rise (SLR), pose a threat to this sensitive ecosystem. Increasing salinity creates challenges for hydraulic transport in mangrove species. Therefore, studies on variation in intra- and interspecific hydraulic traits, and associated xylem anatomy may allow us to understand the adaptation of mangrove species to changing environmental conditions. in this study, we examined how xylem and hydraulic traits vary among dominant tree species such as Exocecaria agallocha, Xylocarpus moluccensis, and Heritiera fomes growing under climatically identical but heterogeneous environmental conditions in the Sundarbans. Although potential conductivity (KP) and leaf-specific conductivity (KL) showed species-specific variation, a notably greater conductivity was found in the low salinity zone (LSZ), which had lower vessel wall reinforcement (t/b)2. Xylem and hydraulic traits exhibited mostly strong phylogenetic signals, whereas pairwise relationships between traits were phylogenetically independent. The study species had distinct hydraulic characteristics, where vessel diameter was strongly related to the variation in KP and KL. Furthermore, the study species exhibited a weak trade-off between hydraulic efficiency and safety. A higher frequency of smaller vessel diameters in light-demanding E. agallocha indicates greater hydraulic safety under stressful conditions than in shade-tolerant H. fomes, followed by X. moluccensis. Although species characteristics place broad bounds on xylem traits, the combined effects of salinity, nutrient availability, and tree size modulate vessel diameter, which leads to hydraulic conductivity variation. The contrasting safety in terms of vessel diameter in mangroves suggests an important role in adaptation to salinity and reveals an underlying mechanism of tree growth and species distribution in the Sundarbans.

Methods
Sampling was done in the three salinity zones (LSZ, MSZ and HSZ) of the Bangladesh Sundarbans (21°30′ – 22°30′ N, 89° 00′ – 89°55′ E). Total 218 wood cores were collected (at a height of 1.3 m from the ground) from three widely distributed species, such as Exocecaria agallocha (105), Xylocarpus moluccensis (47) and Heritiera fomes (66). The DBH (cm) and tree height (m) were measured using a diameter tape and a Suunto clinometer (Suunto, Vantaa, Finland), respectively. Afterwards, the wood cores were placed in polythene bags separately for laboratory analyses.

We used the outer part of the wood core (near bark) from each tree for microtomy to minimize the age effect. Wood sections of 15 μm in thickness were cut using a sliding microtome (Microm, Fisher Scientific, Walldorf, Germany). Afterwards, the sections were stained in a solution of 0.1% safranin (Merck KGaA, Darmstadt, Germany) and 50% ethanol. The sections were subsequently washed for 5 minutes in 50, 75, 96, and 100% ethanol series and subsequently mounted on microscopic slides. Observations were performed under a light microscope, and images were captured using a camera system connected to the microscope. The microscopic images were analyzed using Fiji ImageJ software.

We used five 500 μm resolution images for vessel density (VD) measurements for each tree using the following equation: 

VD = number of vessels/area (mm2)

The vessel diameter, D (µm) was measured on the images. We randomly measured the diameters of 100 vessels for each tree.

The grouping index (GI) was calculated as the ratio of the total number of vessels to the total number of vessel groupings in each image.

The proportion of water-conducting vessel area (F) and vessel composition index (S, mm4) for each tree were also calculated following the following equations:

F = VA * VD

S = VA/VD

where VA indicates the mean vessel area, and VD indicates the mean vessel density.

Based on the Hagen–Poiseuille law, the hydraulically weighted diameter (DH, µm) was calculated using a statistic that weights the vessel lumen size.

DH = ƩD5/ƩD4

Hagen–Poiseuille’s law was also used to determine potential conductivity (KP, kg m−1 MPa−1 s−1) based on hydraulic diameter (DH) and vessel density (VD).

KP = () * DH4 * VD

where η is the viscosity index of water (1.002 × 10−9 MPa s at 20 °C)

ρ is the density of water at 20 °C (998.21 kg m−3)

D is the vessel diameter (µm)

Leaf-specific conductivity (KL, kg m–1 MPa–1 s–1) was calculated as the ratio of potential conductivity (KP) to leaf area (LA).

The double wall thickness (μm) was measured for 50 pairs of interconnected vessels on each tree to determine the mean thickness. The vessel wall reinforcement (t/b)2 indicates the capacity of vessels to resist water pressure deficit and was calculated as the ratio of the paired vessel wall thickness (t) and the diameter of the vessel closest to the hydraulic mean diameter (b).

All trees with DBH ≥ 4.5 cm (at a height of 1.3 m from the ground) were measured to calculate stand characteristics such as tree density (stem ha−1). The DBH (cm) and tree height (m) were measured using a diameter tape and a Suunto clinometer (Suunto, Vantaa, Finland), respectively. We also quantified the coefficient of variation of DBH (CoV DBH) and tree height (CoV H), which indicates tree size inequality in the sample plots. The third order fully expanded leaf pair from the apex of the plagiotropic (lateral) branches was selected for sampling. We measured the fresh area of the leaves using images (captured by a digital camera, Nikon D5500, Nikon, Tokyo, Japan) in Adobe Photoshop CS (Adobe, San Jose, California, USA) following Sarker et al. (2021). 

The soil samples from each PSP were collected at a depth of 15 cm using a 5 cm diameter cylindrical soil core sampler for laboratory analysis (Sarker et al., 2016). The collected composite soil samples were used for these analyses. The soil salinity (EC, mS cm-1) was measured using a digital conductivity meter (Extech 341350A-P Oyster) at a 1:5 ratio in a suspension of soil in water (Hardie & Doyle, 2012). A digital pH meter (Extech RE300 ExStik) was used to measure the soil pH. The soil texture (silt percentage) was analyzed using the hydrometer method (Gee & Bauder, 1986). Highly dynamic regional hydrology influences sediment deposition in the Sundarbans and, consequently, at variable elevations, even within a small distance (< 1 km) (Sarker et al., 2016). Five elevation readings were randomly extracted from each PSP from the digital elevation model (DEM), and these readings were averaged for each PSP to reduce the errors associated with the DEMs (Sarker et al., 2021).

We selected resource variables such as soil NH4, P, and K because of their important influences on mangrove tree growth and development (Reef et al., 2010). Soil NH4 was measured following the Kjeldahl method (Bremner & Breitenbeck, 1983). We followed the molybdovanadate method to measure the soil P concentration (Ueda & Wada, 1970). The total soil K was determined using an atomic absorption spectrophotometer (AA-7000).