Data from: Improved use of deep soil water promotes radial growth during drought
Abstract
Increasingly frequent seasonal droughts under rapid climate change threaten the growth stability of subtropical mixed forests across the mid-subtropical monsoon region. In this study, we combined stable hydrogen and oxygen isotope analyses with high–precision radial growth monitoring to explore how water source use influences the radial growth of four subtropical tree species. Three contrasting functional strategies were identified based on rooting depth and water uptake plasticity. First, the shallow water-dependent species (Cupressus funebris Endl.) featured a root system concentrated in the upper soil layer and obtained 38.3% of its water from shallow layers. During the drought period, declining surface soil water availability was accompanied by low stomatal conductance and photosynthetic rates, and the species experienced net stem shrinkage of 1.1 mm. Second, the deep-rooted species (Michelia maudiae Dunn and Abies fabri Craib) exhibited a dual rooting strategy: deep taproots accessed stable water in the 80-100 cm soil layer, which contributed 67.2-75.8% of water uptake during drought, while shallow fine roots captured sporadic rainfall in the upper soil layers. During the drought period, these species maintained relatively high water potentials and positive cumulative radial growth, with annual radial growth 1.9-2.5 times that of C. funebris. Third, the balanced-strategy species (Liquidambar formosana Hance) displayed a relatively uniform root distribution throughout the soil profile, enabling flexible water acquisition across soil layers and sustaining the highest cumulative radial growth during the drought period. Structural equation models revealed that access to stable water sources, particularly deeper reserves as surface soils dried, sustained radial growth indirectly by maintaining stomatal conductance and photosynthesis. This pathway provides a mechanistic explanation for contrasting drought adaptation among subtropical tree species.
Dryad DOI: https://doi.org/10.5061/dryad.h70rxwf13
Summary
This dataset supports the manuscript Improved use of deep soil water promotes radial growth during drought. The study examined how water-source depth was associated with leaf water status, gas exchange, and stem radial growth in four co-occurring subtropical tree species in Caoyutang National Forest Park, Zhejiang Province, China.
The study species were Liquidambar formosana, Michelia maudiae, Abies fabri, and Cupressus funebris. They represent three water-use strategies: a balanced strategy (L. formosana), a deep-rooted strategy (M. maudiae and A. fabri), and a shallow water-dependent strategy (C. funebris).
The dataset contains the numerical data used for the main figures describing:
- proportional uptake of water from five soil layers under soil-water-deficit and wet conditions;
- predawn and midday leaf water potential;
- net photosynthetic rate, stomatal conductance, transpiration rate, and leaf cellulose oxygen-isotope enrichment; and
- half-hourly cumulative stem radial growth and soil water content during 2024.
Water-source, leaf-water-potential, and physiological measurements were collected during the 2023 growing season. The continuous radial-growth and soil-water-content series cover 1 January to 31 December 2024.
Files
Data.xlsx
Data.xlsx contains one worksheet named All_Data. Four data sections are arranged vertically in the worksheet. Each section has an English header row. Blank columns separate related data blocks and do not represent missing observations.
Study design
Three 20 m × 20 m monitoring plots were established in the study area. Five dominant or co-dominant trees of each species were monitored in each plot, giving 15 trees per species for radial-growth and gas-exchange measurements. Leaf water potential was measured on three trees per species in each plot, giving nine tree-level observations per species and measurement period.
Soil-water-deficit conditions were defined as at least seven consecutive rainless days with volumetric soil water content below 8% in the 0–20 cm layer. Wet conditions were defined as the day after rainfall exceeding 20 mm when volumetric soil water content in the 0–20 cm layer exceeded 25%.
Water-source contributions were estimated using MixSIAR with both δ²H and δ¹⁸O as tracers. Five soil layers were retained as separate potential water sources: 0–20, 20–40, 40–60, 60–80, and 80–100 cm. Values represent the proportional contribution of each layer to xylem water.
Description of Data.xlsx
Section 1: Water-source contribution
- Section title: row 1
- Header row: row 2
- Data rows: rows 3–6
- One data row is provided for each of the four tree species.
- Columns A–K contain soil-water-deficit values.
- Column L is a blank separator.
- Columns M–W contain wet-condition values.
- Within each condition, soil layers are ordered as 0–20, 20–40, 40–60, 60–80, and 80–100 cm.
- Each soil layer is represented by adjacent mean and standard-deviation columns.
- Water-source contributions are expressed as percentages of total plant water uptake. The five mean contributions within each species and condition sum to approximately 100%, allowing for rounding.
Section 2: Leaf water potential and regression coefficients
- Section title: row 8
- Header row: row 9
- Leaf-water-potential data rows: rows 10–18
- Nine tree-level observations are provided for each species.
- For each species and moisture condition, predawn water potential (
Psi_pd) is followed by midday water potential (Psi_md). - Columns A–K contain soil-water-deficit measurements for LF, MM, AF, and CF, separated by blank columns.
- Columns L and X are blank separators.
- Columns M–W contain wet-condition measurements for LF, MM, AF, and CF, separated by blank columns.
- Leaf water potential is expressed in MPa. Negative values are valid observations and must not be treated as missing data.
The regression-coefficient table is in columns Y–AC and rows 10–13:
| Column | Variable | Description |
|---|---|---|
| Y | Species |
Tree species |
| Z | R2_Psi_pd_vs_SWS |
R² for predawn water potential versus shallow-water contribution |
| AA | R2_Psi_pd_vs_DWS |
R² for predawn water potential versus deep-water contribution |
| AB | R2_Psi_md_vs_SWS |
R² for midday water potential versus shallow-water contribution |
| AC | R2_Psi_md_vs_DWS |
R² for midday water potential versus deep-water contribution |
SWS denotes the 0–20 cm soil-water source and DWS denotes the 80–100 cm soil-water source.
Section 3: Photosynthetic physiology
- Section title: row 19
- Each variable block contains 15 individual-tree observations.
- Within every block, paired columns are ordered by species as LF, MM, AF, and CF.
- Within each species pair, the first column is the tree-level mean under soil-water-deficit conditions and the second column is the wet-condition value.
- Blank columns separate species pairs.
| Variable block | Header row | Data rows | Unit |
|---|---|---|---|
A — net photosynthetic rate |
21 | 22–36 | μmol CO₂ m⁻² s⁻¹ |
gs — stomatal conductance |
37 | 38–52 | mol H₂O m⁻² s⁻¹ |
Tr — transpiration rate |
53 | 54–68 | mmol H₂O m⁻² s⁻¹ |
Delta18O — leaf cellulose oxygen-isotope enrichment above source water |
69 | 70–84 | ‰ |
Section 4: Radial growth and soil water content
- Section title: row 86
- Header row: row 87
- Data rows: rows 88–17,655
- The section contains 17,568 half-hourly records covering the 2024 calendar year.
- The left block contains cumulative stem radial growth for four species.
- Column J is a blank separator.
- The right block contains volumetric soil water content for five soil depths.
| Column | Variable | Description | Unit or format |
|---|---|---|---|
| A | Date_growth |
Calendar date associated with the radial-growth record | YYYY-MM-DD |
| B–C | LF_Rg_mean, LF_Rg_SD |
Mean cumulative radial growth and SD for LF | mm |
| D–E | MM_Rg_mean, MM_Rg_SD |
Mean cumulative radial growth and SD for MM | mm |
| F–G | AF_Rg_mean, AF_Rg_SD |
Mean cumulative radial growth and SD for AF | mm |
| H–I | CF_Rg_mean, CF_Rg_SD |
Mean cumulative radial growth and SD for CF | mm |
| J | — | Blank separator | — |
| K | Datetime_SWC |
Date and time associated with the soil-water-content record | YYYY-MM-DD hh:mm |
| L–M | SWC_0_20_mean, SWC_0_20_SD |
Mean and SD of volumetric soil water content at 0–20 cm | % v/v |
| N–O | SWC_20_40_mean, SWC_20_40_SD |
Mean and SD of volumetric soil water content at 20–40 cm | % v/v |
| P–Q | SWC_40_60_mean, SWC_40_60_SD |
Mean and SD of volumetric soil water content at 40–60 cm | % v/v |
| R–S | SWC_60_80_mean, SWC_60_80_SD |
Mean and SD of volumetric soil water content at 60–80 cm | % v/v |
| T–U | SWC_80_100_mean, SWC_80_100_SD |
Mean and SD of volumetric soil water content at 80–100 cm | % v/v |
Date_growth contains the calendar date only. Consecutive rows within each date retain the original half-hourly sequence, but clock time is not explicitly stored in that column. Datetime_SWC contains the complete date and time. Later blank cells in the soil-water-content block indicate that no value was supplied in the source table.
Variable and abbreviation dictionary
| Abbreviation | Definition | Unit or coding |
|---|---|---|
| LF | Liquidambar formosana | species code |
| MM | Michelia maudiae | species code |
| AF | Abies fabri | species code |
| CF | Cupressus funebris | species code |
| A | Net photosynthetic rate | μmol CO₂ m⁻² s⁻¹ |
| gs | Stomatal conductance | mol H₂O m⁻² s⁻¹ |
| Tr | Transpiration rate | mmol H₂O m⁻² s⁻¹ |
| Psi_pd | Predawn leaf water potential | MPa |
| Psi_md | Midday leaf water potential | MPa |
| Delta18O | Oxygen-isotope enrichment of leaf cellulose above source water | ‰ |
| Rg | Cumulative stem radial growth | mm |
| SWC | Volumetric soil water content | % v/v |
| SWS | Shallow water source, corresponding to the 0–20 cm soil layer | proportional contribution or % |
| DWS | Deep water source, corresponding to the 80–100 cm soil layer | Proportional contribution or % |
| SD | Standard deviation | same unit as the corresponding mean |
| R² | Coefficient of determination | unitless |
Missing data and data-quality notes
- Empty cells represent unavailable or non-applicable values. No numeric sentinel such as
-999is used. - Negative leaf-water-potential values and negative radial changes are valid measurements.
- Blank separator columns are intentional formatting elements and should not be interpreted as variables.
- Multiple soil-water-deficit measurements were averaged within each tree before comparison with the single post-rainfall wet-condition measurement.
- Summary values may differ slightly from totals or values recalculated from displayed numbers because of rounding.
Sharing and access information
The associated article is intended for publication in Functional Ecology. The Dryad dataset DOI is https://doi.org/10.5061/dryad.h70rxwf13. The article DOI can be added after it becomes available.
No external data sources are required to interpret Data.xlsx. Study-site climate and soil information and full sampling and analytical protocols are described in the associated manuscript and Supporting Information.
Code and software
Data.xlsx can be opened with Microsoft Excel, LibreOffice Calc, or other software supporting the Office Open XML spreadsheet format. No proprietary Origin project file is required to access the deposited data.
The analyses described in the associated manuscript used:
- R version 4.3.0;
- R packages
lme4andlmerTestfor linear mixed-effects models; - MixSIAR version 3.1.7 for dual-isotope water-source partitioning; and
- SPSS version 16.0 for repeated-measures ANOVA and descriptive statistics.
Analysis scripts are not included in this dataset.