Forest structure and biomass
Stem density significantly differed among the three mangrove forests (p<0.001). The KO plantation had the highest stand density, while its average stem diameter and height were significantly lower than those of the other two forests (p<0.001; Table 1). Natural colonization of K. obovata increased the stand density of KS markedly from 1367 to 2600 stem ha-1 over the history of forest. Tree height and basal area of S. apetala showed no significant differences between the monoculture and mixed forest, while such parameters of K. obovata in the KO plantation were significantly lower than those of S. apetala in both SA and KS.
Table 1 Forest structure and biomass of different mangrove forest types in Hanjiang River Estuary. Data are mean±SE, n=3 to 5. Different letters indicate significant differences among the three plantations (p<0.05).
Parameters
|
Forest type
|
SA
|
KO
|
KS
|
Density (stems ha-1)
|
1367±145a#
|
9533±63c#
|
2600±557b
|
DBH (cm)
|
14.5±0.9b#
|
8.8±0.3a#
|
16.3±0.9b*
|
Tree height (m)
|
8.5±0.3b#
|
4.8±0.07a#
|
8.8±0.3b*
|
Basal area (m2 ha-1)
|
26.0±3.0a#
|
62.7±4.5b#
|
27.8±2.8a*
|
Aboveground biomass (Mg ha -1)
|
67.1±7.7a
|
195.8±23.7b
|
71.9±11.7a
|
Ground layer biomass (Mg ha -1)
|
2.9±0.6b
|
1.5±0.3a
|
2.8±0.5b
|
Root biomass (Mg ha -1)
|
3.7±0.2a
|
34.0±3.6c
|
11.9±1.3b
|
Total vegetation biomass per unit stem (kg)
|
53.9±3.9a
|
24.3±1.5c
|
33.3±3.2b
|
Different superscripts in the same column indicate significant difference across forest types (p<0.05).
*Data only refers to S. apetala species in the mixed KS forest.
#Data source: He et al. (2018)
The highest mean total vegetation biomass occurred in KO (231.3±14.7 Mg ha-1), followed by KS and SA at 86.6±8.2 and 73.7±5.4 Mg ha-1, respectively. However, the total vegetation biomass per unit stem of the two forests with S. apetala present were both significantly higher than that of K. obovata monoculture (p<0.001; Table 1). The aboveground and ground layer biomass significantly differed across forest types (p<0.05; Table 1), while this profile also applied to total root biomass, which ranged from 3.7±0.2 to 34.0±3.6 Mg ha-1 (p<0.001; Table 1). In SA, live root biomass was negatively correlated with soil depth (p=0.002; Fig. 2a). Similarly, root necromass in SA significantly decreased with soil depth (p=0.003; Fig. 2b). No significant difference was detected in live root biomass among soil depths in KO, while the root necromass of this species showed a similar distributional pattern in depth as that of SA (p=0.01; Fig. 2b). In KS, live root biomass did not differ among soil depths, while root necromass increased with soil depth (p<0.001; Fig. 2b). The total live root biomass was 1.1±0.2, 17.8±3.9 and 3.4±0.6 Mg ha-1 in SA, KO and KS, respectively, accounting for 29.7%, 52.3% and 28.7% of total root biomass. There were significant differences between both plantation types and soil depths (p<0.001). Forest types and soil depth also interacted significantly in terms of live root biomass in the three mangrove forests (p=0.015; Table 2). The overall root necromass of KO and KS were 7.3±1.4 and 9.5±1.4 Mg ha-1, respectively, which were significantly higher than that of SA. There were significant differences of root necromass between both forest types and soil depths (p=0.004 and 0.008, respectively; Table 3), also with significant interaction effects (p<0.001; Table 2).
Table 2 F values of two-way ANOVA testing the differences in live roots biomass, roots necromass and soil variables among different mangrove forest types / sites and soil depth in Hanjiang River Estuary, south China.
Dependent variables
|
Sources of variance
|
Forest types/Sites
|
Soil depth
|
Forest types /Sites × Soil depth
|
Live roots biomass (Mg ha-1)
|
33.454**
|
5.058**
|
7.429*
|
Root necromass (Mg ha-1)
|
5.127*
|
4.162*
|
0.738**
|
Organic carbon storage in live roots (MgC ha-1)
|
34.316*
|
5.812**
|
2.804**
|
Organic carbon storage in dead roots (MgC ha-1)
|
15.063**
|
2.227
|
5.162**
|
Organic carbon storage in soil (MgC ha-1)
|
2.659*
|
34.167**
|
6.629**
|
* p<0.05 (n=75 to 85); ** p<0.001 (n=75 to 85).
Organic carbon stocks and accumulation
The total vegetation organic carbon storage of these three mangrove forests was significantly different, with KO (85.8±1.4 MgC ha-1) being 2.5 and 3.0 times higher than those of KS and SA, respectively. The organic carbon storage in aboveground biomass of KO was significantly higher than those of SA and KS (p<0.001; Table 3). Ground layer organic carbon storage accounted for 0.7%-4.1% of the total organic carbon storage, and the ground layer organic carbon storage peaked at KS (p=0.019; Table 3). For belowground biomass, the maximum value was 15.8±0.8 MgC ha -1 at KO, which was significantly higher than those of the other two forests (p<0.001; Table 3). Ratios of organic carbon storage in aboveground to belowground biomass may provide an indication of the proportion of C allocated to above- and belowground components. These ratios were 21.9 in SA, 6.4 in KO and 5.3 in KS, and were significantly different. Organic carbon storage in live roots varied with depth and the mean values were 0.3±0.06, 6.7±0.8 and 1.3±0.2 MgC ha-1 in SA, KO and KS, respectively (Fig. 3). Significant differences in live root organic carbon storage among the three forest types and soil depths were detected (p<0.001; Table 2). In SA and KO, the organic carbon storage in dead roots significantly decreased with soil depth (p=0.002 or 0.009; Fig. 3a and b). In contrast, organic carbon storage in dead roots in KS increased with soil depth (p<0.001; Fig. 3c). Organic carbon storage in live and dead roots of KO was significantly higher than those of SA and KS (p<0.001; Table 2). Plantation type and soil depth have significant interactive effects on organic carbon storage in live and dead roots of the three mangrove plantations (p<0.01; Table 2).
Table 3 Organic carbon storage in different component of different habitats in Hanjiang River Estuary. Data are mean±SE, n=3 to 5.
Parameters
|
Habitats
|
SA
|
KO
|
KS
|
MF
|
AGOC (MgC ha-1)
|
26.1±0.8b
|
73.6±1.3a
|
27.3±0.7b
|
NA
|
BGOC (MgC ha-1)
|
1.2±0.1c
|
11.5±1.0a
|
5.2±0.5b
|
NA
|
AGOC:BGOC
|
21.9a
|
6.4b
|
5.3c
|
NA
|
GLOC (MgC ha-1)
|
1.5±0.3a
|
0.7±0.1b
|
1.6±0.2a
|
NA
|
SOC (MgC ha-1)
|
7.8±0.5c#
|
15.8±0.8a#
|
11.9±2.4b
|
4.7±0.9d#
|
TOC (MgC ha-1)
|
36.6±1.3c
|
101.6±1.4a
|
46.0±3.0b
|
4.7±0.9d
|
TOC per unit stem (kgOC)
|
26.8±8.9a
|
10.7±2.2c
|
17.7±5.4b
|
NA
|
Superscripts in the same column indicate significant difference across forest types (p<0.05). AGOC, BGOC, GLOC, SOC and TOC refer to aboveground, belowground, ground layer, soil, and total organic carbon storage, respectively.
#Data source: He et al. (2018)
The mangrove forests (0.96-3.3%) showed significantly higher soil organic carbon concentration than the mudflat (0.55%) 12 years after the mangrove planted and subsequent forest growth. The soil organic carbon (SOC) concentration (0-100 cm) of the mangrove forests varied between 0.96±0.3 % (He et al. 2018) at SA and 3.3±0.4 % (He et al. 2018) at KO, with the SOC concentration of both KO and KS (2.1±0.4 %) being significantly higher than that of SA (p<0.001). The SOC concentration significantly decreased with soil depth for KO and SA but no significant trend was evident for KS and MF. Soil bulk density (SBD) showed the opposite trend to SOC concentration. The adjacent unvegetated mudflat had the highest SBD value for the entire 1 m soil column (0.94±0.08 g cm-3; He et al. 2018) among the four study sites, which was close to that of SA (0.89±0.04 g cm-3; He et al. 2018). The mean SBD (0-100 cm) of KO (0.45±0.03 g cm-3; He et al. 2018) and KS (0.64±0.05 g cm-3) were significantly lower than those of MF and SA (p<0.001), indicating much finer sediment was found in these forests. The mean organic carbon storage in soil of KO was 15.8±0.8 MgC ha-1, 2.01, 1.33 and 3.35 times higher than those of SA, KS and MF, respectively. Soil organic carbon storage was also significantly affected by site and soil depth, with a significant interaction effect (p<0.01; Table 2 and 3 and Fig. 3).
The total organic carbon storage of different habitats was estimated by summing the vegetation organic carbon storage, ground layer organic carbon storage, and soil organic carbon storage. In general, total organic carbon storage was highest in KO amongst all habitats, then decreasing significantly in the order KS, SA and MF (p<0.001; Table 3). However, this trend was reversed in the individual tree level (Table 3).
Organic carbon allocation patterns in different forests
Total organic carbon storage in all mangrove forests was positively correlated with organic carbon storage in aboveground biomass, belowground biomass, and soil (p<0.05; Fig.4). Moreover, in SA and KS, a significant positive correlation was found between organic carbon storage in litter and total organic carbon storage, while no such relationship was detected in KO, suggesting that roots contributed mostly to forest total organic carbon stock in KO (p<0.01; Fig.4).
Due to their vital contribution to soil organic carbon storage, the correlation among litter, root and soil organic carbon had also explored. The highest value of organic carbon storage in litter was found in KS (0.6±0.07 MgC ha-1) and the lowest in KO (0.15±0.08 MgC ha-1). Soil organic carbon density in all forests was significantly positively correlated to organic carbon storage in litter, and the slope of the regression line of SA and KS were significantly steeper than that of KO (p<0.01; Fig.5). Similarly, in all forests, a significant positive correlation was detected between organic carbon storage in soil and roots. The slopes of the regression lines are all significantly different, suggesting that the contribution of root organic carbon to soil organic carbon is dependent on stand composition (p<0.001; Fig.5).