The littoral zone ecosystem of the Three Gorges Reservoir (TGR) has become significantly degraded by annual cycles of prolonged winter flooding and summer drought. For purposes of flood control and sediment management, the water level in the reservoir is lowered by 30 m during the summer monsoon season and raised again to 175 m above sea level each year at the end of the monsoon period. To explore an effective way to promote biodiversity and associated ecosystem services, we examined Taxodium distichum as a species for afforesting the littoral zone. Sapling growth variations were measured after two rounds of winter flooding. Dominant influence factors were determined by redundancy analysis. Herb community similarities between the experimental afforested areas and nearby control areas were assessed to detect the ecosystem influence of the experimental afforestation. 94.5% of saplings planted at elevations above 168 m survived. All measured growth indices (tree height, diameter at breast height, crown width and foliage density) decreased as the flood depth increased. Completely submerged saplings had a mean dieback height of -0.65 m. Greater initial foliage density led to increased tree height and stem diameter. Shannon-Wiener indices were not significantly different between plots in experimental and control areas, but the low similarity of herb communities between experimental and control areas (0.242 on average) suggested that afforestation would enrich plant community structure and improve littoral zone ecosystem stability. Because littoral zone afforestation provides several ecosystem services (habitat, carbon sink, water purification and landscaping), it is a promising revegetation model for the TGR.
Citation: Li B, Du C, Yuan X, Willison JHM, Xiao H (2016) Suitability of Taxodium distichum for Afforesting the Littoral Zone of the Three Gorges Reservoir. PLoS ONE 11(1): e0146664. https://doi.org/10.1371/journal.pone.0146664
Editor: Manuel Reigosa, University of Vigo, SPAIN
Received: June 16, 2015; Accepted: December 21, 2015; Published: January 15, 2016
Copyright: © 2016 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by the National Natural Science Foundation of China (grant number:51278505 http://www.nsfc.gov.cn/publish/portal1/). This funder will afford the publication fee. It was also supported by the Postdoctoral Foundation of China (grant number:2014M550458, http://jj.chinapostdoctor.org.cn/V1/Program1/Default.aspx) and Special Postdoctoral Science Fund of Chongqing (grant number:Xm2014098, URL:http://zhuanji.cqhrss.gov.cn/u/zhuanji/). These funders support for the preparation work of this manuscript.
Competing interests: The authors have declared that no competing interests exist.
Vegetation is an important element of the functional ecology of reservoir littoral zones and consequent provision of ecosystem services. It improves water quality by filtering agricultural nutrient runoff and trapping sediments [1,2], and reduces flood risks by intercepting runoff and enhancing evaporation via transpiration [3,4]. Littoral zone plants can reduce soil erosion and reinforce bank stability . Vegetation communities in the littoral zone provide habitat for fish , birds  and biodiversity in general. In light of these benefits, restoring vegetation to damaged reservoir ecosystems is widely regarded as important [8,9].
The Three Gorges Dam (TGD) constructed on the Changjiang (Yangtze River) at Yichang City in China created the Three Gorges Reservoir (TGR) which runs more than 600km upstream to Jiangjin County of Chongqing . The TGR management strategy is popularly expressed as “storing clear water and releasing muddy”: the water level is drawn up to 175 m above sea level in winter and is drawn down to 145 m above sea level in summer. On the one hand, this strategy has provided the benefits of flood control, hydropower generation, improved navigation and sediment release; on the other hand, the 30 m water level fluctuation has produced a huge littoral zone which is largely bare of vegetation. Many plant species have disappeared from the littoral zone due to lack of adaptation to prolonged winter flooding followed by extended summer drought, and ecosystem functions such as maintenance of plant cover and resistance to erosion have been significantly degraded as a result [11–13]. Some ruderal plant species have propagated rapidly in the altered littoral zone of the TGR (LZTGR), such as outbreaks of Xanthium sibiricum in 2008 and 2009 . For these reasons, it is important to find appropriate methods for improving ecological functions and services in the bays and minor tributaries of the reservoir.
Ecosystem restoration by rehabilitating vegetation in the LZTGR has been extensively studied in recent years. One group of perennial non-woody plants (Acorus calamus, Alternanthera philoxeroides, Arundinella anomala, Hemarthria compressa, Triarrhena sacchariflora, Vetiveria zizanioides, Cynodon dactylon and Cyperus rotundus) has been identified as suitable for combating soil erosion [15,16], and a group of agricultural hydrophytes (Nelumbo nucifera, Sagittaria trifolia var. edulis, Eleocharis dulcis, Zizania latifolia, Trapa bispinosa and Oryza sativa) has been reported as promising species for alleviating the conflict between ecosystem conservation and agricultural production . In addition, some woody species such as Salix variegata , Salix rosthornii , Distylium chinense, Populus x canadensis , Pterocarya stenoptera , Morus alba , Nyssa aquatica , Taxodium ascendens and Taxodium distichum  have been examined as candidate species for revegetation of the LZTGR. Of these, Nyssa and Taxodium are native to the United States, Populus species are widespread in China (Populus x canadensis is a horticultural hybrid), and the other species are native to China.
T. distichum was first introduced into regions of China having suitable humid monsoon climatic conditions at least 100 years ago and is now used widely in wetland landscaping. The species has excellent flood tolerance in the littoral zones of reservoirs in the United States [26,27], and although it has been proposed as a promising candidate for revegetation projects in the LZTGR , studies reported to date have been conducted in greenhouses [28,29] or in regions of the TGR mostly subjected to summer flooding . Experiments are needed on the acclimation, performance, and habitat effects of T. distichum seedlings and saplings in typical water-fluctuation conditions of the LZTGR itself.
In this paper we report experimental afforestation with T. distichum conducted in the context of an ecological engineering model for improving ecosystem functions in the LZTGR. The model is derived from field experiments and named 'littoral woods engineering' (LWE) . We have addressed the following questions. 1) To which water-fluctuation elevation ranges are saplings of T. distichum adapted in the TGR, and how well do they perform at these elevations? 2) How does the status of T. distichum saplings prior to submersion affect their subsequent performance in the LZTGR? 3) What are the effects of this afforestation on the understory herb communities and on LZTGR ecosystem stability?
Materials and Methods
The study site lies in the experimental zone of Pengxi River Wetland Natural Reserve, on the banks of the Baijia Stream associated with the Pengxi River, a tributary of the middle reaches of the Changjiang, lying to its north. The experimental zone was expressly established to permit and encourage experiments of the sort described in this paper. The planting site is about 291 km upstream from the TGD, close to a small research station which was set up to facilitate research on the reservoir drawdown zone (latitude 31°09′ N, longitude 108°34′ E, Fig 1). The China Three Gorges Corporation implemented land clearing in the LZTGR when the TGD was under construction and almost all the trees and shrubs were removed from the LZTGR , including this site. Because the study site is directly affected by seasonal water-level fluctuation of the TGR, it is characterized by prolonged winter flood and extended summer drought. As a result, many pre-existing local plants have not adapted to the alternating flood-drought environment and have disappeared . Like other regions in the LZTGR, the ecosystem of the study area has been characteristically unstable since 2008 when the water level was first raised to more than 173 m above sea level .
On March 28th, 2009, 168 four-year-old T. distichum trees, grown from seed, were obtained from Lushan Seedling Company in Jiujiang City, Jiangxi Province, China. The average tree height (TH) was about 2.5 m and average diameter at breast height (DBH) was about 2.0 cm. With the permission of Pengxi River Wetland Natural Reserve Management Bureau, all the trees were planted in the study area 4 days after being dug from the donor field. Each was transplanted with an attached earth ball, about 30 cm in diameter.
Nine fallow fields lying at elevations between 168 m and 173 m above sea level were established as experimental plots for planting the T. distichum trees (Fig 1). The 168 trees were planted with 4-m spacing between individuals and numbered sequentially. The plots were in two clusters and were selected to be as similar as possible to each other with respect to physiographic conditions and soil conditions, with the intent that two sets of variables could be monitored: hydrological variables due to elevation among the plots, and the variable initial biological conditions of the trees. Statistical and ordination analyses were conducted to determine the dominant factors that influenced tree growth using an experimental approach similar to that of Jimenez et al.  and Li et al. . To detect possible effects of T. distichum afforestation on understory vegetation, we surveyed the herb communities in 4 plots within the afforested experimental areas, and a set of 4 control areas lying at similar elevations within the Baijia Stream drawdown zone nearby (Fig 1). The control areas had not been influenced by the afforestation work. The understory investigation was conducted about 30 months after the plantations had been established.
Sampling and data analysis
The following growth indices were measured on each tree in August 2009 and August 2011: TH, DBH, tree top elevation (TTE, the sum of plot altitude and TH), living crown width (CW, mean value measured in two opposite directions), and foliage cover (FC). For FC we measured the proportion of sky occluded by foliage when looking up at the canopy from beneath the tree . This was done on a per tree basis by the same observer. Hydrologic data from the Wanzhou station (about 289 km upstream from the TGD) were recorded continuously and used to estimate the water level variation of the study site (Fig 2).
The duration of continuous flooding was more than four months at 168 m above sea level, and there were about 225 accumulated days of flooding at that elevation.
Surveys of understory vegetation were conducted in late August and early Sept 2011. In the afforested experimental area, 3 square sample plots of 1 m2 were set randomly in each of the 4 largest experimental plots (Table 1, LWP-4, LWP-5, LWP-6, LWP-9, containing 116 trees in total), and for comparison 4 sampling sites containing the characteristic dominant plant communities of the contrast region were selected and 3 square sample plots of 1 m2 were examined within each of the 4 sampling sites. Number of individuals, height, and coverage of each plant species in each sample plot were recorded separately.
The redundancy analysis (RDA) method  was used to evaluate the growth responses of the trees to the following influence factors: TH, TTE, DBH, CW, FC, altitude (Alt), continuous flooding duration (FLD) and accumulated flooding duration (Acc-FLD).
Species diversities of herb communities were characterized by Richness index S: S = s (total number of species), and Shannon-Wiener diversity index H’ : , where pi is the proportion of the ith species. Plant communities were distinguished according to importance values (IV) [37,38]: IV = (Dr + Cr + Hr)/3, where Dr, Cr and Hr are respectively the relative density, coverage and height of the ith species in sampling plots. A symmetric square matrix of similarities among sampling plots was constructed using the abundance-based Bray-Curtis coefficient: , where yij (or yik) is the abundance of the ith species in the jth (or kth) sample . The RDA ordination was conducted using the CANOCO software package (version 4.5) , while other statistical analyses were performed with SPSS 17.0 software.
In August 2009 (about 5 months after transplantion), 163 trees had survived (survival rate 97.02%), with average parameters: TH 2.68 m, DBH 2.11 cm, CW 0.76 m, and FC 78.16%. Two years later, in August 2011, 154 trees had survived after two rounds of submersion (survival rate of 94.48%), with the following changes in parameters: TH 0.37±0.03 m, DBH 0.96±0.05 cm, CW 0.37±0.04 m and FC 4.28±1.26%. There were significant differences (Mann-Whitney U Test, p<0.05) between trees in partly and completely submerged situations with respect to variation of TH (var-TH), DBH (var-DBH) and CW (var-CW), but not of FC (var-FC) (Fig 3). In 2011, surviving trees that had been completely submerged had dieback rate of 6.35%, and mean dieback height (measured from top of dead stem to top of living tissue) of -0.65 m.
Significant differences between the two conditions were analyzed using the Mann-Whitney U test. Box plots with different letters (a, b) were significantly different at the 0.05 level. "·" indicates data considered mild outliers, and "*" indicates data considered extreme outliers.
RDA ordination (growth response)
Results of Monte Carlo permutation tests showed highly significant effects (p = 0.002 with 499 permutations) of the influence factors on the growth responses of the T. distichum trees. As shown in the RDA summary (Table 2), the influence factors explain about 26.4% of the total variation of plant growth, and the first two axes taken together displayed most (91.4%) of the variation.
The growth variation influence-factor biplot diagram derived from the RDA ordination (Fig 4) showed that the dominant influence factors on tree growth were FC, FLD, Acc-FLD, Alt and TTE (positively correlated with Alt, r = 0.9743). Foliage cover (FC) was positively correlated with variations in var-TH and var-DBH. The Alt variable was positively correlated with all the growth parameters, especially var-CW. By contrast, the flood variables (FLD and Acc-FLD) were negatively correlated with all growth parameters. The morphological variable, FC, was positively correlated with var-DBH and var-TH, negatively correlated with var-FC, and almost uncorrelated with var-CW.
Growth variations are shown as arrows with solid arrowheads, and influence factors are shown as arrows with empty arrowheads. The approximate correlation between growth variation and influence factor is equal to the cosine of the angle between the corresponding arrows.
Shannon-Wiener (H’) and species richness (S) indices of biodiversity are summarized in Table 3. H’ indices of understory herb communities in the afforested experimental area were distributed within the range 1.474±0.172–2.652±0.201, in contrast with the range 1.814±0.476–2.526±0.264 in control plots. The S indices in the afforested experimental areas were 4.000±0.577–8.667±0.882, in contrast with 5.000±1.155–9.667±0.667 in control plots. Results of One-Way ANOVA analysis indicated that there were no significant differences in H’ between most plots, with the exception of plot LWP-9 compared with plots LWP-4 and Con-3, but there were more plots which had significant differences in species richness (S), such as plot Con-1 and Con-3, Con-1 and LWP-4, Con-4 and LWP-9.
Eight herb species were found only in the afforested experimetal area, while 12 species were found only in the control plots. The IV indices of species in each plot were calculated and a similarity matrix of Sjk among sampling plots was constructed for further analysis of community structures. According to the IV numbers in Table 4, the principal plant communities in the afforested experimental area were the Cynodon dactylon+Bidens pilosa, Oplismenus compositus, Setaria viridis and Bidens pilosa associations. By contrast, the most widespread plant communities in control plots were the Cynodon dactylon, Xanthium sibiricum+Cynodon dactylon, Conyza japonica+Bidens pilosa associations. The similarity matrix (Table 5) indicated that communities in the control plots were very dissimilar to most of those in the plots of the afforested experimental area (with the exception of plot LWP-5). Similarities between herb communities in the control plots were generally higher than those of experimental plots.
Growth variation of Taxodium distichum
T. distichum is native to the southern region of the United States. Within its natural range, it has long been a component of the mix of arboreal species selected for projects designed to restore ecological functions to forest-depleted bottomlands  and damaged ecosystems such as mine sites . It has also been planted in littoral zones of artificial reservoirs, both within its native range  and beyond its native range in the western region of the United States . T. distichum is adapted to specific natural flood pulse conditions which limit its capacity to establish self-sustaining populations . Natural recruitment has been reported to occur following prolonged reservoir drawdown , but not in the absence of drawdown . As a result of limited recruitment capacity, the species has not invaded beyond its natural range and even within its native habitat can be difficult to establish as naturally regenerating populations because the optimal conditions for natural spread may not occur annually . Although it has been grown for many years in China, it has not become invasive and is not listed as an invasive species of concern .
While Taxodium distichum is well recognized as a flood-tolerant species in its native habitat, the environmental conditions in the southeastern United States are not similar to those in the LZTGR, which is characterized by unusually prolonged winter flooding, often followed by extended summer drought. Despite the difference in environment, T. distichum performed well in the experiment reported here. According to previous reports, T. distichum seedlings tolerate complete submergence for no more than a month during the growing season [48,49], but in our study transplanted saplings survived at least 3 months of continuous flooding, some of which were completely submerged.
We found that T. distichum survived in the LZTGR at elevations above 168 m. As shown in Fig 3, the growth parameters (TH, CW and DBH) of all partly-submerged saplings (i.e. having tops above the water surface throughout the winter flood period) increased, despite a tendency for foliage cover (FC) to decline. By contrast, completely submerged saplings grew more slowly on average (Fig 3). The TH and CW of some individuals declined due to die-back from the tip (Fig 3). Greater depth and period of flooding led to slower sapling growth rates in the LZTGR (Fig 4), which corresponds with findings for growth of T. distichum at the margins of the Mississippi River . Although the average growth rates of completely submerged saplings were lower than partly submerged saplings, a few exceptional individuals were apparently stimulated by submergence (Fig 3).
Analysis revealed that the status of the T. distichum saplings before they were flooded significantly affected subsequent variation in their growth. Although var-FC was negatively correlated with FC, it appears that greater FC led to better sprouting and enhanced photosynthesis, which in turn led to better assimilation of materials and energy, greater TH and DBH, and therefore greater ability to overcome flooding in the long term by means of the tree top rising above flood water.
Flood tolerance by T. distichum has been attributed to growth in diameter and the formation of aerial roots and knee roots [51,52]. In our experiment, few aerial roots and knee roots were observed. The reason for this is not clear, but Yamamoto  found that knee roots form in an alternating (rather than an extended) state of oxidation and reduction. Ethylene that is produced and accumulates in the plant under anoxic conditions promotes increase of stem diameter and the development of aerenchyma , which permits transportation of oxygen from the atmosphere to the roots , thus helping flood-tolerant plants to cope with anaerobiosis.
Influence of afforestation on the understory herb communities
Previous investigation of the vegetation in the LZTGR suggested that plant communities at the same altitude range in any local area are similar . For this reason, we compared the understory herb communities of the afforested experimental plots with herb communities at the same elevation in nearby unmanaged areas.
Average plant species richness per unit area (1 m2) was less than 10 in both the afforested experimental area and the control area. The perennial grass Cynodon dactylon was a dominant species in all of the control plots, but annual plants (O. compositus, S. viridis, B. pilosa, D. sanguinalis) were dominant in most of the experimental plots, with the exception of LWP-5. In previous studies, C. dactylon has been recommended as a flood-tolerant species that is useful for alleviating soil erosion in the LZTGR and improving its aesthetic appearance . C. dactylon is clearly valuable for ecological restoration of the LZTGR where trees are not suitable, notably in the main transportation corridor of the reservoir, but a diversity of revegetation approaches is more appropriate in the many bays and minor tributaries of the reservoir. In these bays and minor tributaries, soil erosion is of less concern and sediment often accumulates during the winter flood periods, leading to colonization of the LZTGR by ruderal weeds when the water level falls .
Prolonged flooding not only affected the vitality and growth rate of the saplings in the experiment, but also brought about changes in the understory community. This finding is consistent with reports on stand composition and structure in typical forested wetlands , but drought tolerance may be more significant than flood tolerance in determining which annual plants dominate, and reproduce within, the understory communities. Successful colonists may have seeds that can be dispersed via water , lie dormant through the winter flood period, and germinate quickly once flood water recedes. The dispersal characteristics of species are important considerations for effective plant establishment  and should be taken into account when selecting candidate species for revegetation of the LZTGR.
Potential functions of afforestation
Afforestation of the LZTGR has many potential functions in improving the ecosystem quality. It could provide habitat for various animals living in and around the littoral zone . Firstly, almost all the shrubs and trees in the LZTGR were removed when the TGD was built, but through afforestation we can try to reintroduce well-adapted trees and shrubs and in that way rebuild plant community structures and improve habitat quality . Littoral zone afforestation may provide habitat for resident birds  and create foraging and sheltering places for migratory waterfowl in winter. Secondly, littoral woods can act as carbon sinks when carbon is fixed as organic matter in the form of root, bole and branch through photosynthesis [63–65]. Thirdly, littoral woods with understory plant communities create a barrier which can intercept non-point source pollutants carried by upland surface runoff. Lastly, afforestation can improve the aesthetic appearance of the LZTGR, which would be popular in the many towns situated along the banks of the reservoir (Fig 5).
Proposals and recommendations
The study reported here was an initial attempt to carry out LWE in the LZTGR. In order to properly understand the potential role of afforestation in restoring LZTGR ecosystems, further research is required to optimize afforestation, as follows.
- Afforestation influences the understory vegetation , and it is reasonable to expect an effect in the case of littoral zone, but by what mechanisms does afforestation affect plant community structure in the understory?
- Sustained attention and long term monitoring are needed to discover whether saplings that died back (in the experiment described here) will survive in the long term. When selecting and testing the application of other flood-tolerant candidate species, testing for tolerable flood-depth ranges should be conducted in the field using well-designed experimental protocols .
- Several herb, shrub and tree species have been shown to tolerate flooding in the LZTGR, such as Cynodon dactylon , Vetiveria zizanioides , Calamagrostis arundinacea, Salix variegata , Salix rosthornii , and “Zhongshanshan 118” (a Taxodium hybrid used in silviculture) . It will be useful to construct experimental complex littoral-woods ecosystem models composed of trees, shrubs and grasses, and test these under field conditions.
- It may be argued that the leaf litter from littoral woods would decompose under water and lead to net release of greenhouse gases. Would this be the case? Mitsch et. al.  suggested that most wetlands are net carbon sinks, even when greenhouse gas emissions were considered. In the LZTGR, leaf litter is generated mainly in winter and the relatively low water temperature will decrease methanogen activity. For this reason, the amount of greenhouse gases emissions resulting from littoral zone afforestation is uncertain. Littoral woods are complex systems: on one hand, some greenhouse gases will be emitted, but on the other hand woody plant growth will lead to storage of atmospheric carbon in the form of biomass [71,72]. Furthermore, litter fall into the TGR ecosystem will contribute dissolved organic carbon and nutrients to the TGR ecosystem [73,74], and these play important roles in biogeochemical reactions while also supplying nutrients and energy to microbial food webs [75,76]. Therefore, the advantages and disadvantages of littoral zone afforestation in general need further experimentation and modelling.
- Since water-level variation is the dominant factor affecting plant community structure in the LZTGR, is it possible to create a more robust littoral-zone ecosystem by adjusting the water-level regulatory regime of the Three Gorges Reservoir? The Glen Canyon Dam Adaptive Management Program and the Missouri River Recovery Program [77,78] provide examples of achieving a balance between environmental protection and water resource development by means of adaptive management and stakeholder engagement.
The results of this research indicate that the T. distichum saplings had excellent tolerance of winter flooding, with survival rate of more than 94% after two rounds of annual winter flooding when planted at elevations of 168m and higher above sea level. All the growth indices measured (CW, TH, DBH and FC) decreased as the flood depth increased. The growth of the T. distichum saplings in the LZTGR was also significantly influenced by their status prior to submersion, particularly FC.
The low levels of plant biodiversity in the control plots suggests that ecosystem restoration via natural recolonization is insufficient for rapidly restoring ecosystem health to the LZTGR and that interventions, such as afforestation, are indicated. Although there were no significant differences between most H’ indices of the herb-community plots in the T. distichum afforested experimental area and control area, similarity analysis indicated that herb community complexity was increased by the afforestation, and could therefore be helpful for enhancing the stability of the LZTGR ecosystem. We conclude that the afforestation method described here has potential for improving ecosystem quality, but that further experimental research is required before afforestation with T. distichum can be proposed for wide application in the LZTGR. This experimentation might include assessment of the potential for T. distichum to establish self-sustaining populations (which could be beneficial), or to become invasive beyond the sites for which it has been selected (which could be deleterious). The economic acceptability of T. distichum for local people, notably farmers, should also be examined. Continued monitoring of the biological communities that establish in association with populations of T. distichum, both at this site and others in the LZTGR, should continue so that its ecological functions can be properly understood. Littoral zone afforestation experiments involving the creation of multi-aged multi-species woods that more closely mimic natural populations are clearly indicated, as are experimental afforestations of diverse sites which vary with respect to environmental conditions such as soil, slope, aspect, and hydrological variables.
S1 Table. Water level records of Wanzhou Sation during July 1, 2009 to August 31, 2011.
We appreciate support from the Pengxi River Wetland Natural Reserve Management Bureau and the Pengxi River Wetland Research Station. We thank Kang Liu for his excellent picture of the LWE in Hanfeng Lake (Fig 5). The field assistance of Yazhou Huang is gratefully acknowledged.
Conceived and designed the experiments: CD XY. Performed the experiments: BL. Analyzed the data: BL HX. Contributed reagents/materials/analysis tools: XY. Wrote the paper: BL JHMW.
- 1. Riggsbee JA, Wetzel R, Doyle MW. Physical and plant community controls on nitrogen and phosphorus leaching from impounded riverine wetlands following dam removal. River Res Appl. 2012; 28: 1439–1450.
- 2. Awad J, van Leeuwen J, Abate D, Pichler M, Bestland E, Chittleborough DJ, et al. The effect of vegetation and soil texture on the nature of organics in runoff from a catchment supplying water for domestic consumption. Sci Total Environ. 2015; 529: 72–81. pmid:26005751
- 3. Tabacchi E, Lambs L, Guilloy H, Planty-Tabacchi AM, Muller E, Decamps H. Impacts of riparian vegetation on hydrological processes. Hydrol Process. 2000; 14: 2959–2976.
- 4. FISRWG. Stream Corridor Restoration: Principles, Processes, and Practices. Federal Interagency Stream Restoration Working Group. USDA Natural Resource Conservation Service; 1998 Oct. SuDocs No.: A 57.6/2:EN3/PT.653. GPO Item No.: 0120-A.
- 5. Butt MJ, Waqas A, Mahmood R. The combined effect of vegetation and soil erosion in the water resource management. Water Resour Manag. 2010; 24: 3701–3714.
- 6. Gois KS, Antonio RR, Gomes LC, Pelicice FM, Agostinho AA. The role of submerged trees in structuring fish assemblages in reservoirs: two case studies in South America. Hydrobiologia. 2012; 685: 109–119.
- 7. Godinho C, Rabaca JE, Segurado P. Breeding bird assemblages in riparian galleries of the Guadiana River Basin (Portugal): the effect of spatial structure and habitat variables. Ecol Res. 2010; 25: 283–294.
- 8. de Lima JA, de Santana DG, Nappo ME. Initial behavior of species in revegetation of the gallery forest in the Mandaguari farm in Indianopolis, MG. Revista Arvore. 2009; 33: 685–694.
- 9. Matic V, Dukanovic G. Application of erosion-control materials and spontaneous vegetation in the protection of reservoirs in southern and eastern Serbia. Arch Biol Sci. 2009; 61: 475–482.
- 10. New T, Xie Z. Impacts of large dams on riparian vegetation: applying global experience to the case of China's Three Gorges Dam. Biodivers Conserv. 2008; 17: 3149–3163.
- 11. Willison JHM, Li RX, Yuan XZ. Conservation and ecofriendly utilization of wetlands associated with the Three Gorges Reservoir. Environ Sci Pollut Res Int. 2013; 20: 6907–6916. pmid:23288679
- 12. Yuan XZ, Zhang YW, Liu H, Xiong S, Li B, Deng W. The littoral zone in the Three Gorges Reservoir, China: challenges and opportunities. Environ Sci Pollut Res Int. 2013; 20: 7092–7102. pmid:23296972
- 13. Wang Q, Yuan XZ, Willison JHM, Zhang YW, Liu H. Diversity and above-ground biomass patterns of vascular flora induced by flooding in the drawdown area of China's Three Gorges Reservoir. PLOS ONE. 2014; 9(6): e100889. pmid:24971514
- 14. Wang Q, Liu H, Yuan XZ, Sun R, Wang JX. Pattern and biodiversity of plant community in water-level-fluctuation zone of Peng Xi River after impoundment of Three Gorges Reservoir. Journal of Chongqing Normal University (Natural Science). 2009; 26: 48–54.
- 15. Wang HF, Zeng B, Qiao P, Li Y, Luo FL, Ye XQ. Survival and growth response of Vetiveria zizanioides, Acorus calamus and Alternanthera philoxeroides to long-term submergence. Acta Ecologica Sinica. 2008; 28: 2571–2580.
- 16. Xu SJ, Zeng B, Lei ST, Su XL. Root features of several flooding-tolerant plants and their roles in enhancing anti-erodibility of the soil in Three Gorges Reservoir Region. Acta Pedologica Sinica. 2011; 48: 161–167.
- 17. Li B, Yuan XZ, Xiao HY, Chen ZL. Design of the dike-pond system in the littoral zone of a tributary in the Three Gorges Reservoir, China. Ecol Eng. 2011; 37: 1718–1725.
- 18. Su XL, Bo Z, Qiao P, Ayiqiaoli, Huang WJ. The effects of winter water submergence on flowering phenology and reproductive allocation of Salix variegata Franch. in Three Gorges Reservoir Region. Acta Ecologica Sinica. 2010; 30: 2582–2592.
- 19. Ai LJ, Yu JH, Zhang YL. Influence of flooding on community characterization of the species Salix rosthornii Seemen in water level fluctuation zone of the Three Gorges Reservoir. J Nanjing Forestry Univ (Nat Sci Edition). 2013; 37: 23–28.
- 20. Liu ZB, Cheng RM, Xiao WF, Guo QS, Wang N. Effect of off-season flooding on growth, photosynthesis, carbohydrate partitioning, and nutrient uptake in Distylium chinense. PLOS ONE. 2014; 9(9):e107636. pmid:25222006
- 21. Yang F, Wang Y, Chan ZL. Perspectives on screening winter-flood-tolerant woody species in the riparian protection forests of the Three Gorges Reservoir. PLOS ONE. 2014; 9(9): e108725. pmid:25265326
- 22. Li CX, Wei H, Geng YH, Schneider R. Effects of submergence on photosynthesis and growth of Pterocarya stenoptera (Chinese wingnut) seedlings in the recently-created Three Gorges Reservoir Region of China. Wetl Ecol Manag. 2010; 18: 485–494.
- 23. Liu Y, Willison JHM. Prospects for cultivating white mulberry (Morus alba) in the drawdown zone of the Three Gorges Reservoir, China. Environ Sci Pollut Res Int. 2013; 20: 7142–7151. pmid:23757029
- 24. Wang YH, Luo R. Light-response of an introduced tree species (Nyssa aquatica) under the environment of high temperature in Three Gorges Reservoir Region. Journal of Sichuan Forestry Science and Technology. 2013; 34: 10–13.
- 25. Li CX, Zhong ZC, Geng YH, Schneider R. Comparative studies on physiological and biochemical adaptation of Taxodium distichum and Taxodium ascendens seedlings to different soil water regimes. Plant Soil. 2010; 329: 481–494.
- 26. Keeland BD, Conner WH. Natural regeneration and growth of Taxodium distichum (L.) Rich in Lake Chicot, Louisiana after 44 years of flooding. Wetlands. 1999; 19: 149–155.
- 27. Keeland BD, Young PJ. Long-term growth trends of baldcypress (Taxodium distichum (L.) Rich.) at Caddo Lake, Texas. Wetlands. 1997; 17: 559–566.
- 28. Li CX, Zhong ZC. Influences of mimic soil water change on the contents of malic acid and shikimic acid and root-biomasses of Taxodium distichum seedlings in the hydro-fluctuation belt of the Three Gorges Reservoir Region. Acta Ecologica Sinica. 2007; 27: 4394–4402.
- 29. Li CX, Zhong ZC. Nutrient dynamics in soils of Taxodium distichum seedlings under simulated soil water changes in the hydro-fluctuation belt of the Three Gorges Reservoir Region. Scientia Silvae Sinicae. 2008; 44: 124–129.
- 30. Xian XD, Feng YL, Willison JHM, Ai LJ, Wang P, Wu ZN. Restoring ecosystem services to littoral zones of rivers in the urban core of Chongqing, China. Environ Sci Pollut Res Int. 2015; 22: 12576–12584. pmid:25907628
- 31. Wang XR, Cheng RM, Xiao WF, Guo QS, Feng XH, Wang RX. Relationship between standing vegetation and soil seed bank in water-level-fluctuating zone of Three Gorges Reservoir at the beginning after charging water. Acta Ecologica Sinica. 2010; 30: 5821–5831.
- 32. Jimenez MN, Pinto JR, Ripoll MA, Sanchez-Miranda A, Navarro FB Restoring silvopastures with oak saplings: effects of mulch and diameter class on survival, growth, and annual leaf-nutrient patterns. Agroforest Syst. 2014; 88: 935–946.
- 33. Li Y, Haerdtle W, Bruelheide H, Nadrowski K, Scholten T, von Wehrden H, et al. Site and neighborhood effects on growth of tree saplings in subtropical plantations (China). Forest Ecol Manag. 2014; 327: 118–127.
- 34. Holland EP. Inferring changes in foliar mass and area from foliage cover: a mechanistic model. Austral Ecol. 2013; 38: 121–130.
- 35. van den Wollenberg AL. Redundancy analysis an alternative for canonical correlation analysis. Psychometrika. 1977; 42: 207–219.
- 36. Magurran AE. Ecological diversity and its measurement. New Jersey: Princeton University Press; 1988. p. 179.
- 37. Curtis JT, Mcintosh RP. An upland forest continuum in the prairie-forest border region of Wisconsin. Ecology. 1951; 32: 476–496.
- 38. Wang YS, Shangguan TL. Discussion on calculating method of important values. Journal of Shanxi University (Nat. Sci. Ed.). 2010; 33: 312–316.
- 39. Clarke KR. Nonparametric multivariate analyses of changes in community structure. Aust J Ecol. 1993; 18: 117–143.
- 40. Ter Braak CJF, Smilauer P. Canoco reference manual and canodraw for windows user's guide: software for canonical community ordination (version 4.5). New York: Microcomputer Power; 2002.
- 41. Haynes RJ. The development of bottomland forest restoration in the lower Mississippi River alluvial valley. Ecological Restoration. 2004; 22(3): 170–182.
- 42. Clewell AF. Restoration of riverine forest at hall branch on phosphate-mined land, Florida. Restor Ecol. 1999; 7: 1–14.
- 43. Conner WH, Day JW. Diameter growth of Taxodium-distichum (L.) Rich and Nyssa-aquatica L. from 1979–1985 in 4 Louisiana swamp stands. Am Midl Nat. 1992; 127: 290–299.
- 44. Skeesick DG. Blue river reservoir ecosystem restoration. Warmwater fisheries symposium 1, USDA forest service; 1991 Aug. Report No.: RM-207. pp. 149–154.
- 45. Middleton BA. The flood pulse concept in wetland restoration. In: Middleton BA, editor. Flood pulsing in wetlands: Restoring the natural hydrological balance. New York: John Wiley & Sons, Inc.; 2002. pp. 1–10.
- 46. Myers RS, Shaffer GP, Llewellyn DW. Baldcypress (Taxodium-distichum (L) Rich) restoration in Southeast Louisiana—the relative effects of herbivory, flooding, competition, and macronutrients. Wetlands. 1995; 15: 141–148.
- 47. Xu HG, Qiang S, Genovesi P, Ding H, Wu J, Meng L, et al. An inventory of invasive alien species in China. NeoBiota. 2012; 15: 1–26.
- 48. Hosner JF. The effects of complete inundation upon seedlings of 6 bottomland tree species. Ecology. 1958; 39: 371–373.
- 49. Day RH, Doyle TW, Draugelis-Dale RO. Interactive effects of substrate, hydroperiod, and nutrients on seedling growth of Salix nigra and Taxodium distichum. Environ Exp Bot. 2006; 55: 163–174.
- 50. Hoeppner SS, Shaffer GP, Perkins TE. Through droughts and hurricanes: tree mortality, forest structure, and biomass production in a coastal swamp targeted for restoration in the Mississippi River deltaic plain. Forest Ecol Manag. 2008; 256: 937–948.
- 51. Yamamoto F. Effects of depth of flooding on growth and anatomy of stems and knee roots of Taxodium-distichum. IAWA Bulletin. 1992; 13: 93–104.
- 52. Tang LZ, Huang BL, Kikuo H, Hiroto T. Ecological adaptation mechanisms of roots to flooded soil and respiration characteristics of knee roots of Taxodium Ascendens. Journal of Plant Ecology (Chinese Version). 2008; 32: 1258–1267.
- 53. Xiao Y, Jie Z, Wang M, Lin G, Wang W. Leaf and stem anatomical responses to periodical waterlogging in simulated tidal floods in mangrove avicennia marina seedlings. Aquat Bot. 2009; 91: 231–237.
- 54. Wang GB, Cao FL. Formation and function of aerenchyma in baldcypress (Taxodium distichum (L.) Rich.) and Chinese tallow tree (Sapium sebiferum (L.) Roxb.) under flooding. S Afr J Bot. 2012; 81: 71–78.
- 55. Sun R, Yuan XZ, Liu H, Chen ZL, Zhang YW. Floristic composition and species diversity of plant communities along an environment gradient in drawdown area of Three Gorges Reservoir after its initial impounding to the water level of 156 m. Chinese Journal of Ecology. 2011; 30: 208–214.
- 56. Chen FQ, Zhang JX, Zhang M, Wang JZ. Effect of Cynodon dactylon community on the conservation and reinforcement of riparian shallow soil in the Three Gorges Reservoir Area. Ecological Processes. 2015; 3: 10–1186.
- 57. Ernst KA, Brooks JR. Prolonged flooding decreased stem density, tree size and shifted composition towards clonal species in a central Florida hardwood swamp. Forest Ecol Manag. 2003; 173: 261–279.
- 58. Middleton BA. Hydrochory, seed banks, and regeneration dynamics along the landscape boundaries of a forested wetland. Plant Ecol. 2000; 146: 169–184.
- 59. Middleton BA. Soil seed banks and the potential restoration of forested wetlands after farming. J Appl Ecol. 2003; 40: 1025–1034.
- 60. Senzota R, Mbago F. Impact of habitat disturbance in the wetland forests of East Usambara, Tanzania. Afr J Ecol. 2010; 48: 321–328.
- 61. Faulkner SP, Bhattarai P, Allen Y, Barras J, Constant G. Identifying baldcypress-water tupelo regeneration classes in forested wetlands of the atchafalaya basin, louisiana. Wetlands. 2009; 29: 809–817.
- 62. Watson J, Whittaker RJ, Dawson TP. The importance of littoral forest remnants for indigenous bird conservation in southeastern Madagascar. Biodivers Conserv. 2005; 14: 523–545.
- 63. Bernal B, Mitsch WJ. Comparing carbon sequestration in temperate freshwater wetland communities. Global Change Biol. 2012; 18: 1636–1647.
- 64. Mckee SE, Seiler JR, Aust WM, Strahm BD, Schilling EB, Brooks S. Carbon pools and fluxes in a tupelo (Nyssa aquatica)-baldcypress (Taxodium distichum) swamp 24-years after harvest disturbances. Biomass Bioenerg. 2013; 55: 130–140.
- 65. Giese LAB, Aust WM, Kolka RK, Trettin CC. Biomass and carbon pools of disturbed riparian forests. Forest Ecol Manag. 2003; 180: 493–508.
- 66. Newmaster SG, Bell FW, Roosenboom CR, Cole HA, Towill WD. Restoration of floral diversity through plantations on abandoned agricultural land. Can J Forest Res. 2006; 36: 1218–1235.
- 67. Tan SD, Zhu MY, Dang HS, Wang Y, Zhang QF. Physiological responses of bermuda grass (Cynodon dactylon(L.) Pers.) to deep submergence stress in the Three Gorges Reservoir Area. Acta Ecologica Sinica. 2009; 29: 3685–3691.
- 68. Lei ST, Zeng B, Yuan Z, Su XL. Changes in carbohydrate content and membrane stability of two ecotypes of Calamagrostis arundinacea growing at different elevations in the drawdown zone of the Three Gorges Reservoir. PLOS ONE. 2014; 9(3): e91394. pmid:24608821
- 69. Yin YL, Yu CG, Hua JF, Huan JJ, Han LW, Qi BY, et al. A trial on the silviculture of Taxodium hybrid ‘zhonshanshan118’ planted in the hydro-fluctuation belt of the Three Gorges Reservoir within the Wanzhou District Area of Chongqing City. China Forestry Science and Technology. 2014; 28: 110–114.
- 70. Mitsch WJ, Bernal B, Nahlik AM, Mander U, Zhang L, Anderson CJ, et al. Wetlands, carbon, and climate change. Landscape Ecol. 2013; 28: 583–597.
- 71. Bridgham SD, Megonigal JP, Keller JK, Bliss NB, Trettin C. The carbon balance of North American wetlands. Wetlands. 2006; 26: 889–916.
- 72. Mitsch WJ, Gosselink JG. Wetlands. New York: John Wiley & Sons; 2007. pp. 201–206.
- 73. Chow AT, Dai J, Conner WH, Hitchcock DR, Wang J. Dissolved organic matter and nutrient dynamics of a coastal freshwater forested wetland in Winyah Bay, South Carolina. Biogeochemistry. 2013; 112: 571–587.
- 74. Goni MA, Teixeira MJ, Perkey DW. Sources and distribution of organic matter in a river-dominated estuary (Winyah Bay, SC, USA). Estuar Coast Shelf S. 2003; 57: 1023–1048.
- 75. Waiser MJ, Robarts RD. Photodegradation of DOC in a shallow prairie wetland: evidence from seasonal changes in DOC optical properties and chemical characteristics. Biogeochemistry. 2004; 69: 263–284.
- 76. Schiff SL, Aravena R, Trumbore SE, Dillon PJ. Dissolved organic-carbon cycling in forested watersheds—a carbon isotope approach. Water Resour Res. 1990; 26: 2949–2957.
- 77. Berkley J. Opportunities for collaborative adaptive management progress: integrating stakeholder assessments into progress measurement. Ecol Soc. 2013;18(4):69.
- 78. Susskind L, Camacho AE, Schenk T. A critical assessment of collaborative adaptive management in practice. J Appl Ecol. 2012; 49: 47–51.