Branches and Leaves Growth Patterns in Myricaria laxiflora Remnant Populations Are Affected by Human-altered Water Level Fluctuations


 Background: The construction of the Three Gorges-Gezhou Dam altered the water level fluctuation pattern in the downstream habitat of the endangered species Myricaria laxiflora. This study investigated how branch and leaf growth traits of M. laxiflora remnant populations changed to adapt the environmental stress caused by human-altered water level fluctuations.Results:Due to such disturbance, branch and leaf growth traits of M. laxiflora populations exhibited significant differences across water level conditions. The number of secondary branches, plant height, and leaf number of the plants in the upper area of the habitat were significantly higher than that in the middle and lower river bank areas. The longest secondary branch length of the plants in the upper and middle region was significantly higher than that in the lower region. The branch and leaf volume of plants in the middle region was significantly higher than that in the upper and lower region. The maximum water content of plants in the middle and lower region was significantly higher than that in the upper region. Principal component analysis showed that the branch and leaf traits of plants changed with decreasing water level toward to decreasing plant height, leaf number and the number of secondary branches, and increasing maximum water content of branch and leaves. Conclusions: The phenotypic plasticity of M. laxiflora plants in branch and leaf growth traits alleviates the impacts of human-regulated water level fluctuations. However, the above ground growth of M. laxiflora plants distributed at the middle and lower areas of the fluctuation zone is still negatively affected.


Background
Plant growth traits are core morphological and physiological attributes that re ect the performance of plants in resource acquisition, processing and conservation, the relationship among plants, and the interaction between plant and environment (Roderick et al., 1999;Violle et al., 2007;Reich et al., 2014). Phenotypic plasticity is the ability of plant genotypes to make corresponding changes and produce different phenotypes to cope with the spatiotemporal heterogeneity (Bradshaw, 2006). Branch and leaf growth traits include leaf number, leaf area, branch length, stem height, branch dry mass, leaf dry mass, nitrogen content, stoma density, etc. (Zhou et al., 2018). Leaves and young branches are, in general, highly sensitive organs to external environmental conditions. The branch and leaf growth traits of plants in heterogeneous ecological environment often change to improve their resource acquisition and utilization e ciency. This adaptive change is bene cial for plants to maintain their normal physiological activities, growth and development, and improve the survival and competitiveness of species (Bernard-Verdier et al, 2012;Collins et al, 2016). The phenotypic plasticity of branch and leaf traits is commonly used as key parameters to evaluate not only the effect of environmental change on plant growth, but also the plant adaptability to environmental changes (Gleason et al., 2017;Liu et al., 2019).
Water level may vary spatially and temporally resulting in differential uctuations. Water-level uctuations are important ecological processes that in uence the ecological environment of riparian plant habitat (Spencer et al., 2004;Miao et al., 2017). Riparian plants have adapted to the submergenceemergence pattern brought by natural/normal water-level uctuations and the accompanying changes in soil water content (Blom et al., 1994;Li et al., 2015;Campbell et al., 2016). Under the expected water level uctuation pattern, although the leaf growth traits of riparian plants vary in space and time to a certain extent (Li et al., 2013;Chen and Xie, 2009), plant growth will not be severely affected as the change of branch and leaf characters is within the phenotypic plasticity (Campbell et al., 2016;Gao et al., 2017).
However, plant branch and leaf growth traits and tness might be signi cantly affected when the riparian environment is altered by the construction of a dam (Li et al., 2013;Dalke et al., 2018). They might increase stem height and single leaf area, and reduce leaf number to adapt the stress of increasing ooding time and depth caused by a dam (Chen et al., 2010;Smaouiet al., 2011). If the atypical uctuations prolong and become severe, the growth of riparian plants is impaired leading to population dieback (Van der Sman and Van Tonger, 1988;Bijarchi et al., 2011).
Myricaria laxi ora (Franch.) P.Y. Zhang et Y.J. Zhang is a shrub species of the Tamaricaceae family. M. laxi ora measures 1-1.5 m in height. Young branches are very small and green. Racemes are terminal and measure 6-12 cm in length. Capsules are narrowly conical and measure 6-8 mm in length. Seeds measure 1-1.5 mm in length and are dispersed by the wind and river water ow (Chen and Xie, 2007;Guan et al., 2020a). M. laxi ora is mainly distributed in the riparian zone between Yichang at the Yangtze River and Yibin in the Jinsha River. The Three Gorges Reservoir Region is the core distribution region of this plant (Chen et al., 2005;Chen and Wang, 2015). The construction of the Three Gorges dam has signi cantly increased the water level of the reservoir and submerged the habitat of M. laxi ora, leaving only a few river islands in the downstream of the Three Gorges-Gezhouba Dam and Xiluodu-Xiangjia Dam cascade hydropower stations with M. laxi ora remnant populations (Chen and Xie, 2009;Chen and Wang, 2015). Before the construction of the Three Gorges-Gezhou Dam cascade hydropower station, M. laxi ora was usually completely submerged by ooding from June-August every year. When the water level gradually decreased in early September, M. laxi ora plants emerged and quickly restored their vegetative growth, and then began to ower and set fruits within a week (Chen and Xie, 2009). After the construction of the Three Gorges-Gezhou Dam cascade hydropower station, because human regulation on river discharge is conducted to ful ll the needs of ood prevention and power generation, the water level uctuation pattern in the habitat of M. laxi ora remnant populations was signi cantly altered: the emergence time of habitats is delayed, average periods of emergence was shortened from 220 days to 190 days; the ooding retreat rapidly during the growth restart period; also, and the water level in the winter dry season dramatically decreased Duan et al., 2016;Sun et al., 2007). We investigated the branch and leaf growth traits in populations across a water level condition during the growth restart period after emergence to understand the pattern of branch and leaf growth traits changing with water level gradient and the phenotypic plasticity. Based on the effect of the Three Gorges Dam construction on the pattern of water level uctuations in the habitat of M. laxi ora remnant populations and the aquatic ecological environment, we discussed the subsequent effect on the above ground growth of plants.

Differences in branch and leaf growth traits among various water level conditions
The branch and leaf traits of M. laxi ora exhibited signi cant differences across the water level conditions (P < 0.05) (Fig. 2). Plant height, the number of secondary branches, the longest secondary branch length, and leaf number all increased from the lower region to the upper region. There was a signi cant increase of 47.1%, 41.6%, and 15.9% in plant height, the number of secondary branches, and leaf number of plants distributed at the upper region compared to the middle region (P < 0.05), and a signi cant increase of 60.6%, 59.3% and 33.7% compared to the lower region (P < 0.05), respectively. The longest secondary branch length of plants distributed at the upper region had not signi cant difference with middle region, but both of them were signi cantly higher than that of plants distributed at the lower region; branch and leaf volume, and the maximum branch and leaf water content initially increased and then decreased from the lower region to the upper region. Branch and leaf volume of plants distributed at middle region signi cantly increased by 25.9% and 29.9%, respectively, compared to the upper and lower regions (P < 0.05). The maximum branch and leaf water content of plants did not differ signi cantly between middle region and upper region (P > 0.05), but both of them were signi cantly higher than lower region (P < 0.05).
Changes in branch and leaf trait syndrome along the water level condition PCA showed that plant height, leaf number, and the number of secondary branches had a higher contribution to PC 1, whereas the longest secondary branch length and branch and leaf volume had a higher contribution to PC 2 (Table 1). PC1 re ected a gradient of branch and leaf traits, from low to high values of plant height, leaf number, secondary branch number, longest secondary branch length and soil water content and high to low values of the maximum branch leaves water content. PC2 re ected a gradient of branch and leaf traits, from low to high values of branch and leaf volume, longest secondary branch length and leaf number and high to low values of secondary branch number, plant height and the maximum branch leaves water content. PCA ordination showed that the branch and leaf functional symptoms of plants differentiated signi cantly among different water level conditions (Fig. 3). Higher values of the maximum branch leaves water content appeared in the lower region compared to the upper region, which showed higher values of leaf number, plant height, and number of secondary branches. Therefore, as the water level and soil water content decreased, M. laxi ora shifted their growth investment towards a decrease in plant height, number of leaf and secondary branches, and an increase in the maximum branch leaves water content.

Discussion
Responses in branch and leaf functional traits to the water level gradient Water level uctuations are important ecological processes that in uence the ecosystem in water leveluctuating regions (Matthew et al., 2019). Under the in uence of water level uctuations, the ecological environment at the water level-uctuating region exhibits clear spatial heterogeneity, where environmental factors including emergence period, emergence time and soil water content are signi cantly different across the water level conditions . Caused by the environmental heterogeneity, branch and leaf growth traits of riparian plants also exhibit spatial variations across the water level conditions (Spencer et al., 2004;McCoy-Sulentic et al., 2017;Duan et al., 2017). Lawson et al. (2015) investigated the relationship between growth traits in riparian plant communities and water level uctuations. Their results showed that plant functional trait diversity was highly and positively correlated with metrics describing both ooding disturbance and patterns of water availability.
Differences in branch and leaf growth traits due to environmental heterogeneity is phenotypic plasticity by which plants respond to and adapt to environmental changes (Munguia-Rosas et al., 2019;Lajoie and Vellend, 2018 ). Miles et al. (2017) studied the branch and leaf growth traits in riparian plants of the Colorado River and found that as the inundation gradients shifted from being frequently to less frequently inundated, high speci c leaf area (SSA) and low stem speci c gravity (SSG) shifted to low SSA and high SSG. In this study, the branch and leaf growth traits of M. laxi ora remnant populations exhibited spatial variations across the water level gradient. In the upper water level-uctuating region that had an early emergence time and longer emergence period, it is easier for plants to produce su cient numbers of branches and leaves. Plants in this region had the highest values in plant height, number of secondary branches, the longest secondary branch length and leaf number, but the lowest value in the maximum water content. On the contrary, in the middle and lower water level-uctuating regions, where the emergence time was delayed, with a short emergence period and fast decline in water level, it is di cult for plants to produce leaves and branches. Plants in this region had the highest values in branch and leaf volume, and the maximum branch and leaf water content, but the lowest value in plant height, number of secondary branches, the longest secondary branch length and leaf number.
The effect of changes in water level uctuation patterns on branch and leaf functional traits Plants can adjust morphological and physiological traits to maximize the consistency of their phenotypes with environment and reduce the negative effects of environmental changes on plant growth and reproduction (Sultan, 2005;Ellers et al., 2010). This phenotypic plasticity enhances plant tolerance and facilitates plants to establish more diverse habitats (Sultan, 1995). The change of water level uctuations might affect the plant growth, propagation and seedling regeneration of the M. laxi ora remnant populations. An investigation on the M. laxi ora remnant populations by Bao et al. (2010) revealed that the remnant population structure had an extremely low ratio of young seedlings, indicating a population degradation caused by the hindered population regeneration. In our recent study on the growth and reproduction characteristics of M. laxi ora remnant populations, sexual reproduction, photosynthetic activity and seedling establishment were signi cantly reduced due to the changes in the water level uctuation pattern (Chen et al., 2019;Guan et al., 2020a;Guan et al., 2020b). This study further elucidated the effects of regulated water-level uctuations on the formation and changes in M. laxi ora branch and leaf growth traits. The change of branch and leaf growth traits facilitate the remnant M. laxi ora populations to cope with environmental stress. However, generally the changes of branch and leaf growth traits are clearly unfavorable to the growth of plants distributed at the middle and lower regions of the uctuation zone. Because M. laxi ora have extremely small leaves, their young branches (green-colored) are also important photosynthetic organs. A decrease in the number of branch and leaves will affect the photosynthetic capability and the accumulation of organic matters of plants (Guan et al., 2020b), which further in uences the plant growth and sexual reproduction after ood (Chen et al., 2019: Guan et al., 2020a. This could explain why M. laxi ora populations continue to degenerate under the regulated water level uctuation pattern (Bao et al., 2010).

Conclusions
The construction of the Three Gorges-Gezhou Dam alters the water level uctuation pattern in the downstream habitat of the endangered species Myricaria laxi ora, which causes the emergence time of habitats delayed and the ooding retreat rapidly. We investigated the changes of M. laxi ora remnant populations in branch and leaf growth traits across water level condition during the growth restart period. The effect of environmental change on plant growth and plant adaptability to environmental changes were evaluated and analyzed. Due to such disturbance, branch and leaf growth traits of M. laxi ora populations exhibited signi cant differences across water level conditions. The branch and leaf traits of plants changed with decreasing water level toward to decreasing plant height, leaf number and the number of secondary branches, and increasing maximum water content of branch and leaves. The phenotypic plasticity of M. laxi ora plants in branch and leaf growth traits alleviates the impacts of human-regulated water level uctuations. However, the above ground growth of M. laxi ora plants distributed at the middle and lower areas of the uctuation zone is still negatively affected. To protect this endangered species, the current ow regulation by the large-scale hydropower station could be adjusted to move the reservoir retention time earlier. The habitat exposure time would be moved earlier, and the rate of water level decline would be reduced, to satisfy the needs for plant growth and development.

Study site
The eld experimental site was the Yanzhiba (111°19'26''E, 30°38'57''N), an island located downstream of the Three Gorges-Gezhou Dam cascade hydropower station on the Yichang section along the Yangtze River mainstream. Approximately 10,000 M. laxi ora plants are distributed at this habitat, which is a relatively large remnant population according to available reports (Chen and Wang, 2015). This distribution site has humid subtropical climate with an annual average temperature of 16.9 °C, extreme maximum temperature 41.4 °C, extreme minimum temperature of -9.0 °C, annual average precipitation of 1164.1 mm. The site has sandy soil. Common vegetation on the river beach is shrubs dominated by M. laxi ora and Salix variegata communities.
Due to the typical seasonal water-level uctuations in the Yangtze River, M. laxi ora remnant populations are submerged by ooding between June-August every year. From September to May of the following year when water level is relatively low, M. laxi ora emerges and grows. Based on the characteristics in the submergence and emergence at different altitudes of the habitat, we divided the research area into three water level conditions: upper region (altitude (above the sea level) ≥ 45.1 m), middle region (altitude 45.1 m-42.8 m), and lower region (altitude ≤ 42.8 m) (Chen et al., 2019). A sample transect was established at each water-level condition. In each sample transect, a 5 m × 5 m quadrat was set up every 10 m, with a total of 10 quadrats for measuring changes in branch and leaf traits of M. laxi ora in each elevational belt during growth recovery.

Measurements of branch and leaf traits
On Days 10, 20, 30, 40, and 50 after plant emergence, one M. laxi ora plant with uniform size was selected from each quadrat for the investigation of branch and leaf growth traits. We measured the following six parameters: plant height, number of secondary branches, the longest secondary branch length, leaf number, branch and leaf volume, and branch and leaf water content (Xiang et al., 2019;Li et al., 2018 ).
Plant height was measured using a measuring tape, from the green leaf on top of the plant to the base.
The number of secondary branches, leaf number, and the longest secondary branch length were obtained by randomly selecting a primary branch from each sample plant, counting the number of secondary branches grown from the rst branch, and measuring the length of the longest secondary branch. Then, three secondary branches were randomly selected from each sample branch, from which leaf number was determined. In each water-level uctuating region at each time point since growth restart, 30 secondary branches were measured. The M. laxi ora secondary branches used in the measurements were harvested, placed on ice, and transferred to the laboratory for the measurements of branch and leaf volume, and branch and leaf water content. The branches harvested were placed in water and kept in dark at 5 °C for 12 h. Then branches were taken out of the water and blotted using lter paper. The fresh weight of the branches was measured on an electronic balance (0.0001 accuracy). Then, the branches were placed in a burette (accuracy 0.01 mL) to determine branch and leaf volume. Finally, the branches were placed at 105 °C for 15 min for enzyme deactivation, and dried at 60 °C to a constant weight. Branch and leaf dry weight was recorded and used to calculate the maximum water content of branch and leaf using the following equation: Maximum branch and leaf water content = (Branch and leaf fresh weight -Branch and leaf dry weight) / Branch and leaf fresh weight × 100% (Pérez-Harguindeguy et al., 2013).

Measurements of the aquatic ecological environment
From the exposure of the habitat, water level information at the test site was monitored and recorded every day (using the water level of Yangtze River at Yichang section as a reference, published at http://www.cjh.com.cn). Based on our on-site investigation, combining the altitude changes of the Yangtze River water level, we analyzed the water level changes at each sample transect and the emergence status (Fig. 1). The upper water level-uctuating region was emerged on September 8, 2018; the middle and lower regions emerged on September 16, 2018 and November 3, 2018, respectively. During this period, from September 16 to October 14, 2018, due to the rising of water level in the Yangtze River, plants in the middle region were submerged twice. In comparison with that before the dam was built, the emergence time of the habitat under human regulation was postponed for nearly two months.
Concurrently with branch collections, soil water content was also measured. Three locations were randomly selected in each quadrat of each sample transect; soil samples were collected using a cutting ring (15-cm depth) and transferred to the laboratory. The fresh weight and dry weight of soil were weighed to calculate soil water content. Thirty soil samples were collected at each sample transect.

Statistical analysis
After In-transformation of data and means were compared with LSD test, one-way ANOVA and multiple comparisons (post hoc Tukey) were performed by setting water level condition as independent variable, and branch and leaf growth traits as dependent variables. The signi cance of differences among treatments for each parameter and the pattern of how branch and leaf growth traits varied across water level conditions were examined and analyzed respectively. Pearson correlation analysis was conducted to analyze and test the correlation between M. laxi ora branch and leaf growth traits and soil water content. Principal component analysis (PCA) of branch and leaf growth traits was conducted to assess the effects of the regulated water level uctuations on the branch and leaf functional symptoms, and reveal the response of M. laxi ora remnant populations in growth to the regulated water level uctuations. All data analyses were performed using SPSS 19.0. Water level uctuations and changes in soil water content across water level conditions within the sample plot at the Yichang section of the Yangtze River