Phenotypic plasticity of four Chenopodiaceae species with contrasting saline–sodic tolerance in response to increased salinity–sodicity

Abstract It is unknown whether phenotypic plasticity in fitness‐related traits is associated with salinity–sodicity tolerance. This study compared growth and allocation phenotypic plasticity in two species with low salinity–sodicity tolerance (Chenopodium acuminatum and C. stenophyllum) and two species with high salinity–sodicity tolerance (Suaeda glauca and S. salsa) in a pot experiment in the Songnen grassland, China. While the species with low tolerance had higher growth and allocation plasticity than the highly tolerant species, the highly tolerant species only adjusted their growth traits and maintained higher fitness (e.g., plant height and total biomass) in response to increased soil salinity–sodicity, with low biomass allocation plasticity. Most plasticity is “apparent” plasticity (ontogenetic change), and only a few traits, for example, plant height:stem diameter ratio and root:shoot biomass ratio, represent “real” plasticity (real change in response to the environment). Our results show that phenotypic plasticity was negatively correlated with saline–sodic tolerance and could be used as an index of species sensitivity to soil salinity–sodicity.

It is worth noting that congeneric species are ideal materials for comparing plant plasticity, but it is difficult to find congeneric species with contrasting tolerance to abiotic stress. Therefore, dominant species are often used as an alternative (Chmura et al., 2017;Valladares et al., 2002).
When responding to biotic or abiotic environmental variation, plants often change in terms of their phenotypic and physiological activity, ultimately adjusting their growth and biomass allocation.
Biomass distribution and the adjustment of growth characteristics are key factors determining the growth rate and performance of species under different environmental conditions (van Hees, 1997;Jarčuška, 2009). However, evidence suggests that growth traits are more sensitive to the environment than biomass allocation (Curt, Coll, Prevosto, Balandier, & Kunstler, 2005), which is mostly ontogenetic and varies little with environmental availability (Reich, Tjoelker, Walters, Vanderklein, & Bushena, 1998). Plasticity caused by ontogenetic shifts under varying environmental conditions is referred to as "apparent" plasticity, while plasticity that involves variation in environmental conditions that affects the rate of plant development is referred to as "real" plasticity (Weiner, 2004). The understanding of "apparent" and "real" plasticity would help in the investigation of the difference between growth and biomass allocation plasticity.
Over half a billion hectares of the land surface area around the world has been salinized (Chartres, 1993;Gupta & Abrol, 1990;Rengasamy & Olsson, 1991;Wang et al., 2010). This soil salinitysodicity restricts the distribution of plant species, and no species can establish under the most extreme conditions of salinitysodicity. Due to long-term overgrazing and the reclamation of natural grassland, secondary bare saline-sodic patches are very common on the Songnen Plain, China, resulting in low community diversity, with many plant communities consisting of single species, for example, Suaeda glauca, S. salsa, or Chloris virgata (Jiang, He, Wu, & Zhou, 2010;Zheng & Li, 1999). So far, most studies comparing phenotypic plasticity between tolerant and intolerant species have focused on shading, flooding, and drought stress (Couso & Fernández, 2012;Lazarus, Richards, Gordon, Oki, & Barnes, 2011;Valladares et al., 2005), and there have been very limited efforts addressing saline-sodic stress. To our knowledge, only one study has compared the phenotypic plasticity of early and late successional congeneric species in response to saline-sodic soil (Huang et al., 2015); therefore, plant plasticity in response to saline-sodic stress is still largely unknown.
In this study, we aimed to evaluate plasticity in growth traits and biomass allocation between species with low and high saline-sodic tolerance over a salinity-sodicity gradient in the Songnen grassland of northern China. This study includes four plant species from the family Chenopodiaceae: two species with low saline-sodic tolerance (Chenopodium acuminatum Willd. and Chenopodium stenophyllum Koidz.) and the other two with high saline-sodic tolerance (Suaeda glauca (Bunge) Bunge. and Suaeda salsa (L.) Pall.), as found previously (Guan, Lin, Zhou, & Yu, 2013;Ma, Lv, Li, & Liang, 2014;Yang, Shi, & Wang, 2008). The following questions were addressed: (a) Does plasticity differ between high and low saline-sodic tolerant species? (b) Are the growth traits more sensitive than biomass allocation in response to salinity-sodicity? and (c) Is the plasticity largely ontogenetic?

| Plant species and growth conditions
Four plant species, including two saline-sodic sensitive species (Chenopodium acuminatum and C. stenophyllum) and two salinesodic tolerant species (Suaeda glauca and S. salsa) from the same family, were included in this study. All four species are annuals found in the Songnen grassland, with C. acuminatum and C. stenophyllum often found in soils with low saline-sodic levels, while S. glauca and S. salsa dominate soils with high saline-sodic levels (Guan et al., 2013;Li & Yang, 2004). Seeds of all four species were collected from wild plants growing in the Songnen grassland in 2011.
The field trial was conducted in 2012 at the Changling Ecological Research Station for Grassland Farming (ERSGF), Chinese Academy of Sciences (44°33′ N, 123°31′ E, 145 m a.s.l.), located in the southern Songnen grassland, China. This region has warm, humid summers and cold, dry winters, with a mean temperature of 23°C in July and −20°C in January. The mean annual rainfall is 410 mm, with 80% occurring between July and September.
Plastic pots (25 cm in diameter × 25 cm in depth) were buried in the field, with the top edge of the pots approximately 1 cm above the ground. Thirty-to-forty seeds per pot were sown on 4 May 2012, and thinned to one seedling per pot at the two-leaf stage 2 weeks after sowing. To eliminate positional effects and avoid potential contamination from leachate, each treatment combination as a group was rotated clockwise every 2 weeks; meanwhile, the pots within each treatment were also rotated randomly. All pots were well watered throughout the experiment.
The experiment was a two-factorial (four species, five salinitysodicity levels) completely randomized design. There were five salinity-sodicity treatments, including non-, low, moderate, high, and extreme saline-sodic levels, as shown in Table 1, with 30 replicates, giving a total of 600 pots. Soil was taken from the top 30 cm of five different patches at the natural grassland of ERSGF, with a radius of 7-8 m for each patch. To minimize the large variation in soil salinity-sodicity, all five patches were located within 1 km of one another. Soils taken from each patch were air-dried, passed through a 4-mm sieve, and then mixed thoroughly. Information on the soil nitrogen, pH, electric conductivity, and exchangeable sodium percentage is shown in Table 1.

| Plant measurements
Twenty intact plants from each treatment per species were destructively harvested on 20 August 2012. The roots were washed carefully, and the plants were separated into roots, stems, leaves, and reproductive organs. All samples were dried in an oven at 80°C for 48 hr, and the dry weights were recorded.
Based on previous studies, fitness-related traits (including growth and allocation traits) were analyzed to evaluate the phenotypic plasticity of the species showing contrasting tolerance to salinity-sodicity (Javid, Ford, & Nicolas, 2012;Kuehny & Morales, 1998;Portsmuth & Niinemets, 2007;Shi & Sheng, 2005). Growth traits, including plant height (absolute height), total biomass, stem diameter (basal stem diameter), root length (taproot length), and root diameter (taproot diameter), were measured. The ratios of plant height:stem diameter, plant height:root length, and root length:root diameter were calculated. The dry biomass of the root, leaf and stem, total shoot (including leaves, stems, and reproductive tissues), nonstem (including roots, leaves, and reproductive tissues), and nonleaf (including roots, stems, and reproductive organs) components were measured. Allocation traits, including root mass ratio (root/total biomass), stem mass ratio (stem/total biomass), leaf mass ratio (leaf/total biomass), shoot:root mass ratio (shoot/root biomass), and root:leaf mass ratio (root/leaf biomass), were calculated.
To evaluate the phenotypic plasticity of growth and biomass allocation traits, the phenotypic plasticity index was calculated using the following equation (Valladares, Sanchez-Gomez, & Zavala, 2006;Valladares et al., 2000): where PI is the phenotypic plasticity index and T is the trait mean values in each soil treatment.

| Statistical analysis
To overcome the assumptions of normality and homoskedasticity, permutation multivariate analysis of variance (PerMANOVA, vegan package in R software, number of permuted data sets = 999) and permutation analysis of variance (PerANOVA, lmPerm package in R software) were performed to test the main effects of species and saline-sodic treatment and their interaction on plant traits (Anderson, 2001). Species and saline-sodic treatment were treated as fixed factors. Differences among the mean values of plant traits and differences in plasticity among the species were determined using the least significant difference (LSD) tests. A t-test at p = 0.05 was performed to examine the difference between growth and allocation plasticity.
Allometric analysis was conducted for all calculated traits (ratio of two traits). Nevertheless, the allometric analysis for both root mass ratio and shoot:root ratio is the comparison between the shoot mass and root mass. The allometric relationship was described by log y = b + a × log x, where a is the scaling exponent (slope) and b is the allometric coefficient or "scaling factor" (yintercept). Differences in shifts of the slope and in the elevation of slopes (y-intercept) among the different soil treatments were assessed using standardized major axis regression (SMA, also known as reduced major axis, RMA, SMATR package in R software) Warton, Wright, Falster, & Westoby, 2006). SMA analyses are appropriate for summarizing the relationship between two variables in terms of a single slope (Wright, Reich, Cornelissen, Falster, Groom, et al., 2005).
In SMATR, the heterogeneity between SMA slopes is tested via a permutation test. Differences in SMA slope, elevation (intercept), and plant size (i.e., a shift along the common slope) were estimated (Wright, Reich, Cornelissen, Falster, Garnier, et al., 2005).
In the extreme saline-sodic soil, the two Chenopodium species did not germinate, while the two Suaeda species survived; thus, the data from the extreme saline-sodic soil treatment were not analyzed, and the data are shown in the Supporting Information (Table S1). The different letters are not statistically significant in columns. * The abandoned cropland. † In the ecotone between dunes and grassland.

| Plant trait response
Plant height, total biomass, stem diameter, root length, and root diameter significantly decreased with an increasing level of soil salinity-sodicity (Table 2 and Supporting Information Table S1). A significant interaction between species and soil salinity-sodicity was found for plant height (p = 0.005), stem diameter (p = 0.012), root length (p < 0.001), and root diameter (p < 0.001). The species with low tolerance (C. acuminatum and C. stenophyllum) had a higher reduction in their response to soil salinity-sodicity than the highly tolerant species (S. glauca and S. salsa) ( A significant interaction between species and soil salinitysodicity was found for the root mass ratio (p < 0.001), stem mass ratio (p = 0.002), shoot:root mass ratio (p < 0.001), and root:leaf mass ratio (p = 0.002). For example, with increasing soil salinity-sodicity, low-tolerance species showed a higher increase in shoot:root mass ratio and a higher decrease in root allocation and root:leaf mass ratio than those of high-tolerance species. The leaf mass ratio differed significantly among species (p < 0.001). No effect of salinity-sodicity or two-way interaction was found for the leaf mass ratio (Tables 2   and Supporting Information Table S1).

| Plasticity of growth and allocation traits
The species with low tolerance (C. acuminatum and C. stenophyllum) exhibited higher growth, allocation, and total plasticity compared with the highly tolerant species (S. glauca and S. salsa) (Figure 1). No difference was found in growth, allocation, or total plasticity between the species with similar tolerance. For S. glauca, the growth plasticity was significantly higher than the allocation plasticity (p = 0.005), while no such difference was found for other three species (p > 0.05) (Figure 1).

| Allometry of growth and allocation traits
Significant positive relationships were found between all paired traits in all species except the relationship between plant height and root length in S. salsa and between root length and root diameter in S. glauca (  Figure 2), and the differences among the different saline-sodic treatments were significant for all three species (all p < 0.05 for slope homogeneity), but not for S. salsa (p > 0.05 for both slope homogeneity and shift in elevation). For C. acuminatum and C. stenophyllum, the allometric scaling slope of root biomass vs. shoot biomass decreased with increasing soil salinitysodicity (Table 3; Figure 3), and the differences among the different saline-sodic soil treatments were significant (p < 0.001 and p = 0.002, respectively), while the differences were not significant for S. glauca and S. salsa.

| Growth difference among four species in response to soil salinity-sodicity
Our study found that soil salinity-sodicity greatly inhibited plant growth, such as plant height, biomass, and stem diameter, which was consistent with the findings of others (Hooks et al., 2018;Javid et al., 2012;Kuehny & Morales, 1998;Shi & Sheng, 2005). When the soil salinity-sodicity increased from the non-to high level, the two species with low tolerance rapidly decreased all their growth traits and root mass ratio, possibly due to their higher threshold levels that induce osmotic effects and ionic imbalance in plant tissues (Ahmad, Ghafoor, Akhtar, & Khan, 2013;Maas, 1987), while the two highly tolerant species had the ability to maintain higher F I G U R E 1 Phenotypic plasticity in four species. The four species include C.a: Chenopodium acuminatum, C.s: Chenopodium stenophyllum, S.g: Suaeda glauca, and S.s: Suaeda salsa. Growth traits include the following: 1: plant height, 2: total biomass, 3: stem diameter, 4: root length, 5: root diameter, 6: plant height:stem diameter ratio, 7: plant height:root length ratio, and 8: root length:diameter ratio. Structural traits include the following: 9: root mass ratio, 10: leaf mass ratio, 11: stem mass ratio, 12: shoot:root mass ratio, and 13: root:leaf mass ratio. For each compartment, bars sharing the same letters are not significantly different at p = 0.05 for the four species. p Values were determined using a t-test for mean differences between growth and allocation plasticity TA B L E 3 Results of standardized major axis regression (SMA) analysis of pairwise combinations of traits for each species in different saline-sodic soils Y and X

| Plasticity in species with low and high salinesodic tolerance
Plants have a remarkable capacity to adjust their morphological and physiological traits in response to abiotic conditions through acclimation or, more broadly, phenotypic plasticity (Bhattarai et al., 2017;Delagrange, Messier, Lechowicz, & Dizengremel, 2004;Sultan, 2000;Valladares & Niinemets, 2008). For the growth and allocation traits explored in this study, the species with salinesodic tolerance exhibited higher phenotypic plasticity in response to soil salinity-sodicity than the highly tolerant species. Our finding was in agreement with previous studies on shade stress where species with low tolerance exhibited greater plasticity than highly tolerant species (Chmura et al., 2017;Portsmuth & Niinemets, 2007;Valladares & Niinemets, 2008). The lower plasticity of highly saline-sodic tolerant species suggests that tolerance is associated with phenotypic stability and a conservative resource-use strategy even when resources are temporarily abundant (Balaguer et al., 2001;Valladares et al., 2002). Our results show that saline-sodic tolerant species exhibit lower phenotypic plasticity than salinesodic sensitive species. This suggests that phenotypic plasticity is negatively correlated with fitness-related traits and is inconsistent with previous findings that phenotypic plasticity can be regarded as an index of species sensitivity to soil salinity-sodicity (Couso & Fernández, 2012).
This study found that the highly tolerant species exhibited less allocation plasticity compared with growth plasticity, while the species with low tolerance exhibited higher growth and allocation plasticity. This implies that the highly tolerant species (S. glauca and S. salsa) have the ability to adapt to abiotic stress by adjusting their growth traits, which is a more conservative strategy in terms of biomass allocation compared with the species with low tolerance (C. acuminatum and C. stenophyllum). Our findings that growth traits are more sensitive than biomass allocation are more applicable to the highly tolerant species than the species with low tolerance.

| "Apparent" versus "real" plasticity
Our results show that growth and allocation plasticity are a result of ontogenetic changes or environmental factors. Across 28 pairs of allometric relationships (Table 2), only 25% of the plasticity was "real" plasticity, that is, resulting from the soil salinity-sodicity (heterogenetic slope), while 75% of the plasticity was "apparent" plasticity, that is, resulting from the allometric coefficient (heterogenetic elevation) and the plant size (shifting along a common slope and elevation). The allometric results suggest that the phenotypic plasticity was mostly due to differences in individual size in the different saline-sodic soils (Weiner, 2004 (Enquist, 2002;Li, Weiner, Zhou, Huang, & Sheng, 2013;Thomas, Martin, & Mycroft, 2015). The phenotypic plasticity for most traits in this study was ontogenetic, that is, "apparent" plasticity, but for plant height versus stem diameter and root versus shoot biomass, the species with low tolerance adjusted both their plant size and growth pattern in response to soil salinity-sodicity. This suggests that the plasticity in growth traits and biomass allocation is largely ontogenetic.

| CON CLUS IONS
In summary, soil salinity-sodicity inhibited plant growth in species with both low and high tolerance and had more significant negative effects on root allocation in species with low tolerance than highly tolerant species. Growth and biomass allocation plasticity were found to be negatively correlated with saline-sodic tolerance. For highly salinesodic tolerant species, growth traits are more sensitive to salinitysodicity than allocation traits, while both growth and allocation traits are sensitive to soil salinity-sodicity in species with low tolerance. Thus, the species with low saline-sodic tolerance have greater plasticity in response to environmental variation, while the highly saline-sodic tolerant species maintained high fitness in unfavorable environments. Our results show that the plastic response could be either "apparent" (75%) or "real" (25%). To elucidate the actual mechanistic differences among species with different soil salinity-sodicity tolerances, we therefore strongly recommend that future studies should address whether phenotypic plasticity is "apparent" or "real" plasticity.

YH is supported by the National Key Research and Development
Program of China (2016YFC0500606), the National Science Foundation of China (31570454), and a grant from the Youth Innovation Promotion Association, the Chinese Academy of Sciences (2016210).

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R CO NTR I B UTI O N S
YH and DZ designed the study; YH collected the data; YH, GF, and JP analyzed the data. YH and JP wrote the manuscript, and all authors contributed to the final version of the paper. All authors read and approved the final manuscript.