Differential Effects of Increasing Salinity on Germination and Seedling Growth of Native and Exotic Invasive Cordgrasses

Soil salinity is a key environmental factor influencing germination and seedling establishment in salt marshes. Global warming and sea level rise are changing estuarine salinity, and may modify the colonization ability of halophytes. We evaluated the effects of increasing salinity on germination and seedling growth of native Spartina maritima and invasive S. densiflora from wetlands of the Odiel-Tinto Estuary. Responses were assessed following salinity exposure from fresh water to hypersaline conditions and germination recovery of non-germinated seeds when transferred to fresh water. The germination of both species was inhibited and delayed at high salinities, while pre-exposure to salinity accelerated the speed of germination in recovery assays compared to non-pre-exposed seeds. S. densiflora was more tolerant of salinity at germination than S. maritima. S. densiflora was able to germinate at hypersalinity and its germination percentage decreased at higher salinities compared to S. maritima. In contrast, S. maritima showed higher salinity tolerance in relation to seedling growth. Contrasting results were observed with differences in the tidal elevation of populations. Our results suggest S. maritima is a specialist species with respect to salinity, while S. densiflora is a generalist capable of germination of growth under suboptimal conditions. Invasive S. densiflora has greater capacity than native S. maritima to establish from seed with continued climate change and sea level rise.


Introduction
Salt marshes are highly stressful environments where halophytes are subjected to high mortality risk [1]. In these habitats, soil salinity is one of the key environmental factors determining vegetation distribution, partially by limiting seed germination and seedling establishment [2]. These phases are crucial in the life cycle of halophytes [3][4][5]. The general behaviour of halophytic seeds in the presence of salt is well documented [6]. Seeds of most halophytes show optimal germination in freshwater, differing in their germination responses to higher salinities [6][7][8]. High salinities usually inhibit germination of halophytes, however, some seeds maintain viability and are able to germinate when osmotic stress decreases [9][10][11].  Different letters indicate significant differences among treatments for each trait (Mann-Whitney U test, in italic, for Kruskal-Wallis or Tukey's HSD test, in non-italic, for one-way ANOVA, p < 0.05). Table 1. Germination percentage (G), T 50 of germination (T 50 G), recovery germination percentage after salt exposure (RG), T 50 of germination recovery (T 50 R), and seed viability percentage (V) for native Spartina maritima from the Gulf of Cádiz (Southwest Iberian Peninsula) in six salt treatments. Data show mean ± Standard Error (SE) (n = 3-4). Different letters indicate significant differences between treatments (Mann-Whitney U test, in italic, for Kruskal-Wallis or Tukey's Honest Significant Difference (HSD) test, in non-italic, for Analysis of Variance (one-way ANOVA), p < 0.05). df (degrees of freedom). Increasing salinity decreased the germination speed (higher T 50 G) for seeds sourced from all marsh elevation zones and locations. Additionally, S. densiflora seeds produced at LM germinated 27% faster in freshwater than those from MM and HM (one-way ANOVA, F = 11.03, degrees of freedom (df) = 2, p < 0.0005) ( Table 2).

Germination Responses after Salinity Exposure
The germination rate of Spartina maritima seeds increased, and the speed of germination decreased significantly during recovery after exposure to higher salinities. T 50 R (days necessary to reach 50% of the final germination percentage in recovery assays) was reduced after recovery from all levels of salinity exposure than seeds that were germinated in freshwater. The T 50 reduction indicating an increase in germination speed was most extreme during recovery from 0.15 M NaCl exposure when germination accelerated almost nine times (Table 1). In contrast to S. densiflora, no dormant seeds were recorded for S. maritima during the recovery experiment. S. maritima seed viability was c. 76%, without showing significant differences among salinities ( Table 1).
As in the case of S. maritima, S. densiflora seeds had increased germination rates during recovery following exposure to higher salinities, and this result held for seeds sourced from every location and within-marsh habitat elevation zone ( Figure 1). S. densiflora seeds tended to have increased germination speeds (lower T 50 R) during recovery following exposure to higher salinities ( Table 2). Seeds from LM exposed previously to hypersalinity showed lower recovery germination percentage (78% ± 3%) and lower T 50 (5 ± 1 days) than seeds from MM and HM (c. 90% and 7 days, respectively) ( Figure 1, Table 2). Dormant S. densiflora seeds in freshwater were c. 7% for every habitat and location and tended to decrease with increasing salinity exposure ( Figure 1). Seed viability in control treatments were higher for seeds of S. densiflora from LM and MM (c. 76%) than those from HM (54% ± 4%) (one-way ANOVA, F = 12.95, df = 2, p < 0.0001). On the other hand, seed viability was not affected by salt treatments in seeds from LM, while in seeds from MM and HM the viability decreased as salinity increased (Table 2). Regarding the three source population locations, there were no significant differences in seed viability at any salinity level for seed sourced from Almendral, whereas viability was reduced at higher salinities for the other two study locations ( Table 2).

Initial Seedling Growth Responses to Salinity
S. maritima seedlings had similar cotyledon and first leaf lengths at every salinity level ( Figure 2). However, the radicles of S. maritima were four times longer at 0.15 M NaCl than radicles that emerged from seeds exposed to other salinity concentrations (Kruskal-Wallis test, H 3,59 = 8.66, p < 0.05). Cotyledon, first leaf and radicle of S. maritima seedlings growing in freshwater all increased in length in recovery after seeds had been exposed to salinity higher than 0.15 M NaCl ( Figure 2). 2). However, the radicles of S. maritima were four times longer at 0.15 M NaCl than radicles that emerged from seeds exposed to other salinity concentrations (Kruskal-Wallis test, H3,59 = 8.66, p < 0.05). Cotyledon, first leaf and radicle of S. maritima seedlings growing in freshwater all increased in length in recovery after seeds had been exposed to salinity higher than 0.15 M NaCl ( Figure 2). . Different letters indicate significant differences between salinity treatments (Mann-Whitney U test, in italic, for Kruskal-Wallis or Tukey's HSD test, in nonitalic, for one-way ANOVA, p < 0.05). Asterisks and plus sign indicate significant differences . Different letters indicate significant differences between salinity treatments (Mann-Whitney U test, in italic, for Kruskal-Wallis or Tukey's HSD test, in non-italic, for one-way ANOVA, p < 0.05). Asterisks and plus sign indicate significant differences compared to control treatment (0.00 M) for each seedling parameter (Mann-Whitney U test, asterisks, for Kruskal-Wallis or Tukey's HSD test, plus sign, for one-way ANOVA, p < 0.05). Triangle indicates that seedlings were dead before 15 days after germination, so they were not measured.
The cotyledon and first leaf length of S. densiflora were reduced as salinity increased in seeds sourced from every study location and elevational habitat ( Figure 3). The length of cotyledon and first leaf were shorter for seedlings from LM seed that germinated at salinity higher than 0.30 M NaCl ( Figure 3a). In contrast, seedlings from MM and HM seeds displayed this reduction in size at a lower salinity concentration of 0.15 M NaCl (Figure 3b,c). Also, in contrast, salinity had a positive effect on radicle growth at 0.30 M NaCl for seeds coming from LM, though radicle length was reduced at salinities higher than 0.45 M NaCl (Kruskal-Wallis test, H 5,205 = 84.93, p < 0.0001) (Figure 3a). This shorter radicle length trait was expressed in salinities higher than 0.30 M NaCl for seeds from MM   . Different letters indicate significant differences among salinity treatments (Mann-Whitney U test, in italic, for Kruskal-Wallis or Tukey's HSD test, in non-italic, for one-way ANOVA, p < 0.05). Asterisks and plus sign indicate significant differences compared to control treatment (0.00 M) for each seedling parameter (Mann-Whitney U test, asterisks, for Kruskal-Wallis or Tukey's HSD test, plus sign, for one-way ANOVA, p < 0.05).

Discussion
Our hypotheses were partially confirmed by the outcomes of the experiments. As expected, germination rates of seeds from native S. maritima expressed high tolerance to salinity, but tolerance of this native species was lower than that of salinity levels tolerated by invasive S. densiflora. In this circumstance, S. densiflora has a higher capacity for phenotypic plasticity in response to salinity. However, invasive seedling trait responses to salinity suggest a lower tolerance to salinity during the initial 15 days of seedling growth than the more vigorous initial growth of native S. maritima.
Some salinity responses were common to both studied cordgrasses. For example, high seed viability after salinity exposure was recorded for both Spartina species, as it has commonly been observed for many halophyte species [9]. Also, elevated salinity concentrations inhibited and delayed germination for both studied Spartina species [8]. This seed quiescence prevents seed germination under stressful conditions [1,5]. Decreases in germination rates have been reported for the congener Spartina alterniflora Loisel. at salinities higher than 0.20-0.40 M NaCl in the native and invasive range [34,35], however some authors recorded high germination (>90%) even at hypersalinity [15]. Germination of Spartina ciliata Brongn. from Brazil was reduced at salinities higher than 0.20 M NaCl, totally inhibited at seawater concentration, and speed of germination increased after salinity exposure [36]. In addition, exposure to salinity followed by recovery after freshening accelerated germination for S. maritima and S. densiflora in our study, which has been reported previously for other halophyte species from a range of functional groups [8,[37][38][39]. The observed stimulation of germination speed and rate after salinity exposure can provide windows of opportunity for seeds to germinate and quickly establish when salinity is sporadically and temporarily reduced by precipitation events or other sources of freshwater inflow [10,40].
While both studied cordgrasses showed high tolerance to salinity and shared some common germination and initial seedling trait responses, each also expressed distinctly contrasting responses. Germination rates and speed of germination under increasing NaCl concentrations indicated S. densiflora had higher germination tolerance to salinity than S. maritima. S. maritima germination was completely inhibited and S. densiflora was able to germinate at hypersalinity. Moreover, germination percentage decreased from 0.15 M NaCl up for S. maritima and from 0.30 M NaCl up for S. densiflora. As in our study, invasive S. densiflora in Humboldt Bay (California) showed reductions in seed germination at salinities higher than 0.30 M NaCl [30]. Some authors recorded total germination inhibition at 1.00 M NaCl and at 0.70 M NaCl for invasive S. densiflora in the Gulf of Cádiz [29,31]. The great salinity tolerance of S. densiflora was also reflected on its rapid germination after being pre-treated at increasing salinities, whereas this study is the first to document the opposite response for native S. maritima. Seed quiescence of S. densiflora did not alter its initial seedling growth, in contrast to responses of S. maritima, in which seedlings were longer after saline pre-treatments compared to control. In addition to the capacity for seed quiescence under stressful conditions, the response of S. densiflora indicates a degree of physiological seed dormancy ( < 10%). In addition, the radicle and first leaf of S. densiflora always emerged earlier than those of S. maritima. This suggests S. densiflora seeds were able to remain dormant in stressful saline environments without damaging the quality of the embryo, and then had the capacity to germinate later when salinity stress was reduced, providing multiple opportunities for establishment [1,10,39].
Its higher germination tolerance to salinity, faster germination after being pre-treated at increasing salinities and the presence of seed dormancy help to explain that S. densiflora is able to invade a wide range of habitats along the intertidal gradient [21] including hypersaline saltpans [3]. Invasive S. densiflora germination showed higher salinity tolerance than native S. maritima, but the autochthonous species showed higher salinity tolerance in relation to early seedling growth. Negative effects of salinity on seedling growth have been reported previously for invasive S. densiflora in Humboldt Bay (California) where seedling height decreased at salinities higher than 0.20 M NaCl [30], for invasive S. alterniflora in China [35], where shoot height decreased from 0.20 M NaCl up and radicle length from 0.10 M NaCl up, and for native S. ciliata in Brazil [36], with shoot and radicle being smaller at salinities higher than from 0.05 M NaCl.
Besides the general comparison between both Spartina species, contrasted responses to salinity in seed viability, germination rate and speed, seed dormancy and seedling growth were also recorded among native and invasive species and among S. densiflora populations along the intertidal gradient. These differences in germination and seedling trait responses between S. maritima and S. densiflora and for seeds sourced among contrasting S. densiflora habitats could be attributed to local adaptation to contrasted environments [41], or to pre-adaptive conditioning determined by the maternal environment during seed development [42][43][44]. Salinity responses from different S. densiflora locations along the Odiel-Tinto Estuary (grouping LM, MM and HM elevations at each location) may support the pre-adaptive conditioning hypothesis. Supporting this idea, seed viability was high and independent of exposure to salinity concentrations for seeds from Almendral, but it decreased as salinity increased for seeds from Bacuta and Calatilla locations. Almendral is the nearest location to the coastline (12,500 m), whereas Bacuta and Calatilla are located along a tidal gradient 1800 m and 5500 m inland from Almendral, respectively. Thus, seeds ripening in low elevations and in locations closer to the coastline are more frequently exposed to tidal flooding and salt spray than those at higher elevations and more inland locations on the intertidal gradient [45]. Differentiated environmental conditions may acclimate seeds to salt stress in LM and closer to the sea, protecting their embryo from being killed due to ion toxicity at high salinities [2]. Furthermore, the invasive S. densiflora populations have low genetic diversity in North American and European marshes [46,47], which also supports that differences recorded along the intertidal gradient would likely be due to phenotypic plasticity rather than to genetic adaptation. Pre-adaptive conditioning determined by maternal stress conditions could increase survivorship and germination under high salinities, which may suppose an advantage for offspring in conditions similar to those experienced by the parents [42][43][44]. Moreover, salinity acclimation would facilitate survivorship during hydrochorus dispersal of buoyant seeds with sea water currents [48]. Other authors have observed that salt tolerance in halophyte germination is related to the duration and intensity of their exposure to salts in field conditions [9,11,32,49]. As in our study, Iris hexagona Walter and Suaeda aralocaspica (Bunge) Freitag and Schütze growing in high salinities produced seeds that had higher germination rates and speeds of germination when seeds were exposed to different salinity concentrations than seeds produced in low salinity environments [37,50]. Furthermore, other environmental factors such as temperature, photoperiod, soil moisture and nutrients availability can influence seed viability, germinability and dormancy [1,[51][52][53][54].

Study Area and Plant Material
The plant propagules evaluated in this study were sourced from the Odiel Marshes by the Gulf of Cádiz, in the Southwest Iberian Peninsula. The coast of the Gulf of Cádiz is mesotidal and the mean sea level in this area is +1.85 m relative to Spanish Hydrographic Zero (SHZ). The tides are semidiurnal and have a mean range of 2.10 m and a mean spring tidal range of 2.97 m, representing 0.40-3.37 m above SHZ. This area is under a Mediterranean climate with Atlantic influence, with +18.2 • C as the annual mean temperature [23]. Native vegetation in salt marshes along the Gulf of Cádiz has been described in previous works [23,33,55].
Inflorescences in fruiting stage were randomly collected from S. maritima and S. densiflora tussocks in the Odiel Marshes. Since native S. maritima colonizes mainly low elevations in the tidal frame [23], inflorescences were collected in August 2017 from a low marsh at the location known locally as Ludovico (37.174341N, -6.931643W; See a site description in previous work [23]) (Figure 4). S. densiflora invades LM, MM and HM in the Gulf of Cádiz [21], so its inflorescences in fruiting stage were collected from those three habitats at each of three different locations distributed from close to the estuary inlet to more inland areas (Almendral: 37.209699N, -6.953506W; Bacuta: 37.218836N, −6.964066W; Calatilla: 37.250382N, −6.969434W) in November 2016 (Figure 4). Marsh habitats were distinguished based on tidal influence and soil characteristics [33]. LM were defined between Mean High-Water Neap (MHWN) and Mean High Water (MHW), MM went from Mean High Water (MHW) to Mean High Water Spring (MHWS), and HM from Mean High Water Spring (MHWS) to Highest Astronomical Tide (HAT) [56]. In all sampled salt marshes, S. densiflora has become very abundant, displacing native vegetation [21]. Spikelets containing caryopses were randomly selected from collected inflorescences and stored in paper bags in dark and dry conditions at +5 • C until use.  [56]. In all sampled salt marshes, S. densiflora has become very abundant, displacing native vegetation [21]. Spikelets containing caryopses were randomly selected from collected inflorescences and stored in paper bags in dark and dry conditions at +5 °C until use.

Salinity Germination Experiment
Before sowing, in September 2017 in the case of S. maritima and in December 2016 in the case of S. densiflora, the spikelets of both Spartina species were surface sterilized in 5% (v/v) sodium hypochlorite for 10 min to prevent fungal contamination and then rinsed with distilled water [25,39]. Four replicates with 25 spikelets each were sown, for each habitat and location, on two layers of autoclaved filter paper watered with six different salt treatments (sodium chloride puriss pro analysis >99.5%, Sigma-Aldrich; 0.00 (control), 0.15, 0.30, 0.45, 0.60 and 0.75 M NaCl) in Petri dishes (9 cm diameter) sealed with adhesive tape (Parafilm™) to avoid desiccation. Sodium chloride was chosen as the salt to be investigated since it is by far the most prevalent major salt dissolved in the Odiel estuary water [57]. This salinity range was chosen to include salinities from freshwater (0.0 M NaCl) to sea water (0.60 M NaCl), and hypersalinity (0.75 M NaCl). The dishes were maintained during 2 months under controlled-environmental conditions in a plant grow room, at temperatures between

Salinity Germination Experiment
Before sowing, in September 2017 in the case of S. maritima and in December 2016 in the case of S. densiflora, the spikelets of both Spartina species were surface sterilized in 5% (v/v) sodium hypochlorite for 10 min to prevent fungal contamination and then rinsed with distilled water [25,39]. Four replicates with 25 spikelets each were sown, for each habitat and location, on two layers of autoclaved filter paper watered with six different salt treatments (sodium chloride puriss pro analysis >99.5%, Sigma-Aldrich; 0.00 (control), 0.15, 0.30, 0.45, 0.60 and 0.75 M NaCl) in Petri dishes (9 cm diameter) sealed with adhesive tape (Parafilm™) to avoid desiccation. Sodium chloride was chosen as the salt to be investigated since it is by far the most prevalent major salt dissolved in the Odiel estuary water [57]. This salinity range was chosen to include salinities from freshwater (0.0 M NaCl) to sea water (0.60 M NaCl), and hypersalinity (0.75 M NaCl). The dishes were maintained during 2 months under controlled-environmental conditions in a plant grow room, at temperatures between 20 • C and 25 • C and a 12 h light/12 h dark photoperiod. Radiation was provided by fluorescent lamps that produced a photosynthetic photon flux density of 60 µmol m −2 s −1 . During this time, germination was recorded every 2 or 3 days. A seed was considered germinated when the coleoptile emerged.

Post-Salinity Exposure Recovery Experiment
Spikelets that did not germinate during the 2 months salinity exposure trials were rinsed with distilled water and sown in new Petri dishes with distilled water to assess post-salinity exposure recovery. Germination was recorded every 2 or 3 days for 2 months. Seed viability of the spikelets that did not germinate during the recovery experiment was tested using the Tetrazolium test [58]. For this purpose, the embryo was incised with a scalpel and submerged in a 1% aqueous solution of 2,3,5 triphenyl tetrazolium chloride at 25 • C in darkness for 24 h. Then, red-stained viable embryos were counted through a magnifying glass.
The percentage of viable seeds (germinated plus dormant seeds) was calculated for each Petri dish. The germination rates (percentage) for viable seeds at different salinities, and the recovery germination percentage after salt exposure were then calculated. Seeds that did not germinate during the salinity treatments, but germinated in the recovery experiment, were considered quiescent seeds. Seed dormancy percentage was calculated for each Petri dish using the number of viable seeds that did not germinate at the end of the recovery experiment. In addition, the days necessary to reach 50% of the final germination percentage was calculated for each Petri dish in both the germination experiment (T 50 G) and the recovery experiment (T 50 R) [25,39].

Initial Seedling Growth
To evaluate the effects of salinity exposure and post-salinity recovery on initial seedling growth, the cotyledon, first leaf and radicle length of 1-7 seedlings per Petri dish (n = 4 Petri dishes per treatment) were measured under a magnifying glass using a ruler [59]. These data were recorded 15 days after germination in both experiments to assess initial growth of both S. maritima and S. densiflora seedlings.

Statistical Analysis
Statistical analyses were carried out with STATISTICA 8.0 (StatSoft Inc., USA) considering significant results when p ≤ 0.05. Deviation to the mean was calculated as Standard Error (SE). The normality of the data series was tested with Kolmogorov-Smirnov test and the homogeneity of variance using the Levene test. When data or their transformations (using √ x, 1/(x + 1) or arcsine(x) functions) had a normal distribution and presented homeostasie, differences in germination parameters and seedling measurements between different salinities were analysed using one-way analysis of variance (ANOVA) and Tukey's Honest Significant Difference (HSD) test as post-hoc test. If data series did not have a normal distribution or homogeneity of variance after transformation, we evaluated response differences using a non-parametric Kruskal-Wallis H-test and a Mann-Whitney U post-hoc test.

Conclusions
Together, our results provide new information on seed germination and early seedling life stage characteristics of native Spartina maritima in comparison to responses of co-occurring invasive South American S. densiflora to increasing estuarine salinity changes driven by global warming and sea level rise. At these life stages, critical to survival and establishment, S. maritima displayed a specialist strategy by germinating primarily under salinity concentrations that support survival and optimal initial seedling growth. This strategy is in accordance with field observations with S. maritima colonizing mostly only stressful low salt marshes. In contrast, invasive S. densiflora behaved as a generalist species [21] and showed the capacity to germinate and produce seedlings under a wide range of salinity concentrations. However, it presented its optimum seedling growth at freshwater and light brackish conditions, with sub-optimum seedling growth at higher salinities. This behaviour is in agreement with S. densiflora adult individuals showing high phenotypic plasticity for many traits and opportunistically colonizing a wide range of habitats along the intertidal gradient, though often as sub-optimal phenotypes for the conditions [46]. In view of these results, invasive S. densiflora seems to be better prepared than native S. maritima to tolerate salinity changes provoked by climate change and sea level rise [16]. Conservation priorities to provide a future habitat for S. maritima and other native tidal wetland flora should consider preservation of undeveloped uplands for accommodation space above current high water levels for estuarine marsh transgression with sea level rise, and immediate implementation of invasive plant management, as successional development of new tidelands will favour aggressive alien colonizers such as S. densiflora [60]. Considering this work, wetland restoration strategies should consider seed and seedling life stage responses under changing environmental conditions to support recruitment and establishment of S. maritima.