Developing Growth Promotion Strategies for Cressa truxillensis to Improve Success of Restoration Activities

Wetlands provide important ecosystem services such as supporting biodiversity, filtering water, sequestering carbon, and protecting against storm surge (Zedler and Kercher 2005; Duarte et al. 2013; Benson et al. 2019). Anthropogenic impacts have resulted in a global loss of approximately half of the world's wetlands (Zedler and Kercher 2005). Wetland losses in California have exceeded 90% in the last 200 years, thus creating a continued need for wetland research, conservation, and restoration or enhancement to prevent and slow further loss (Allen and Feddema 1996). Wetlands have increasingly faced threats by invasive species, anthropogenic impacts, and changes in the climate; however, restoration by revegetation with native plants provides one possible solution to returning wetland ecosystem services. Revegetation approaches include transplantation of plants or rhizomes, planting plugs, and direct seeding (Godefroid et al. 2011; Kettenring and Tarsa 2020). Understanding the optimal conditions to promote seed germination and plant establishment of target species is of substantial importance to wetland restoration and management.

A seed-based approach to vegetative restoration has the advantage of being less expensive and logistically less challenging to implement (Kettenring and Tarsa 2020). However, there is a range of soil conditions and hydrology that are likely to affect in situ field germination (e.g., nutrient availability, soil grain size and porosity, invasive species presence, etc.), and the downside is that many plant species exhibit extremely low seed germination rates, especially due to abiotic stresses such as salt or the presence of a hard seed coat that results in dormancy and prevents imbibition (Almansouri et al. 2001; Baskin and Baskin 2014, 2020). To break dormancy, scarification can break the seed coat and allow a seed to imbibe and eventually germinate (Baskin and Baskin 2014). A better understanding of effective scarification techniques for individual species that are targets of restoration projects is essential for the success of a seed-based approach.

A plant's tolerance to abiotic stress can also be manipulated, with numerous examples of plant growth promoting bacteria (PGPB) forming beneficial relationships with host plants (Cassán et al. 2009; Enebe and Babalola 2018). PGPB can reside within the seed, as well as on and inside plant roots, promoting plant growth and resilience via biochemical properties, resistance to disease, and acting as a buffer to abiotic stressors (Eida et al. 2018). Determining optimal seed germination to account for these inhibitory factors can be important for seeding plants during restoration, leading to a higher percentage of germinated seeds. With a growing need for restoration and improving seeding success of suitable plants for revegetation, this study aimed to inform effective growth strategies of the California native wetland plant, alkali weed, Cressa truxillensis.

Cressa truxillensis, commonly known as akali weed, is a facultative wetland plant native to the western United States and commonly found in many southern California wetlands (Lichvar 2012). Cressa truxillensis is known to have a late spring bloom period with hallmark characteristics including a high salinity tolerance and high calcium carbonate tolerance (Jepson and Hickman 1993). Studies on C. cretica, another species in the genus, found that the plant could grow in soils with an upper range of 22% calcium carbonate and saline concentrations up to 400 mM (Kabir et al. 2010; Sivasankaramoorthy et al. 2011). Although often found in wetlands, the plant can also thrive as a native in many arid and alkaline habitats, making it an important focus for the conservation or enhancement of multiple habitat types. The Cressa genus has very low germination rates in its naturally occurring habitats, which has been attributed to a hard seed coat and suboptimal environmental conditions (Etemadi et al. 2020). This research project aimed to inform the current state of knowledge regarding the use of this species in revegetation projects by investigating C. truxillensis scarification techniques and through characterization of its microbial endophytes.

Different scarification methods were applied to seeds of C. truxillensis to determine optimal conditions to increase the number of seeds that imbibe and germinate. Seeds for the scarification experiment were obtained from S&S Seeds (www.ssseeds.com). C. truxillensis seeds are ovular, brown, and 2-4 mm in length (Austin 1998). Four approaches were taken to scarify C. truxillensis seeds, including three mechanical and one chemical approach (Table 1). Each treatment had five replicate batches of 20 seeds. Chemical scarification was performed by immersing seeds in 0.5 mL of concentrated sulfuric acid in a centrifuge tube for 45 min. The seeds were then rinsed with sterile distilled water five times to remove all traces of the acid. The three methods of mechanical scarification were scratching individual seeds against sandpaper, rubbing seeds in a batch between sandpaper, and nicking the seed coat with a razor blade. For scarification of individual seeds by sandpaper, each seed was scratched across 220-grit sandpaper for 2.5 cm. For the batch treatment, seeds were randomly dispersed between two pieces of 220-grit sandpaper, a book placed on top to apply even pressure, and a circular motion applied to rub the seeds between the sheets of sandpaper for 10 passes. To scarify seeds with a razor blade, each seed was nicked once to break the surrounding seed coat. Batches of untreated seeds were placed in petri dishes to serve as controls. For all treatments and the controls, seeds were placed in autoclaved petri dishes containing filter paper saturated with 7 mL of sterile deionized water. The petri dishes with seeds were wrapped in parafilm to prevent evaporation and then placed in the dark at 24°C. The number of germinated seeds within each dish was counted after seven days.

Bacteria associated with C. truxillensis were identified, as many plant-associated bacteria have plant-growth promoting properties. C. truxillensis seeds were collected from the ground, within C. truxillensis and Salicornia pacifica plant communities in the non-tidal high salt marsh and transition habitats of the western area B marsh at the Ballona Wetlands Ecological Reserve (Los Angeles, CA, 33.963025°, -118.447015°). No more than 10% of available seed was collected from within individual polygons that were identified throughout the appropriate habitat areas, ensuring that no seed banks were impacted during the collection. A total of 150 seeds were collected from the Ballona Reserve. Seeds were scarified and surface sterilized with 95% ethanol for five minutes, followed by 10% bleach for 10 min, and then rinsed with five washes of sterile deionized water. The water from each wash step was plated onto Tryptone Yeast extract agar (TY; 5 g/L tryptone, 3 g/L yeast extract, 0.0662 g/L CaCl₂·2H₂O, 15 g/L agar) and incubated at 30°C to rule out contamination by a lack of microbial growth on the media. Following sterilization, seeds were crushed individually in 5 mL of sterile saline (0.9% NaCl) using a sterile mortar and pestle. Dilutions of the bacterial suspensions were plated onto TY agar and incubated at 30°C (Siddikee et al. 2010). Dilutions were made so that individual colonies would be easily identified and isolated when grown on TY agar. Two individual bacterial strains, C15 and C3G, were selected to represent the different colony morphologies seen on Yeast extract Mannitol agar (YM; 0.5 g/L yeast extract, 10 g/L mannitol, 0.5 g/L KH₂PO₄, 0.5 g/L K₂HPO₄, 0.2 g/L MgSO₄·7H₂O, 0.2g/L NaCl, 15 g/L agar). A simple stain with safranin and a spore stain were performed on each strain and observed by brightfield microscopy (Zeiss Axioscope 5).

The bacterial strains were characterized for biochemical properties associated with plant growth promotion and their ability to tolerate salt and heavy metal stress. Strains were tested for nitrogen-fixation by growth on nitrogen-free Jensen's media (Jensen 1942) after 7 d. Cellulase activity was tested by growing the strains on Carboxymethyl Cellulose (CMC) agar (Kasana et al. 2008), flooding the plates with Gram's iodine, and checking for the presence of clear halos around the streaked bacteria. The production of exopolysaccharide was determined by growing bacteria on YM agar and visually identifying the presence of a mucoid colony morphology. Phosphate solubilization was tested by visually identifying the presence of clear halos around bacteria on Pikovskaya's agar (Pikovskaya et al. 1948) after 7 d of growth. Auxin production was determined by culturing bacteria in TY supplemented with 1 mg/ml tryptophan and adding Salkowski reagent (Ehmann 1977; Gordon and Weber 1951) after growth. The presence of red coloration indicated the production of indole acetic acid (IAA). For analysis of bacterial tolerance to abiotic stressors, bacteria were streaked on TY agar that had been supplemented with zinc (50, 100, 200, 400, 750, or 1000 µM ZnSO₄), cadmium (50, 100, or 200 µM CdCl₂) or salt (1%, 2%, 3%, 4%, or 5% NaCl), incubated for seven days at 30°C, and growth compared to that seen on TY. All assays were done in triplicate.

To identify bacterial strains C3G and C15, DNA was obtained from each strain using a Chelex extraction (Walsh et al. 2013). The DNA was used to amplify the 16S rRNA gene by the polymerase chain reaction using universal primers 27F and 1492R (Weisburg et al. 1991). The PCR products were sequenced, and the sequences were compared to that of bacterial type strains using tools of the Ribosomal Database Project (Cole et al. 2014) and NCBI BLAST (Altschul et al. 1990). Closely related sequences were retrieved from the GenBank database. A phylogenetic tree of the 16S rRNA gene was generated using MEGAX (Kumar et al. 2018; Stecher et al. 2020) and the maximum-likelihood algorithm. Bootstrap analysis was done for statistical support using 1000 re-samplings with gamma distribution (G) and the Tamura 3-parameter (T92) model (Tamura 1992). The following new sequences were deposited in the GenBank database: 16S rDNA of Paenibacillus sp. strain C15 (OK178930) and Paenibacillus sp. strain C3G (OK178931).

For the seed germination assays including salt stress and bacterial inoculation, C. truxillensis seeds were scarified and surface sterilized as described above. Bacterial strains C3G and C15 were grown overnight in TY media, centrifuged and the bacterial pellet washed once with 10 mM MgSO₄ to remove any residual TY, and then centrifuged again and the bacteria resuspended in 10 mM MgSO₄ to an absorbance at 600 nm of 0.1 (Montañez et al. 2012). C. truxillensis seeds were incubated for 1 hr in the bacterial suspension or 10 mM MgSO₄ as a control. The 35 seeds were transferred in triplicate to a petri-dishes containing 0.8% agar (Sigma A1296) with either 0, 150, or 300 mM NaCl. The number of seeds germinated were counted after 7, 14, and 21 days. Wet weight of germinated seedlings was taken at 21 d.

Cressa truxillensis control seeds exhibited a low average percentage of germination when seeds were not scarified (5.8% ± 3.4%) (Fig. 1). Chemical scarification in concentrated sulfuric acid for 45 min resulted in the greatest average germination (54% ± 6.1%), substantially higher than the unscarified control, although mechanical scarification of individual seeds via sandpaper (34% ± 9.8%) and nicking with a razor blade (29% ± 3.4%) also improved average germination (Fig. 1). However, batch scarification did not show substantial improvement in the percentage of germination (11% ± 6.1%) as compared to the control treatments.

Results indicated that the low salinity treatment (150 mM NaCl) had no impact on percent seed germination of C. truxillensis at 7 days (61% ± 1.2%) as it was comparable to the control (59% ± 5.1%) (Fig. 2). However, the high salinity treatment (300 mM NaCl) caused a reduction in germination (8% ± 0.7%) (Fig. 2). By 21 d the percentage of germinated seeds did not vary between saline treatments (Fig. 2). Although the final germination percentages were similar, seedlings exposed for 21 d to 150 mM or 300 mM NaCl showed a reduction in growth compared to unexposed seedlings (Fig. 3A and 3B). In addition, the higher the concentration of salt, the greater the reduction in seedling size and mass (Fig. 3A and 3B).

Bacteria were isolated from within C. truxillensis seeds that had been collected from the Ballona Wetlands. Although bacterial colonies looked similar on TY, when transferred to YM agar they showed two distinct colony morphologies. Strains C15 and C3G were chosen as representatives of each colony type (Fig. 4A and 4B). Most notably, the texture was a differentiating characteristic between strains C3G and C15; on YM agar, strain C15 had a mucoid appearance indicating the presence of exopolysaccharide while strain C3G did not (Fig. 4B). Microscopy of stained bacteria showed that both strains looked similar and were bacillus in shape (Fig. 4C and 4D). In addition, both strains were found to produce endospores (Table 2). The 16S rRNA gene sequences of strains C3G and C15 were 100% identical to each other and identified them as species of Paenibacillus, with 99.71% (1393/1397 n.t.) identity to and phylogenetically grouping with P. peoriae DSM8320 (Fig. 5).

Strains C15 and C3G were tested for biochemical properties associated with plant growth-promoting bacteria. Both strains grew on nitrogen-free media (Table 2), indicating that they were able to fix nitrogen, similar to many previously characterized Paenibacillus strains (von der Weid et al. 2002). In addition, both strains had cellulase activity, produced auxin, and solubilized phosphate (Table 2). Strains C15 and C3G were tested for growth at varying concentrations of zinc, cadmium, and sodium chloride in order to look at the impact of various abiotic stressors. C3G grew with up to 400 µM ZnSO₄, however C15 grew even in the presence of 1 mM ZnSO₄ (Table 2). The strains also showed varying tolerance to cadmium. C3G was sensitive to cadmium, but C15 grew in up to 50 µM CdCl₂ (Table 2), highlighting the higher tolerance C15 has to heavy metals. Both strains behaved similarly to salt stress and could grow in as much as 4% NaCl but were inhibited at 5% NaCl (Table 2). When C. truxillensis seeds were inoculated with either strain of bacterium, no difference was seen in the percentage of seeds that germinated, whether or not salt was present (150 and 300 mM) (Fig. 2). However, in the absence of salt, inoculation with C3G resulted in a slight increase in seedling wet weight (Fig. 3B).

The seed coat can be a major barrier to germination for many plants. How to overcome this seed coat-imposed dormancy, or “hardseededness”, and thus increase the seed's permeability to water, has been the basis of numerous studies (Tadros et al. 2011; Mousavi et al. 2011). Typical scarification strategies include hot water treatment, chemical scarification with acid, or mechanical scarification to nick the seed coat (Rusdy 2017). Though differing in method, chemical and mechanical scarification work to obtain similar results of breaking the seed coat. Mechanical scarification may require more labor on fewer seeds via physical removal or permeation of the barrier, while chemical scarification is effective for larger batches and uses concentrated acid to dissolve part of the seed coat (Majd et al. 2013). The use of sulphuric acid to overcome germination difficulties in seeds has been well-established in the literature as a highly effective scarification technique (Kheloufi et al. 2017). Congruent with studies of C. cretica, which found that naturally occurring seeds have very low germination rates (Etemadi et al. 2020), the non-scarified control C. truxillensis seeds in this study showed a low percentage of germination. This suggests that restoration or revegetation projects seeding with non-scarified seeds may experience low initial germination rates due to seed coat-imposed dormancy, which would delay bringing back native plant cover to an area. An experimental study at the University of Wisconsin comparing hand seeding and transplantation as restoration techniques found hand seeding to be more successful in producing greater diversity and growth of native plants in addition to greater cost and time effectiveness; however, they noted that an extended dormant period for unscarified seeds may present problems with the regrowth of non-native plants (Weiher et al. 2003). Results from this study indicated both mechanical and chemical scarification methods improved C. truxillensis seed germination, showing that scarification is a useful strategy to overcome dormancy. The sulfuric acid scarification technique, which was also effective in scarifying C. cretica (Etemadi et al. 2020), substantially improved germination rates of C. truxillensis compared to other treatments. Additionally, the sulfuric acid scarification method allowed for many seeds to be scarified at the same time in a batch, meaning this technique could be a feasible option for preparing large quantities of seeds prior to seeding for restoration or revegetation activities, particularly because of the reduced time and effort. If the percentage of seeds that are germinating at each restoration site can be increased using pre-scarified seeds, restoration projects may be able to see more rapid success in reaching native plant cover objectives.

Bacterial endophytes reside intracellularly or intercellularly within the plant, often acting as beneficial PGPB and helping the plant survive abiotic stress (Rashid et al. 2012; Egamberdieva et al. 2017). An increasing number of studies have shown bacteria residing as endophytes in seeds as well (Li et al. 2019; Khalaf and Raizada 2016), with these bacteria likely derived from the parent plant and contributing to increased fitness and survival of the seed in its environment (Li et al. 2019; Mastretta et al. 2009; Zhang et al. 2010). Paenibacillus sp. are known to be PGPB for numerous plant species and have been found residing as seed and root endophytes (Eida et al. 2018; Khalaf and Raizada 2016). Paenibacillus has been studied in Arabidopsis thaliana and found to protect the plant from abiotic stressors and plant pathogens (Hong et al. 2016). Further studies have tagged the bacteria with green fluorescent protein and found that the bacteria form biofilms and colonize root tips of plants (Timmusk et al. 2005). The C. truxillensis bacterial strains isolated in this study were identical to each other in 16S rRNA gene sequence and were most similar to Paenibacillus peoriae. Although the strains showed variation in exopolysaccharide production and tolerance to cadmium, they both showed plant growth promoting traits such as nitrogen fixation, cellulase production, phosphate solubilization, and auxin production. Consistent with this, after inoculation of C. truxillensis seeds, one of the strains resulted in an enhancement of seedling mass. In addition, Paenibacillus produce endospores. Bacterial endospores are a dormant form of bacteria that can withstand harsh chemical and thermal environments and allow the bacteria to survive stressful conditions, thus making them ideal for inoculant formulations. If these Paenibacillus strains were to be used for plant growth promotion in the field, their endospores could be explored as a form of inoculant as this could increase inoculant shelf life due to the ability of endospores to survive variable conditions (Kim et al. 2010).

Studies have shown that plant associated bacteria, including strains of Paenibacillus, can help plants better cope with heavy metal stress (Kumari and Thakur 2018; Sukweenadhi et al. 2018; Eida et al. 2020). The Paenibacillus strains from this study showed tolerance to zinc and cadmium. Previous studies in the Ballona Reserve showed that zinc and cadmium concentrations are relatively high, presumably due to urban runoff and other anthropogenic impacts (Johnston et al. 2012). Thus, it is not surprising that bacterial strains isolated from the Ballona Wetlands have increased tolerance to these heavy metals. A study of C. truxillensis at a salt marsh in Patagonia, Argentina, found that C. truxillensis was sensitive to heavy metals, with morphology of the leaves altered depending on concentrations of pollutants such as zinc (Pollicelli et al. 2018).

Halophytes, which include C. truxillensis, are plants that can tolerate NaCl levels greater than 200 mM (1.17%). This is consistent with the study findings for C. truxillensis, which could germinate with at least 300 mM NaCl. This finding, as well as the decreased seedling growth seen with increasing salt concentrations, is analogous to what has been reported for C. cretica (Etemadi et al. 2020). Recent studies suggest that plant-associated microbes can play a key role in the ability of halophytes to grow in high salinity (Etesami and Beattie 2018) but may also help non-halophytes tolerate salt stress. Paenibacillus sp. JZ16, isolated from inside the roots of the halophyte Zygophyllum simplex, was found to promote salinity tolerance of Arabidopsis (Eida et al. 2018). The Paenibacillus strains from this study grew in both the absence of salt as well as in up to 4% NaCl. Both strains also produced auxin, a trait frequently associated with salt tolerant PGPB (Dodd et al. 2010). It is thought that the altered hormonal signaling, which can influence lateral root development (Yang et al. 2009), contributes to the ability of these bacteria to increase the plant's fitness in highly saline environments (Siddikee et al. 2010; Tiwari et al. 2011). Other studies have found that exopolysaccharide may allow bacteria to promote plant growth under saline stress (Abbas et al. 2019). Since C. truxillensis is known to grow in saline soils, inoculation with an exopolysaccharide producing strain may allow for optimized plant growth in high saline conditions. However, although one of the Paenibacillus strains in this study showed some growth promotion of seedlings in the absence of salt, neither strain, whether or not it produced exopolysaccharide, alleviated the negative impact salt had on seed germination or growth under the laboratory conditions tested. Future studies might look at whether these bacterial strains have a beneficial impact on plant growth with salinity stress using greenhouse and field conditions.

Many studies have focused on the use of PGPB in agriculture; however, few studies have explored applications of the microbial community on restoration and native plant revegetation projects (Ahn et al. 2007). The application and results of this study could inform habitat restoration projects throughout the range of C. truxillensis, including at the Ballona Wetlands. Wetland restoration projects that have focused on returning native cover to an area have quickly discovered a vast lack of available peer reviewed literature when it comes to best practices for species-specific cleaning, storage, and breaking dormancy, despite breaking seed dormancy being a widely known requirement for revegetation (Kettenring and Tarsa 2020; Barton et al. 2016). Scarification techniques recommended by this study could be used by practitioners working on habitat restoration projects with the goal of improving native plant cover. Past restoration efforts within the Ballona Wetlands have focused on reducing anthropogenic uses and removal of invasive plants within the area (Johnston et al. 2021); however, the findings from this study could allow for a unique opportunity to combine knowledge of revegetation techniques with the microbial community for successful revegetation with C. truxillensis. Application of these findings at a restoration project would allow for a greater knowledge base around the use of PGPB for revegetation, as well as development of best practices for germinating wild seeds.

Utilizing the findings from this study, a few key recommendations can be made for future revegetation and restoration projects. Regarding improving germination, sulfuric acid scarification should be used on seeds prior to deployment: this method resulted in the greatest average percent germination and would also allow for large batches of seeds to be scarified together, reducing energy and time. Seeds can be inoculated with PGPB's prior to seeding, such as the ones identified in this study, to increase plants' protections against heavy metals and provide biochemical advantages. Finally, considerations should be taken to deploy seeds in habitats that include some freshwater hydrology, especially in salt marsh soils. Findings suggested that high salinity had a delaying effect on germination of C. truxillensis seeds, and freshwater sources may ameliorate this delay from salt stress to provide maximum germination.

Future directions with these findings will aim to test scarification methods and microbial applications in situ for applications within restoration projects. Scarification techniques from this study should also be explored for other native plants commonly seeded in restoration and revegetation projects, including rare species. To further contribute to the available knowledge informing habitat restorations, pre-scarified seeds could be dispersed within an area to measure the success and feasibility of scarification methods in varying field conditions. Based on findings in the Pollicelli et al. (2018) study that C. truxillensis has sensitivity to heavy metals, future studies might look at whether the Paenibacillus strains from this study are able to increase the tolerance of C. truxillensis to certain heavy metals. The endophytic nature and presence of plant growth promoting traits of the two strains isolated in this study suggest they are PGPB, although the strains did not alleviate the impact of salt stress under the conditions tested in this study. Future studies can further assess the conditions in which these strains might promote plant growth and assist C. truxillensis withstand abiotic stress, and if this would be helpful when planting seeds collected from other locations that might not carry the same microbes. Formulating inoculants for the plants and measuring growth could give definitive answers on how well each strain is able to promote plant growth and other characteristics that may be advantageous for future restoration projects.

This research was supported in part by the William F. McLaughlin Chair in Biology to M.R.L. H.L., K.A., and C.E. received support through Loyola Marymount University's Coastal Research Institute and a grant from the U.S. Environmental Protection Agency. The authors would also like to thank The Bay Foundation and California Department of Fish and Wildlife for support.