Promoting success in thin layer sediment placement: effects of sediment grain size and amendments on salt marsh plant growth and greenhouse gas exchange

Thin layer sediment placement (TLP) is a method to mitigate factors resulting in loss of elevation and severe alteration of hydrology, such as sea level rise and anthropogenic modifications, and prolong the lifespan of drowning salt marshes. However, TLP success may vary due to plant stress associated with reductions in nutrient availability and hydrologic flushing or through the creation of acid sulfate soils. This study examined the influence of sediment grain size and soil amendments on plant growth, soil and porewater characteristics, and greenhouse gas exchange for three key US salt marsh plants: Spartina alterniflora, Spartina patens, and Salicornia pacifica. We found that bioavailable nitrogen concentrations (measured as extractable NH4+-N) and porewater pH and salinity were found to have an inverse relationship with grain size, while soil redox was more reducing in finer sediments. This suggests that utilizing finer sediments in TLP projects will result in a more reduced environment with higher nutrient availability, while larger grain-sized sediments will be better flushed and oxidized. We further found that grain size had a significant effect on vegetation biomass allocation and rates of gas exchange, although these effects were species-specific. We found that soil amendments (biochar and compost) did not subsidize plant growth but were associated with increases in soil respiration and methane emissions. Biochar amendments were additionally ineffective in ameliorating acid sulfate conditions. This study uncovers complex interactions between sediment type and vegetation, emphasizing limitations of soil amendments. The findings aid restoration project managers in making informed decisions regarding sediment type, target vegetation, and soil amendments for successful TLP projects.

Methods
Coastal marsh plant taxa, including Spartina alterniflora (synonym Sporobolus alterniflorus), Spartina patens (synonym Sporobolus pumilus), and Salicornia pacifica, were obtained from restoration nurseries (Native West Nursery, San Diego, CA & Pinelands Nursery, Columbus, NJ and propagated during the 2018 growing season in a roof-top greenhouse in Philadelphia, PA (39.9539°, -75.1878°) in benthic sediments like those used in thin layer sediment placement projects. Plants were tempered over two weeks to a final salinity of 20‰, using a mixture of water collected from a local marsh at Barnegat Bay, NJ (39.7483°, -74.1931°) and distilled water. Plants were exposed to ambient light conditions under 15% shade cloth, and the greenhouse was outfitted with several fans for temperature moderation. Following a 3x4 factorial design replicated four times, these three plant species were propagated in four types of homogenized sediments of contrasting textures over the course of a growing season (130 days; 22 June – 29 Oct 2018). To replicate the way plant plugs are planted in the field in restoration projects post sediment application,  plugs (5cm x 5cm x 9cm) were obtained from restoration nurseries, and those which were relatively homogenous in the amount of biomass present were planted into larger containers (10cm x 10cm x 24cm).

Plants were exposed to simulated once-daily tides where plants were flooded to a depth of 5 cm for four hours, and the soil was drained to 16.5 cm below the sediment surface for twenty hours. For reference, this inundation time (17%) corresponds to that considered 'regularly flooded.' Sediment texture of soil source material was analyzed for all sediment types using a laser granulometer (LS 13-320, Beckman Coulter, Brea, CA) after pretreatment. Average particle size distributions were post-processed with Gradistat.v8 software, including bin aggregation to texture classes and statistical description.

Photosynthesis, community respiration, net ecosystem exchange, and methane emissions were measured once from 20 to 29 July 2018 using an LGR ultraportable greenhouse gas analyzer (ABB, San José, CA) in a 20L chamber. Measurements of net ecosystem exchange were collected during five-minute incubations in a transparent chamber, and respiration was determined by similar incubations with the chamber covered with black-out material. Photosynthesis was calculated as the sum of respiration and net ecosystem exchange. The Ideal Gas Law (PV = nRT) was used to convert linear changes in CO2 and CH4 concentrations within the chamber during each incubation period to fluxes standardized to the surface area of the plant pots.

Porewater was sampled three times (17 August, 18 September, 24 October 2018), using a Rhizon sampler from a depth range of 0-5 cm. Porewater pH was measured using a benchtop Thermo Orion A111 pH meter, and porewater salinity was measured using a YSI pro30 conductivity and salinity meter. At the end of the growing season, aboveground and belowground biomass of the plants was determined by harvesting, washing, and drying the plant samples at 60 °C to a constant weight. Belowground root material was extracted by washing the container sediment over a 2 mm sieve. Soil redox (eH) was measured at a depth of 5 cm at harvest using a benchtop Oakton oxidation-reduction potential electrode. Sediment samples were collected at harvest and processed for KCl extractable ammonium-nitrogen (NH4+-N), with ammonium concentrations analyzed using the phenate method (EPA Method 350.1; APHA 2012). Saturated hydraulic conductivity (Ksat) was measured (for S1-4) using a Decagon KSAT (Decagon Services, Pullman, WA) using the falling head method.

We conducted short incubation experiments to assess the δ13C of the CO2 emitted from amended soils to help determine whether these emissions could be attributed to plant respiration (~δ 13C=-16 to -12‰ for C4 grasses) vs. remineralization of the carbon in biochar or compost (~δ 13C=-30 to -25‰ associated with C3 plant material). We sampled the headspace of the chamber containing plant at the beginning and end of 30-m incubations using 60 mL luer lock syringes outfitted with a stopcock, where the gas was evacuated and stored in 0.1 L Cali-5-bond gas pillows. Carbon dioxide was subsequently analyzed for δ 13CO2 using a benchtop Picarro (Santa Clara, CA, USA) G2201i isotope and gas concentration analyzer.

Biochar and Compost Treatments

S. alterniflora was grown in coarse sand presumed to have low nutrient levels, with and without soil amendments: softwood biochar (10% v/v), compost (10% v/v), and with both softwood biochar (10% v/v) and compost (10% v/v) to match a paired field study (Raposa et al., 2023). Biochar amendments were a commercially available softwood biochar (Blacklite Pure, Pacific Biochar, Santa Rosa, CA; Pseudotsuga menziesii feedstock). Compost feedstock included manure, livestock products, aged pine bark, coir, and worm castings (Planting Mix Compost Blend, Organic Mechanics Soil Company, Modena, PA). Plants were propagated under identical conditions as the first experiment over a growing season (22 June – 29 Oct 2018), with 16 total units (n=4 per treatment). Plant biomass, CO2 and CH4 emissions, porewater salinity and pH, KCl-extractable NH4+-N, and eH measures were conducted.

Additional samples of S. pacifica were grown in two types of benthic sediments prone to acidification (S6, S7) with and without a 10% (v/v) addition of softwood biochar and without tidal flooding. S. pacifica was propagated for 181 days; 22 June – 19 December 2018. Each treatment was replicated four to six times for a total of 22 experimental units. As described above, plant total biomass, porewater salinity and pH, KCl-extractable NH4+-N, and eH were measured. In addition, porewater total alkalinity was measured (EPA Method 2320 B; APHA 2012).

Data Analysis

All statistical analyses were performed in R ver. 4.0.3 (R Core Team 2023). Correlation matrices were created to examine the dependency of measured variables. The relationship between sediment texture and edaphic parameters (eH, NH4+-N, Ksat, and porewater pH and salinity) was tested using a non-parametric Kruskal-Wallis test with Bonferroni-correction. Significant interactions (p<0.05) were followed by a post hoc non-parametric Dunn’s Multiple Comparison Test.

To determine how sediment texture-related parameters (eH, KCL extractable NH4+-N, soil hydraulic conductivity, and porewater pH and salinity) drive patterns of plant biomass allocation and photosynthesis or CR rates, partial least squares regression (PLSR) was performed due to collinearity of predictors. Each variable was assessed for variable importance in projection (VIP), where VIP scores >1 represent high importance to the regression. Bonferroni-corrected one-way Analysis of Variance (ANOVA) tests were run to determine differences in sediment grain size effects on plant species’ biomass and gas emissions, as well as to test if biochar and compost treatments on low-nutrient sediments would significantly alter sediment eH, NH4+-N, porewater pH, and porewater salinity. In certain cases where normality or homoskedasticity assumptions could not be met, Kruskal-Wallis tests were conducted. Significant effects in the ANOVAs or Kruksal-Wallis tests were followed by a post hoc Tukey’s Honestly Significant Difference test or Dunn’s Multiple Comparison Tests, respectively. To determine the effects of biochar-treatments within non-tidal mesocosms on soil properties (e.g., porewater pH, salinity, eH, NH4+-N, and total alkalinity), Welch’s Two Sample t-tests were run.