We constructed a marsh organ to study Sagittaria lancifolia’s responses to varying levels of inundation and nutrient. The marsh organ was situated in a tidally influenced brackish coastal marsh (salinity 0-7 ppt, mean 2.3 ppt) to facilitate both low to no salinity and tidal inundation. We measured a variety of vegetation traits once or twice monthly, including leaf count, longest green leaf length, leaf width at widest point, stem width, and presence or absence of reproductive structures (flowers) during the growing season. End-of-season plants were collected to quantify live above- and belowground biomass. The temporal measurements of the vegetation morphology and end of season biomass measurements provide complimentary data to comprehensively evaluate how the species responded to inundation and nutrient. This contrasts to most other marsh organ experiments that focused on end-of-season biomass measurements only.
Study Site
The marsh organ experimental site was located in the eastern channel of the lower Pascagoula River, Mississippi, US (Figure 1). An ideal site possessed both low salinity (to support the habitat requirements for the desired study species) and tidal influence (to provide variation in inundation times for marsh organ levels). This selected site just south of Interstate10 has a small salinity range that is fresher than the mouth of the river (0-7 ppt as compared to 10-25 ppt) to support growth of Sagittaria lancifolia. The organ was constructed and placed at N 30.434°, W -88.559° in the eastern channel (Figure 2).
The marsh platform at the site, while small, was large enough to support the dimensions of the organ (Figure 2). The low-marsh edge contains mixtures of Spartina alterniflora (10%), Schoenoplectus tabernaemontani (5%), and Sagittaria lancifolia (85%). Further upslope, the mid-marsh is primarily filled with Juncus roemerianus. Old, dead Taxodium disctichum and Pinus sp. spot the high marsh, which likely perished during saltwater intrusions from previous large storm events, such as Hurricane Katrina. A pair of HOBO water logger (HOBO U20L-04, Onset Computer Corporation) were used to measure the water levels since November 2019 with one sitting on the water bottom and the second being suspended above the high water to record compensational atmospheric pressure. Data obtained were used to aid in the design of the marsh organ levels and to incorporate percent inundation time for the duration of the organ experiment.
Marsh Organ Design
The marsh organ design includes six different elevation levels (rows) with 8 replicate pipes in each row (Figure 3). Each row had an estimated inundation level of 100%, 80%, 60%, 40%, 20%, and 5% based on the recorded water levels from July to November 2019 using the HOBO logger on site and a nearby NOAA tidal gauge (https://tidesandcurrents.noaa.gov/ports/ports.html?id=8741533), with consideration of precipitation, storm surges, and seasonal changes of water levels.
Marsh organ construction was completed in July 2020. We used 15 cm (6 in) diameter PVC pipes at a standard length of 61 cm (24 in). We screwed pipes together according to their respective estimated heights in a matrix in the field. Rows one through four were pushed into the sediment, and we capped the bottoms of the two highest rows with nylon meshing to contain the sediment but allow natural vertical water flow for the higher rows (rows five and six). A wooden frame was constructed around the pipe matrix for stability against river flow, tides, and storm surges. We added and packed local sediment to each pipe to ensure plants would grow at the top opening of the pipe and not subside into the pipe. We then collected Sagittaria lancifolia individuals from locations nearby and transplanted one plant into each of the PVC pipes of the marsh organ in July 2020. We also considered settling of sediment and added extra sediment to the tops of the pipes to account for any gradual compaction.
We measured plant traits (initial condition) at the time of transplanting (July 14, 2020), and every two to four weeks afterwards. In more detail, we visited the marsh organ on July 21st, July 31st, August 17th, September 4th, September 30th, October 14th, and November 11th in 2020. We let the plants acclimate to the new environment for one and a half months before beginning nitrogen addition treatment (September 4). In each row, four randomly chosen plants received repeated nitrogen supplements and four did not. We buried 45.5 grams of hydrated ammonium calcium nitrate double salt (15.5% of NO3-N and 1.0% of NH4-N) at a 5-cm depth in the sediment within each PVC pipe. This addition yields 7.06 g N per PVC pipe with 6.60 g NO3-N (362.0 g/m2) and 0.46 g NH4-N (25.2 g/m2). The amount of added NH4-N is similar to the nitrogen applying ammonium chloride solution in one previous marsh organ study (25 g/m2, Langley et al., 2013) but the total nitrogen is about 15 times that study. As larger than 99% of inorganic nitrogen in the marsh porewater is in the format of ammonium, this resulted in a trial to examine the effects of similar ammonia addition but large amounts of extraneous nitrate nitrogen input on the plants. As we are particularly interested in studying the freshwater diversions on salt marsh vegetation, the elevated nitrate nitrogen is relevant as it dominates the inorganic nitrogen in river flows. We began adding supplements on September 4th and at each of the two subsequent visits (September 20 and October 14) until the plants were harvested on November 11, 2020.
Monitoring and Biomass Collection
We conducted morphological measurements every two to four weeks, including leaf counts, longest green leaf length, leaf width at widest point, stem width, and presence-absence of reproductive structures (flowers). We ignored leaves that were fully senesced at the time of measurement in the leaf counts. Since each stem of S. lancifolia terminates in only one leaf, leaf count is equal to stem count. At the conclusion of one growing season, we sampled half of the plants by removing above- and below-ground biomass from a random selection of control and nutrient-added pipes. Biomass was separated into live and dead material and dried in an oven at 70°C for three days. To obtain below-ground biomass measurements after harvest, we cleaned and sorted collected samples by hand. Samples were filtered through a 1 mm mesh to remove inorganic matter under running water. We placed washed roots back into their respective bags and stored them at 4°C.
Washed roots were separated into live and dead biomass. Color, turgidity, and buoyancy were all characteristics we used to identify live biomass. Biomass that sank into the water, was dull in coloration, or was soft (loss of turgidity) was considered dead. We placed the roots into pre-weighed aluminum trays. We then dried the roots in an oven at 70°C (~ 3 to 5 days) until constant weight (Wu et al., 2020). Samples were weighed again to obtain dry weight.
Statistical Analyses
To derive effects of inundation and nutrient addition on plant morphological traits, we applied (generalized) linear mixed-effects models with percent of inundation time, presence or absence of nutrient addition, initial condition, and days since planting as fixed factors, and individual plant as the random factor. To account for the nonlinear effect of inundation on belowground, aboveground, and below-above ratios of harvested live biomass at the end of growing season, we also applied a quadratic term of percent of inundation as one of the fixed factors. We developed separate mixed-effects models for each of the morphological traits, including leaf counts (Poisson distribution), longest green leaf length, width at widest point, stem width, and presence-absence of reproductive structures (binomial distribution). We evaluated the models with different combinations of fixed factors and selected the best model based on Akaike Information Criterion (AIC).