Coastal habitat-forming species provide protection and essential habitat for fisheries but their ability to maintain these services are under threat from novel stressors including rising temperatures. Coastal habitat restoration is a powerful tool to help mitigate the loss of habitat-forming species, however, many efforts focus on reintroducing a single, imperiled species instead of incorporating alternatives that are more conducive to current and future conditions. Seagrass restoration has seen mixed success in halting local meadow declines but could begin to specifically utilize generalist seagrasses with climate change-tolerant and opportunistic life history traits including high reproduction rates and rapid growth.

Here, we built on decades of successful eelgrass (Zostera marina) restoration in the Chesapeake Bay by experimentally testing seed-based restoration potential of widgeongrass (Ruppia maritima) – a globally distributed seagrass that can withstand wide ranges of salinities and temperatures. Using field experiments, we evaluated which seeding methods yielded highest widgeongrass survival and growth, tested if seeding widgeongrass adjacent to eelgrass can increase restoration success, and quantified how either seagrass species changes restored bed structure, invertebrate communities, and nitrogen cycling.

We found widgeongrass can be restored via direct seeding in the fall, and that seeding both species maximized total viable restored area. Our pilot restoration area increased by 98% because we seeded widgeongrass in shallow, high temperature waters that are currently unsuitable for eelgrass survival and thus, would remain unseeded via only eelgrass restoration efforts. Restored widgeongrass had higher faunal diversity and double animal abundance per plant biomass than restored eelgrass, whereas restored eelgrass produced three times greater plant biomass per unit area and higher nitrogen recycling in the sediment.

Synthesis and applications: Overall, we provide evidence that supplementing opportunistic, generalist species into habitat restoration is a proactive approach to combat climate change impacts. Specifically, these species can increase trait diversity which, for our study, increased total habitat area restored - a key factor to promote seagrass beds' facilitation cascades, stability, and grass persistence through changing environments. Now, we call for tests to determine if the benefits of restoration with generalist species alone or in conjunction with historically dominant taxa are broadly transferrable to restoration in other marine and terrestrial habitats.

Experimental evaluation of widgeongrass restoration techniques

Our study was conducted in Broad Bay in the Lynnhaven River system in Virginia Beach, VA, USA (36.90418 latitude, -76.03084 longitude) – a heavily human-influenced system (Sisson et al., 2010) where practitioners are actively managing the area to improve water and habitat quality through restoration of multiple habitats including seagrass. Our site was chosen based on a combination of available leasing space and environmental conditions including sediment characteristics adhering to Chesapeake Bay’s long-term, successful eelgrass restoration protocols (Marion and Orth 2010a, 2010b; Moore et al., 2014). Additionally, Lynnhaven River shorelines were historically vegetated with widgeongrass and eelgrass, but meadows experienced declines in the early 2000s and the last observation of seagrass was an eelgrass bed in 2012 (https://www.vims.edu/research/units/programs/sav/access/maps/). However, in May 2020, small patches of natural widgeongrass were observed in Broad Bay indicating an improvement in water quality conditions just 2.5 km away from our study site (Partick et al., 2021). As a reference site, we simultaneously conducted additional plots of each of our experimental treatments (described below) along the southern shoreline of the Goodwin Islands (37.22104 latitude, -76.38951 longitude), part of the Chesapeake Bay National Estuarine Research Reserve (CBNERR-VA) where both widgeongrass and eelgrass are currently established (Moore et al., 2001). This location was selected because it has similar site characteristics to the main study site, and it is known to be viable seagrass habitat. Thus, failure to establish widgeongrass in bare sediment at this reference site would indicate a methodological issue with the seeding technique rather than an environmental issue. The overall study occurred from October 2020 – through April 2022. The experiment occurred from October 2020 to August 2021 and was followed by the Lynnhaven River pilot restoration planted in October 2021 with data collected up until April 2022 (Figure 1; Table S1).

Widgeongrass and eelgrass seeds were collected and stored using similar procedures described in Orth et al., 2003,  Ailstock et al., 2010, and Orth et al., 1994 and specific modifications for our study are in our supplemental appendix. At each site, we established 25, 4 m² experimental plots at least 5 m apart from one another in October 2020 and assigned one of five treatments with five replicates each: control (no seeds), eelgrass seeded in fall, widgeongrass seeded in fall, widgeongrass seeded in spring, or widgeongrass seeds that were given a 48-hour freshwater shock prior to seeding in spring (Figure 1). The 48-hour freshwater shock mimics spring freshets which are a natural cue for widgeongrass germination (Ailstock et al., 2010). Fall treatment plots were seeded in October 2020 and spring treatment plots were seeded in March 2021 for both our experimental and reference site. Our two seeding times coincide with lower Chesapeake Bay eelgrass seeding (fall) and upper Chesapeake Bay widgeongrass seeding or planting efforts (spring) (Chesapeake Bay SAV Restoration Methods: Literature review: https://greenfinstudio.com/wp-content/uploads/2021/11/SAV-Lit-Syn_Final.,pdf).

For each experimental plot minus controls which were left as bare sediment, we broadcast seeded approximately 500 seeds within the 1 m² center, i.e., hand sprinkled seeds in the water column just above the sediment, allowing a 1 m buffer area to measure grass established from our seeds (Figure 1d). To determine the best seeding technique for widgeongrass, we compared multiple seagrass establishment indicators, i.e., initial seedling estimates (April 2021): plot percentage survival through the end of the first growing season (June and July 2021 for eelgrass and widgeongrass, respectively), total plot areal cover, and shoot density. Importantly, widgeongrass’ initial seedlings or shoots are delicate to hand manipulation which hindered our ability to verify visually if singular, isolated shoots in April 2021 were one seedling or two seedlings directly adjacent to one another; to not overestimate, isolated shoots were considered one initial seedling for widgeongrass. Lastly, at Broad Bay, Lynnhaven River in October 2020, we also transplanted wild widgeongrass and eelgrass shoots in five, 1 m² plots per species to aid in identifying site suitability, i.e., transplant survival during the entirety of our study, because there were no natural occurring beds of these species within 1 km. The transplanted shoots were collected from the same donor beds for our seed collection.

Ambient environmental conditions were monitored using an YSI EXO2 multi-parameter water quality sonde station within 50-130 m from our experimental plots collecting water temperature (^oC), turbidity (NTU), salinity, and depth (m) from July 2020 to December 2021 except December 2020 to March 2021(CBNERR-VA VIMS, 2022). In Broad Bay, water temperature data loggers (Onset® HOBO®) were stationed along the border of potential suitable restoration area and along three mean low tide depth values (0.5 m, 1 m, and 1.5 m) to help determine environmental barriers or lethal stressors for grass establishment (Figure 1e). We used 25^○C and 30^○C as our upper water temperature limits for eelgrass as these water temperatures have been shown to be stressful and lethal, respectively, for eelgrass in the lower Chesapeake Bay (Shields, et al., 2019). We also monitored water clarity using Secchi discs as a potential environmental stressor for widgeongrass with no set threshold as widgeongrass is known to need high light availability (Moore et al., 2014; Batuik et al., 2000).

Comparing restored seagrass bed structure, associated invertebrate community, and ecosystem function

Seagrass bed structural differences between our experimental treatments was determined by differences in grass areal coverage, shoot density, and canopy height. Measurements were collected monthly starting March 2021 until our final data collection at time of peak biomass – late June 2021 for eelgrass and late July 2021 for widgeongrass. Because we seeded our spring treatments prior to lower Chesapeake Bay’s spring seedling emergence, all treatments had an equal growing season duration, and we did not need to account for any experimental seeding time difference between our fall and spring treatments. We measured the 4 m² area surrounding our 1m² plot (Figure 1) by subdividing it into 16, 0.25 m² subplots. For each subplot, we recorded shoot density using a haphazardly placed 0.01 m² quadrat and measured five randomized shoot heights to the nearest centimeter. Because we seeded the center 1 m² of the 4 m² plot but collected data over a 4 m² area to observe growth from seeds that were locally dispersed by wave action before settling into the sediment, our final seagrass plot area coverage (m) was calculated as the total number of subplots with at least 5% seagrass cover at peak growth season, multiplied by 0.25 m².  To measure mean shoot density and canopy height where most seeds settled, we took the mean of the four subplots (1 m²) that had the highest seagrass percent cover.

To measure differences in habitat provision between widgeongrass and eelgrass, epi-benthic and -phytic invertebrate fauna were collected from our fall widgeongrass and eelgrass treatments and control to measure differences in community composition, i.e., effective Shannon guild diversity, abundance, and biomass. This sampling was conducted after all habitat characteristic data was collected in June 2021. For each plot, we placed a mesh bag 20 cm in diameter over the seagrass to the benthos and collected the top benthic surface (< 2 cm into the sediment) and aboveground plant material and froze each sample until later processed in the laboratory. All invertebrates were counted and identified to one of the following groups or guilds based on taxonomy as well as specific habitat and/or niche use: sponge (Porifera), free squirt tunicate (Tunicata), mud snail (Ilyanassa obsolete), Bittium spp. snail, mobile amphipod (Amphipoda), sessile amphipod (Caprellidae), worms (Annelida), shrimp (non-Brachyura Decapoda), and blue crabs (Callinectes sapidus). Fauna and plant materials were dried for four days at 60 ^○C and weighed to the nearest 0.1 mg. Faunal community composition metrics were expressed on a per sample basis and calculated as a ratio to the aboveground plant biomass.

To compare differences in primary production between seagrass species, we collected one biomass core from each plot during their respective peak growth seasons. Cores were 15 cm in diameter and went 10 cm into sediment or where no more roots and rhizomes were observed. Core location was representative of the four subplots with the highest seagrass cover. In the laboratory, plant material was separated into above- and below-ground biomass with epiphytic algae scraped off the blades. Seagrass material was dried for four days at 60 ^○C and weighed to the nearest 0.1 mg.

To measure selected microbial nitrogen cycling processes, sediment cores were collected in triplicate from our fall widgeongrass and eelgrass treatments and control in fall October 2020, spring April 2021, and peak growth season which was June 2021 for eelgrass and July 2021 for widgeongrass. At each of the plots, cores were collected within one of the four subplots that had the highest seagrass percent cover. Core collection tubes were fashioned from plastic 50 mL falcon tubes to collect the top 5 cm of sediment and placed on ice until returned to the lab and homogenized. To measure the rates of denitrification (DNF; nitrogen removal) and dissimilatory nitrate reduction to ammonia (DNRA; nitrogen recycling), sediment slurry incubation experiments were conducted as described by Song and Tobias (2011) and Fortin et al., (2021); details are reported in the supplementary appendix.

Seagrass Restoration Pilot 

For our large-scale restoration pilot at Broad Bay, Lynnhaven River, suitable habitat to grow seagrass was constricted by various environmental characteristics observed during the experiment including water temperature, clarity, depth, and physical disturbances within available space for restoration. Overall, our pilot restoration area totaled 7,761 m² with grass species seeded into two monocultures directly adjacent to one another (Figure 1e). In October 2021, we dispersed 90 widgeongrass seeds per m² over a total area of 3,837 m² close to the shoreline with a mean low tide depth from 30-50 cm and 50 eelgrass seeds per m² over a total area of 3,924 m² in areas with a mean low tide depth from 50-150 cm. The shallower zone where we seeded widgeongrass, temperatures above 25 ^○C were observed for 133 days and reached 30 ^○C for total of 40 days whereas our area selected to grow eelgrass encountered above 25 ^○C for 119 days and reached 30 ^○C for total of 30 days.

To compare and test the scalability of our experimental outcomes to the larger spatial scale of our restoration efforts, in April 2022 we collected initial seedling percentages, i.e., the proportion of seedlings to seeds dispersed, a seagrass establishment indicator to compare our pilot restoration to 69 restorations throughout the Chesapeake Bay from 2015-2020 (Orth et al., 2020). Because the restoration pilot overlapped with our experimental area, we excluded our experimental and transplant plots to avoid overestimation of April 2022 initial seedling percentages.

Statistical Analyses 

Data were analyzed with R version 4.2.1 (Team 2022). To describe the effect of our experimental restoration technique treatments on bed structure measurements we fit three linear models allowing treatment to predict final bed area, shoot density, and canopy height. We did not include data from our Goodwin Island site into our analyses because wild widgeongrass and eelgrass runners from adjacent meadows colonized our plots, confounding our experimental treatments. To quantify seagrass species effects on fauna, we fit three linear models allowing treatment to predict effective Shannon guild diversity using the hillR R package (Chao et al., 2014) , abundance, and biomass. To understand changes in primary production and nitrogen cycling, we fit three linear models allowing species identity to predict final total plant biomass as well as sediment nitrogen recycling (DNRA) and removal (DNF). One-way ANOVAs were conducted to assess main effects and Tukey’s test were used for post-hoc multiple comparisons. Normality of the residuals and heterogeneity of variances were checked prior to data analyses using Kolmogorov-Smirnov normality test as well as the Performance R package (Lüdecke et al., 2021).