An evaluation of the efficacy of shell hash for the mitigation of intertidal sediment acidification

Abstract

Our objectives were twofold: (1) to determine whether the addition of shell hash to intertidal sediments would mitigate porewater acidification and (2) whether its effectiveness was dependent on the type of sediment as described by organic matter (OM) and particle grain size (PGS). Field experiments were conducted at two sites within Burrard Inlet, British Columbia; Maplewood Mudflats (MM), high in OM and silt and Whey-ah-Wichen/Cates Park (WAW), low in OM and an equal PGS among very coarse, coarse, fine sand, and silt. Shell hash was added to triplicate treatment plots matched with triplicate controls at each site and porewater pH measured at flood and ebb tide over eight tidal cycles. Sampling occurred during June and July when tidal cycles were at their maximum inundation and exposure. Porewater pH was significantly greater for ebb versus flood tide and also between sites with MM significantly lower (7.59) as compared to WAW (8.03). Although pH was not mitigated by the shell hash, for WAW, variation in pH was reduced as compared to MM, as indicated by coefficients of variation over the 6-week sampling period. We suggest that the application of shell hash to reduce the impact of ocean acidification (OA) on intertidal sediments will be site dependent. The combined processes of eutrophication in sediments with high OM and respiration of infauna, especially at high densities, could act in concert with OA to create an intertidal region unsuitable for bivalve larvae settlement and development.

INTRODUCTION

The Tsleil-Waututh, meaning the “People of the Inlet,” have occupied the lands and waters of Burrard Inlet for at least 3000 years as indicated by their knowledge, stories, and rich concentration of archeological and cultural heritage sites. These sites and artifacts are indicative of the intensity of Tsleil-Waututh's continuous history, use, and management of the area (Morin, 2015). Traditionally, the Tsleil-Waututh economy and diet were derived from local marine resources. Shellfish were a key component of the Tsleil-Waututh diet, especially during periods of low finfish availability (e.g., salmon, herring) and during the winter. Burrard Inlet is within what is now known as the Lower Mainland of coastal British Columbia (BC), Canada. The inlet is surrounded by the city of Vancouver and neighboring municipalities and contains the largest port in Canada. It is heavily influenced by freshwater outflows of the Fraser River and is part of the Salish Sea, a highly productive estuarine ecosystem that encompasses the coastal waters and surrounding watersheds of the Strait of Georgia, the Strait of Juan de Fuca, and Puget Sound in southwestern BC and northern Washington (Pierson, 2011).

Nearly 150 years of cumulative anthropogenic effects in Burrard Inlet have reduced the availability and health of culturally important species (Pierson, 2011). The loss of habitat through rapid urbanization (and associated stormwater and wastewater inputs), shipping (commercial and pleasure craft), and industrial pollution have combined to degrade the health of the Inlet including a significant reduction in species abundance and diversity (Pierson, 2011). The federal government permanently banned bivalve shellfish harvesting in the BC Lower Mainland in 1972 due to contamination and sanitation concerns. Tsleil-Waututh Nation (TWN) has a goal to restore conditions that would once again support the harvest of wild traditional marine resources from the inlet as well as enhance the ecological services provided by shellfish within the inlet (Pierson, 2011).

An emerging threat to the inlet is the impact of ocean acidification (OA). The International Panel on Climate Change (IPCC) predicts atmospheric carbon dioxide (CO₂) could reach 720–1000 ppm by 2100 (IPCC, 2014) leading to a 100%–150% increase in ocean acidity resulting in significant impacts on calcifying organisms (Bednarsek et al., 2012; Orr et al., 2005). Coastal regions such as those within the Salish Sea may be more vulnerable to the impacts of acidification due to a number of co-occurring events (Riche et al., 2014). These include poorly buffered freshwater inputs such as from the Fraser River, and anthropogenic stressors, such as nutrient loads. Nutrients loads, for example, induce increased rates of respiration and eutrophication, which can, in turn, lower pH (Feely et al., 2010; Riche et al., 2014).

Benthic and pelagic marine calcifying organisms are particularly vulnerable to acidification because they require carbonate (CO₃²⁻) to produce their shells (Gazeau et al., 2013; Green et al., 2009). Acidification reduces the availability of CO₃²⁻ and makes it more difficult for organisms to form biogenic CaCO₃ causing early mortality and shell deformation in some species (Timmins-Schiffman et al., 2013). In addition to the direct effects of OA on carbonate chemistry, Wang and Wang (2020) have summarized how OA can compromise the most common behaviors in marine invertebrates. For Mollusca, these include (but are not limited to) settlement, burrowing rates, and feeding rates. The consequences of OA on intertidal invertebrates in terms of actual numbers have been reported by Petraitis and Dudgeon (2020); in the last two decades, five intertidal invertebrate species in the western North Atlantic have declined. Declines in mussels (Mytilus edulis) and the common periwinkle (Littorina littorea) were attributed to warming sea temperatures, whereas declines in the tortoise shell limpet (Testudinalia testudinaslis) and the dog whelk (Nucella lapillus) were attributed to aragonite saturation state and rates of shell calcification. For shellfish populations along the coast of BC, declines of up to 80% have been observed since 1978 (Gazeau et al., 2013).

For juvenile bivalves, once settled, the quality of the porewater at the sediment water interface plays a significant role in survivorship as the external shell remains in direct contact with sediment porewater (Green et al., 2009). For example, within 4 days of exposure to corrosive porewater conditions, Green et al. (2009) showed substantial pitting of the shell ostracum in juveniles. Until the development of a siphon increases burrowing capacity, juveniles are restricted to the upper millimeters of sediment, just below the sediment–water interface where the lowest pH of coastal marine sediments typically occurs (Green et al., 2009).

Silburn et al. (2017) have provided an overview of the factors that can influence the pH at the sediment–water interface across a range of sediment characteristics and seasonal variability. They note that the porewater pH of surficial marine sediments, which can range from 6.5 to 8.2 in the upper surficial sediment (Widdicombe et al., 2011), is influenced by a number of biogeochemical processes such as the cycling of carbon (organic matter [OM]) nutrients (e.g., nitrogen, phosphorous) and trace elements (e.g., iron and silicate). The cycling of these compounds is integrated with key biogeochemical processes such as respiration and photosynthesis which, in turn, can both increase and decrease pH at the sediment–water interface (Silburn et al., 2017). Furthermore, porewater pH can be depth dependent decreasing with depth as the sediment becomes increasingly anaerobic as a result of the respiration of OM. The depth dependence of porewater pH will also be influenced by sediment porosity. Larger sediment grain size allows for the flushing of the sediments with oxic surface waters which buffers and, in turn, increases porewater pH. However, the increasing acidity of surface waters due to OA could serve to overwhelm natural processes leading to acidifying conditions at the sediment–water interface, the region critical for the survival of juvenile bivalves. A possible mitigation strategy then could be natural buffering of surficial sediments through the application of shell hash, a natural source of calcium carbonate.

There have been previous studies that have demonstrated that the addition of shell hash to offset sediment acidity has been successful in aiding the recruitment and survival of juvenile clams (Greiner et al., 2018; Jackley et al., 2016). In their study on the habitat effects of macrophytes and shell on carbonate chemistry and juvenile clam recruitment, survival and growth, Greiner et al. (2018) found that porewater pH was higher in the presence of shell hash. This led the authors to propose that with increasingly corrosive conditions shell hash may help provide refugia under future ocean conditions. Recent studies on ancient clam gardens of the First Nations in coastal BC have also shown that shell hash has a positive effect on bivalve recruitment. Jackley et al. (2016) found that clam recruits (0.5–2 mm in length) tended to be greater in clam gardens as compared to unaltered beaches. The authors attributed this finding to the addition of shell hash by ancient people, which remains on the landscape today. However, Beal et al. (2020) studied the effects of shell hash on the recruits of two infaunal bivalve species and found that neither bivalve species responded positively to the presence of shell hash additions. Beal et al. (2020) recommended that fishery managers should focus attention on mitigating effects due to predators rather than spreading shell hash to buffer intertidal sediments.

Such contradictory results with respect to the ability of shell hash to offset sediment acidification may, in part, be due to the existing geochemical conditions of the sediments. For example, eutrophic sediments with greater nutrient loads and OM content and finer grain size would have greater rates of respiration and thereby lower pH as compared to sediments low in OM and comprised of coarse sediment. Natural sediment processes could therefore mask any potential mitigative effects of added shell hash.

Our objective was therefore to determine whether the addition of shell hash could indeed offset acidification of the upper layers of intertidal sediments and to assess whether this offset was dependent on sediment physiochemical characteristics, specifically grain size and OM. To meet this objective, we compared two intertidal regions with distinct sediment features: (1) Maplewood Mudflats (MM), a protected muddy-silt beach and (2) Whey-ah-Wichen (WAW), a cobble-sand exposed beach, both located within Burrard Inlet, BC. It is our hope that such a comparison will add some insight into the efficacy of shell hash as a means to mitigate sediment acidification.

METHODS

Study site

This study was conducted within the shared ancestral and unceded territories of the səl̓ilwətaɁɬ (Tsleil-Waututh; TWN), xʷməθkʷəy̓əm (Musqueam), and Sḵwx̱wú7mesh (Squamish) Coast Salish Nations. The investigation was informed by elements of Indigenous knowledge and traditional management techniques with the intent of restoring or sustaining traditional food sources and ecosystem services under changing coastal conditions in a culturally appropriate way.

A complete description of the Inlet, as well as an expression of TWN's sacred and legal stewardship obligations, can be found in TWN (2015, 2017). The inlet lacks a sill at the seaward edge and is surrounded by mountains on the north shore and relatively flat lands on the south shore. The shoreline and upland areas have been extensively modified through port, industrial, and urban development. The inlet has six distinct basins; one of the basins, Indian Arm, is a true fjord, while the other basins are relatively shallow. From 1954 to 1974, the pH of the Inner Harbor remained stable at 7.8–8.1. Recent data indicate that pH has become increasingly variable and declined to 7.3–7.9 (Marliave et al., 2011; Nijman & Swain, 2001) which has the potential to significantly affect shellfish. Bivalve species native to this study area include littleneck clam (Protothaca staminea), butter clam (Saxidomus giganteus), bent-nosed clam (Macoma nasuta), horse clam (Tresus capax), heart cockle (Clinocardium nattallii), and Olympia oyster (Ostrea lurida). Non-native species include the Manila clam (Venerupis philippinarum), varnish clam (Nuttalia obscurata), soft shell clam (Mya arenaria), Pacific oyster (Crassostrea gigas), and blue mussel (Mytilus edulus).

The two sites selected for study were located within the Central Harbor of the Inlet: MM and WAW (Figure 1a,b). Both sites are significant to TWN and are in close proximity to the TWN village. MM (Figure 1c) is an intertidal mudflat and salt marsh that remains critical to the Burrard Inlet ecosystem, despite legacy impacts of industrial use and development. MM has always been a critical TWN traditional harvesting site, and is now a protected conservation area. WAW (Figure 1d) is one of the largest and most significant ancient village sites of TWN, and is now a public park which operates under a cultural co-management agreement between TWN and the District of North Vancouver. MM is predominantly mud and silt (Figure 1c), whereas by contrast WAW is sand to cobble (Figure 1d).

FIGURE 1

(a) Location of the study sites within Burrard Inlet, Vancouver, British Columbia. (b) Location of the two sites, Maplewood Mudflats (MM) and Whey-ah-Wichen (WAW) within the central harbor of Burrard Inlet as indicated by two blue dots. (c) MM showing high organic matter sediments. (d) WAW showing cobble to sand sediments

Sediment sampling of biotic and abiotic attributes

Bivalve community

A shellfish survey was designed following Gillespie and Kronlund (1999). At each of the two sites, transects were placed at regular intervals to cover the maximum available intertidal height of shellfish habitat, which was 1280 m² at WAW and 225 m² at MM. Quadrats (0.25 m²), 36 at WAW and 24 and MM, were randomly placed along each of the transects and sediment dug and recovered to a maximum depth of 30 cm. Sediment was placed into a 1-mm metal sieve washed with sea water and all bivalves removed. At each site, bivalves were identified, counted, and wet weight recorded. All biota were returned to the sediment post-sampling.

Particle grain size and OM

Sediment cores were collected from each triplicate plot per sampling week. A 10-cm long polyvinylchloride (PVC) pipe (4.5 cm diameter) was inserted into the sediment and manually plugged at the bottom end. Cores were divided into 0–3, 3–6, and 6–10 cm segments, placed into separately labeled Ziploc bags, stored in a cooler and then transferred to a freezer until analysis. Particle grain size (PGS) and OM contents were used to describe site sediment characteristics. OM was determined through loss on ignition, the difference in dry weight before and after igniting 1 g of dried sediment (60°C for 48 h) at 450°C for 5–10 h. PGS was determined by sieving 100 g of each sediment core segment through four mesh sizes: 1.0, 0.5, 0.25, and 0.064 mm, very coarse sand, coarse sand, fine sand, and silt, respectively (Wentworth size classification). Sediment from each sieve was collected and weighed (wet weight) separately, dried at 60°C for 48 h and weighed again to obtain particle size distribution on a dry weight basis. Both OM and PGS are expressed as a % dry weight.

Interstitial porewater pH

Porewater collection targeted extreme interstitial water quality conditions by capturing the large ebb and flood tides that occur during June and July. Sampling immediately after ebb tide captured conditions resulting from metabolic processes occurring during tidal inundation. Sampling as flood tides covered the plots captured conditions occurring within the sediment during low-tide exposure. Triplicate treatment (i.e., shell hash) and triplicate control (i.e., no shell hash) plots (3 m × 3 m) were established at each site, for six plots per site. A combination of oyster shell hash and naturally occurring, weathered shell hash was used for the treatment plots. Weathered shell hash was collected from the upper intertidal region of WAW, with permission from TWN. Equal volumes of oyster and weathered shell hash for a total of 2.26 kg/m² (based on Green et al., 2009) were raked into the surface sediment to a depth of 5 cm of treatment plots.

To measure porewater pH (total scale), an interstitial water sampler (IWS) (Bendell et al., 2014) was inserted in the center of each plot. The IWS was constructed of PVC pipe with holes drilled on all sides to allow for porewaters to flood into the PVC pipe. To capture porewater in the upper most sediment layers (from 0 to 6 cm), and to ensure that enough sample was extracted to allow for pH determination, each IWS had a sampling portal at 3 and 6 cm. This allowed for the extraction of replicate samples of ca. 10 ml of porewater from each IWS. Following extraction, pH and temperature were measured within 10 min with an Accumet AP85 which was calibrated daily. A maximum of 12 (3 control, 3 treatment × 2 replicate samples/IWS) porewater samples were taken at each site for each tidal period. Eight flood and ebb tides were captured for each site (Figure 3a,b with sample size provided for each tide cycle over each whisker box).

Statistical analysis

All data analyses were performed using Sigma Plot.

Bivalve community

Species richness was measured as the number of species found at each site.

Sediment attributes: PGS and OM

Percentage data were arcsine transformed after Sokal and Rohlf (1981). A two-way ANOVA within each site with core depth (0–3, 3–6, and 6–10 cm) and sieve size (0.063, 0.025, 0.5, and 1.00 mm, silt, fine sand, coarse sand, and very coarse sand, respectively) was first applied to determine whether PGS was dependent on the two factors. For both sites, PGS size was independent of core depth (F = 0.138; p = 0.872 and F = 0.00 and p = 1 for MM and WAW, respectively); hence, core depth was pooled and a one-way ANOVA on PGS within site was applied. Within site, PGS was dependent on sieve size (F = 155.47, p < 0.001 and F = 123.34, p < 0.001 for MM and WAW, respectively) (Table 1). For MM, PGS fine and coarse sand were not statistically different (Table 1), with very coarse sand and silt comprising the main sediment components (Table 1). By contrast, for WAW, PGS was more equally distributed among the four sediment components with very coarse sand > the fine sand > silt > coarse sand (Table 1). To compare sediment attributes between the two unique sites, a simple student's t test was applied to transformed data (Table 2). Organic matter and % silt were significantly greater at MM as compared to WAW. By contrast, coarse and fine sand was greater at WAW as compared to MM (see Table 2 for t and p values). The Holm-Sidak method was applied for all comparison procedures with an overall significance level = 0.05.

TABLE 1. Differences in percentage of particle grain size (PGS) and source of variation within sites in Burrard Inlet, British Columbia: Maplewood Mudflats (MM) and Whey-ah-Wichen/Cates Park (WAW)

Site, PGS, and source of variation	N	Mean	SE	df	SS	MS	F	p
MM
Very coarse sand	67	55.1	2.64
Coarse sand	67	6.9	0.65
Fine sand	67	8.46	0.88
Silt	67	29.5	2.21
Source of variation
Between groups				3	102301.8	34100.6	155.5	<0.001
Residual				264	57904.6	219.3
Total				267	160206.4
WAW
Very coarse sand	26	40.4	1.6
Coarse sand	26	17.6	0.73
Fine sand	25	31.42	1.08
Silt	24	12.7	0.95
Source of variation
Between groups				3	12180.95	4060.3	123.34	<0.001
Residual				97	3193.12	32.9
Total				100	15374.1

TABLE 2. Differences in particle size distribution between the sites, Maplewood Mudflats (MM) and Whey-ah-Wichen (WAW): percentage of organic matter and percentage of particle grain size (mm) (mean ± SE)

Site	N	Organic matter	Very coarse sand (1 mm)	Coarse sand (0.5 mm)	Fine sand (0.25 mm)	Silt (<0.064 mm)
MM	66	3.3 ± 0.27	55.0 ± 2.6	6.9 ± 0.64	8.4 ± 0.8	29.5 ± 2.2
WAW	26	2.4 ± 0.24	40.0 ± 1.6	17.7 ± 0.73	31.4 ± 1.1	12.7 ± 1.0
t/p		1.97/0.05	0.96/0.34	−9.65/0.001	−15.85/0.001	1.94/0.05

• Note: N is sample size (values sum to greater than 100% due to rounding to one decimal place). Results of the student's t test, t and p are given within the last row of the table. Organic matter and % silt are significantly greater at MM as compared to WAW. By contrast, coarse and fine sand are greater at WAW as compared to MM.

Interstitial porewater pH

Table 3 presents a summary of the data for each site, tide, and treatment. Data were tested for normality (Shapiro–Wilk) and equal variances prior to the application of the ANOVA tests. Both assumptions of the ANOVA were generally met, so data were not transformed for analysis. A two-way ANOVA with treatment (addition of shell hash vs. controls with no shell hash) and tide was first applied to interstitial porewater pH within each site. Porewater pH was dependent on tide but not treatment (Table 4). Values for control and treatments were pooled and a two-way ANOVA with site and tide as the two factors was applied. pH was both site and tide dependent (Table 5). Coefficients of variation (a measure of relative variability, ratio of the standard deviation to the mean × 100) were applied to compare the relative variation in ebb and flood tides between the two sites MM and WAW. The Holm-Sidak method was applied for all comparison procedures with an overall significance level = 0.05.

TABLE 3. Summary of mean interstitial porewater pH sample size (N) and SE for control (3) and treatment (3) plots for the two sites, Whey-ah-Wichen (WAW) and Maplewood Mudflats (MM)

Site	Treatment	Tide	N	pH	SE
WAW	Control	Flood	43	8.16	0.062
WAW	Treatment	Flood	44	8.16	0.033
WAW	Control	Ebb	37	7.8	0.062
WAW	Treatment	Ebb	40	7.93	0.047
MM	Control	Flood	46	7.69	0.032
MM	Treatment	Flood	48	7.66	0.033
MM	Control	Ebb	32	7.53	0.043
MM	Treatment	Ebb	36	7.42	0.043

• Note: Sample loss at the time of sampling (e.g., tides flooding or ebbing too quickly for sample collection) resulted in different N's for each overall sampling period.

TABLE 4. Within site interstitial porewater pH; treatment and tide for Whey-ah-Wichen (WAW) and Maplewood Mudflats (MM). Treatment (addition of shell hash) was not significantly different from controls (no shell hash added)

Site, pH, and source of variation	Mean	SE	df	SS	MS	F	p
WAW
pH flood control	8.18	0.048
pH ebb control	7.82	0.052
pH flood treatment	8.16	0.05
pH ebb treatment	7.93	0.05
Source of variation
Treatment			1	0.09	0.09	0.88	0.347
Tide			1	3.56	3.55	35.51	<0.001
Treatment × Tide			1	0.02	0.02	2.02	0.16
Residual			159	15.92	0.1
Total			162	19.68	0.12
MM
pH flood control	7.69	0.03
pH ebb control	7.54	0.04
pH flood treatment	7.66	0.03
pH ebb treatment	7.42	0.04
Source of variation
Treatment			1	0.21	0.21	3.78	0.053
Tide			1	1.58	1.58	28.19	<0.001
Treatment × Tide			1	0.09	0.09	1.67	0.2
Residual			156	8.75	0.06
Total			159	10.67	0.07

• Note: Treatment (addition of shell hash) was not significantly different from controls (no shell hash added). Interstitial porewater pH was dependent on tide with ebb tide being significantly less than flood tide. As intertidal porewater chemistry is reset at each tide, each sampling effort is an independent measure of porewater pH.

TABLE 5. Interstitial porewater pH and source of variation for Whey-ah-Wichen (WAW) and Maplewood Mudflats (MM) at ebb and flood tides

Site, pH, and source of variation	Mean	SE	df	SS	MS	F	p
WAW pH ebb	7.86	0.033
WAW pH flood	8.16	0.031
MM pH ebb	7.47	0.035
WAW pH flood	7.66	0.03
Source of variation
Site			1	15.74	15.74	184.82	<0.001
Tide			1	4.96	4.96	58.19	<0.001
Site × Tide			1	0.132	0.132	1.55	0.214
Residual			320	27.25	0.08
Total			323	47.79	0.14

• Note: Both factors are significantly different from each other with WAW having a great porewater pH than MM and at both sites, flood tide being greater in pH as compared to ebb sites.

RESULTS

Bivalve community

Six species of bivalves were collected from each site and included the following: N. obscurata (varnish clam), C. nattallii (cockle), S. giganteus (butter clam) M. Arenaria, (softshell clam), V. philippinarum (Manila clam), M. nasuta (bentnose macoma clam), and P. staminea (littleneck clam). The number of bivalve species recovered from the two sites was the same (six at each). The bivalve biomass sampled at WAW was 1.27 kg/m² with the majority of the biomass comprised of the varnish clam and the cockle (Figure 2). By contrast, the biomass recovered at MM was 0.23 kg/m² with the most prevalent species observed being softshell and Manila clams (Figure 2).

FIGURE 2

Bivalve community composition at Whey-ah-Wichen and Maplewood Mudflats

Sediment characteristics

Sediments of the two sampling sites were distinct based on OM and PGS. MM had significantly greater amounts of OM as compared to WAW (Table 2) Particle size distribution was also site dependent; MM was comprised of a PGS primarily of very coarse sand and silt (85%), whereas WAW was more evenly distributed among the four size classes with very coarse and fine sand comprising the majority (71%) of the sediment (Table 2). Greater amounts of silt found at MM as compared to WAW is consistent with also greater amounts of OM.

Interstitial porewater pH

Intertidal pH was both tide and site dependent (Table 3, Figure 3a–d). Flood tide was greater in pH as compared to ebb tide and intertidal sediments of WAW were greater in pH as compared to MM The addition of shell hash to the intertidal sediments did not influence porewater pH. However, it did reduce the amount of variation in intertidal pH at WAW, although not at MM as indicated by coefficients of variation (Figure 4).

FIGURE 3

Boxplots of pH versus ebb and flood tides for Maplewood Mudflats (MM) and Whey-ah-Wichen (WAW). (a) and (c) are each individual sampling event, (b) and (d) are all sampling events combined. Numbers refer to the number of samples for each flood and ebb tide. Boxplots are median (line across box), lower and upper quartile (box), minimum and maximum values (error bar) and outliers (circle). Sample size provided above each boxplot

FIGURE 4

Coefficient of variation for porewater pH for each site, and control and treatments (with shell hash), for flood and ebb tide

DISCUSSION

Two key observations were made in this study under field conditions: (1) significant differences in intertidal pH occurred between the two sites and (2) significance differences in pH between ebb and flood tide occurred at both sites. Also observed was that the addition of shell hash to intertidal sediments at WAW reduced the variation in porewater pH for both flood and ebb tides as compared to MM where variation remained the same despite the addition of shell hash. These combined observations support others (Jackley et al., 2016) that shell hash could be useful for the mitigation of sediment acidification as a result of OA. However, the degree of success will be dependent on the site, existing habitat, and treatment.

Porewater pH was significantly lower within sediments of MM as compared to WAW. Sediments at MM were higher in OM and silt content as compared to WAW. Sediments enriched with OM and other nutrients such as ammonium provide an environment for enhanced microbial activity producing high rates of OM remineralization. As noted, this, in turn, leads to increases in CO₂ in porewaters (Drylie et al., 2019) which results in lowering the pH within sediments.

We found that under field conditions, the addition of hash did not directly buffer porewater pH at either MM or WAW. However, at WAW, shell hash did reduce the variability in porewater pH. At this site, coefficients of variation for porewater pH were 5.0% and 2.6% for control and treatment flood tides and 4.9% and 3.6% for control and treatment ebb tides. No reduced variability was found for MM porewater pH. Sediments at WAW were comprised of an even distribution of PGS of very course, course, fine sand, and silt. By contrast, MM had greater amounts of OM and silt. The even distribution of PGS at WAW, that is, a site-specific sediment characteristic, may have facilitated the flow of seawater through the sediments, allowing for greater contact with the shell hash, hence providing a buffering effect on porewater pH values.

In contrast to the findings of Jackley et al. (2016) and Greiner et al. (2018), Beal et al. (2020) have reported that the addition of weathered shell hash to sediments within Caso Bay, ME, to offset corrosive conditions, had no positive effect on bivalves as measured by changes in density and size over a 3-year period. Moreover, the addition of hash had no effect on porewater pH. These apparently contradictory findings provide a good example, as demonstrated in our study, of the site-specific nature of the efficacy of shell hash to mitigate low intertidal pH. Although sediment characteristics (i.e., % OM, grain size distribution) were not provided in Beal et al. (2020), it is noted in their methods that sediment grain size varied between medium to very fine sand and were comprised of soft-bottom mudflats. Sediment characteristics for Casco Bay have been reported in Ramboll Environ. (2017) and indicate that the sediments of the bay are dominated by silt and clay (62 ± 30%) and sand (35 ± 28%) and samples rarely include larger grain sizes, such as pebbles and shells (1.4 ± 6%) and gravel (1.2 ± 4%). Furthermore, sediments average 3.1 + 2% total organic matter throughout the bay. These sediment geochemical characteristics are similar to our MM site, where the addition of shell hash also did not have an effect on porewater pH. As in our findings, silty sediments high in OM are likely to be the least amenable to mitigation by the addition of shell hash. Sites, such as WAW which are low in OM and are more porous in nature, may be responsive to shell hash treatments.

In addition to its role in offsetting corrosive conditions within surficial sediments, shell hash treatments have many other known benefits. Applying shell hash to beaches can concentrate environmental settlement cues, provide a substrate for juvenile attachment, enhance topographic relief and habitat complexity, provide refuge from predation, and provide insulating effects (Dumbauld et al., 2009; Thompson, 1995). Indeed, supplementing coastal sediment with shell material is not a novel concept. First Nation communities created conditions to recruit naturally occurring shell hash (e.g., barnacle and mussel shell fragments), through the construction of clam gardens, or discarded clam shells from harvests on managed shellfish beds (Deur et al., 2015; Lepofsky et al., 2015) and commercial shellfish growers have used gravel and crushed oyster shell to enhance spat recruitment (Thompson, 1995).

Intertidal sediments are influenced by a number of dynamic environmental and biological factors such as weather events (e.g., storms and extreme heat events) tides, benthic respiration, and eutrophication. The interplay among these natural factors all serve to influence the physio-chemical characteristics of the sediments which, in turn, can alter parameters such as porewater pH. This dynamic interplay is most often in balance within the intertidal environments. However, OA can override these natural processes resulting in the lowering of intertidal sediment pH to the point where it is no longer hospitable for those organisms that depend on this habitat for some portion of their life cycle. Our study, therefore, investigated the possibility of using a natural buffering agent, shell hash, to offset or at least slow down the negative impacts of OA on these ecosystems.

Our findings indicate that site-specific sediment characteristics need to be fully understood to assess the suitability of shell hash as management options of mitigating corrosive conditions in intertidal sediments. Sediments high in OM coupled with benthic respiration during ebb tides, combined with surface water of a pH of less than 7.6, will work together to make the intertidal region inhospitable for larvae settlement and development and the combination of these three factors will need to be considered when assessing a site for mitigation. By contrast, sediments with greater porosity and hence faster water flow which would allow for more contact of the shell hash with the overlying acidic seawater could, in turn, provide a greater buffering effect and therefore may be more amenable to mitigation.

The timing of our field experiments should also be considered. We applied shell hash to sediments and measured the response of porewater pH during the most extreme tidal events of the year. During this time of the year, the intertidal region is exposed and inundated for the longest (16 h) and shortest (8 h) periods of time, respectively. Microbial activities that would lower porewater pH and the period of time that the sediments would have been infused with seawater would be at the highest and lowest of the year. These extreme conditions could have dampened the effects that the shell hash may have had on porewater pH. Further studies that include the length of time that sediments are exposed and inundated (e.g., the spring and fall equinox as compared to the winter and summer solstice) would be useful in assessing if the addition of shell hash is indeed a viable option to offset sediment acidification.

ACKNOWLEDGMENTS

This research was conducted within the unceded traditional territories of Musqueam Indian Band, Squamish Nation, and the Tsleil-Waututh Nation. The authors gratefully acknowledge the following: Tsleil-Waututh Nation leadership, staff and community members, and supervisory committee members, Dr. Audrey Dallimore and Dr. Leslie King. We would like to acknowledge the Natural Science and Engineering Research Council of Canada (NSERC) for the funding opportunity. This study was first published as part of the requirements for a MSc degree from Royal Roads University; B. A. Doyle (2017). Investigating a mitigation strategy for acidic sediment conditions in support of First Nation Bivalve restoration initiatives: the effect of shell hash on porewater pH and carbonate saturation states in Burrard Inlet, British Columbia, Canada. http://hdl.handle.net/10170/1042 (B. A. Doyle, 2017).

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Open Research

DATA AVAILABILITY STATEMENT

Data (Doyle & Bendell, 2022) are available from Summit: http://summit.sfu.ca/item/21722.