Perched marshes in Lunenburg County differ in exposure to prevailing waves (from most to least exposed: Crescent Beach, Blue Rocks, Feltzen South). We sampled the middle marsh, recording replicates’ dried masses. In picking stage 1 (PS1), the first 400 foraminifera, whether agglutinated or calcareous-walled, were picked and identified. From these data we calculated the Shannon Function H and compared the between-locale signals of benthic foraminifera using a relative-dominance index, ACW, for agglutinated and calcareous-walled foraminifera. Crescent Beach was a distinct locale with much medium-grained quartz sand and abundant allochthonous calcareous foraminifera (hence ACW < 1). The other sites’ ACWs ranged from ∼50–100. There was a significant negative ACW:residue mass correlation. In picking stage 2 (PS2), the remaining agglutinated foraminifera were picked. The Feltzen South perched marsh yielded only Miliammina fusca, perhaps reflecting high %N and %C. Meanwhile, Trochammina inflata was abundant at Crescent Beach and Blue Rocks.

Studies of modern benthic foraminiferal distributions in Nova Scotia have concentrated on large marshes that cover a wide altitudinal range, between about mean sea level and highest high tide (Deonarine, 1979; Price, 1980; Scott & Medioli, 1980a, 1980b). These studies, inspired by earlier work by Scott (1976) in southern California, were used to develop training sets for deciphering Holocene sea-level change (Scott et al., 1995).

Scott & Medioli (1978) distinguished three benthic foraminiferal zones in the marsh at Chezzetcook Inlet, SE Nova Scotia. These were, from top to bottom: High Marsh Zone IA, with monospecific Jadammina macrescens (Brady) (=Entzia macrescens), at 95–100 cm above mean sea level; Middle to High Marsh Zone IB, with Tiphotrocha comprimata (Cushman and Brönnimann) and E. macrescens, at 75–95 cm above mean sea level; Low to Middle Marsh Zone IIA, with Trochammina inflata (Montagu) and Miliammina fusca (Brady), at 65–75 cm above mean sea level; and Low Marsh Zone IIB, with M. fusca, Ammotium salsum (Cushman and Brönnimann), and Cribroelphidium excavatum (Terquem), at -20–65 cm above mean sea level.

In temperate floral terms, the middle marsh is a floral zone of monospecific Spartina patens (salt-meadow cordgrass) that at Chezzetcook Inlet ranges from 70–80 cm above mean sea level and encompasses the uppermost part of the benthic foraminiferal Zone IIA and the lowermost part of Zone IB (Scott & Medioli, 1980a, fig. 14). Thus, the S. patens floral zone supported an entirely agglutinated fauna. These alphanumeric foraminiferal zones have been employed worldwide, being detectable despite the presence locally of other species, especially in tropical areas (e.g., Barbosa et al., 2005; Horton & Edwards, 2006; Horton et al., 2003; Scott et al., 1990, 1996).

Porter et al. (2015) documented the relationships between environmental factors and intertidal plant species at sites with a range of tidal magnitudes (<2 to >14 m) throughout Nova Scotia. They found that, although elevation had previously been suggested to drive vegetation types in salt marshes in the region, lateral variations in salinity can differentiate vegetation types at the same elevation. The S. patens association is characterised by high pore-water salinity (20.0 ± 0.17), intermediate elevation, and intermediate inundation times (351.7 ± 9.5 min per flooding event). At Lawrencetown, on Nova Scotia's Atlantic coast, Porter et al. (2015) found monospecific S. patens middle marsh to occupy an altitudinal range of 0.15 m, compared with the 0.10 m of Scott & Medioli (1978), and to live at sites with a mean sediment organic matter concentration of 19.4 ± 4.67%. These monospecific beds of S. patens are easily identifiable in the field. When stressed by waves or tides, the weak stem base in S. patens bends (Silberhorn, 1976). Where this happens to numerous, adjacent individuals, the stems intertwine to give a characteristic swirled pattern, colloquially called cowlicks (Pike, 2018; Roman et al., 1984).

Wilson et al. (2018) examined the total benthic foraminiferal communities in the intertidal LaHave River Estuary, SW Nova Scotia. They sampled where possible from among swirled S. patens beds, so as to constrain altitude. These beds are only patchily developed along the LaHave estuary's banks. Replicates were taken from four sites along the estuary's eastern bank: Miller Point Peace Park, Dayspring, Upper LaHave, and East LaHave (Fig. 1), one replicate being tested for %C and %N. Those washed replicates picked clean of foraminifera yielded only (in rank order of overall abundance) Entzia macrescens, Trochammina inflata, and Miliammina fusca. The mean foraminiferal density at Miller Point Peace Park was significantly different from the other sites, which acted as a group. Despite S. patens’ narrow environmental limits with regard to pore-water salinity and inundation, the foraminiferal assemblage varied along the LaHave River estuary. The most upstream assemblage was dominated by E. macrescens, the most downstream by M. fusca. The difference in the dominant species may have been related to organic matter concentration, which has a wide variance between S. patens beds (Porter et al., 2015), resulting in a lateral transition between Zones I and II of Scott & Medioli (1978). The transformed mean proportions per site of E. macrescens and T. inflata were not significantly correlated with %C or %N, but those of M. fusca were positively correlated with both.

There are along the rocky coastline of SW Nova Scotia small, isolated patches of marsh vegetation that are separated from the open sea by bare bedrock, rather than by marsh supporting lower, more seaward floral zones. One of us (BW) found three such areas, here termed ‘perched marshes’, occupied by swirled, monospecific beds of S. patens. These were sampled to assess whether the sediment in these small, perched marshes contains the tests of intertidal benthic foraminifera, and to compare the benthic foraminiferal associations between them and within the LaHave Estuary.

The perched marshes sampled for this study are at Crescent Beach (44.232°N, 64.385°W), Blue Rocks (44.354°N, 64.234°W), and Feltzen South (44.332°N, 64.285°W) (Figs. 1, 2). The investigated sites are located in Lunenburg County, Nova Scotia, and all were sampled in early October, 2019. The sites differed in exposure (cf. Krumhansl & Scheibling, 2011, table 1), being more or less exposed to, or protected from, waves driven by the prevailing south to south-westerly winds (see Walker et al., 2006).

Crescent Beach is the SSW-facing part of a 2 km-long, 75 m-wide tombolo that connects George Island to the Nova Scotian mainland and, being exposed to the prevailing winds, experiences moderate to high wave energy. The beach south of a metalled road constructed along the tombolo comprises medium-grained quartz sand with abundant allochthonous calcareous foraminifera. The perched marsh sampled here is situated at the eastern end of the beach, from which it is separated by narrow expanses of greywacke. The marsh vegetation is sparse and comprises mostly Spartina alterniflora, with a small patch of swirled S. patens. While the surface beach sand adjacent to the perched marsh is light grey, the sandy sediment in the marsh is pale brown. This difference is presumed to reflect the concentration of fresh organic matter in the 20 m2 perched marsh.

Although the coastline at Blue Rocks generally trends NW–SE, the perched marsh is on a semi-sheltered inlet trending NNE–SSW. The marsh here was small, with swirled S. patens being restricted to a 1 m2 patch shoreward of a larger area of S. alterniflora from which it was separated by bare rock. Both vegetated areas were much obscured by a 30 cm-thick layer of the seaweed Fucus spiralis that had been washed in by Hurricane Dorian early in September, 2019, but had not been removed by subsequent wave or tide action. The sediment here was mostly mud but with some sand and gravel.

Feltzen South was the most protected shore sampled, facing NE. The perched marsh here was the largest sampled, being bordered to seaward by a rock pool and separated from the sea by a 15 m-wide expanse of greywacke. The entire perched marsh comprised swirled S. patens. Various other rocky shores were explored (e.g., the eastern bank of the LaHave Estuary downstream of Lower Lahave, Gaffe Point, Green Bay). However, no instances of perched marshes with swirled S. patens beds were found there.

Four replicates were taken at each of the three perched marshes. All were collected from among the swirled, monospecific S. patens beds using a 15 cm-long, metal, sharpened push core 2 cm in diameter, from which the top 1 cm was saved, giving a sample of 3.14 ml. Tubes of larger diameter were tested but found unable to core the poorly consolidated sand at Crescent Beach. The replicates were each washed over a 106 µm mesh, this size being chosen to remove mud and fine sand and to allow comparison of the assemblages recovered with those reported from the LaHave Estuary by Wilson et al. (2018). The washed material was oven dried at 90°C for 30 minutes. The masses of the dried residues were recorded in grams. A surface scrape of indeterminate volume was collected from the open beach at Crescent Beach, ∼1 m from the western edge of the perched marsh. This was processed in the same manner, but its mass not recorded.

Because the middle marsh foraminiferal community typically consists entirely of agglutinated species, and we found calcareous specimens in some of our replicates, they were picked in two stages. For picking stage 1 (PS1), somewhere between a pinch and the entire residue, depending on the residue volume, was spread evenly across a picking tray. Where possible the first 400 foraminiferal specimens encountered, whether calcareous or agglutinated, were picked, sorted and mounted on a card Plummer slide. Where <400 specimens were present, all were picked. In stage 2 (PS2), the remainder of each residue was spread across a picking tray and all remaining agglutinated foraminifera were picked. The numbers of agglutinated foraminifera from PS1 and PS2 were added to give the total number of agglutinated foraminifera per replicate. The foraminifera were identified to species level using standard references (Parker, 1948; Scott et al., 2001; Dabbous & Scott, 2012).

The Shannon Function H is represented by the following equation:
in which pi is the proportional abundance of the ith species and was calculated using the data from PS1 for those marsh replicates yielding >34 specimens. An agglutinant:calcareous-walled foraminiferal index (ACW) was calculated using the results from PS1 only, using
where Nag is the number of specimens with agglutinated walls and Ncw the number with calcareous (hyaline or porcellaneous) walls. This was calculated for the seven replicates yielding >34 specimens.

The residue masses from the perched marsh replicates ranged from 1.9–9.8 g. An ANOVA showed that the mean mass from at least one raised marsh differed significantly (F3,9 = 21.43, p = 0.0004). Scheffé’s test showed that the mean mass from the Crescent Beach perched marsh (forumla = 7.2 g) was significantly different from and greater than those from the perched marshes at Blue Rocks (forumla = 2.6 g) and Feltzen South (forumla = 2.3 g). The mean residue masses from these last two locales did not differ significantly.

From PS1, 1854 foraminifera (calcareous + agglutinated) were picked from the 12 perched marsh replicates (range = 0–408, forumla = 154.5 per replicate, s.d. = 181.2) and 401 foraminifera from the bare sand sample from Crescent Beach (Table 1). Recovery differed markedly between the individual perched marshes (Crescent Beach marsh, 1586 specimens, forumla = 396.5; Blue Rocks marsh, 232 specimens, forumla = 58; Feltzen South marsh, 36 specimens, forumla = 9). Of the four replicates from Feltzen South, one was barren while two others yielded only single specimens of calcareous-walled Bulimina elongata and Rosalina sp.

Using the data from PS1, we tested for differences among all the locales (perched marshes, Crescent Beach bare sand) since the average numbers of specimens per species, NS, across all replicates were distinct: on the bare beach at Crescent Beach, NS = 11; on the marsh at Crescent Beach, NS = 42; at Blue Rocks, NS = 6; and at Feltzen South, NS = 1. There being replicates with a considerable number of zeroes, we transformed the raw data for the total recovery (calcareous + agglutinated) to the square root of specimen numbers per species per replicate plus unity. We ran a general linear model with the dependent variable of group (locale). Results were F3,148 = 5.832 with p < 0.001, this test having a posterior power of 0.948. Scheffe's test showed that Crescent Beach perched marsh was significantly different from each of the remaining locales with respect to NS, while there was no detectable difference among the remaining three locales of the Crescent Beach bare sand and the Blue Rocks and Feltzen South perched marshes.

Among the 1655 calcareous-walled foraminiferal specimens obtained from the replicates, only 11 were from planktonic species. Ten of these came from the Crescent Beach perched marsh. Of the total perched marsh recovery from PS1, only 199 foraminifera (10.7% of total recovery) had agglutinated walls, these being from five species: Eggerelloides advena (Cushman), E. scaber (Williamson), Lepidodeuterammina ochracea (Williamson), Miliammina fusca (Brady), and Trochammina inflata (Montagu).

Values of H were calculated using the data from PS1 for those eight replicates with ≥34 agglutinated + calcareous specimens (Crescent Beach marsh, 4 replicates; Blue Rocks marsh, 3 replicates; Feltzen South, 1 replicate). They ranged between H = 0.82–1.94. Student's t test showed the mean values of H at Crescent Beach (forumla = 1.81) and Blue Rocks (forumla = 1.23) to differ significantly (t = 2.02, p = 0.05, df = 5). The value of H = 1.35 calculated for the single replicate from Feltzen South was closer to the mean diversity for the Blue Rocks marsh than for the Crescent Beach marsh. Values of H, however, were not significantly correlated with residue mass (r = 0.51, p = 0.20).

The ACW ranged from 0.25–97.44, with all values on the Crescent Beach perched marsh being <1.0 (Fig. 3). For those eight residues with >34 specimens for which ACW was calculated, ACW and residue mass were inversely and significantly correlated (r = –0.765, p = 0.027; Fig. 4). This correlation reflects the dominance of recovery from PS1 of the Crescent Beach perched marsh replicates by the calcareous-walled species Cibicides ex gr. refulgens (42.8–52.0% per replicate) with lesser calcareous-walled Cribroelphidium excavatum (4.5–14.1%) and Cribroelphidium magnum (8.0–19.4%) ( = Elphidium excavatum forma magna Miller et al., 1982). Elsewhere the agglutinated species were dominant. No specimens of Cribroelphidium spp. or Elphidium spp. were found with cytoplasm coloured by sequestered chloroplasts.

During picking stage 2 (PS2), we picked all remaining agglutinated foraminifera from the replicates (Table 2) and summed these with those picked in PS1, giving the total number of agglutinated foraminifera per replicate. The number of agglutinated specimens recovered ranged from 0–109 per replicate, with a total of 397 specimens across all replicates. Although only 17 agglutinated specimens were recovered from Feltzen South, all in one replicate, our ANOVA did not find any difference in the mean recovery from the three locales (F2,9 = 2.142, p = 0.17). However, the species’ distributions were notable. The Feltzen South perched marsh yielded only Miliammina fusca, which was abundant at Blue Rocks (65 specimens) but rare (4 specimens) at Crescent Beach. Meanwhile, the mean number of Trochammina inflata on the perched marshes at Crescent Beach (forumla = 49.75) and Blue Rocks (forumla = 25.75) did not differ significantly (t = 0.88, p = 0.21, df = 6).

Scott & Medioli (1978) found the middle marsh zone of monospecific S. patens at Chezzetcook Inlet (70–80 cm above mean sea level) to yield only agglutinated foraminifera, it encompassing the uppermost Zone IIA (Trochammina inflata and Miliammina fusca) and the lowermost Zone IB (Tiphotrocha comprimata and Entzia macrescens; see also Wilson et al., 2018). Farther east, at Lawrencetown, Porter et al. (2015) found monospecific S. patens to have an altitudinal range of 0.15 m. With the exception of the bare beach sand at Crescent Beach, all our sampled material was obtained from swirled S. patens beds, the swirled appearance resulting from wave action stress. They are concluded to have been taken at about the same altitude relative to mean sea level.

The exposed Crescent Beach perched marsh of this study was significantly different from each of the other locales (Crescent Beach bare sand, Blue Rocks perched marsh, Feltzen South perched marsh), among which there was no detectable difference, although the recovery from the Crescent Beach bare sand sample consisted of calcareous foraminifera only. The replicates from the perched marsh at Crescent Beach were likewise dominated by calcareous species. Cibicides ex gr. refulgens, the dominant species here, lives in open marine water, where it attaches to firm substrates in high energy areas (Murray, 2006). Miller et al. (1982) found abundant C. magnum in the nearshore turbulent zone at Chezzetcook Inlet, Nova Scotia, while Nathan et al. (2014) found C. subarcticum (Cushman), common at Crescent Beach, to occur on the inner shelf in the Gulf of Maine. Many Cribroelphidium spp., which were abundant at this site, when live contain kleptoplasts that colour the cytoplasm, often green (Pillet et al., 2011; Nathan et al., 2014; Jauffrais et al., 2018; Jesus et al., 2022). Cribroelphidium excavatum retains the chloroplasts from diatoms (Correia & Lee, 2000). We did not find any specimens of any Cribroelphidium species with coloured cytoplasm. We conclude that the calcareous association at the Crescent Beach perched marsh is allochthonous, being washed inshore by waves driven by the prevailing winds. Some of this input may have come from the lower marsh Zone IIB within the perched marsh, which supported Spartina alterniflora and may have had a live population of C. excavatum. However, the majority of the calcareous association was of inner neritic aspect. This resulted in low values of ACW (<1.0) at Crescent Beach. Our conclusion regarding shoreward transport of the calcareous component is supported by the close comparison between the calcareous assemblage in the perched marsh replicates and the bare sand sample at this site. The very low abundance of planktonic foraminifera at this site confirms derivation of the allochthonous sand-sized fraction from inner neritic depths (cf. Hayward, 1979; van der Zwaan et al., 1990; de Rijk et al., 1999). The other, more sheltered locales (Blue Rocks, Feltzen South) yielded few allochthonous, calcareous-walled specimens.

The mean residue masses, consisting of sand, lithic fragments and foraminifera, differed between the locales, being greatest on the Crescent Beach perched marsh. Despite the input of allochthonous specimens, however, there was no significant correlation between the Shannon Function H and replicate residue mass for those replicates with ≥34 specimens. There was, however, a significant correlation between replicate residue mass and the ACW, which was lowest at the most exposed perched marsh at Crescent Beach. This suggests that the ACW has potential as a tool for quantifying the degree of exposure on rocky coastlines. It is more quantitative than the more traditionally used Ballantine Exposure Scale (BEI, see Ballantine, 1961), which is determined by comparing the abundance of a selection of seaweeds and calcareous organisms with a diagram (see Wilson & Hayek, 2019). Furthermore, whereas the BEI was developed in Wales and is thus limited to NW Europe, the ACW can be applied to perched marshes on rocky coasts at all temperate latitudes, all of which support a similar community of agglutinated foraminifera (Scott & Medioli, 1986). The ACW differs from the Marine Influence Index (MII) of Jorissen et al. (2022) in that the ACW, here applied to vegetated perched marshes, is based on foraminiferal wall type, not the physico-chemical aspects of estuaries or unvegetated mudflats.

In PS2 of our study, all remaining agglutinated foraminifera were picked from the perched marsh replicates, and their numbers combined with those from PS1. Wilson et al. (2018) examined foraminiferal assemblages in swirled S. patens beds at four sites along the estuary banks. They found the assemblages to change along the estuary's length. The upstream-most site was dominated by Entzia macrescens, with considerably fewer T. inflata and Miliammina fusca, suggestive of a transition upstream to the High Marsh Zone I of Scott & Medioli (1978). The downstream-most site yielded a mix of M. fusca and T. inflata, indicative of the Low to Middle Marsh Zone IIA of Scott & Medioli (1978). Crescent Beach perched marsh, which yielded primarily Trochammina inflata, is situated at the mouth of the LaHave Estuary. The Crescent Beach perched marsh assemblage thus differed from the most downstream assemblage reported by Wilson et al. (2018), confirming that the benthic foraminiferal association accompanying swirled S. patens varies geographically.

The agglutinated perched marsh assemblages furthermore differed from one another. Only M. fusca was recovered from the sheltered marsh at Feltzen South. This distribution might indicate that M. fusca is more common at more sheltered locations. It might also indicate that there were especially high values of %C or %N at this locale (cf. Wilson et al., 2018). This species was abundant also at the moderately sheltered Blue Rocks marsh, where it was admixed with T. inflata (Low to Middle Marsh Zone IIA). However, it was rare at the exposed Crescent Beach marsh. In contrast T. inflata was equally abundant at Blue Rocks and Crescent Beach, but not recovered from Feltzen South, perhaps suggesting an association with more exposed locales.

Studies of intertidal foraminifera around Nova Scotia have typically concentrated on large expanses of marsh from a range of floral zones. However, smaller patches of marsh, whether along estuary banks or on rocky coastlines, can yield data of interest. This study examined samples from expanses of swirled monospecific Spartina patens on perched marshes facing the open sea but differing in their degree of exposure. The most exposed site yielded not only intertidal agglutinants (primarily Trochammina inflata) but also large numbers of allochthonous, presumably inner neritic, calcareous foraminifera. The ratio of agglutinants to calcareous-walled foraminifera was examined using ACW, an agglutinant:calcareous-walled foraminiferal index that might be used as an indicator of coastal exposure. Miliammina fusca was most abundant at the most sheltered site. Comparison with assemblages in the LaHave Estuary suggests that the common occurrence of this species at Feltzen South might reflect either lower wave energy or higher levels of nitrogen or organic carbon at this site.

This work did not receive any funding. Thanks are due to Jacqueline Attong-Wilson for help with fieldwork. We also thank the two anonymous reviewers for their insightful and helpful comments and observations.