Jobos Bay, southeastern Puerto Rico, experiences strong environmental gradients between an historically impacted coastal-plain and oligotrophic Caribbean waters. The coastal zone is dynamic both seasonally and interannually. During 2018–2019, water quality, sediments, and benthic-foraminiferal assemblages were assessed from fore- and back-reef sites off three cays that separate Jobos Bay from Caribbean waters. Temperature and salinity reflected seasonal variations, inorganic nitrogen indicated terrestrial runoff, and sediment texture reflected influence of winds and wave energy. Foraminiferal assemblages in the fore-reef were dominated by Amphistegina while taxa such as Quinqueloculina and Discorbis dominated back-reef sediments. Low test densities reflected the influence of wave energy in predominantly siliciclastic sediments. Interannual differences in sediment textures and assemblages collected during comparable months reflected timing of storm passages. The prevalence of algal symbiont-bearing taxa in fore-reef sites indicated suitable water quality for reef accretion. This study of Jobos Bay foraminiferal assemblages provides baseline data on seasonal and interannual variability.

Background

Tropical coral reefs are among Earth’s most vital marine ecosystems. Such reefs provide a wide array of ecosystem services that contribute to environmental and societal interests (e.g., Moberg & Folke, 1999; Eddy et al., 2021). Despite their importance, reefs are threatened by a wide range of global and local environmental stressors, both natural and anthropogenic (e.g., Beyer et al., 2018; Woodhead et al., 2019). Increasing concentrations of carbon dioxide in the atmosphere are driving ocean acidification and climate change, including rising sea-surface temperatures (e.g., Hoegh-Guldberg et al., 2007, 2018). Terrestrial runoff increases after tropical storms and hurricanes, resulting in nutrient pollution and water-quality changes (e.g., increased water turbidity) on coral reefs (Takesue et al., 2021). Human influences from agriculture, industrial, and coastal development practices contribute to increases in turbidity, nutrients, and other chemical pollution (e.g., Wooldridge & Done, 2009; Lesser, 2021). In the past 50 years, coral-bleaching events associated with warmer sea-surface temperatures have increased from rare occurrences in the 1970s, to two significant events in the Caribbean in 1983 (Jaap, 1985) and 1987 (Lang et al., 1992), to multiple mass bleaching events since the worldwide event in 1997–1998 (e.g., McClanahan, 2022; Shlesinger & van Woesick, 2023).

In Puerto Rico, coral-reef ecosystems directly influence the economies of coastal communities, primarily through tourism and recreational activities. At the same time, coral reefs are threatened by human activities (e.g., Hernández-Delgado & Ortíz-Flores, 2022) that promote terrigenous sediment influx (Ramos-Scharrón et al., 2015; Ramos-Scharrón, 2021), nutrient pollution (Larsen & Webb, 2009), sewage pollution (Hernández-Delgado et al., 2011), non-point pollution sources (Bonkosky et al., 2009), and direct and indirect impacts from recreational activities (Webler & Jakubowski, 2016).

Benthic foraminifers have been used for more than 50 years to provide insight into environmental changes, pollution, and climate change (e.g., Alve, 1995; Frontalini & Coccioni, 2011; Barbosa et al., 2012). They are useful as bioindicators because of their diversity in estuarine, coastal, and marine habitats, combined with their production of agglutinated or calcium carbonate (CaCO3) shells (commonly known as tests). The kinds of shells found in sediments can provide information about past environmental conditions (Barragán Montilla & Sanchez Quiñonez, 2021), types and rates of changes, and present conditions in a habitat (e.g., Oliver et al., 2014; Mathes et al., 2022). A variety of metrics (e.g., species richness and other diversity indices) are commonly used to evaluate foraminiferal assemblages (e.g., Murray, 1973, 2006; Hayek & Buzas, 2010; Gonzales et al., 2022). In addition, several indices specifically using benthic foraminifers to evaluate the health of coastal environments have been developed (e.g., Hallock et al., 2003; Schönfeld et al., 2012; El Kateb et al., 2020; Prazeres et al., 2020).

Two studies (Donnelly, 1993; Oliver et al., 2014) off La Parguera, Puerto Rico, specifically contributed to the development and application of indices using benthic foraminifers in coral reef ecosystems. Hallock et al. (2003) used samples from La Parguera (Donnelly, 1993) and the Florida reef tract (Cockey et al., 1996) to develop a water-quality index known as “Foraminifera in Reef Assessment and Monitoring”, abbreviated as the FoRAM Index or FI. The FI reflects how reef water quality influences the occurrences of benthic foraminifers. Oliver et al. (2014) subsequently compared the FI with other environmental metrics, including coral cover, fish communities, and macrobenthic fauna, for La Parguera waters, finding that reefs exposed to more significant human disturbances had lower FI values (range of 2–3) indicative of deterioration, while those farther away (with less human disturbance) had higher FI values (4), indicating better water quality for coral survival and reef accretion.

Study Site: Jobos Bay National Estuarine Research Reserve

Jobos Bay National Estuarine Research Reserve (JBNERR), offshore from Salinas, Puerto Rico, is part of the U.S. National Estuarine Research Reserve system (Fig. 1A). Jobos Bay, located on the southeast coast, is the second largest estuary in Puerto Rico. The expansion of coastal urban areas around the watershed exerts tremendous ecological pressure on the bay. Illegal coastal construction brings land-based contaminants to the bay, affecting water quality and biota (e.g., Alegría et al., 2016; Martínez-Colón et al., 2021). In addition, JBNERR inner regions continuously receive warm water from a nearby Thermoelectric Power Plant, a predominant source of thermal pollution, runoff, and oil spill contamination. Former agriculture and pharmaceutical industries within Salinas and Guayama municipalities remain among the non-point sources of chemical pollution. As the distance from the shoreline to the reefs is on the order of 1–3 km, gradients from turbid, nutrient-enriched terrestrial runoff to very clear, nutrient-poor Caribbean waters are very strong and shift with weather changes, especially seasonally. Thus, the coral reefs of Jobos Bay are subject to the influence of local stressors.

Seasonality in Puerto Rico, as in many tropical environments, is associated with rainfall. Puerto Rico typically experiences a cool dry season from December through March, followed by a rainy season starting in April. Tropical storm systems are common, especially in August and September, producing heavy rainfall, as well as strong winds, waves, and currents. For example, in 2017, two major storms, hurricanes Irma (Category 4) and María (Category 5) devastated Puerto Rico. Rainfall associated with Hurricane María resulted in as much as 55 cm of rainfall in the upland watershed and 11–22 cm in the immediate vicinity of Jobos Bay (https://www.weather.gov/sju/maria2017).

Coral reefs are vital ecosystems for the protection of Jobos Bay and their management is a priority for JBNERR. Practices to better manage reef ecosystems subject to increasing stressors, including increasing pollution, ocean acidification, and climate change, are needed for effective resource conservation. The JBNERR is an appropriate system to utilize indices based on foraminiferal assemblages, including the FI, as long-term bio-monitoring tools for the reserve. Characterizing the parameters influencing foraminiferal assemblages can be key to understanding which ecosystems can persist in the reserve and more broadly along Caribbean coastlines.

The objectives of this paper were: a) to summarize environmental conditions seasonally and spatially (i.e., fore reef and back reef) at selected sites at JBNEER in 2018 and 2019; b) to collect and evaluate foraminiferal assemblages at those sites; and c) to test the hypothesis that assemblages reflect seasonal changes (rainy vs. dry seasons) in environmental parameters. The original data set is available in Rosa Marín (2022).

Field Sampling

The samples collected at JBNERR were from just offshore Cayo Morillo, Cayo Pájaros, and Cayo Caribe (Fig. 1). At Cayo Morillo and Cayo Pájaros, two transects were sampled off each cay: one each in the fore reef and back reef. At Cayo Caribe, three transects were sampled: two at the back-reef and one at the fore-reef. Overall, each transect included two to three stations between 2–5 m and 10–15 m, for a total of 17 stations among all seven transects (Fig. 1B–D). The goal was to conduct sampling in 2018 and 2019 at intervals representing rainy (August, September) and dry seasons (March, December), though weather conditions prevented fore-reef sampling in March and December 2019 (Table 1).

At each station, in-situ water-quality parameters were measured at the surface and bottom of the water column using a multiparameter sonde (YSI-EXO2®). The parameters measured were temperature (°C), salinity (measured as conductivity and reported unitless), dissolved oxygen (%DO) and pH. Two 3.8 L surface seawater samples were filtered on site through a pre-combusted 0.44 µm glass microfiber filter (Whatman glass microfiber filters, Grade GF/F) using a pump-line system. If weather conditions did not allow for on-site filtration, water samples were collected in Nalgene polypropylene containers and filtered at the JBNERR laboratory facilities. The filters were stored in aluminum foil and frozen (−80°C) until further analysis for total chlorophyll and total suspended solids (TSS).

In addition, three filtered water subsamples (0.5 L each) were stored in autoclaved and pre-combusted amber bottles and frozen at −18°C for nutrient analysis including orthophosphate (PO43−), ammonium (NH4+), and nitrate+nitrite (NOx). For sediment analyses, approximately 50 g were collected per station and the samples were kept in Nalgene polypropylene containers (0.5 L). Depending on coral coverage, the sediment samples were collected either by SCUBA diving or using a petite-ponar grab. All sediment samples were stored at 4°C until analyzed in the laboratory for grain-size, carbonate content (CaCO3%), and foraminiferal assemblages.

Seawater Nutrient Analyses

Seawater samples were analyzed for [PO43−], [NH4+], and [NOx]. Specific colorimetric methods were used for each nutrient and the absorbance was measured using a SpectraMax M5 spectrophotometer (Molecular Devices, San Jose, CA). Orthophosphate analyses followed the modified protocol of ammonium molybdate/potassium antimonyl tartrate/ascorbic acid method of Strickland & Parsons (1972) for microplate assay. Five standards (2.5 µM, 5 µM, 10 µM, 15 µM, and 20 µM) were created using the primary standard (KH2PO4) and deionized (DI) water. To complete the assay, 1.4 mL of each standard and water sample were added to each well in a 48-microwell plate followed by 140 µL of the working reagent. The [PO43−] (µM) was determined by measuring the absorbance at 885 nm.

The modified salicylate-hypochlorite method from Bower & Holm-Hansen (1980) was used to determine [NH4+] in seawater samples. Four standards (5 µM, 10 µM, 25 µM, 50 µM) were created using the primary standard (NH4Cl) with DI-water for a 100 mL (final volume) used for the calibration curve. The [NH4+] (µM) was determined by measuring the absorbance at 625 nm. The [NOx] was determined calorimetrically using the procedure of Schnetger & Lehners (2014). Five standards were created using the KNO3 as primary standard. Then the NOx reagent (450 µL) and 540 µL of each standard or sample were added to microcentrifuge tubes (2 mL). Each microtube was manually mixed for 10 seconds, then heated at 45°C ± 5°C for 60 min in an oven, and subsequently mixed again for 10 additional seconds. The [NOx] (µM) in the samples was determined by measuring the absorbance at 540 nm. The [NH4+] and [NOx] together indicated relative availability of dissolved inorganic nitrogen (DIN).

Total Chlorophyll and Suspended Solids

The total chlorophyll concentrations were determined following the fluorescence method of Arar & Collins (1997). Each filter from each sample was manually broken down into pieces and placed in a 15 mL polypropylene centrifuge tube with 1 mL of 90% acetone solution. An additional 10 mL of the acetone solution was added prior to centrifugation (1,000 rpm for 10 minutes) to separate the solution containing chlorophyll from the filter debris. To measure the fluorescence of each sample, a dilution of 2:1 (2 mL of the 90% acetone solution and 1 mL of the chlorophyll solution sample) was made in centrifugal glass vials (2 mL). Then, fluorescence was measured using a Fluorometer TD-700 (Turner Designs) and recorded as absorbance (FSU).

The method of Van der Linde (1998) was used to determine total suspended solids (TSS) in seawater. Prior to field work and water filtration, filters were pre-combusted at 550°C for four hours in a furnace to determine the dry and empty weights. After filtration, the wet filters were dried at 60°C for 24 hours and then re-weighed. To determine the TSS (mg/L) content in the water sample, weight difference was calculated and divided by the total volume filtered for each sample.

Sediment Composition

Muddy samples were processed by both wet and dry sieving; sandy samples were only dry sieved. For the muddy samples, 10–20 g (depending on available sample size) were wet sieved (<63 µm) to remove the mud (silt + clay) particles. After oven drying (60°C) for 24 hours, the samples were re-weighed to determine the amount of mud content by mass difference. The residual sediments (>63 µm) were dry sieved through 2-mm, 1-mm, 500-µm, 250-µm, 125-µm and 63-µm mesh sieves. For sandy samples, 10 g of sediment was dry sieved using the same mesh sieves as above. The Wentworth grade scale was used for grain-size classification (Wentworth, 1922).

To determine calcium carbonate (CaCO3%) content, each dry sediment subsample (1 g) was placed in a pre-cleaned and pre-weighed crucible. To ensure complete dryness, the subsamples were dried in an oven at 105°C for 24 hours and, after cooling in a desiccator to room temperature, crucibles were re-weighed. Each subsample was “baked” in a muffle furnace at 550°C for four hours and, after cooling in a desiccator to room temperature, crucibles were re-weighed. Finally, CaCO3% was determined by heating in a muffle furnace at 1000°C for one hour and subsequently re-weighing after cooling in a desiccator to room temperature.

Benthic Foraminiferal Analyses

A micro-splitter was used to reduce the bulk sediment into an evenly representative subsample for removal of foraminiferal specimens. A 1-g subsample is recommended in the protocols of Hallock et al. (2003) and Prazeres et al. (2020). However, given the predominance of siliciclastic sediments and scarcity of foraminiferal specimens, 3 g of subsample were used for analysis. The 3 g were wet sieved to discard any mud-sized (<63 µm) sediments and the remaining sediment (>63 µm) was dried in the oven at 60°C for 24 hours. Using a stereomicroscope, the foraminiferal specimens were picked from the sediment using a 0/18 paint brush. Specimens were picked from each subsample until a maximum of 200 individuals were collected or the entire 3 g were examined. The work of Loeblich & Tappan (1987), the World Register of Marine Species (https://www.marinespecies.org/index.php), and other sources were used for taxonomic classification.

Foraminiferal density (FD = #/g) was calculated as the number of individuals found divided by the mass of sediment examined (3 g). Species richness (S) is the number of species identified in a sample. To calculate the Shannon-Wiener Index [H(S)] the following equation was used:
formula

(e.g., Hayek & Buzas, 2010). The Paleontological statistics software package for education and data analysis (PAST; Hammer et al., 2001) was used to calculate to the diversity indices using raw foraminiferal counts.

To apply the FoRAM Index (FI), the foraminiferal specimens were divided into three functional groups defined by Hallock et al. (2003), as modified by Carnahan et al. (2009). These functional groups are: symbiont-bearing (s), stress-tolerant (o), and other small taxa (h). To calculate the FI, the following equation was used:
formula

where Ps = the proportion of symbiont-bearing specimens using the total number of foraminiferal specimens (T) in the sample, Ph = the proportion of OSF (other small foraminifers) in the sample, and Po = the proportion of STF (stress-tolerant foraminifers) in the sample. For example, Po = STF/T represents the number of stress-tolerant benthic foraminifers divided by the total specimens found in that sample; Ph and Ps were calculated similarly. After the calculation, the criteria for the FI values (Hallock et al., 2003) were used to indicate if water-quality conditions support coral-reef accretion in JBNERR.

Environmental Parameters

Data for temperature, salinity, pH, and dissolved oxygen collected at the times of sampling were summarized by month, year, and reef location (Tables 1, 2). Temperatures ranged from ∼26°C in March, to highs of 29–30°C in August/September for both years. Salinities in March 2018, which is near the end of the dry season, were >37. In the late-summer rainy season, salinities ranged from ∼34.6–35.6, with similar salinities persisting into December in 2018 and March 2019 (Tables 1, 2). The pH values also showed a seasonal trend, with intermediate values (8.1–8.2) in March, slightly lower values coupled with the lower salinities in late August/September, and highest values in December 2018 (8.3–8.5). Dissolved oxygen was generally near saturation or supersaturated.

Nutrient analyses indicated that during the late-summer rainy season available DIN was likely in excess, while [PO43−] was below detection. In March, [PO43-] was detectable and DIN was lower, indicating a possible shift to nitrogen limitation (Tables 1, 2). Chlorophyll concentrations and total suspended solids are indicators of water transparency and variability. Total chlorophyll concentrations were relatively consistent within years (average: 2018 = 7.0 × 10−4 mg/L; 2019 = 9.0 × 10−4mg/L). Concentrations in August/September were lower and less variable than in other seasons. Suspended solids in general were higher in 2019 (10.0–17.0 mg/L) versus 1.0–8.0 mg/L in 2018. Concentrations of suspended solids in December (both years) were lower than in other months (Tables 1, 2).

Grain-size analysis revealed that back-reef sites generally were dominated by fine sediments, while sediments on the fore-reef were mostly granules and coarse sand ( Appendix A). Calcium carbonate (CaCO3%) in sediment varied somewhat by season, with lower values in August/September and highest in December (Tables 1, 2).

Benthic Foraminiferal Assemblages

From a total of 64 surface-sediment samples, 4,753 foraminiferal tests were picked and a total of 117 taxa were identified ( Appendix B). From the 64 samples, 45 back-reef samples and 19 fore-reef samples were analyzed. In the back-reef a total of 3,533 foraminiferal tests were picked and a total of 115 taxa were identified. In the fore-reef a total of 1,220 foraminiferal tests were counted from which 96 taxa were identified. In general, Amphistegina gibbosa (Am. gibbosa) tests dominated in fore-reef samples, while smaller rotaliids such as Rotorbinella rosea and Discorbis spp., and miliolids, especially Quinqueloculina spp., dominated in back-reef samples (Figs. 2, 3). In March and December 2019, inclement weather prevented sediment sampling at fore-reef stations.

Back-reef stations showed strong variability across sampling events. In March 2018, R. rosea made up 19% of the tests counted, with Quinqueloculina spp. making up 15%, and two symbiont-bearing taxa [Am. gibbosa 11% and Archaias angulatus 6% (subsequently Ar. angulatus)] found with similar abundances as the two common stress-tolerant genera (Elphidium 7% and Ammonia 6%; Fig. 2-A). Other foraminiferal taxa with relative abundances >5% comprised 36% of those found and 19% comprised accessory taxa that were defined as any taxa for which relative abundance was >1% in the assemblage. In September 2018, two rotaliids, D. mira and R. rosea, together comprised 36% of the specimens, followed by Am. gibbosa (13%), Quinqueloculina spp. (11%), Ammonia (7%), and Ar. Angulatus (5%) (Fig. 2-C). Other foraminiferal taxa with relative abundance <5% comprised the remaining 28%, from which 12% were accessory taxa. Quinqueloculina spp. (26%) dominated the back-reef samples in December 2018, followed by Triloculina spp. At 12%. In addition, Am. gibbosa and R. rosea were found at 6–7%, others comprised of taxa (49%) that did not exceed 5% the total (Fig. 2-E).

Smaller taxa dominated the back-reef samples in March 2019, making up 57% (Quinqueloculina spp. 33%, Pyrgo spp. 18%, and Discorbis spp. 6%); symbiont-bearing Am. gibossa (8%) and the stress-tolerant Ammonia (7%) were also common. The other foraminiferal taxa that individually did not exceed 5% comprised 29% of specimens counted (Fig. 2-B). In August 2019, Quinqueloculina spp. again dominated (34%), followed by Am. gibbosa (15%), Discorbis spp. (10%), and Pyrgo spp. (7%). Other foraminiferal taxa present at <5% comprised the remaining 34% of those found (Fig. 3-D). In December 2019, Discorbis spp. (24%) and Quinqueloculina spp. (24%) co-dominated, while symbiont-bearing taxa included Am. gibbosa (13%) and Ar. angulatus (5%; Fig. 2-F). Stress-tolerant Ammonia beccarii (6%) and Elphidium spp. (3%) were also present.

In fore-reef stations, the proportion of Am. gibbosa in March 2018 was 37%, followed Quinqueloculina spp. (15%) and R. rosea (14%; Fig. 3-A). The proportion of Am. gibbosa in September 2018 was 39%, with two rotaliids, R. rosea (26%) and Discorbis spp. (10%) together comprising 36% of the specimens, followed by symbiont-bearing Sorites marginalis (5%; Fig. 3-B). The proportion of Am. gibbosa was lowest in December 2018 (29%), followed by Quinqueloculina spp. (23%), Discorbis spp. (10%), and Ar. angulatus (9%; Fig. 3-C). In August 2019, Am. gibbosa (49%) dominated, followed by rotaliids and miliolids, including Quinqueloculina spp. (19%) and Discorbis spp. (10%; Fig. 2-D).

Benthic Foraminiferal Ecological Indices

Densities of foraminiferal specimens in the sediments were low (Tables 3, 4), consistent with relatively low proportions of CaCO3 (Tables 1, 2). Back-reef sites averaged from 3 to 67 individuals per gram, fore-reef stations averaged from 1 to 46 individuals per gram. Consistent with low densities, species richness per sample was low; most of the 127 species were rare. In back-reef samples, mean species richness ranged from a low of 11 in September 2018, down from the high of 24 in March 2018, then increasing to 23 in December 2018 (Table 3). In contrast, the mean species richness in fore-reef stations was lowest in March 2018 and highest in December 2018 (Table 4).

Mean Shannon Diversity 0506Index (SDI) values for back-reef sites in 2018 ranged from 1.9 in September 2018 to 2.9 in December 2018 (Table 3). Mean SDI values in fore-reef sites were lowest in March 2018 (1.2) and again highest in December 2018 (1.7). Note that fore-reef stations were not sampled in March and December 2019 because of inclement weather conditions. The mean FoRAM indices reflected the consistent presence of symbiont-bearing foraminifers in both back-reef and fore-reef samples. The FI in the back-reef sites were in the 3–4 range, while fore-reef values were in the 6–8 range.

Seasonal Variations in Environmental Parameters and Ecological Indices

The original objectives of this study were to assess foraminiferal assemblages and environmental conditions seasonally and spatially around cays that define Jobos Bay, Puerto Rico, to test the hypothesis that those assemblages reflect seasonal changes (rainy vs. dry seasons). The choice to sample in December, at the beginning of the dry season, and March, at the end of the dry season, was originally planned to characterize dry season conditions. In contrast, August and September are commonly the peak of the rainy and tropical storm season.

In March both years, the back-reef environmental parameters and the foraminiferal assemblages were generally similar (inclement weather prevented fore-reef sampling in 2019). The highest [PO43-] were recorded in March, indicating possible inorganic nitrogen (DIN) limitation ([NOx] and [NH4]). Temperatures were the coolest, though salinity was notably lower in 2019 (37 in 2018, 35 in 2019; see Table 2).

In back-reef and fore-reef sites, temperature was highest in August and September, salinity was intermediate (34.6–35.8), and DIN was in excess, while [PO43-] was below detection in 2018 (no [PO43-] data are available for 2019), possibly indicating effects of runoff. The foraminiferal assemblages in 2018 were least diverse (10–11 species per sample, SDI = 1.9). In contrast, in August 2019, density was still low (14–26 individuals/g), but species richness (13–17) and SDI (2.4) in the back-reef were notably higher (see Fig. 2-C, D). Moreover, in 2018 the back-reef samples were dominated by R. rosea, a species often found in higher energy settings, while in 2019 the assemblages were dominated by Quinqueloculina spp. In the fore reef, samples were dominated by Am. gibbosa and R. rosea, both species that occur in hydrodynamic environments (e.g., Triffleman et al., 1991; Crevison et al., 2006).

In December 2018, salinities were lowest in both back-reef and fore-reef sites (∼34.4), indicating the influence of terrestrial runoff. The foraminiferal assemblages in the back-reef sites were quite diverse, with an average of 23 species per sample and the highest mean SDI (2.86). In December 2019, in contrast, salinities were the highest found, higher even than March 2018, and species richness, SDI, and assemblage composition were also more similar to March assemblages.

Fortunately, no major hurricanes directly hit Puerto Rico in 2018 or 2019. The only notable storm reported to affect southeast Puerto Rico in 2018 occurred in July. The remnants of Hurricane Beryl passed south of Puerto Rico, delivering heavy rain on the east side of the main island. Thus, sampling in September 2018 occurred after the only major storm in 2018 (https://www.nhc.noaa.gov/archive/2018/). In 2019, sampling occurred in mid-August, prior to Hurricane Dorian passing just east of Puerto Rico in late August, with winds and heavy rain. Roughly a month later, Tropical Storm Karen also passed along the east coast of Puerto Rico (https://www.nhc.noaa.gov/archive/2019/). Thus, the differences in the foraminiferal assemblages between September 2018 and August 2019 may reflect differences in timing of storm activity. Sampling in September 2018 occurred after Hurricane Beryl’s passage and may reflect the influence of wind and storm waves, with lower species richness and high relative abundances of hyaline taxa. Porcelaneous taxa such as Quinqueloculina spp. are often small, relatively fragile, and readily winnowed by wave action (e.g., Wetmore, 1987). Those are the taxa that are much more prevalent in samples from August 2019, which were collected before major storm passage.

The differences in assemblages between December 2018 and 2019 show the opposite trend of the late summer samples. In 2018, after a less hydrodynamic but rainy late summer and autumn, the foraminiferal assemblages were much more diverse and dominated by Quinqueloculina spp. and other smaller miliolids. In 2019, after a more active tropical storm season, hyaline taxa made up half the assemblage, though Quinqueloculina spp. were also abundant.

The effects of storm activity may also provide insight into differences in March assemblages. The back-reef assemblages in March 2018 were >40% hyaline taxa. The relative abundance of R. rosea (19%) and relatively high species richness may reflect residual effects of the extreme 2017 hurricane season, when two major storms, hurricanes Irma (Cat. 4) and María (Cat. 5) devastated Puerto Rico. In March 2019, after a quiet storm season in 2018, relative abundance of hyaline taxa was ∼20% of the assemblages, and proportions of porcelaneous Quinqueloculina were double those in 2018.

Additional environmental parameters must be considered, especially when considering sediment winnowing. High proportions of CaCO3 and low proportions of mud in August 2019, along with coarser sediments including coral and shell fragments, reflect hydrodynamic events. But overall, sediments of Jobos Bay and reefs are mixed biogenic carbonates and terrigenous siliciclastics, with the latter dominating, especially in the back reef. Because siliciclastics are generally much harder (6–7 on Mohs scale) compared to biogenic carbonates (3–4), wave winnowing also predominantly breaks down more fragile carbonate grains, including the tests of smaller foraminifers (e.g., Wetmore, 1987).

The FoRAM Index

The “Foraminifera in Reef Assessment and Monitoring”, abbreviated as the FoRAM Index or FI, was developed as an indicator of whether water quality is sufficiently oligotrophic to support carbonate sediment accumulation by organisms that host algal endosymbionts, such as reef-building corals and larger benthic foraminifers. Hallock et al. (2003) used foraminiferal assemblages in samples from La Parguera, PR (Donnelly, 1993) and the Florida Reef Tract (Cockey et al., 1996) to develop this index. Oliver et al. (2014) applied the FI in La Parguera, concluding that foraminiferal assemblages were the most responsive indicator to gradients of anthropogenic influence. Though developed based on western Atlantic and Caribbean foraminiferal assemblages, the FI has since been applied worldwide (Prazeres et al., 2020, and references therein).

Overall, the FI values were 3–4 for back-reef sites and 6–8 for fore-reef sites of the three Jobos cays. Given the proximity of terrestrial influence, these values indicate that the influence of Caribbean waters allows water quality to support Am. gibbosa, which is a prevalent symbiont-bearing foraminiferal species throughout the western Atlantic and Caribbean. In addition, temporal changes in the overall assemblages suggest that wave action associated with tropical storm activity may help dilute and possibly remove pollutants associated with terrestrial runoff and groundwater discharge.

Two additional factors explain the persistence of Am. gibbosa as a key member of the foraminiferal assemblages, despite intermittent unfavorable conditions: dormancy and propagules. For example, Ross & Hallock (2019) reported that individuals of this species can survive and recover after 20 months in aphotic conditions. And more importantly, Alve & Goldstein (2002, 2003, 2010) demonstrated that tiny propagules, primarily produced by sexual reproduction, provide a key dispersal mechanism for benthic foraminifers. Amphistegina propagules undoubtedly are carried in warm western Atlantic and Caribbean waters, and when they settle into a suitable habitat (or into an intermittently suitable habitat), they can develop, grow, and asexually reproduce.

Hallock et al. (2003) discussed the limitation of applying the FI in shallow water because high-energy environments generally have well-sorted sediments, from which smaller heterotrophic taxa have been broken down and winnowed out. Prazeres et al. (2020) discussed ways to standardize the FI protocol, suggesting the elimination of sites shallower than 3 m and those with predominantly coarse textures (>2 mm). Prazeres et al. (2020) further recommended against using samples containing <50 specimens per gram of sample. Although terrigenous sediments are mentioned in previous studies, the dilution and probability of destruction of carbonate sediments were not addressed.

Thus, sediments from the vicinity of Jobos reefs have a variety of characteristics that don’t fit the protocols for use of the FI as recommended by Prazeres et al. (2020). A variety of depths can be encountered in a transect <100 m long. The carbonate fractions are clearly diluted by terrigenous siliciclastic sediments, which is why 3-g samples were assessed rather than the standard 1-g. In such sediments, even 5-g samples would not be unrealistic.

The JBNERR presents an excellent example of why study sites often don’t fit “standard” definitions. The reefs border a coastal-plain estuary from which both surface and ground waters discharge. The coastal area has a long history of anthropogenic disturbance, though since establishment of the reserve, much of the coastline is vegetated by mangroves. The reefs are only a few kilometers offshore from the muddy mangrove shoreline. And, as noted above, the sediments, even around the reefs, are predominantly siliciclastic.

Indeed, foraminiferal densities in fore-reef and back-reef sites of Jobos cays limit the FI interpretation. Nevertheless, the foraminiferal assemblages showed the predominance of small heterotrophic taxa in finer sediments from the back-reef sites and symbiont-bearing taxa from fore-reef sites with coarser grains and higher proportions of CaCO3. In addition, the changes in proportions of hyaline taxa such as R. rosea, along with Am. gibbosa, provided additional insight into the influence of hydrodynamics in the Jobos Bay system.

The essential characteristic that allows the reefs to be present at Jobos Bay is the proximity of clear, oligotrophic Caribbean water. The salinities recorded, even in back-reef areas, consistently exceeded 34 and mostly exceeded 35. The proximity of, and exchange with, Caribbean waters provides the reefs, cays, and the back reefs near the cays with waters that support foraminifers and corals that host algal symbionts and are important producers of carbonate sediments. The variability in the foraminiferal assemblages recorded in this time series indicate that tropical storms enhance flushing of back-reef and lagoon waters, limiting eutrophication and allowing the foraminifers hosting algal symbionts to maintain sufficient presence that the FI values of 4 predominated in back-reef sites, with FI ≥6 in fore-reef sites.

Finally, this work demonstrates the value of time-series studies that provide essential insight into variability that naturally occurs in tropical coastal waters. The original hypothesis was that assemblages in samples collected during dry-season months of December and March would contrast with those found during the rainy season. The reality is more complex, and the influence of tropical-storm systems is an essential consideration.

Thanks to Jobos Bay National Estuarine Research Reserve for allowing us to do fieldwork and contribute to the needs of the reserve. Thanks to Ángel Dieppa, Milton Muñoz Hincapié, Luis D. Ortiz Serrano, and Aitza Pabón for their collaboration in the fieldwork, instruments, and facilities. We are grateful to Johannys Jiménez Collazo, Lexa Medero, Zakiya, Marcus, Kyle, and Dante for their assistance in data collecting and processing. ARM is grateful to the GeoLatinas organization for the space to test and solidify ideas and for support. This research was possible by funding from Sea Grant Puerto Rico (#NA18OAR4170089) and NOAA Educational Partnership Program with Minority Serving Institutions Cooperative Agreement (#NA16SEC4810009).

Appendix