Effect of seawater temperature, pH, and nutrients on the distribution and character of low abundance shallow water benthic foraminifera in the Galápagos
Effect of seawater temperature, pH, and nutrients on the distribution and character of low abundance shallow water benthic foraminifera in the GalÂapagos
Alexander F. Humphreys 0 1 2
Jochen Halfar 0 1 2
James C. Ingle 0 2
Derek Manzello 0 2
Claire E. Reymond 0 2
Hildegard Westphal 0 2
Bernhard Riegl 0 2 3
0 Funding: A Natural Sciences and Engineering Research Council of Canada Discovery grant to JH , grant number:1303409
1 Department of Chemical and Physical Sciences, University of Toronto Mississauga , Mississauga, ON , Canada , 2 Department of Geological Science, Stanford University, Stanford, California, United States of America, 3 Atlantic Oceanographic and Meteorological Laboratory, National Oceanic and Atmospheric Administration, Miami, Florida, United States of America, 4 Leibniz Centre for Tropical Marine Research (ZMT) , Bremen, Germany , 5 Department of Geosciences, University of Bremen , Bremen , Germany
2 Editor: Alex Dickson, Royal Holloway University of London , UNITED KINGDOM
3 Department of Marine and Environmental Sciences, Halmos College of Natural Sciences and Oceanography, Nova Southeastern University , Dania Beach, FL , United States of America
In order to help predict the effects of anthropogenic stressors on shallow water carbonate environments, it is important to focus research on regions containing natural oceanographic gradients, particularly with respect to interactions between oceanography and ecologically sensitive carbonate producers. The GalaÂ pagos Archipelago, an island chain in the eastern equatorial Pacific, spans a natural nutrient, pH, and temperature gradient due to the interaction of several major ocean currents. Further, the region is heavily impacted by the El Niño ÐSouthern Oscillation (ENSO) and the GalaÂ pagos exhibited widespread coral bleaching findings are coupled with reports of unusually low abundances of time-averaged benthic foraminiferal assemblages throughout the region. Foraminifera, shelled single-celled protists, are sensitive to environmental change and rapidly respond to alterations to their surrounding environment, making them ideal indicator species for the study of reef water quality and health. Here, statistical models and analyses were used to compare modern shallow water benthic foraminiferal assemblages from 19 samples spanning the GalaÂ pagos Archipelago to predominant oceanographic parameters at each collection site. Fisher α analysis and FORAM-Index (FI; a single metric index for evaluating water quality associated with reef development) implied a combined impact from ENSO and upwelling from Equatorial Undercurrent (EUC) waters to primarily impact foraminiferal abundances and drive assemblage patterns throughout the archipelago. For instance, repeated ENSO temperature anomalies might be responsible for low foraminiferal density, while chronically high
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Competing interests: The authors have declared
that no competing interests exist.
nutrients and low aragonite saturation and low pHÐinduced by EUC upwelling and La Niña
anomaliesÐlikely inhibited post-ENSO recovery, and caused foraminiferal assemblages to
exhibit a heterotrophic dominance in the southern archipelago. What resulted are low FI
values in the southern collection sites, indicating environments not conducive to endosymbiont
development and/or recovery.
Carbonate-producing organisms are vital to a host of shallow water marine ecosystems
throughout the world, and with predictions for anthropogenic warming and its associated
oceanic changes, there is a need to better understand distribution, physiology, and environmental
interactions of carbonate producers ([1±3] and others). This is particularly relevant for high
nutrient tropical systems, which produce sediments more suggestive of extra-tropical
environments. These environments potentially represent natural laboratories for the study of ancient
depositional environments, as well as future environmental conditions . Among shallow
water carbonate producers, benthic foraminifera are of particular relevance. These unicellular
eukaryotic protists, commonly housed within external calcium carbonate or agglutinated tests
(shells), are among the most common organisms found within the carbonate sediments of all
major reef systems on the planet [3, 5±7]. The distribution of these ecologically sensitive marine
organisms is driven by a host of interacting environmental factors including temperature, pH,
salinity, nutrients, water motion, light intensity, depth, sediment texture, food, substrate, and
taphonomic processes ([7±9] and others). Additionally, many species thrive within specific
ecological niches [7, 10] and exhibit rapid responses to changing biotic and abiotic factors on
temporal and spatial scales ([7±9] and others). As a result of these sensitive biophysical interactions,
benthic foraminifera are widely considered to be primary indicators of the health of reefal
environments, and useful tools for interpreting environmental change [11, 12]. This physiochemical
sensitivity makes benthic foraminifera potentially highly vulnerable to temperature and nutrient
anomalies associated with periodic disturbance events like the El NiñoÐSouthern Oscillation
(ENSO) . This is particularly true in the GalaÂpagos Archipelago, a moderate to high nutrient,
high CO2, tropical environment in the eastern tropical Pacific (ETP), often strongly impacted
by ENSO. Interest in the GalaÂpagos is magnified by reports of a near absence of benthic
foraminifera in its shallow water environments, relative to other ETP locations [13, 14]. Here, we use
statistical models and analyses to investigate the interaction between shallow water benthic
foraminifera of the GalaÂpagos and the major local oceanographic parameters. Of particular
relevance is the notion of time averaging, which is the mixing of grains of different ages prior to
permanent burial in the geologic record [
]. Hence, any analyses on shallow water assemblages
must be assessed not according to individual events, but rather to the long-term oceanographic
conditions in which the sediments developed. In summary, this manuscript serves to highlight
the environmental forces driving the diversity and distribution patterns of benthic species
within GalaÂpagos foraminiferal assemblages, and to uncover potential causes for the low
foraminiferal abundances. Additionally, we discuss future changes in shallow water foraminiferal
abundance under a global change scenario.
The GalaÂpagos Archipelago (Fig 1) lies roughly 1000 km due west from the Ecuadorian
mainland (Between 1Ê40'N-1Ê25'S and 89Ê15'W-92Ê00'W; , straddling the equator in the center
2 / 25
Fig 1. Map showing collection islands discussed in this study. Map details relative locations of samples represented
by respective symbols within cluster analysis (Fig 2).
of a system of currents and countercurrents , which cause the region to span a climatic
and oceanographic transition zone with resultant variable physio-environmental
conditions . The southern GalaÂpagos are directly influenced by the equatorial undercurrent
(EUC) which shoals from the west, resulting in elevated nutrient levels, from mesotrophic
conditions in the southeastern archipelago to eutrophic conditions at Isabela [13, 14], as
well as chronically depressed pH throughout the southern islands . As a direct result
of this high nutrient and high CO2 EUC, the southern GalaÂpagos contains the highest
natural ambient CO2 and lowest aragonite saturation (Oarag) of any modern tropical surface
ocean . Further, the GalaÂpagos are strongly influenced by ENSO, a climatic and
oceanic event that recurs approximately every 3±7 years. The El Niño phase of ENSO is
associated with higher than normal ocean temperatures in the region, while the La Niña phase
brings periods of higher-than-normal nutrient and low pH upwelling from the EUC .
Fluctuations in EUC and ENSO force the species and habitats in the GalaÂpagos to face
cyclical shifts in climate , which have had dramatic effects on the carbonate producing
species of the region. For example, the stronger-than-average 1982±1983 and 1997±1998
ENSO events resulted in widespread, heat-related coral degradation throughout the
southern islands [
]. Likely, foraminifera would be similarly impacted as corals , but
no such information exists yet for the GalaÂpagos region.
3 / 25
128 samples were collected between 1 and 46 meters water depth (see Table 1 for depth details)
using SCUBA and Van Veen grabs, from seven sites spanning the geographic extent of the
archipelago and stored in plastic collection bags. In the lab, samples were washed of organic
material using filtered water over a 2mm and 63μm sieve stack to split the samples into gravel
and sand fractions, respectively. Washed samples were dried in an oven at 50ÊC and the bulk
sand fraction of each sample was further split (using a sediment splitter) into subsamples for
foraminiferal picking (no benthic foraminifera were identified in the gravel fraction).
Foraminifera were picked using ultra-fine brushes from sediment picking trays and placed in 60-cell
cardboard micropaleontology slides for identification. Species identifications were performed
using McCulloch [24±26], Cushman and McCulloch [27±30], and Lalicker and McCulloch
] as primary references, as well as a number of other sources (see references [23, 32±42]).
Work in the Galapagos was permitted by the authorities of Parque Nacional Galapagos (permit
# 030-13PNG) and conducted under the auspices of the Charles Darwin Research Station.
Furthermore, the field studies in this project did not involve endangered or protected species.
The preservation of foraminiferal tests tends to be higher within shallow water carbonate
environments than in siliciclastic and organic-rich settings, so the sediment collection sites for
this project were selected for their high carbonate sediment productionÐa commonly used
methodology for foraminiferal analysis [21±23]. In addition to higher foraminiferal
abundance, these settings are also typically characterized by superior test preservation, compared to
organic-rich and siliciclastic settings [21±22]. Benthic foraminifera have been reported to be
nearly absent from shallow water carbonate rocky reef settings in the GalaÂpagos , so
sample collection for this project focused on similar soft substrate (white sand) carbonate
production sites (identified using satellite imagery) . However, even with these sampling
criteria, unusually low foraminiferal abundances (Avg. 0.7% of total carbonate fraction following
removal of terrigenous data) obtained from sediment thin section point count analysis on the
128 samples at 300 counts-per-sample (See  for details on sediment thin section
methodology), and poor benthic foraminiferal test preservation throughout the GalaÂpagos, made
picking and species identification challenging. Thus, in order to obtain statistically significant
numbers of benthic foraminifera (300+ tests per sample), test collection concentrated on
samples with sediment thin section point counts indicating test presence (88 samples; ). From
this sample subset, species identification was completed at each sampling site for samples with
a picking rate of greater than 15 tests per hour. For Darwin, Baltra and Santa FeÂ, which had
few samples to choose from (5±7 samples), test-collection was attempted for all samples within
these sites. This methodology resulted in 19 samples, representing 7 collection sites spanning
the GalaÂpagos (Fig 1).
FORAM-Index (FI) was calculated for averaged samples at each island where samples had
been analyzed, as a proxy for the general state of coral reef health throughout the GalaÂpagos
Archipelago [11±12]. FI is used to determine whether water conditions within marine habitats
are capable of supporting algal symbiosis, and to assess the impact of environmental stressors
on coral habitats . The FI is based on foraminiferal shell presence, and does not rely on
coral populations, which allows for rapid and relatively simple assessments of reef
11, 12, 43
]. It is also a useful method for comparison to other sedimentological
assessments of reefal habitats, such as the coral reef turn on/turn off zone (CRTTZ; [
calculation relies on the relative abundances of symbiont-bearing, heterotrophic and
opportunistic functional groups [
] and is constructed using the following equation from Hallock
et al. :
in which FI is the FORAM Index, Ps is the number of symbiont-bearing species/T, Po is the
proportion of the opportunistic taxa/T, Ph is the proportion of smaller heterotrophic taxa/T,
and T is the total number of foraminifera counted [11, 12]. See Hallock et al.  for details.
A hierarchical agglomerative cluster analysisÐChord distance, Ward linkageÐwas
performed on the Hellinger transformed species count data [
] from the top 28 species,
representing ~75% of all benthic foraminiferal production. Hellinger transformation (y’ij) is defined
where y is abundance, yij is the abundance of species j in sample i, and i+ is the sum of values
over row i [
]. The transformation ensures that the samples are being compared according to
their specific abundances, without giving undue importance to double zero counts throughout
the data [45±47]. The double zero problem arises because of the nebulous interpretation of an
absent species in a dataset. Species presence at two sites indicates a similarity between the sites.
However, a species absence may result from the two sites lying above or below the optimal
niche zone for that species or, alternatively, one site could be above and the other below the
optimal niche value [
]. The decision to transform the data was made to reduce the skewing
effects of the high number of zero values within our foraminiferal count results, especially due
to samples with strong species dominance. Furthermore, the transformation shows the cluster
analysis on a closely comparable level to the canonical correspondence analysis (CCA) results,
5 / 25
which keep χ2 distances between sample sites intact, and are not affected by the double zero
problem (see below).
In order to relate the distribution of the benthic foraminiferal communities to the
overarching environmental constraints, a canonical correspondence analysis (CCA) was performed on
the count data for the predominant 28 foraminiferal species using the `vegan' package in R
]. CCA reduces the variables to a few digestible combinations, with species, sites, and
environmental variables presented in a triplot. The χ2 distance between sample sites is preserved,
which removes the skewing effects of double-zeros in the data ([
]; See below for
explanation), and species and environmental variables are represented as points and vectors on the
biplots, respectively. A benefit of CCA is that the ordering of species along the canonical axes
follows a pattern related to their ecological optimum [
]. Environmental parameters used in
the construction of the CCA were the result of data exploration of a larger master set of
regional physical and environmental data (see Appendix for master dataset), in order to
remove covariant variables. These data included remotely sensed Moderate-resolution
Imaging Spectroradiometer (MODIS) environmental chlorophyll-a (Chl-a as a proxy for nutrients;
]), Seaviewing Wide Field of view Sensor (SeaWiFS) sea surface temperature (SST) data
from NASA Giovanni (https://giovanni.gsfc.nasa.gov/giovanni/; [
]), sea surface salinity
(SSS) estimates from The Simple Ocean Data Assimilation (SODA; [
]), as well as in situ pH
and aragonite saturation data from Manzello , Manzello et al. [
] as well as depth
(Table 1). For a description of oceanographic statistics, see Table 2. In addition to satellite
oceanographic data collection, a temperature data logger was placed on Darwin Reef at a
depth of 12m, from 26 November 2016 to 23 April 2017 to collect temperature data at 0.5 hour
intervals in order to compare relatively long term satellite averages to short-term reefal
Species richness and Fisher α diversity indices were calculated for all known foraminifera
in our samples in order to better understand the relationship between foraminiferal
production and diversity throughout the GalaÂpagos. Foraminiferal tests were identified as hyaline
perforate, imperforate porcellaneous, and agglutinated, according to the classification of Loeblich
& Tappan (1984) [
] and plotted as ternary diagrams alongside average Fisher α indices and
pH values in order to assess geographic patterns in wall structure types.
In addition to cluster analysis and CCA, a univariate regression tree analysis ([
package in R [
]) was constructed by using samples (binned according to their respective
cluster locations) as response variables, and the oceanographic controls (including sample
depth) as explanatory variables. The purpose of this analysis was to look for any possible
overarching environmental controls influencing the clustering of samples.
FORAM-Index (FI) values for each site, as well as their associated ecological interpretations, are
plotted in Fig 2A. Interpretations were based on the definitions of Hallock et al. , where
FI > 4 is indicative of environmental conditions suitable for symbiotic organisms required for
Mean: mean of monthly values
Mean Anomaly: per-month mean of all monthly anomalies over all months)
(monthly anomalies: annual mean of mean monthly values minus each monthly value)
6 / 25
Fig 2. Comparisons of reef assessment analyses. (A) Average FORAM±Index values plotted against average percent
coral abundance in sediments for each collection site. (B) Average mean Chlorophyll-a (nutrient proxy) plotted against
average percent coral abundance in sediments for each collection site, showing relative position with respect to coral
reef turn on-turn off zone (CRTTZ; ). Abundances based on thin section point count data for 128 samples. D,
Darwin; B, Baltra; SF, Santa FeÂ; ES, Española; FL, Floreana; SC, San CristoÂbal; IS, Isabela. Islands color coded
according to strong symbiont activity and reef production (green), marginal symbiont activity and no true reef
production (blue), and low symbiont production and low coral production (red) according to their respective scales.
true reef development, as well as coral recovery. FI values between 2 and 4 indicate
environments that are marginal to reef growth and not suitable for coral recovery. FI values < 2 are
considered to indicate stressed environments, which are not likely to sustain reef growth.
Although coral can exist in environments with FI<2, it is less common, not extensive, and not
reef forming [11±12]. Fig 2A shows site-averaged FI values plotted against sample-averaged
percent coral content from thin section point counts (128 samples; see Humphreys et al.  for
details) and reveals FI values ranging from 1.2 (Floreana) to 4.2 (Darwin). The benthic
foraminiferal community at Darwin falls within the low end of FI values for true reef development and
post-stress recovery (green dots). FI values place Santa FeÂ and Baltra within the low marginal
categorization, with no post-stress recovery (blue dots). All other southern sites had
foraminiferal communities that indicate water conditions unsuitable for symbiotic activity (red dots; Fig
7 / 25
2A). Further, Fig 2A shows a photozoan sedimentary signature at Darwin (FI>4; 55% coral), a
mixture of photozoan and heterozoan sediments at Baltra and Santa FeÂ (2<FI<4; heterozoan/
photozoan transition defined as >20% corals (or others) with a majority of heterozoan
carbonate producers) and diminished coral abundances in study islands with FI < 2; Española (14%),
San CristoÂbal (4.5%), Floreana (12%), and Isabela (0.9%; Fig 2A). Hence, FI results reveal a
majority of the southern GalaÂpagos collection sites to be unsuitable for extensive endosymbiont
development (FI<2) in corals and larger foraminifera. These findings run contrary to those
based on sample site mean chlorophyll values (Chl-a; ), which places Baltra, Santa FeÂ,
Española, and Floreana within the coral reef turn on/ turn off zone (CRTTZ; blue dots), which
demarcates the Chl-a maximum (~0.3mg/m3) for coral reef growth (Fig 2B). Sites with Chl-a
values below the CRTTZ (green dots) readily develop reef framework, while sites with Chl-a
values above the CRTTZ (red) are not conducive to coral development.
Agglomerative hierarchical cluster analysis (Chord distance,Ward linkage; data Hellinger
transformed), comparing the composition and abundances of the 28 predominant species
throughout the GalaÂpagos Archipelago, revealed 5 clusters, (#1±5), separated into two major
groups (Group I and Group II, composed of clusters 1±3 and clusters 4±5, respectively; Fig 3).
Cluster 1 contains foraminiferal assemblages from the coral reef at Darwin Island. Cluster 2
contains samples from Baltra channel between Baltra and Santa Cruz islands. Cluster 3
contains samples from Santa FeÂ, Española, and Isabela. Clusters 4 and 5 are composed of samples
from San CristoÂbal and Floreana islands (Fig 3).
Canonical correspondence analysis (CCA)
Canonical correspondence analysis (CCA) results constructed from the percent abundance
data of the 28 most abundant species, which make up roughly 75% of the total population of
foraminifera (Table 1; for count data on all species encountered, see S1 Table (supplementary
section)) were plotted as a triplot (Fig 4; for CCA permutation tests, see S1 Fig (supplementary
section)). Samples were represented by their respective cluster analysis symbol (Fig 3), species
were represented as abbreviations (see Table 3) and color coded according to functional
group. The CCA and cluster analysis (Figs 4 and 3, respectively) revealed two major groupings
along the CCA1 axis, with the higher diversity Group 1 (Clusters 1±3) plotting to the right,
along positive CCA1 values, and the lower diversity Group 2 (Clusters 4±5) plotting to the left
along negative CCA1 values (Fig 4).
Cluster 1, representing the reef setting at Darwin, was strongly associated with mean sea
surface temperature (μSST) and aragonite saturation (Oarag), particularly for the larger
symbiont-bearing Amphisorus hemprichii and Sorites marginalis. Cluster 2, composed of samples
from the south-central site of Baltra channel (Fig 1), showed a positive correlation
between μSST, Oarag and the symbiont-bearing foraminifera Amphisorus hemprichii, Borelis
sp., as well as the opportunistic Elphidium macellum. Instead, the heterotrophic Cibicidoides
schmitti, Parahauerina displicata and fractured tests from Neohauerina or Hauerina species
were (mostly) positively aligned with increased mean sea surface temperature anomaly (μSST.
An) and pH. Cluster 3, composed of samples from Santa FeÂ, Española and Floreana, was
heavily dominated by heterotrophic taxa, and showed a strong relationship with mean
chlorophyll anomaly (μChl.An), mean chlorophyll (μChl) as well as a moderate mean sea surface
salinity anomaly (μSSS.An) influence (Fig 4). This cluster was strongly associated with high
quantities of unidentified agglutinated specimens (Usp) 1±4, the heterotrophic rotolids
Rotorbinella mira galapagosensis, Rotorbinella mira clarionensis, Miniacina barringtonensis, Cibicides
8 / 25
Fig 3. Cluster analysis. Cluster analysis (Chord Distance; Ward Linkage) on Hellinger-transformed count data for 28 foraminifera species making up top
75% of production. Sample sites from individual clusters are marked by grey symbols for corresponding plots within CCA.
fletcheri, Cibicides lobatulus, as well as the miliolids Quinqueloculina galapagosensis, and
Quinqueloculina blackbeachensis (Fig 4). It should be noted that Usp 1±4 were entered as a group
into the statistical analyses due to the high abundance of these agglutinated tests (5% of total
contribution) and their removal would have resulted in a skewing of the statistical results.
Cluster 4, comprising samples from San CristoÂbal, showed a strong correlation to increasing
mean sea surface salinity (μSSS) and depth, a strong inverse correlation with increased μSST,
Oarag and pH, and was primarily defined by the opportunistic Elphidium postulosum and the
heterotrophic Poroeponides cribrorepandus, Sphaerogypsina globulus and a species of the genus
Miniacina. Cluster 5, which included all samples from the north coast of Floreana (Fig 1),
9 / 25
Fig 4. Canonical correspondence analysis (CCA) triplot results. CCA constructed from percent abundance data of 28 most abundant species, making up roughly 75%
of total population of foraminifera. (A) optimal display for cluster sample (grey symbols) interpretation; (B) optimal display of foraminifera species interpretation. CCA
results represent constrained ordination of foraminiferal population numbers, and not principal components analysis (PCA) of environmental variables at each site.
Thus, triplots display how foraminiferal community is organized with respect to environmental parameters [
]. Environmental parameters are plotted as vectors
(black), samplesÐlabeled with their respective cluster symbolsÐare represented as points, and abbreviated species names (Table 1 caption) are plotted as points,
colorcoded according to their respective functional group affiliation (blue, symbiont bearing; red, opportunistic; green, heterotrophic). Permutation tests show high
significance for CCA axes in question (S1 Fig).
showed no positive association with any of the explanatory environmental variables tested, but
indicated a negative correlation with μSSS.An (Fig 4). Cluster 5 was strongly grouped with the
heterotrophic rotalid Poroeponides cribrorepandus with notable contributions from the rotalid
Gypsina vesicularis and the opportunistic rotalid Elphidium crispum subcrispum.
Foraminiferal diversity and total contribution
Foraminiferal species richness and Fisher α diversity indices, plotted against abundance data
from thin section point counts (see Humphreys et al.  for details), are represented in Fig
5A and 5B, respectively. Both, species richness and Fisher α indices showed strong negative
correlations with respect to total foraminiferal production at each site. For instance, Baltra
island foraminifera, with an average richness of 56.5 species (range: 53±60) represented an
average contribution of 0.07% to total carbonate production at that site, while Floreana island
foraminifera, with an average richness of 20.4 species (range: 6±31) represented an average
contribution of 2.84% to total carbonate production at that site. It should be noted that Darwin
Island foraminifera, which exhibited as many species as Baltra taxa, had a total foraminiferal
production amounting to 0.5% of total carbonate sediment, which is comparable to the
southern island sites of Santa FeÂ, Española, and Isabela (Fig 5). Further, diversity aligns with the
clusters in the cluster analysis, with the highest sample diversity in Cluster 1 and Cluster 2
samples, moderate diversity in Cluster 3 samples, and lowest mean diversity in Group II Cluster 4
and Cluster 5 samples (Fig 3-top).
10 / 25
Oceanographic influence of GalaÂ pagos shallow water benthic foraminifera
Ternary diagrams of wall structure types
Samples from Darwin and Baltra were comosed of a majority of porcellaneous forms, with few
agglutinated species. San CristoÂbal and Floreana were heavily dominated by hyaline
foraminifera with few agglutinated taxa (Fig 6A). Falling between these endmember groups, samples
from Santa FeÂ, Española, and Isabela contained a majority hyaline species and follow a line of
increasing agglutinated test material (Fig 6A-circled). Agglutinated test percentages were
higher at the island of Santa FeÂ (Avg 17%; range: 15±20%) than the surrounding southeastern
collection sites, and highest in two samples from Isabela (47% and 59%, respectively). In
general, there was a transition toward higher rotalid dominance along declining pH and Fisher α
diversity gradients (Fig 6A).
PLOS ONE | https://doi.org/10.1371/journal.pone.0202746
Fig 5. Average percent benthic foraminifera composition plotted against species richness (A) and Fisher α diversity (B). Plots show
an inverse correlation between foraminiferal production and species richness / diversity. Islands are as follows: + Baltra;ÐDarwin; · Santa
FeÂ; ▲Española; × Isabela; ♦ San CristoÂbal; ■Floreana.
Darwin reef temperature
Fig 7 shows five months of in situ temperature at a depth of 12m at 0.5 hour intervals. With an
average temperature 25.9ÊC (compared to 25.6ÊC for 12 year satellite average), Darwin reef
12 / 25
Fig 6. Ternary diagrams for major test structure categories of Loeblich & Tappan (1984) [
]. (A) average
compositions of foraminifera from each sampled island, including mean pH (red) and mean Fisher α indices (blue),
revealing a general increase in hyaline forms along declining pH and Fisher α indices gradients as well as shift toward
higher agglutinated content at Española, Santa FeÂ and Isabela islands with decreasing pH (circled). (B) Ternary
representation of a shift from calcareous to agglutinated dominance along a declining pH gradient as represented in
Dias et al. [
]. Baltra pH values are inferred from regression of all site data.
showed a tropical signature, according to the SST categorizations of Betzler et al. [
However, Fig 7A reveals a significant temperature variabilityÐMin 18.7ÊC; Max 29.7ÊCÐover
the collection period, periodically drawing the reefal setting into warm-temperate
conditions. The temperature anomaly of January 14 to 19, 2017 (Fig 7B) identified the reef setting
13 / 25
Fig 7. Graphs of in situ logger temperature data from Darwin reef taken at a depth of 12m. (A) Full temperature
data set from 26 November 2016±18 April 2017. (B) Detail of temperature fluctuations from 14 Jan± 19 Jan 2017,
showing strong temperature depression resulting from localized shoaling of deeper water masses into reef
at Darwin as subjected to cooler water periods, followed by warm temperate conditions, for
days at a time, which forces the tropical biofacies in the region to experience periodic
14 / 25
Fig 8. Univariate regression tree analysis. Plot shows the dominant controls on the cluster distribution in cluster analysis. The plot was
produced by modeling the response variables (binned results of the cluster analysis) against the explanatory variables (oceanographic controls and
sample depths. The resulting dendrogram reveals long-term average temperature anomalies (Mean.Anomaly..SST) as the dominant
oceanographic influence of cluster splits, followed by sample depth.
Regression tree analysis
The univariate regression tree of the major environmental parameters revealed mean sea
surface temperature anomaly (μSST.An) as well as depth at each (collection) island to be the most
prominent grouping variable of foraminiferal community types. In other words, changes in
foraminiferal communities throughout the GalaÂpagos samples seemed to be most strongly
influenced by μSST.An, as well as collection depth (Fig 8).
While foraminifera in our GalaÂpagos sediment samples were unusually rare for the Eastern
Pacific, the analysis of the encountered assemblages suggested strong environmental
influences that separated the northern, more tropical, parts of the archipelago, from the southern.
Thus, the overall distribution of foraminifera follows general trends also observed in other
calcifiers in the GalaÂpagos ([
] and others) but their rarity is clearly indicative of an at best
marginal environment for these globally important carbonate producers.
Interpretation of cluster analysis and CCA
The close overlap among cluster analysis groupings, site diversity, and CCA (species and site
distributions) demonstrated a clear and unambiguous pattern of environmental influence on
time-averaged GalaÂpagos foraminiferal assemblages (Figs 3 & 4). This was particularly evident
along the CCA1 axis, which divided two overriding cluster groups (Group I, Cluster1±3;
Group II, Clusters 4±5; Fig 4). Long-term warmer mean water temperature (μSST), higher
average aragonite saturation (Oarag), pH, and average SST anomalies (μSST.AN; a proxy for
Holocene El Niño variation in the region) predominantly influenced the major symbiont taxa
in the low-diversity shallow waters of Darwin and Baltra (Figs 1 & 4), which resulted in the
close sample association within the cluster analysis (Fig 3). Likewise, high mean nutrient water
(proxied by Chl) from eastward-shoaling equatorial undercurrent (EUC) flow, as well as
average nutrient anomalies (μChl.An; a proxy for Holocene La Niña in the region) positively
influenced various heterotrophic species (green in Fig 4), particularly among concentrations of
agglutinated taxa, at the southern sites of Isabela, Santa FeÂ, and Española (Figs 1 & 4). A
combination of higher long-term salinity, low mean SST, and low Oarag primarily influenced
15 / 25
predominant heterotrophic and opportunistic taxa in San CristoÂbal, while temporally stable
salinity values (negatively-correlated salinity anomaly) most-strongly aligned with dominant
heterotrophic and opportunistic taxa at Floreana (left cluster in Figs 1 and 4). The relatively
high foraminiferal abundance (mean 1.2%) exhibited in the Group II assemblages (Group I
mean: 0.35%; Fig 5) further set these sites apart within the cluster analysis and CCA. The
coupling of low foraminiferal diversity with high species dominance, as well as the abundant
opportunistic taxa within Floreana and San CristoÂbal assemblages (Figs 3 & 5; Table 3) may
indicate ecological stress at these sites. Hallock et al.  suggested high foraminiferal
opportunism to signal stressed systems. Furthermore, Floreana CCA results may indicate some
asof-yet unidentified local oceanographic influence. Floreana samples were collected in the
vicinity of Corona del Diablo, an eroding volcanic cinder cone, which provides a unique
shallow water environment for a multitude of invertebrate species not found elsewhere in
GalaÂpagosÐincluding a large fungiid coral bed [56±58].
Benthic foraminifera within the context of the FORAM-Index (Fig 2A; ), did not
correspond with the larger nutrient-driven trends of carbonate sediment production in the
GalaÂpagos, as outlined in Fig 2B and discussed in Humphreys et al. . The latter authors found
coral and coralline algae to define most of the time-averaged sedimentological variation
throughout the archipelago along a clear nutrient gradient delineating the coral reef turn
onturn off zone (CRTTZ; ). Specifically, there was a photozoan association in the low
mesotrophic-upper oligotrophic (<CRTTZ, e.g. within the reef building realm) sediments of
Darwin reef, a mixed photozoan-heterozoan association in all sediments from the moderate
upwelling (~CRTTZ) southeastern archipelago (Floreana to San CristoÂbal; Fig 2B), and a
heterozoan association in the waters of western Isabela, which is directly impacted by high
nutrient EUC upwelling (Fig 4B).
In contrast to the nutrient and coral sediment relationship in Fig 2B, the FI (calculated
using time-averaged foraminiferal functional groups; Fig 2A) placed a majority of the low
mesotrophic southeastern sites in proximity to the strongly eutrophic Isabela site (southwest)
(Figs 1 & 2A). Additionally, the FI revealed all southern islands to range from low marginal
(not conducive to symbiont recovery after disturbance), to stressed (not favorable to symbiont
development). By exhibiting a general rise in coral-derived sediment along an increasing FI
gradient (Fig 2A), the GalaÂpagos FI results are in general agreement with the FI trends outlined
in Hallock et al. . However, coral production did not follow the foraminiferal-based FI as
directly as it did nutrients and the CRTTZ (Fig 2). This indicates that significant
oceanographic parameters in addition to nutrients may have been at play in the distribution of
foraminiferal symbiont producers, resulting in the observed alignment of Isabela FI with that of a
majority of southeastern islands.
The FI, which can be used as a predictor of symbiont recovery potential , may shed
light on the impact of ENSO on the GalaÂpagos carbonate systems. At Darwin island, which
had a FI >4 (Fig 2), high rates of coral recovery were observed following the recent 1982/83
and 1997/98 ENSO events [
]. In contrast, minimal to no recovery took place in the low FI
values (<4) of the southern archipelago [
] which indicates marginal to stressed systems.
For example, time-averaged Santa FeÂ carbonates contained moderate quantities of
coralderived sediment (28%; ) and FI values of 2.2Ðindicating low marginal conditions for
coral reef development with a low probability of recovery (Fig 2A). This setting had previously
been shown to exhibit a strong response to recent ENSO events through extensive coral
degradation  and a subsequent shift to a rubble, rhodolith, and sand system [
]. Hence, the
16 / 25
regionally-high coral sedimentary signature at Santa FeÂ (Fig 2A) may have been symptomatic
of the recent stronger-than-normal ENSO-influenced ecological shift, which caused an influx
of degraded coral material into the sediments, and not representative of current coral
production at the site.
FI, in concert with CCA, indicated potentially contrasting outcomes for Baltra and Darwin
foraminifera, and the possibility for ecological shifting among carbonate producers at the
Baltra site, akin to post-1982 Santa FeÂ. For instance, time-averaged foraminiferal assemblages
from Baltra and Darwin followed similar trends within the CCA, with μSST and Oarag
positively corresponding with dominant larger symbiont bearing taxa (Fig 4). However, while
rising mean sea surface temperature is forecasted to continue this century, which could benefit
these larger taxa, ocean pH and aragonite saturation are predicted to further decline in the
coming decades [
], which could negate these benefits. Furthermore, the strong differences
in FI between Baltra and Darwin (Fig 2A), could indicate a more tenuous scenario for
symbiont bearers at the low marginal FI site of Baltra, whichÐlike Santa FeÂÐplots near the FI
threshold (FI = 2) for environments unsuitable for symbiont activity. Unlike Darwin, the FI
values at Baltra, while shaped within the context of Holocene ENSO, indicate time-averaged
foraminiferal assemblages that teeter at the limit for endosymbiont development. Hence these
assemblages are increasingly unlikely to recover from anomalously strong ENSO events,
which are predicted to increase in frequency [
There is currently no consensus for the cause of the low benthic foraminiferal
representation (0.7%) within the time-averaged shallow water sediments of the GalaÂpagos [13, 14].
However, evidence suggests that ENSO may have played a major role in the low abundance of these
sensitive indicator species. While studying the effects of ENSO events on benthic foraminifera
within the bank reefs of northern Bahia, Brazil, Kelmo and Hallock  found that
environmental stress brought on by the stronger-than-normal 1997±1998 ENSO led to dramatic losses
in foraminiferal density in all shallow reef environments. They concluded that the 1997 El
Niño resulted in declines of symbiont bearing taxa, through a combination of elevated
temperature and reduced turbidity, as well as a collapse of heterotrophic taxa due to the depression of
nutrient-controlled food resources. Further, Kelmo and Hallock  found that La
Niña-associated nutrient increases resulted in a rebound in heterotrophic taxa before other forms. These
findings indicate that, while the time-averaged assemblages within our samples were shaped
within the context of late Holocene ENSO variability, ENSOÐparticularly strong ENSO events
like those in 1982±1983 and 1997±1998, which devastated corals throughout the southern
GalaÂpagos islands [
], might have had similar effects on benthic foraminiferal
communities in the region. The long-term patterns of low abundance among time-averaged
foraminiferal populations throughout the GalaÂpagos (including at the Darwin reef site in the far
northern archipelago) hint toward chronic environmental oceanographic stress as an inhibitor
of post-ENSO foraminiferal reboundÐkeeping overall foraminifera numbers in the GalaÂpagos
low. Additionally, the dominant μChl.An signature within the Cluster 3 CCA sites (Fig 4),
which are strongly indicative of La Niña nutrient anomalies in the GalaÂpagos, suggested a
close association among these southern heterotrophic (particularly agglutinated) taxa to
repeated cycles of La Niña nutrient conditionsÐsimilar to those which drove the observed
heterotrophic rebound in Bahia Brazil . Ultimately, these findings may help explain the
geographic transition toward hyaline and/or agglutinated communities in the southern
GalaÂpagos, for these foraminiferal taxa are more resistant to the higher background nutrient
and lower pH conditions.
The low abundances of time-averaged foraminifera in the northern Darwin reef sediments
(avg. 0.45%; 7 samples spanning the reef) were unexpected. However, while Darwin reef is not
as directly impacted by EUC oceanographic effects, it experiences peripheral EUC nutrients
17 / 25
during La Niña, as well as the highest temperature anomalies of any GalaÂpagos island during
ENSO (Table 1). These factors would likely have had significant impacts on foraminiferal
densities during repeated ENSO cycles. Additionally, high resolution in situ data logger
temperature measurements, taken at a depth of 12m on Darwin reef (Fig 7A), revealed this
coraldominant environment to experience strong temperature instability through time. This
indicates that the tropical , and higher pH, Darwin island site (8.07) contains foraminifera that
are repeatedly temperature stressed, which likely contributed to the record of low
(time-averaged) foraminiferal sediment abundance for the site (Fig 5).
Regression tree analysis (Fig 8) offered additional support of Holocene ENSO impacts
on these time-averaged foraminiferal assemblages. Although the effects of elevated
nutrients and low pH in the southern GalaÂpagos cannot be ignored, regression tree analysis
(Fig 8) indicated that foraminifera may be most affected by positive temperature
anomalies over time. When incorporated into the CCA and FI findings, this regression tree
analysis further supported the argument that long-term and repeated exposure to El Niño
(resulting in positive temperature anomalies) served as the primary suppressor of
foraminifera throughout the archipelago, while high nutrient / low pH waters in the southern
sites may have hindered the recovery of some species and resulted in a dominance of
heterotrophic taxa over time. This caused the resultant `marginal' and `stressed' FI values
seen in the southern GalaÂpagos samples. Furthermore, the results supported a previous
finding of an overriding high temperature and low nutrient ENSO signal over all shallow
water carbonate producers throughout the GalaÂpagos .
Benthic foraminifera under naturally suppressed pH conditions
The notably high proportion of broken and abraded tests in GalaÂpagos samples was evident in
all test structure types, including the large hyaline species Poroeponides cribrorepandus (Fig 9).
However, porcellaneous and agglutinated taxa were heavily affected by degradation, with some
species rendered unidentifiable due to extensive test abrasion, fracturing, and dissolution
effects. For calcareous species, these dissolutions patterns likely reflected trends in test
magnesium to calcium ratios (Mg/Ca), with high magnesium calcite skeletons being more susceptible
to dissolution than those with low magnesium calcite skeletons [65±67]. Hence, high-Mg
calcifiers would be the first to be negatively affected by a declining saturation state and ocean pH in
shallow waters [
]. For calcitic foraminifera, porcellaneous benthic taxa tend to have higher
Mg/Ca ratios than hyaline taxa [
]. The regionally high CO2/low pH extremes, from EUC
and periodic La Niña anomalies (resulting in unusually intense periods of EUC upwelling),
create conditions adverse to long-term porcellaneous development and may have allowed
hyaline taxa or diverse agglutinated forms to proliferate. Indeed, this may explain the low densities
(10% of all foraminiferal production) of the high magnesium Quinqueloculina (12±16 mol%
], relative to other shallow water environments of the eastern tropical Pacific and tropical
regions globally. For example, Fajemila et al.  reported more than 90 species of
Quinqueloculina as well as predominances of the genera in near shore habitats of Moorea island, French
Polynesia. Similarly, Quinqueloculina were reported to contribute more than 60% of shallow
water foraminifera at La Paz, Gulf of California, Mexico [
]. The poor preservation of
Poroeponides cribrorepandus, particularly within San CristoÂbal and Floreana samples, likely
stemmed from the high magnesium content of this large hyaline species. For reference,
Blackmon and Todd [
] reported magnesium contents for this species at 13 mol % which were
strikingly high compared to the generally well-preserved Elphidium crispum (3.3 mol %) found
in the same GalaÂpagos samples (low percentages of Elphidium specimens were found in some
southern samples, however; Fig 9).
18 / 25
Oceanographic influence of Galapagos shallow water benthic foraminifera
Fig 9. Scanning electron micrographs of select foraminifera species exhibiting varying degrees of test degradation. 1.
Elphidium crispum subcrispum; 2. Poroeponides cribrorepandus; 3. Sphaerogypsina globulus; 4.Quinqueloculina sp.; 5. Textularia
sp.; 6. Unidentified agglutinated fragment; 7. Cibicides(?) sp.; 8. Elphidium sp.
The poor preservation of agglutinated taxa in GalaÂpagos samples was likely driven primarily
by physical processes over chemical alteration of tests. In an examination of test degradation
patterns in benthic foraminifera from the tropical, intertidal communities from Cleveland
Bay, Australia, Berkeley and colleagues [
] found the dominant alteration process of the
calcareous tests examined to be dissolution, while agglutinated tests showed a more arbitrary
pathway of degradation, related to an initial loss of their organic cement coating, followed by a
predominant physical-mechanical process. However, as agglutinated tests are inherently
weaker than their calcareous counterparts and only a small amount of chemical degradation of
the organic cements and test material is needed to undermine the entire test structure [
must be considered that the high CO2 environments of the southern GalaÂpagos could serve to
further weaken the structural integrity of agglutinated foraminifera, leading to the observed
test-fracturing patterns within these samples. Ultimately, it must be stressed that carbonate
dissolution is complex and potentially caused by a number of processes including corrosive
sediment pore waters and bacterial destruction [
] in the taphonomically active zone
(TAZ), which sediments must pass through prior to permanent burial . It is important for
future investigations to delineate living from dead assemblages if we are to better understand
the pathways to dissolution and fossilization in the regionÐinsights which could also clarify
richness and diversity patterns in GalaÂpagos foraminifera.
The inverse correlation between foraminiferal species richness and foraminiferal
abundance at each collection island (Fig 5A & 5B) was not anticipated. However, it agreed with
previous studies on foraminifera in high nutrient, low pH environments [
]. For example,
benthic foraminiferal assemblages along a transect of declining pH values (comparable to pH
values in the GalaÂpagos) near natural CO2 seeps in Papua New Guinea exhibited an observed
drop in foraminiferal abundance before a decline in diversity [
Low pH is also known to have an additional detrimental effect on the metabolic function of
some symbiont-bearing species [
]. For example, in a study of the influence of reduced pH
on the growth rate of the larger symbiont bearing foraminifera Marginopora rossi, Reymond
et al. [
] reported a drastic reduction in growth through dissolution and inhibition of
precipitated calcite at the site of calcification [
]. Furthermore, rates of photosynthesis in this species
decreased (primarily through a decline in endosymbiont cell density [
]) along a declining
pH gradient, even at pH values similar to those observed in the southern GalaÂpagos
Archipelago. These findings may help explain the low abundance of porcellanous and larger symbiont
In combination with the outcome of CCA (Fig 4), the ternary diagrams reveal an additional
trend among heterotrophic taxa in the assemblages from Española, Santa FeÂ, and Isabela
islands toward a higher percentage of agglutinated species along decreasing pH regimes (Figs 4
& 6A). Similar findings were reported in a study of foraminifera within the low pH waters
surrounding volcanic vents off the island of Ischia, Italy, which found a transition from calcareous
forms to agglutinated taxa along a declining pH gradient ([
]; Fig 6B).
This study represents the first statistical analysis of the shallow water benthic foraminiferal
communities of the GalaÂpagos Archipelago, Eastern Tropical Pacific (ETP), and their
relationship to major regional oceanographic controls. Results indicate long term and repeated ENSO
20 / 25
temperature anomalies to influence low foraminiferal density in GalaÂpagos carbonate
sediments. Naturally low levels of pHÐinduced by La Niña and equatorial undercurrent (EUC)
upwellingÐmay have primarily inhibited post ENSO recovery and, in concert with EUC
upwelling nutrients, resulted in heterotrophic dominance in the southern archipelago. These
oceanographic conditions result in lowered FORAM-Index values in the southern GalaÂpagos
indicating environments not conducive to endosymbiont development. This further supports
the well documented ENSO-induced collapse of coral communities throughout the southern
archipelago following the strong ENSO events of 1982±1983 and 1997±1998. The combined
ENSO-ocean acidification effect, in concert with the predicted increase in the frequency of
strong ENSO [
] and declining ocean pH [
], could result in a further increase of
heterotrophic foraminiferal taxa. Additionally, forecasts have been made for the decline and `ecological
extinction' of benthic foraminifera globally, due to declining ocean pH, by the end of the
]. Hence, the extremely low abundances throughout the GalaÂpagos may signal a system
already well advanced on the path towards ecological extinction with respect to foraminifera.
With benthic foraminifera considered to be important indicators of environmental change,
the herein-presented results help to better understand the complex interactions driving the
unique foraminiferal character of the region, and advance our knowledge of and predictions
for the biogeophysical implications of a high CO2 world.
S1 Fig. Permutation tests for canonical correspondence analysis (CCA) triplot results (Fig
4). Results show high significance for both CCA plots along the displayed axes (bold).
S1 Table. Count data for all foraminifera species encountered in each sample.
This research was funded by a Natural Sciences and Engineering Research Council of Canada
Discovery grant (1303409) to JH, and a Geological Society of America grant (10536±14) to
AH. Our thanks also to Dr. Kenneth Finger of the Museum of Paleontology at the University
of California for discussions concerning the taxonomy of Pleistocene benthic foraminifera
from the GalaÂpagos Archipelago with SEM images of these taxa provided. Additional thanks
to Dr. Alberto Zirino and colleagues for providing Gulf of California pH data.
Data curation: Alexander F. Humphreys, Bernhard Riegl.
Formal analysis: Alexander F. Humphreys.
Funding acquisition: Jochen Halfar, Hildegard Westphal.
Methodology: Alexander F. Humphreys, Jochen Halfar, James C. Ingle, Derek Manzello,
Project administration: Jochen Halfar, Claire E. Reymond, Bernhard Riegl.
Resources: Jochen Halfar, Bernhard Riegl.
Software: Bernhard Riegl.
Supervision: Jochen Halfar, Bernhard Riegl.
21 / 25
Validation: James C. Ingle.
Writing ± original draft: Alexander F. Humphreys.
Writing ± review & editing: Jochen Halfar, James C. Ingle, Derek Manzello, Claire E.
Reymond, Hildegard Westphal, Bernhard Riegl.
22 / 25
23 / 25
24 / 25
1. Iglesias-Rodriguez MD , Armstrong R , Feely R , Hood R , Kleypas J , Milliman JD , et al. Progress made in study of ocean's calcium carbonate budget . Eos, Transactions American Geophysical Union. 2002 ; 83 ( 34 ): 365 ± 75 .
Feely RA , Sabine CL , Lee K , Berelson W , Kleypas J , Fabry VJ , et al. Impact of anthropogenic CO2 on the CaCO3 system in the oceans . Science . 2004 ; 305 ( 5682 ): 362 ±6. https://doi.org/10.1126/science.
1097329 PMID: 15256664
Journal of Eukaryotic Microbiology . 2008 ; 55 ( 3 ): 163 ±9. https://doi.org/10.1111/j.1550- 7408 . 2008 .
00321. x PMID : 18460153
Klicpera A , Michel J , Westphal H . Facies patterns of a tropical heterozoan carbonate platform under eutrophic conditions: the Banc d'Arguin, Mauritania . Facies. 2015 ; 61 ( 1 ): 421 .
Murray JW . Ecology and distribution of benthic foraminifera . Biology of Foraminifera Academic Press, London. 1991 : 221 ± 53 .
Langer MR , Silk MT , Lipps JH . Global ocean carbonate and carbon dioxide production; the role of reef foraminifera . Journal of Foraminiferal Research . 1997 ; 27 ( 4 ): 271 ± 7 .
Kelmo F , Hallock P . Responses of foraminiferal assemblages to ENSO climate patterns on bank reefs of northern Bahia, Brazil: A 17-year record . Ecological Indicators . 2013 ; 30 : 148 ± 57 .
Schafer CT . Monitoring nearshore marine environments using benthic foraminifera: some protocols and pitfalls . Micropaleontology . 2000 ; 46 : 161 ± 9 .
MartÂõnez-ColoÂn M , Hallock P , Green-RuÂõz C. Strategies for using shallow-water benthic foraminifers as bioindicators of potentially toxic elements: a review . Journal of Foraminiferal Research . 2009 ; 39 ( 4 ): 278 ± 99 .
Ward JN , Pond DW , Murray JW . Feeding of benthic foraminifera on diatoms and sewage-derived organic matter: an experimental application of lipid biomarker techniques . Marine Environmental Research . 2003 ; 56 ( 4 ): 515 ± 30 . https://doi.org/10.1016/S0141- 1136 ( 03 ) 00040 - 0 PMID: 12860436 Hallock P , Lidz BH , Cockey-Burkhard EM , Donnelly KB . Foraminifera as bioindicators in coral reef assessment and monitoring: the FORAM index . Coastal Monitoring through Partnerships: Springer; 2003 . p. 221 ± 38 .
Fajemila OT , Langer MR , Lipps JH . Spatial patterns in the distribution, diversity and abundance of benthic foraminifera around Moorea (Society Archipelago, French Polynesia) . PloS one . 2015 ; 10 ( 12 ): e0145752. https://doi.org/10.1371/journal.pone.0145752 PMID: 26710320 Reymond CE , Zihrul K , Halfar J , Riegl B , Humphreys A , Westphal H . Heterozoan carbonates from the equatorial rocky reefs of the GalaÂpagos Archipelago . Sedimentology. 2016 ; 63 ( 4 ): 940 ± 58 .
Humphreys AF , Halfar J , Rivera F , Manzello D , Reymond CE , Westphal H , et al. Variable El Niño± Southern Oscillation influence on biofacies dynamics of eastern Pacific shallow-water carbonate systems . Geology . 2016 ; 44 ( 7 ): 571 ± 4 .
Okey T , Banks S , Born AF , Bustamante RH , Calvopina M , Edgar GJ , et al. A trophic model of a Gal aÂpagos subtidal rocky reef for evaluating fisheries and conservation strategies . Ecological Modelling . 2004 ; 172 ( 2 ): 383 ± 401 .
Glynn PW , Wellington GM . Corals and coral reefs of the Gal aÂpagos Islands : Univ of California Press; 1983 .
Manzello DP . Ocean acidification hotspots: spatiotemporal dynamics of the seawater CO2 system of eastern Pacific coral reefs . Limnology and Oceanography . 2010 ; 55 ( 1 ): 239 ± 48 .
Di Carlo G , d' Ozouville N., Henderson S. , de Koning F., Larrea I. , Ortiz F. , Pidgeon E. , Spurrier L. , Suarez L. Introduction . In: Carlo ILaGD, editor. Climate Change Vulnerability Assessment of the Galapagos Islands. USA: I. Larrea and G. Di Carlo 2011 .
20. Glynn PW , MateÂ JL , Baker AC , CalderoÂn MO . Coral bleaching and mortality in Panama and Ecuador during the 1997±1998 El Niño±Southern Oscillation event: spatial/temporal patterns and comparisons with the 1982±1983 event . Bulletin of Marine Science . 2001 ; 69 ( 1 ): 79 ± 109 .
21. Aller RC . Carbonate dissolution in nearshore terrigenous muds: the role of physical and biological reworking . The Journal of Geology . 1982 ; 90 ( 1 ): 79 ± 95 .
22. Cockey E , Hallock P , Lidz B . Decadal-scale changes in benthic foraminiferal assemblages off Key Largo, Florida . Coral Reefs . 1996 ; 15 ( 4 ): 237 ± 48 .
23. Halfar J , Ingle JC . Modern warm-temperate and subtropical shallow-water benthic foraminifera of the southern Gulf of California, Mexico . Journal of Foraminiferal Research . 2003 ; 33 ( 4 ): 309 ± 29 .
24. McCulloch IA . Qualitative observations on Recent foraminiferal tests with emphasis on the eastern Pacific , Part I : University of Southern California; 1977 : 1 ± 330 .
25. McCulloch IA . Qualitative observations on Recent foraminiferal tests with emphasis on the eastern Pacific , Part II : University of southern California; 1977 : 331 ± 676 .
26. McCulloch IA . Qualitative observations on Recent foraminiferal tests with emphasis on the eastern Pacific , Part III : University of southern California; 1977 :pls. 49 - 248 .
27. Cushman J , McCulloch I. A report on some arenaceous foraminifera . Allan Hancock Pacific Expeditions . 1939 ; 6 ( 1 ):1± 113 .
28. Cushman J , McCulloch I. Some Nonionidae in the collections of the Allan Hancock Foundation. Algunos Nonionidae en las colecciones de la FundacioÂn Allan Hancock . Allan Hancock Pacific Expeditions . 1940 ; 6 ( 3 ): 145 ± 78 .
29. Cushman J , McCulloch I. Some Virgulininae in the collections of the Allan Hancock Foundation , Southern California University, Publications, Allan Hancock Pacific Expeditions. Los Angeles; 1942 ; 6 ( 4 ): 179 ± 230 .
30. Cushman J , McCulloch I. Some Lagenidae in the collections of the Allan Hancock Foundation. Algunos Lagenidae en las colecciones de la FundacioÂn Allan Hancock . Allan Hancock Pacific Expeditions . 1950 ; 6 ( 6 ): 295 ± 364 .
31. Lalicker C , McCulloch I. Some Textulariidae of the Pacific Ocean: Allan Hancock Pacific Expeds . 1940 ; 6 ( 2 ): 115 ±143, pls. 13 ± 16 .
32. Bandy OL . Distribution of foraminifera, radiolaria and diatoms in sediments of the Gulf of California . Micropaleontology. 1961 : 1 ± 26 .
33. Brenner GJ . Results of the Puritan-American Museum of Natural History Expedition to Western Mexico. 14, A zoogeographic analysis of some shallow-water Foraminifera in the Gulf of California. Bulletin of the AMNH; 1962 ; 123 ( 5 ).
34. Crouch RW , Poag CW . Benthic foraminifera of the Panamanian Province: distribution and origins . Journal of Foraminiferal Research . 1987 ; 17 ( 2 ): 153 ± 76 .
35. Cushman JA , Valentine WW . Shallow-water foraminifera from the Channel Islands of southern California: Stanford University Press; 1930 .
36. Finger KL , Hickman CS , James MJ , Lipps JH , Peterson DE , Pitt LI , et al., editors. Pleistocene marine paleoenvironments on the Galapagos islands . GSA Abstracts with Programs; 2007 ; 39 ( 6):399 (see www .ucmp.berkeley.edu/people/klf/GSA_2007_Poster.pdf).
37. Golik A , Phleger FB . Benthonic foraminifera from the Gulf of Panama . Journal of Foraminiferal Research . 1977 ; 7 ( 2 ): 83 ± 99 .
38. Lankford RR , Phleger FB . Foraminifera from the nearshore turbulent zone, western North America . Journal of Foraminiferal Research . 1973 ; 3 ( 3 ): 101 ± 32 .
39. Natland ML , E W., Cruise to the Gulf of California report on the Pleistocene and Pliocene foraminifera . Geological Society of America Memoirs. Part IV . 1940 ; 43 :1± 65 .
40. Segura-Vernis LR , de LB , Schafer CT. Carreño A.L. , ForaminÂõferos y ostracaÂcodos de Laguna Sur, Mexico. Investig. Mar. CICIMAR, lam. I-V . Monitoring nearshore marine environments using benthic foraminifera: some protocols and pitfalls . Micropaleontology 46 161169 . 2000 ; 6 ( 1 ): 195 ± 223 .
41. Uchio T. Ecology of living benthonic foraminifera from the San Diego , California, area (No. 5) . Cushman Foundation for Foraminiferal Research . 1960 ; Spec . Publ. no. 5 : 1 ±72, pls. 1 ± 10 .
42. Walton WR , Baja B . Ecology of living benthonic foraminifera, Todos Santos bay, Baja California . Journal of Paleontology . 1955 ; 29 ( 6 ): 952 ± 1018 .
43. Engle V . Application of the indicator evaluation guidelines to an index of benthic condition for Gulf of Mexico estuaries. Evaluation Guidelines for Ecological Indicators EPA/620/R-99 /005 . 2000 ; 3 :1± 29 .
44. Hallock P. Coral reefs, carbonate sediments, nutrients, and global change . The History and Sedimentology of Ancient Reef Systems . 2001 : 387 ± 427 .
45. Legendre P , Gallagher ED. Ecologically meaningful transformations for ordination of species data . Oecologia . 2001 ; 129 ( 2 ): 271 ± 80 . https://doi.org/10.1007/s004420100716 PMID: 28547606
46. Legendre P , Legendre LF . Numerical ecology, 2nd Edition: Elsevier; 1998
47. Borcard D , Gillet F , Legendre P. Unconstrained ordination . Numerical Ecology with R : Springer; 2011 . p. 115 ± 51 .
48. Oksanen JF , Blanchet G , Friendly M , Kindt R , Legendre P , McGlinn D , et al. vegan: Community Ecology Package. R package version 2 .4-2 https://CRAN.R-project.org/package=vegan. 2017 .
49. NASA Goddard Space Flight Center OEL , Ocean Biology Processing Group. Moderate-resolution Imaging Spectroradiometer (MODIS) Aqua; NASA OB .DAAC, Greenbelt, MD , USA.
50. NASA, Ocean C , Ocean L , Greenbelt MD . Space Flight Processing Group. Seaviewing Wide Fieldofview Sensor (SeaWiFS); NASA OB .DAAC, Greenbelt, MD , USA.
51. National Center for Atmospheric Research Staff (Eds). Last modified 13 Oct 2016 . "The Climate Data Guide: SODA: Simple Ocean Data Assimilation." Retrieved from https://climatedataguide.ucar.edu/ climate-data/soda-simple-ocean-data-assimilation.
52. Loeblich AR Jr, Tappan H. Suprageneric classification of the Foraminiferida (Protozoa) . Micropaleontology . 1984 : 1 ± 70 .
53. Zuur AF , Ieno EN , Elphick CS . A protocol for data exploration to avoid common statistical problems . Methods in Ecology and Evolution . 2010 ; 1 ( 1 ):3± 14 .
54. Ripley B. tree: Classification and Regression Trees . R package version 1 https :// CRANRprojectorgpackagetree. 2016 : 0 ± 37 .
55. Betzler C , Brachert T , Nebelsick J. The warm temperate carbonate province: a review of the facies, zonations, and delimitations . Courier Forschungsinstitut Senckenberg . 1997 ; 201 : 83 ± 99 .
56. Feingold JS . Effects of elevated water temperature on coral bleaching and survival during El Niño disturbance events . 1995 .
57. Feingold JS . Coral survivors of the 1982 ±83 El Niño-Southern Oscillation , Galapagos Islands, Ecuador. Coral Reefs . 1996 ; 15 ( 2 ): 108 .
58. Feingold JS . Responses of three coral communities to the 1997±98 El Niño ±Southern Oscillation: GalaÂ- pagos Islands , Ecuador. Bulletin of Marine Science . 2001 ; 69 ( 1 ): 61 ± 77 .
59. Riegl B , Glynn PW , Wieters E , Purkis S , Dangelo C , Wiedenmann J . Water column productivity and temperature predict coral reef regeneration across the Indo-Pacific . Scientific Reports . 2015 ; 5 ( 1 ).
60. Glynn PW , Riegl B , Purkis S , Kerr JM , Smith TB . Coral reef recovery in the GalaÂpagos Islands: the northernmost islands (Darwin and Wenman) . Coral Reefs . 2015 ; 34 ( 2 ): 421 ± 36 .
61. Glynn PW . State of coral reefs in the GalaÂpagos Islands: natural vs anthropogenic impacts . Marine Pollution Bulletin . 1994 ; 29 ( 1 ±3): 131 ± 40 .
62. Halfar J , Riegl B . From coral framework to rhodolith bed: sedimentary footprint of the 1982/1983 ENSO in the GalaÂpagos . Coral Reefs . 2013 ; 32 ( 4 ): 985 .
63. Orr JC , Fabry VJ , Aumont O , Bopp L , Doney SC , Feely RA , et al. Anthropogenic ocean acidification over the twenty-first century and its impact on calcifying organisms . Nature . 2005 ; 437 ( 7059 ): 681 ±6. https://doi.org/10.1038/nature04095 PMID: 16193043
64. Cai W , Borlace S , Lengaigne M , Van Rensch P , Collins M , Vecchi G , et al. Increasing frequency of extreme El Niño events due to greenhouse warming . Nature Climate Change . 2014 ; 4 ( 2 ): 111 ± 6 .
65. Kleypas JA , Buddemeier RW , Archer D , Gattuso J-P , Langdon C , Opdyke BN . Geochemical consequences of increased atmospheric carbon dioxide on coral reefs . Science . 1999 ; 284 ( 5411 ): 118 ± 20 . PMID: 10102806
66. Morse JW , Arvidson RS , LuÈttge A. Calcium carbonate formation and dissolution . Chemical reviews . 2007 ; 107 ( 2 ): 342 ± 81 . https://doi.org/10.1021/cr050358j PMID: 17261071
67. Sinutok S , Hill R , Doblin MA , Wuhrer R , Ralph PJ . Warmer more acidic conditions cause decreased productivity and calcification in subtropical coral reef sediment-dwelling calcifiers . Limnology and Oceanography . 2011 ; 56 ( 4 ): 1200 ± 12 .
68. Bentov S , Erez J . Impact of biomineralization processes on the Mg content of foraminiferal shells: A biological perspective . Geochemistry, Geophysics, Geosystems. 2006 ; 7 ( 1 ).
69. Blackmon PD , Todd R . Mineralogy of some foraminifera as related to their classification and ecology . Journal of Paleontology . 1959 : 1 ± 15 .
70. Berkeley A , Perry CT , Smithers SG . Taphonomic signatures and patterns of test degradation on tropical, intertidal benthic foraminifera . Marine Micropaleontology . 2009 ; 73 ( 3 ): 148 ± 63 .
71. Freiwald A . Bacteria-induced carbonate degradation: a taphonomic case study of Cibicides lobatulus from a high-boreal carbonate setting . Palaios . 1995 : 337 ± 46 .
72. Murray JW , Alve E. Natural dissolution of modern shallow water benthic foraminifera: taphonomic effects on the palaeoecological record . Palaeogeography, Palaeoclimatology, Palaeoecology. 1999 ; 146 ( 1 ): 195 ± 209 .
73. Uthicke S , Momigliano P , Fabricius KE , Vogel N. High risk of extinction of benthic foraminifera in this century due to ocean acidification . Scientific Reports . 2013 ; 3 .
74. Reymond CE , Lloyd A , Kline DI , Dove SG , Pandolfi JM . Decline in growth of foraminifer Marginopora rossi under eutrophication and ocean acidification scenarios . Global Change Biology . 2013 ; 19 ( 1 ): 291 ± 302 . https://doi.org/10.1111/gcb.12035 PMID: 23504740
75. Dias B , Hart M , Smart C , Hall-Spencer J . Modern seawater acidification: the response of foraminifera to high-CO2 conditions in the Mediterranean Sea . Journal of the Geological Society . 2010 ; 167 ( 5 ): 843 ± 6 .
76. Manzello DP , Enochs IC , Bruckner A , Renaud PG , Kolodziej G , Budd DA , et al. GalaÂpagos coral reef persistence after ENSO warming across an acidification gradient . Geophysical Research Letters . 2014 ; 41 ( 24 ): 9001 ± 8 .
77. Kidwell SM . Taphonomic comparison of passive and active continental margins: neogene shell beds of the atlantic coastal plain and northern Gulf of California . Palaeogeography, Palaeoclimatology, Palaeoecology. 1988 ; 63 : 201 ± 223 .