SYSTEMATIC REVIEW

published: 01 March 2022

doi: 10.3389/fevo.2022.844021

Sea Turtle Epibiosis: Global Patterns

and Knowledge Gaps
Nathan J. Robinson 1,2† and Joseph B. Pfaller 3,4* †
1
Fundación Oceanogràfic, Ciudad de las Artes y las Ciencias, Valencia, Spain, 2 Institut de Ciències del Mar, Spanish
National Research Council (CSIC), Barcelona, Spain, 3 Caretta Research Project, Savannah, GA, United States, 4 Archie Carr
Center for Sea Turtle Research, Department of Biology, University of Florida, Gainesville, FL, United States

Competition for space drives many marine propagules to colonize the external surfaces
of other marine organisms, a phenomenon known as epibiosis. Epibiosis appears to
be a universal phenomenon among sea turtles and an extensive body of scientific
literature exists describing sea turtle-epibiont interactions. When viewed in isolation,
however, these epibiont “species lists” provide limited insights into the factors driving
patterns in taxonomic diversity on a global scale. We conducted an exhaustive literature
review to collate information on sea turtle-epibiont interactions into a global database.
As studies involving meio- and micro-epibionts, as well as plants, are limited, we
exclusively focused on animal, macro-epibionts (>1 mm). We identified 304 studies that
Edited by: included a combined total of 1,717 sea turtle-epibiont interactions involving 374 unique
Adrienne M. S. Correa, epibiont taxa from 23 Higher Taxon categories (full Phylum or select phyla differentiated
Rice University, United States
by Subphylum/Class/Subclass). We found that loggerhead turtles hosted the highest
Reviewed by:
Jeroen Ingels,
taxonomic richness (262 epibiont taxa) and diversity, including representative taxa from
Florida State University, United States 21 Higher Taxon categories, followed by hawksbill, green, olive ridley, leatherback,
Benny K. K. Chan,
Kemp’s ridley, and flatback turtles. In addition, the taxonomic richness for all turtle
Academia Sinica, Taiwan
species except leatherbacks was projected to increase with additional studies. We
*Correspondence:
Joseph B. Pfaller found that taxonomic richness not only varies between species but also between well-
jpfaller@ufl.edu studied populations of loggerhead turtles. Lastly, we assessed biases in the current
† These authors have contributed
literature and identified knowledge gaps for certain species (e.g., Kemp’s ridleys and
equally to this work and share senior
authorship
flatbacks), life stages (e.g., juveniles), habitats (e.g., oceanic habitats), and geographic
regions (e.g., central Pacific, east Atlantic, and east Indian oceans). Our hope is that
Specialty section: this database will serve as a foundational platform for future studies investigating
This article was submitted to
Coevolution, global patterns of the diversity, ecological function, and evolutionary origins of sea
a section of the journal turtle epibiosis.
Frontiers in Ecology and Evolution
Keywords: commensalism, epibiont, parasite, species richness, symbiosis
Received: 27 December 2021
Accepted: 08 February 2022
Published: 01 March 2022
Citation:
INTRODUCTION
Robinson NJ and Pfaller JB
(2022) Sea Turtle Epibiosis: Global
Competition for space drives marine propagules to colonize almost any exposed, undefended
Patterns and Knowledge Gaps. surface in the marine environment (Harder, 2009; Wahl, 2009). While colonization frequently
Front. Ecol. Evol. 10:844021. occurs on inanimate structures (e.g., submerged bedrock and dock pilings), the external surfaces
doi: 10.3389/fevo.2022.844021 of other marine organisms can also provide suitable substrata for settlement—resulting in

Frontiers in Ecology and Evolution | www.frontiersin.org 1 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

a phenomenon known as epibiosis. Epibiosis involves a single while also accounting for differences in sampling effort, could
host species and one or more colonizers called epibionts. Most therefore provide a foundation to analyze broad-scale patterns
epibionts are opportunistic organisms found “free living” in the in sea turtle epibiosis. Such an effort would also help identify
surrounding environment, while others are obligate commensals understudied species or regions, thereby guiding productive
of specific hosts. In complex and captivating cases of epibiosis, directions for future research.
epibionts colonize the bodies of migratory marine megafauna, To enact this important step in the field of sea turtle
such as sea turtles and cetaceans, and are transported as epibiosis, we conducted an exhaustive review of published
“hitchhikers” across marine habitats and even entire ocean basins scientific articles and gray literature (i.e., government reports,
(Ingels et al., 2020). theses and dissertations, and conference presentations) to collate
Sea turtles are renowned for hosting diverse communities information on sea turtle-epibiont species pairs as well as their
of epibionts, including representative taxa from almost every associated metadata. Because studies involving meio- or micro-
Phylum within Animalia (e.g., Frick et al., 1998; Lazo-Wasem epibionts, as well as plants, are limited, we exclusively focused on
et al., 2011; Corrêa et al., 2014). While robust, encrusting animal, macro-epibionts (>1 mm) (Hereafter, we use the term
forms, such as barnacles, tend to be more common and exhibit “epibionts” to exclusively refer to animal macro-epibionts unless
greater diversity (Zardus, 2021), other taxonomic groups that are stated otherwise). As an initial investigation of the information
less resistant to abrasion and dramatic shifts in environmental amassed in this global database, we first quantified and compared
conditions tend to be rarer and less diverse (e.g., Lazo- the taxonomic diversity of epibionts for each sea turtle species
Wasem et al., 2007; Perrault et al., 2015). Because epibiosis to answer two questions. (1) Which sea turtle species hosts
necessitates ecological overlap between host turtles and “free the greatest epibiont diversity? (2) Does the current literature
living” populations of epibionts and/or their propagules, the for each turtle species fully encompass the taxonomic richness
assemblages of epibionts found on sea turtles also tend to reflect of their epibiont communities? Next, we investigated similar
the regions and habitats where host turtles spend time (e.g., Reich questions among three well-studied populations of loggerhead
et al., 2010; Pfaller et al., 2014; Ten et al., 2019). Consequently, turtles (Caretta caretta). Last, we characterized the current
the presence of certain epibiont species or assemblages that epibiont literature for each sea turtle species in terms of turtle life
occupy specific regions (e.g., tropical, temperate, or polar) and/or stage, habitat type, and geographic distribution to assess biases
habitats (e.g., oceanic/pelagic or neritic/benthic) can serve as and identify knowledge gaps for future research.
indicators of the migratory movements and habitat preferences of
sea turtles (Casale et al., 2004; Frick and Pfaller, 2013). Similarly,
the diversity of sea turtle epibionts and the equally diverse ways METHODS
they interact with their hosts, means that the presence or absence
of particular epibiont taxa can also serve as indicators of the hosts’ Database Development
foraging preferences (Pfaller et al., 2014), social or reproductive Between March 2018 and December 2020, we conducted
behavior (Domènech et al., 2017; Robinson et al., 2017a), body a two-tiered literature search to compile all records of sea
condition and/or health status (Lazo-Wasem et al., 2007; Nolte turtle-epibiont interactions. A structured search was conducted
et al., 2020), and more. in Web of Science, Google Scholar, and Sea Turtle Online
Following a rich history of anecdotal reports dating back Bibliography (Archie Carr Center for Sea Turtle Research,
to Darwin (1851, 1854) and Pilsbry (1916), the epibiont University of Florida) using the following Boolean search
communities of sea turtles have received considerable attention terms: epibiont, epibiosis, epifauna, epibiota, and both common
(Frick and Pfaller, 2013). Most of this work has focused on and scientific names of the seven marine turtle species:
animal, macro-epibionts (>1 mm) as they are relatively easy to loggerhead turtle (C. caretta), green turtle (Chelonia mydas),
identify and sample; however, it is increasingly being realized that leatherback turtle (Dermochelys coriacea), hawksbill turtle
sea turtles also frequently host meio- and micro-epibionts as well (Eretmochelys imbricata), olive ridley turtle (Lepidochelys
as various plant species (Robinson et al., 2016; Ingels et al., 2020). olivacea), Kemp’s ridley turtle (Lepidochelys kempii), and flatback
There is now an extensive body of scientific literature describing turtle (Natator depressus). Secondly, an unstructured literature
the epibiotic diversity of sea turtle populations worldwide. search was conducted by reviewing the reference lists of relevant
However, most studies only report the species of epibionts publications and reports from the structured search. We
found on a single sea turtle species at a single locality. Viewed included any peer-reviewed scientific article, thesis/dissertation,
in isolation, these “species lists” provide limited inferences for conference presentation, and official report that contained
understanding the factors driving patterns in epibiont richness information on sea turtle epibiosis. When the same data were
and diversity on a global scale (Lazo-Wasem et al., 2011; Pinou presented in separate publications by the same author/s, we only
et al., 2019; Zardus, 2021). Moreover, because measures of included the data from the original source publication. We did
species richness are inherently connected to sampling effort, not include references published after December 2020 (i.e., those
comparisons of epibiotic diversity among turtle species and published between the completion of our literature search and
regions may be biased by variations in sample size (i.e., the the publication of this article).
number of studies or turtles surveyed) (Robinson et al., 2017b). We constrained our two-tiered literature search to only
Collating these “species lists” along with their associated metadata include records of sea turtle-epibiont interactions from (1)
(e.g., host species and life stage, geographic region, habitat, etc.), turtles surveyed in the wild, (2) animal epibionts (i.e., Kingdom

Frontiers in Ecology and Evolution | www.frontiersin.org 2 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

Animalia), and (3) macro-epibionts (>1 mm). We excluded TABLE 1 | Data parameters and descriptions extracted for each sea
turtle-epibiont interaction identified during an exhaustive review of published
records from turtles reared in captivity (e.g., Crespo-Picazo et al.,
articles and gray literature (i.e., government reports, theses and dissertations, and
2017) to focus on naturally occurring instances of epibiosis. conference presentations), including possible data options for each category.
We excluded epibiotic interactions involving plants (namely
algae) as very few studies have included plants and, those that Parameter Description Data options
have, tend to show relatively low taxonomic diversity. Lastly, Host species Species of host sea Caretta caretta (CC), Chelonia
we focused our review exclusively on macro-epibionts (>1 mm) turtle. mydas (CM), Dermochelys
even though sea turtles are known to host meio- or micro- coriacea (DC), Eretmochelys
epibionts (<1 mm) (e.g., Robinson et al., 2016; Ingels et al., imbricata (EI), Lepidochelys
kempii (LK), Lepidochelys
2020). This was because the collection and identification of meio- olivacea (LO), Natator depressus
or micro-epibionts generally requires specialist equipment (e.g., (ND).
light and/or scanning electron microscopes), while this is not the Epibiont taxon Taxonomic name of See database for full list.
case for macro-epibionts. recorded epibiont
(lowest possible
From each applicable reference, we extracted data on all
taxonomic level),
reported sea turtle-epibiont interactions according to the data reflecting the most
parameters and descriptions listed in Table 1. If a single up-to-date taxonomic
study included more than one host-epibiont pair (e.g., multiple nomenclature from the
epibiont taxa from one host species or multiple host species or World Register of Marine
Species (WoRMS
life stages for one epibiont taxon), each specific host-epibiont pair
Editorial Board, 2021).
was listed on a separate row within the database. At minimum,
Taxonomic rank Taxonomic rank of the Species, Genus, Subfamily,
we recorded the host turtle species, the epibiont taxon (the taxonomic name for the Family, Superfamily, Infraorder,
lowest taxonomic level reported), the study site, and the type of recorded epibiont. Suborder, Order, Infraclass,
survey conducted. When data were available, we also recorded the Subclass, Class, Subphylum,
number of turtles surveyed, the life stage of the sampled turtles, Phylum.
Higher taxon Phylum of the epibiont Annelida (Hirudinea, Polychaeta,
the habitat where turtles were encountered, the percent frequency taxon including further Oligochaeta), Arthropoda
of occurrence per host for each epibiont taxa, the total number of differentiation by (Arachnida, Insecta,
recorded individuals of each epibiont taxon, the turtle body part/s Subphylum, Class, or Malacostraca, Pycnogonida,
on which each taxon was found, and the deposition location for Subclass for select Ostracoda, Thecostraca),
phyla (when applicable). Bryozoa, Chordata (Tunicata and
collected specimens.
Vertebrata), Cnidaria (Anthozoa
To extract further information on each epibiont taxon, we and Hydrozoa), Echinodermata
recorded the Higher Taxon (either full Phylum or select phyla (Asterozoa and Echinozoa),
differentiated by Subphylum/Class/Subclass) and the taxonomic Mollusca (Bivalvia, Gastropoda,
rank (e.g., species, genus, family, etc.) of the taxonomic name and Polyplacophora),
Platyhelminthes, Porifera,
reported in the study. Five phyla were not differentiated further:
Nemertea, Sipuncula.
Bryozoa, Nemertea, Platyhelminthes, Porifera, and Sipuncula. Site Country and/or name of See database for full list.
However, the diversity of taxa within six phyla warranted sampling site.
further differentiation by either Subphylum, Class, or Subclass: Ocean region* Geographic region of Northwest Atlantic (including Gulf
Annelida (Hirudinea, Polychaeta, and Oligochaeta), Arthropoda study. of Mexico), Caribbean, North
Central Atlantic, Northeast
(Arachnida, Insecta, Malacostraca, Pycnogonida, Ostracoda, Atlantic, Mediterranean Sea,
and Thecostraca), Chordata (Tunicata and Vertebrata), Southwest Atlantic, South Central
Cnidaria (Anthozoa and Hydrozoa), Echinodermata (Asterozoa Atlantic, Southeast Atlantic, North
and Echinozoa), and Mollusca (Bivalvia, Gastropoda, and Indian (including Red Sea and
Persian Gulf), Southwest Indian,
Polyplacophora). To ensure that we used the most up-to-date
South Central Indian, Southeast
taxonomic nomenclature and rank for each epibiont taxon, Indian, Northwest Pacific, North
we referred to the World Register of Marine Species (WoRMS Central Pacific (Hawaii), Northeast
Editorial Board, 2021). Pacific, Southwest Pacific
To define the geographic location of each study site, we (including Gulf of Carpentaria and
Melanesia), South Central Pacific
recorded the latitude and longitude (either provided in the
(Polynesia excluding Hawaii),
study or plotted using Google Earth) and the ocean region Southeast Pacific, Indonesian
that best described the study site (see Table 1 for a list of Archipelago, Unknown.
potential ocean regions). Lastly, to further characterize the Regional Regional Management CC-A-NE, CC-A-NW, CC-A-SW,
sea turtle population from which epibionts were sampled for Management Unit, as defined in CC-MED, CC-I-NE, CC-I-NW,
Unit (RMU) Wallace et al. (2010), of CC-I-SW, CC-I-SE, CC-P-S,
each sea turtle-epibiont interaction, we recorded the regional the sampled turtles (see CC-P-N, CM-A-E, CM-A-NW,
management unit (RMU) following designations by Wallace et al. Host Species - Data CM-A-SC, CM-A-SW,
(2010) as these are considered geographic cohorts of turtles on Options for species CM-A-SCR, CM-I-NE, CM-I-NW,
independent evolutionary trajectories. When a study site was acronyms) CM-I-SE, CM-I-SW, CM-MED,
within overlapping RMUs, we used the best available data (e.g., (Continued)

Frontiers in Ecology and Evolution | www.frontiersin.org 3 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

TABLE 1 | (Continued) RMUs (Northwest Atlantic, Mediterranean, and North Pacific).

We selected these specific RMUs because each had been the
Parameter Description Data options
focus of >15 studies. To avoid overestimating the number of
CM-P-E, CM-P-NC, CM-P-NW, taxa when studies identified epibionts to different taxonomic
CM-P-SC, CM-P-SW, CM-P-WC, ranks (e.g., species versus genus level), we counted all potentially
CM-P-WPSEA, CM-P-W*,
equivalent taxa as one taxon. For example, if three different
DC-A-NW, DC-A-SE, DC-A-SW,
DC-I-NE, DC-I-SW, DC-P-E, studies on olive ridley turtles identified Lepas anatifera (species
DC-P-W, EI-A-E, EI-A-SW, rank), Lepas sp. (genus rank), and Lepadidae (family rank), we
EI-A-WCAR, EI-I-NE, EI-I-NW, combined these three taxa into one taxon and only counted
EI-I-SE, EI-I-SW, EI-P-E, them once. We therefore defined taxonomic richness as the
EI-P-NC, EI-P-SC, EI-P-SW,
EI-P-WC, EI-P-WPSEA, LO-A-E,
total number of unique epibiont taxa that could not be further
LO-A-W, LO-I-NE, LO-I-NE(a), hierarchically combined. Because most studies reported epibiont
LO-I-W, LO-P-E, LO-P-E(a), taxa at the species level, the mean percentage of taxa that we
LO-P-W, LK-A-NW, ND-I-SE, hierarchically combined in this way within each host species and
ND-P-SW, UNK. within the three loggerhead RMUs was relatively low (10.2 and
Latitude/ Latitude and longitude of Latitude and longitude
9.0%, respectively).
Longitude sampling site (either provided in coordinates.
the study or plotted using
For each turtle species, as well as the three loggerhead
Google Earth). RMUs, we plotted the number of taxa and proportion
Stage Life stage of turtle sampled (as Juvenile, Adult, Unknown. of taxa documented within each Higher Taxon category
described in each study or (either full Phylum or select phyla differentiated by
based on reported body size at Subphylum/Class/Subclass as detailed in Table 1). To minimize
maturity).
the number of categories displayed in these figures, we
Habitat Habitat type of sampling site. Nesting beach, Neritic, Oceanic,
combined Higher Taxon categories within a given Phylum (e.g.,
Stranding/Dead, Unknown.
Echinodermata – Asterozoa and Echinodermata – Echinozoa)
N (T)* Number of turtles sampled. An integer.
when no individual turtle species was documented hosting eight
Freq (%)* Percent frequency of A percentage.
occurrence per host for the or more taxa within a Higher Taxon category. All remaining
epibiont taxon. Higher Taxon categories with fewer than seven total host-
N (E)* Number of individuals of the An integer. epibiont pairs across all turtle species were also combined into a
epibiont taxon recorded in the single category called “Other taxa.” One exception to these rules
study. was the Higher Taxon category Annelida – Hirudinea, which
Body part* Body part(s) from which the Carapace, Head/Neck, Front includes two globally distributed species of marine turtle leech
epibiont taxon was collected. flippers, Rear flippers, Plastron,
Inguinal Area/Tail, Unknown.
that retained in the diversity plots because of their important role
Deposition* Location where epibiont Not Collected, Museum, Personal
in sea turtle health and disease transmission (Greenblatt et al.,
specimens were curated after Collection, Unknown. 2004; Köhnk et al., 2021).
the study.
Survey type Method of epibiont sampling: Exhaustive/All Taxa, Extrapolating Taxonomic Richness
Did the study sample all Exhaustive/Taxon Specific, To evaluate whether the current scientific literature fully
possible taxa (All Taxa) or only Non-Exhaustive/All Taxa,
encompasses the taxonomic richness of each turtle species, as
focused on a subset of taxa Non-Exhaustive/Taxon Specific.
(Taxon Specific). Also, were the
well as the three selected loggerhead RMUs, we used rarefaction
turtles sampled exhaustively for curves following the Bernoulli product model to estimate the rate
epibionts (Exhaustive) or were at which epibiont taxonomic richness increased with increasing
only a subset of epibionts sample sizes (Colwell et al., 2012). This allowed us to account
sampled (Non-Exhaustive)
for differences in samples sizes when comparing the taxonomic
Primary Reference for data source. See database for full list.
richness between different turtle species and RMUs. Because
reference
rarefaction curves can reasonably extrapolate species richness
Secondary References that also presented See database for full list.
reference(s) these data but were not the up to double or triple the reference sample size (Colwell et al.,
primary data source. 2012), we estimated taxonomic richness after 150 studies for
Asterisks indicate data categories that were included in the global database (Pfaller
each turtle species and after 80 studies for each loggerhead
and Robinson, 2022) but were not analyzed in this study. RMU. We excluded Kemp’s ridley and flatback turtles from the
rarefaction analyses because there were <15 studies available for
these two species and this was not sufficient to provide accurate
life stage and habitat) to select the most likely RMU for the extrapolations of taxonomic richness.
surveyed turtles. We used individual studies as the baseline sampling unit
and built sample-based rarefaction curves instead of individual-
Comparing Epibiont Diversity based rarefaction curves. While it would have been preferential
We collated the total number of distinct epibiont taxa to use individual turtles as the sampling unit for the rarefaction
documented for each turtle species and for three loggerhead analyses, most studies only presented the combined epibiont

Frontiers in Ecology and Evolution | www.frontiersin.org 4 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 1 | Number of All Taxa, Taxon Specific, Exhaustive, and Non-Exhaustive studies by year that included data on sea turtle-epibiont interactions identified
during our two-tiered search of both published scientific articles and gray literature (e.g., government reports, theses and dissertations, and conference
presentations). All Taxa studies attempt to document all possible epibiont taxa, while Taxon Specific studies focus on one or more specific taxonomic groups.
Exhaustive studies sampled all observed epibionts, while Non-Exhaustive only sampled a subset of the observed epibionts.

communities of all turtles sampled and not the unique epibiont study onto global maps that delineated RMUs for each turtle
communities on each turtle. We acknowledge this violates two species following Wallace et al. (2010). We distinguished
key assumptions of rarefaction curves: samples are collected at between studies that reported only the presence of a subset
random from the population and samples are collected with of epibiont taxa (Taxon Specific surveys) from those that
equal effort. Indeed, global epibiont studies were not conducted focused on all potential taxa (All Taxa surveys) because these
at random between different RMUs (see section “Results: differences may bias geographic patterns in taxonomic richness
Identifying Knowledge Gaps”) and the same number of turtles among sea turtle RMUs. For example, studies investigating
were not sampled in each study. For these reasons, the sample- the phylogenetics of a specific sea turtle barnacle will only
based rarefaction curves in this study likely underestimate total report the collection localities of that barnacle species
taxonomic richness. Sample-based rarefaction and extrapolation and not on the other epibiont taxa that may have also
curves as well as their 95% confidence intervals were calculated occurred on the host turtles sampled in those regions (e.g.,
using the program EstimateS V.9.1. Pinou et al., 2013). In contrast, All Taxa studies attempt to
document and report on all epibiont taxa detected on surveyed
Identifying Knowledge Gaps turtles (e.g., Robinson et al., 2017b). Maps were created
To identify knowledge gaps in the literature, we quantified in ArcGIS v10.6.
sampling effort (i.e., the number of studies) for each turtle species
in terms of turtle life stage and habitat type. For turtle life stage,
we collated the number of studies documenting epibionts on RESULTS
juveniles (i.e., non-sexually mature), adults (i.e., sexually mature),
and turtles of unknown size or reproductive status (i.e., not Database Summary
indicated in the study). For habitat type, we collated the number From our literature review, we identified 304 studies that
of studies documenting epibionts on turtles that were intercepted contained data on a combined total of 1,717 sea turtle-epibiont
while nesting on beaches, captured in neritic habitats (<200 m interactions. Across the seven sea turtle species, we recorded
depth), captured in oceanic habitats (>200 m depth), found 374 unique epibiont taxa representing 11 Phyla, separated into
dead/debilitated in the marine environment or washed ashore, 23 Higher Taxon categories. The earliest record included was
and those in which the habitat type was unknown. Studies in published in 1886. Since then, there was a steady increase
which epibionts were surveyed on both juvenile and adult turtles in the annual number of publications reporting on sea turtle
or where turtles were surveyed in more than one habitat type were epibiosis until approximately 2010, after which there was a slight
counted in each applicable category. decline (Figure 1). This increase included both All Taxa and
Additionally, we assessed knowledge gaps in terms of Taxon Specific studies, although Exhaustive studies only began
geographic distribution by plotting every sea turtle epibiont to increase after 1985.

Frontiers in Ecology and Evolution | www.frontiersin.org 5 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

To encourage further exploration of the information amassed Extrapolating Taxonomic Richness

in this database, we have made it freely available in the Dryad Rarefaction curves indicated that loggerhead turtles both host the
Digital Repository1 (Pfaller and Robinson, 2022). Data compiled highest total taxonomic richness of all seven sea turtle species
for several categories listed in Table 1 were not analyzed in and have had the greatest sampling effort in terms of number
this study but were included in the database because they of studies (Figure 3). Extrapolating beyond the current 135
provide important biological and/or methodological information studies documenting the epibiont diversity of loggerheads, the
specific to each sea turtle-epibiont interaction (e.g., frequency rarefaction curve approaches 300 epibiont taxa after 150 studies.
of occurrence, body part, etc.) that may be used in future This suggests that the current scientific literature has not yet
studies. We expect future authors to update this database with fully described the taxonomic richness for loggerhead epibionts
information from new publications (after 2020), as well as any on a global scale.
sea turtle-epibiont interactions that were missed during our The rarefactions curves for loggerheads from the Northwest
literature review. Atlantic, Mediterranean, and North Pacific RMUs also suggested
that further studies will reveal additional epibiont diversity
Comparing Epibiont Diversity (Figure 4). Extrapolating up to 80 studies per RMU indicated
Of the 374 epibiont taxa representing 23 Higher Taxon that the taxonomic richness of Northwest Atlantic loggerheads
categories (either full Phylum or select phyla differentiated could reach an estimated 202 epibiont taxa, while Mediterranean
by Subphylum/Class/Subclass) that were documented on sea and North Pacific loggerheads could reach an estimated 137 and
turtles globally, loggerheads hosted 262 taxa from 21 categories 55 epibiont taxa, respectively. While extrapolations suggest that
(Figure 2 and Supplementary Table 1). Loggerheads from the current studies have not fully described the taxonomic richness
Northwest Atlantic RMU hosted 162 taxa from 18 Higher Taxon within each loggerhead RMU, the lack of overlap between the
categories, while loggerheads from the Mediterranean and North 95% confidence intervals suggests that geographic variation in
Pacific RMUs hosted 85 taxa from 12 categories and 27 taxa taxonomic richness among loggerhead RMUs represents a true
from six categories, respectively (Supplementary Figure 1). biological pattern.
Hawksbills hosted a similar level of diversity as loggerheads Hawksbill turtles hosted the second highest taxonomic
(20 Higher Taxon categories), but the total taxonomic richness richness after loggerheads. However, because this diversity
was lower (87 taxa). The richness and diversity of epibionts was recorded from comparably fewer studies (N = 58), the
hosted by the other five sea turtle species were considerably extrapolated rarefaction curve for hawksbills reached 148
lower than that of loggerheads and somewhat lower than that epibiont taxa after 150 studies (Figure 3). The 95% confidence
of hawksbills: greens (56 taxa from 12 Higher Taxon categories), intervals for hawksbills overlapped with that of olive ridleys.
olive ridleys (51 taxa from 14 categories), leatherbacks (15 taxa Like hawksbills, olive ridleys hosted relatively high diversity from
from 4 categories), Kemp’s ridley (7 taxa from 3 categories), comparably fewer studies (N = 32), causing the extrapolated
and flatbacks (7 taxa from 3 categories) (Figure 2 and rarefaction curve to estimate approximately double the diversity
Supplementary Table 1). (101 taxa) after 150 studies. Additional studies investigating
Among the 13 Higher Taxon categories not combined into the epibiont diversity of hawksbills and olive ridleys globally
“Other taxa,” loggerheads and hawksbills were the only species are needed to unequivocally determine whether the epibiont
that hosted epibiont taxa from all categories (Figure 2 and communities of hawksbill turtles are more taxonomically rich
Supplementary Table 1). Moreover, loggerheads and hawksbills than olive ridley turtles.
hosted roughly similar proportions of taxa from each category, Green turtles, despite having the second most studies
with the predominant categories being Annelida – Polychaeta (N = 111), were recorded hosting significantly fewer epibiont
(i.e., polychaete worms), Arthropoda – Malacostraca (e.g., taxa than hawksbills and approximately the same number of
crabs, shrimps, and amphipods), and Arthropoda – Thecostraca taxa as olive ridleys. Consequently, the extrapolated rarefaction
(e.g., acorn and goose-necked barnacles). Northwest Atlantic curve for green turtles estimated only 64 epibiont taxa after
loggerheads hosted epibiont taxa from 12 Higher Taxon 150 studies, which was only eight taxa higher than is currently
categories not combined into “Other taxa,” while Mediterranean reported (Figure 3). Similarly, epibiont research on leatherbacks
and North Pacific loggerheads hosted epibiont taxa from nine has appeared to reach a plateau: 15 taxa have been recorded
and five categories, respectively (Supplementary Figure 1 and from 35 studies to date, but only 16 total taxa are expected to be
Supplementary Table 1). Green and olive ridley turtles hosted recorded after 150 studies are conducted.
epibiont taxa from all but one category (Echinodermata –
Asterozoa, Echinozoa were not recorded for either host species),
while leatherbacks, Kemp’s ridleys, and flatbacks hosted epibiont Identifying Knowledge Gaps – Life Stage
taxa from four, three, and three categories, respectively. and Habitat
Arthropoda – Thecostraca was the only Higher Taxon category Turtle life stage (adult and/or juvenile) was only reported in
reported from all seven sea turtle species and its predominance 52% of studies (Figure 5A). Within each turtle species, the
(% of taxa) tended to increase as the total epibiont richness of a percentage of studies reporting turtle life stage ranged from 49%
host species decreased. in greens and 84% in olive ridleys. Among studies that reported
turtle life stage, there were strong biases toward adult turtles for
1
http://doi.org/10.5061/dryad.x3ffbg7m8 flatbacks (100%), leatherbacks (86%), and olive ridleys (84%),

Frontiers in Ecology and Evolution | www.frontiersin.org 6 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 2 | Taxonomic composition of the epibiont communities reported from all seven sea turtle species. (A) Number of taxa per Higher Taxon category (either full
Phylum or select phyla differentiated by Subphylum/Class/Subclass) and (B) proportion of taxa per Higher Taxon category. The category “Other taxa” combines taxa
from 10 Higher Taxon categories in which fewer than six total taxa were documented across all turtle species (Loggerhead: Annelida – Oligochaeta, Arthropoda –
Ostracoda, Arthropoda – Pycnogonida, Mollusca – Polyplacophora, Nemertea, Platyhelminthes, Porifera, and Sipuncula; Hawksbill: Annelida – Oligochaeta,
Arthropoda – Insecta, Arthropoda – Ostracoda, Mollusca – Polyplacophora, Platyhelminthes, Porifera, and Sipuncula; Green: Arthropoda – Arachnida and
Platyhelminthes; Olive Ridley: Arthropoda – Pycnogonida, Mollusca – Polyplacophora, Platyhelminthes, and Porifera). See Supplementary Table 1 for numerical
summary of the number of epibiont taxa reported within each Higher Taxon for each sea turtle species.

while significant biases toward adult turtles were not as apparent (Figure 5B). Within each turtle species, the percentage of
for hawksbills (59%), loggerheads (54%), and green turtles (48%). studies that reported habitat type ranged from 50% in flatbacks
Kemp’s ridleys were the only species with a strong bias toward and 92% in olive ridleys. Among studies that reported habitat
juveniles (25% adult). Nevertheless, the two species with the type, there were biases toward nesting beaches for flatbacks
strongest biases (Kemp’s ridleys and flatbacks) also had relatively (100%), leatherbacks (54%), and olive ridleys (53%), reflecting
few studies (seven and two total studies, respectively). their biases toward studies on adult turtles. The habitat types
Habitat type (nesting beach, neritic, oceanic, and/or among studies on leatherbacks and olive ridleys were similar in
stranding/dead) was reported for only 63% of studies proportion: nesting beaches (54 and 53%, respectively), neritic

Frontiers in Ecology and Evolution | www.frontiersin.org 7 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

34% were on nesting beaches, 11% were in oceanic waters, and

17% were from stranded or dead turtles.

Identifying Knowledge Gaps –

Geography and Regional Management
Units
For loggerheads, epibiont studies were geographically
concentrated in the Northwest Atlantic (33%) and Mediterranean
(22%) RMUs, where a high percentage of studies were All Taxa
surveys (42 and 33%, respectively) (Figure 6A). Conversely,
loggerhead studies in the western portion of the North Pacific
RMU were relatively numerous (15%) but were predominantly
Taxon Specific surveys (87%). For green turtles, epibiont studies
were globally distributed but most studies were Taxon Specific
FIGURE 3 | Rarefaction curves to estimate taxonomic richness of epibionts
(79%) (Figure 6B). Both Taxon Specific and All Taxa surveys
for loggerhead, hawkbill, olive ridley, green, and leatherback sea turtles. Solid
lines indicate data curves calculated from previous studies, while dashed lines were geographically concentrated in the East Pacific RMU and
represent extrapolated data. the area of overlap between the Southwest Atlantic and South
Central Atlantic RMUs in southern Brazil and Uruguay. Taxon
Specific surveys were also concentrated in the Northwest Pacific,
Southwest Pacific, and Northwest Indian RMUs, while All Taxa
surveys were also concentrated in the North Central Pacific
RMU (i.e., Hawaii).
For hawksbills, Taxon Specific surveys were more common
(69%) and geographically more homogeneous (Figure 7A),
with the West Pacific/Southeast Asia RMU and East Pacific
RMU having the most Taxon Specific surveys (75 and 100%,
respectively) and only one All Taxa survey each. All Taxa
surveys for hawksbills were geographically concentrated in the
Western Caribbean/United States (100%) and Southwest Pacific
RMUs (71%). For leatherbacks, both Taxon Specific and All
Taxa surveys were geographically concentrated in the Northwest
Atlantic RMU (Figure 7B), which included epibiont studies
conducted in both eastern North America and the Caribbean as
FIGURE 4 | Rarefaction curves to estimate the taxonomic richness of well as western Europe in the Northeast Atlantic. Two All Taxa
epibionts for three loggerhead Regional Management Units: Northwest surveys have been conducted on leatherbacks in the northern
Atlantic, Mediterranean, and North Pacific. Solid lines indicate data curves
portion of the East Pacific RMU. The West Pacific RMU was
calculated from previous studies, while dashed lines represent extrapolated
data.
the subject of several Taxon Specific surveys. However, these
surveys covered a very wide geographic area, ranging from Japan
to Malaysia to New Zealand.
For Kemp’s and olive ridleys, by far the greatest concentration
(22 and 23%, respectively), oceanic (11 and 12%, respectively),
of both Taxon Specific and All Taxa surveys were in the East
and stranding/dead (14 and 12%, respectively). Studies on
Pacific olive ridley RMU (Figure 8A), representing 81% of the All
loggerheads were more balanced between habitat types: nesting
Taxa surveys and 64% of the Taxon Specific surveys for ridleys.
beach (33%), neritic (30%), oceanic (17%), and stranding/dead
Kemp’s ridley studies in the Gulf of Mexico and Northwest
(20%). Studies on hawksbills were biased toward nesting beaches
Atlantic were few with only seven studies but were balanced
(42%) and neritic habitats (44%) and away from oceanic habitats
between Taxon Specific and All Taxa surveys. For flatbacks, only
(only 2%). Studies on greens and Kemp’s ridleys were biased
four studies have been conducted (Figure 8B): one All Taxa
toward neritic habitats (59 and 50%, respectively) and away
survey in the Southeast Indian RMU, and one All Taxa and two
from nesting beaches (19 and 0%, respectively) and oceanic
Taxon Specific surveys in the Southwest Pacific RMU.
habitats (6 and 0%, respectively), but studies on Kemp’s ridleys
also suffered from a small sample size (only three neritic, three
stranding/dead, and three unknown habitats for Kemp’s ridleys). DISCUSSION
Studies on flatbacks were also biased by low sample sizes (two
studies on nesting beaches and two in unknown habitats). The first studies documenting epibionts of sea turtles were
Among all studies from all species, including those that included conducted over a century ago and since then data available on
turtles sampled from more than one habitat type (i.e., those sea turtle epibiosis has grown extensively. To utilize this growing
counted in more than one category), 38% were in neritic waters, body of knowledge, we developed a global database of sea

Frontiers in Ecology and Evolution | www.frontiersin.org 8 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 5 | (A) Number of epibiont studies separated by turtle species and life stage (either adult, juvenile or unknown). (B) Number of epibiont studies separated by
turtle species and habitat type (nesting beach, neritic, oceanic, stranding/dead animals, or unknown).

turtle-epibiont interactions. We compiled data from 304 studies, We demonstrate that additional epibiont diversity remains to be
spanning both published scientific articles and gray literature, documented, even within the most well-studied sea turtle species
and explored global patterns in epibiont diversity. In doing so, and populations, and we identify biases in sampling effort that
we synthesized over 100 years of sea turtle epibiont research. may reveal additional diversity in future studies.

Frontiers in Ecology and Evolution | www.frontiersin.org 9 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 6 | Geographic distribution of all epibiont studies for (A) loggerhead turtles (Caretta caretta) and (B) green turtles (Chelonia mydas). Blue stars represent
studies that report on all potential taxa (All Taxa), while yellow circles represent studies that only focus on specific taxa (Taxon Specific). The geographic outlines for
each Regional Management Unit as represented by differentially colored polygons (based on Wallace et al., 2010).

Frontiers in Ecology and Evolution | www.frontiersin.org 10 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 7 | Geographic distribution of all epibiont studies for (A) hawksbill turtles (Eretmochelys imbricata) and (B) leatherback turtles (Dermochelys coriacea). Blue
stars represent studies that report on all potential taxa (All Taxa), while yellow circles represent studies that only focus on specific taxa (Taxa Specific). The geographic
outlines for each Regional Management Unit as represented by differentially colored polygons (based on Wallace et al., 2010).

Frontiers in Ecology and Evolution | www.frontiersin.org 11 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

FIGURE 8 | Geographic distribution of all epibiont studies for (A) Kemp’s and olive ridley turtles (Lepidochelys kempii and L. olivacea) and (B) flatback turtles (Natator
depressus). Blue stars represent studies that report on all potential taxa (All Taxa), while yellow circles represent studies that only focus on specific taxa (Taxa
Specific. The geographic outlines for each Regional Management Unit as represented by differentially colored polygons (based on Wallace et al., 2010).

Which Sea Turtle Species Hosts the by hawksbill turtles (20 Higher Taxon categories). Nevertheless,
Greatest Epibiont Diversity? the total taxonomic richness reported for hawksbills (87 taxa) is
Based on the current literature and confirmed via the rarefaction still far lower than loggerheads. As for the other five sea turtle
analyses, loggerhead turtles host the most diverse epibiont species, the richness and diversity of their epibionts communities
communities both in terms of total taxonomic richness and were considerably lower than that of loggerheads and somewhat
number of higher taxa represented (262 taxa from 21 Higher lower than that of hawksbills, especially for leatherbacks, Kemp’s
Taxon categories). This level of epibiont diversity is only rivaled ridleys, and flatbacks. Differences in the taxonomic richness and

Frontiers in Ecology and Evolution | www.frontiersin.org 12 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

diversity among the epibiont communities of different sea turtles covering the carapace of “hard-shelled” sea turtles (Wyld and
may be attributed to several, non-mutually exclusive factors. Brush, 1983). This could be one of the primary reasons why
First, because epibiosis necessitates spatial overlap between leatherbacks host lower epibiont richness and diversity than
the ranges of host turtles and their epibionts (Frick and Pfaller, other sea turtles. Among “hard-shelled” sea turtles, the thick,
2013), sea turtle species with wider geographic ranges would overlapping scutes of hawksbills and characteristically “rough”
theoretically encounter a greater diversity of potential epibionts scutes of loggerheads may facilitate epibiont attachment and
and thus host greater richness and diversity. Consistent with persistence to a greater degree than the “smooth” scutes of
this hypothesis, Kemp’s ridleys and flatbacks have the smallest green and ridley turtles and the thin, waxy scutes of flatbacks.
geographic ranges of the seven sea turtle species and host the Moreover, the presence of certain sessile epibionts may also
lowest epibiont richness and diversity. However, among the provide settlement cues for other epibionts, including many
remaining sea turtle species, this hypothesis is not consistent with motile forms (e.g., Arthropoda – Malacostraca). Sea turtles
the observed patterns. Leatherbacks have the widest geographic that provide favorable conditions for these “pioneer” species
range, spanning all tropical to sub-polar oceans of the world may develop more diverse epibiont communities that begin
(James et al., 2006), yet host comparably low epibiont richness to resemble the faunal assemblages found in the surrounding
and diversity. In contrast, loggerheads, greens, hawksbills, and environment (e.g., nesting loggerheads in the Northwest Atlantic
olive ridleys have comparable circumglobal ranges, yet exhibit RMU; Frick et al., 1998). Many of these hypotheses have yet
considerable differences in epibiont richness and diversity. to tested empirically but are likely important biological factors
Moreover, for example, North Pacific loggerheads have a vastly driving differences in epibiont richness and diversity.
wider geographic range than Mediterranean loggerheads yet host Overall, it is unlikely that one single factor that explains why
lower epibiont richness and diversity. Clearly, geographic range some sea turtles host greater epibiont richness and diversity than
size is not the only factor driving differences in epibiont richness others. Instead, a complex suite of factors including geographic
and diversity among sea turtle species or populations. range, habitat use and behavior, surface properties, and likely
Second, sea turtles occupying a wider array of habitats others we have not discussed here, collectively drive the observed
during their life cycle may overlap with a greater diversity of patterns. To evaluate the relative importance of various factors,
potential epibionts and thus host greater richness and diversity. we recommend more detailed comparisons between the epibiont
In support of this hypothesis, loggerheads, which host the communities of turtles from (1) different RMUs of the same
greatest epibiont diversity, not only exhibit an extended oceanic species and (2) different species with overlapping RMUs. Current
developmental phase (Bolten, 2003; Avens et al., 2013) but data may be sufficient to compare some RMUs (e.g., Northwest
are also well known for a behavioral polymorphism in which Atlantic, Mediterranean, and North Pacific loggerhead RMUs).
some individuals transition to neritic foraging areas while other However, for most other RMUs, the acquisition of more data is
remain oceanic (Hatase et al., 2010; Rees et al., 2010; Vander needed to make such comparisons.
Zanden et al., 2010). In contrast, greens and hawksbills exhibit
a more truncated oceanic developmental phase and primarily
inhabit neritic habitats (Vander Zanden et al., 2013; Martinez-
Does the Current Literature Fully
Estevez et al., 2021), while leatherbacks and olive ridleys occupy Encompass the Taxonomic Richness of
primarily oceanic habitats (Fossette et al., 2010; Pikesley et al., Sea Turtle Epibionts?
2013). Moreover, not all habitats will contain equally diverse Despite over 100 years of research describing the epibiont
communities of potential epibionts. For example, the coastal diversity of sea turtles, the current literature has not yet
hard-bottom and coral reef habitats occupied by loggerheads fully encompassed the taxonomic richness of most sea turtle
and hawksbills not only support high invertebrate biodiversity species (Figure 3) or even three well-studied loggerhead RMUs
but also exhibit intense competition for space, both of which (Figure 4). Rarefaction analyses indicate that the rate at which
may contribute to greater epibiont diversity in these habitats. taxonomic richness has increased with additional sampling effort
In contrast, the oceanic habitats occupied by leatherbacks and (number of studies) has not plateaued for any species or RMU,
olive ridleys and the coastal seagrass meadows occupied by green except leatherbacks.
turtles support lower biodiversity and less intense competition for For the “hard-shelled” or Cheloniid sea turtles, additional
space. As such, our results support the idea that the number and studies are expected to continue to reveal more undiscovered
type of habitats occupied by sea turtles, rather than the size of taxonomic richness in all species and regions. That said, some
their geographic range, is likely a more important factor driving species are projected to be greater sources of additional richness
differences in epibiont richness and diversity. than others. Green turtles, which have already been included in
Third, sea turtles presenting more suitable conditions for over 100 studies, host relatively modest epibiont diversity, and
epibiont settlement would be more likely to host greater epibiont appear to be approaching a plateau in taxonomic richness around
richness and diversity. Especially for sessile epibionts, the surface 70 taxa globally. Conversely, the taxonomic richness of hawksbill
properties (e.g., chemical signals, rugosity, and wettability) of and olive ridley turtles, which have received comparably less
the skin and carapace of different sea turtle species may provide attention than green turtles, are projected to almost double
more (or less) favorable conditions for larval attachment and after 150 studies, adding an estimated 61 and 50 new epibiont
subsequent growth. Leatherbacks, in particular, have uniquely taxa, respectively. Even within the most well-studied sea turtle
smooth skin covering their carapace and this may provide a RMUs, like Northwest Atlantic and Mediterranean loggerheads,
less favorable substrate for epibionts than the rigid keratin dozens of new epibiont taxa are expected to be found with

Frontiers in Ecology and Evolution | www.frontiersin.org 13 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

additional sampling effort. These patterns suggest that addressing Two issues regarding life stage and habitat type emerged
the knowledge gaps identified in this study (see below), as well as from our assessment of the sea turtle epibiont literature. First, a
conducting further sampling in well-studied RMUs, will both be significant percentage of studies that report sea turtle epibionts
productive and interesting areas of future research. did not indicate the life stage (48%) or habitat type (37%)
Epibiont research on leatherbacks has appeared to reach a of the sampled turtles (Figure 5). We strongly encourage
plateau in taxonomic richness: only 16 taxa are projected after 150 researchers interested in reporting epibionts from sea turtles to
studies, only one higher than currently described. While there also collect and provide these important pieces of metadata.
are still sampling gaps for leatherbacks in terms of life stage, Efforts to gain insights from the epibionts of specific turtles
habitat type, and geography (see below), their oceanic lifestyle and understand global patterns in sea turtle epibiosis require
and inhospitable surface properties (i.e., leathery skin) have these important data. Second, excluding Kemp’s ridleys and
likely constrained the diversity of their epibiont communities flatbacks, between 12 and 20% of the studies that report habitat
to a limited number and type of taxa. Indeed, the epibionts type came from stranded/dead turtles. Because stranded and/or
of leatherbacks tend to either be oceanic/pelagic specialists dead turtles may acquire epibiont taxa after their debilitation
or taxa found only on sea turtles, including two species of or death, their epibiont communities may not be characteristic
barnacle (Platylepas coriacea and Stomatolepas dermochelys) that of the surrounding population of healthy turtles. Instead,
are essentially exclusive to leatherbacks (Zardus, 2021; this study). the processes involved in the development of their epibiont
Future studies on leatherbacks are therefore unlikely to reveal communities may be quite different. For this reason, inferences
additional undocumented diversity. gleaned from epibiont taxa found on stranded/dead should be
made with caution.

What Knowledge Gaps in Life Stage and

Habitat Type Were Identified? What Knowledge Gaps Among Regional
As with many aspects of sea turtle biology, the relative ease of Management Units Were Identified?
accessing adult female turtles on nesting beaches has created Because sea turtle RMUs are defined by their shared geography,
biases in epibiont research as well. Although nesting females critical habitats, and evolutionary trajectory (Wallace et al., 2010),
represent a relatively small proportion of the total individuals individual RMUs should also be considered the basic unit of
in any sea turtle population (Heppell et al., 2003), the majority sampling for epibiont research. Indeed, we demonstrated that
of epibiont studies that report life stage and/or habitat for epibiont communities vary dramatically not only between sea
leatherbacks, olive ridleys, and flatbacks come from adult turtle species but also between conspecific RMUs. Developing a
turtles (>80% of studies) and nesting beaches (>50%). While holistic checklist of sea turtle epibionts on a global scale would
approximately half of epibiont studies that report life stage for therefore require sampling turtles from all RMUs.
loggerheads, greens, and hawksbills come from adult turtles, When initially assessing the sea turtle epibiont literature for
there is a better balance of studies between nesting beaches and knowledge gaps, it was evident that the two turtle species with
neritic habitats for loggerheads and hawksbills and an emphasis the smallest geographic ranges and fewest RMUs were also
toward neritic habitats for green turtles. Only Kemp’s ridleys have the most understudied. Kemp’s ridleys and flatbacks have been
yet to be surveyed for epibionts on nesting beaches. While nesting the subjects of seven and four epibiont studies, including just
beaches provide an excellent opportunity to initiate epibiont three and two All Taxa studies, respectively. Compared to other
research in many understudied RMUs, the full richness and turtle species, far more sampling effort would be needed to
diversity of epibionts in those RMUs will not be discovered until fully describe the epibiont communities of these hosts. However,
in-water studies (both neritic and oceanic) are also conducted. when comparing among RMUs, the one global RMU of Kemp’s
Relative to their abundance, juvenile turtles have received ridleys (Northwest Atlantic) and the two global RMUs for
considerably less attention than adult turtles, especially for flatbacks (Southeast Indian and Southwest Pacific) have in fact
leatherbacks and olive ridleys. Similarly, the number of non- received more attention than many RMUs for the other “well-
breeding turtles inhabiting neritic and oceanic foraging areas is studied” sea turtle species. Most notably, the following RMUs
always far greater than the number of breeding females (Heppell have never been the focus of a single epibiont study: Northeast
et al., 2003), yet the percentage of epibiont studies conducted Indian loggerheads, Northeast Indian greens, and East Atlantic
at in-water sites only exceeds that of nesting beaches for green hawksbills. Moreover, unlike the RMUs for Kemp’s ridley and
turtles. The difficulty of sampling turtles in oceanic habitats is flatback turtles, many RMUs for the other turtle species have also
evident from the relatively low percentages of epibiont studies never been the focus of an All Taxa study (e.g., Northwest Indian
conducted in these habitats. Nevertheless, the turtle species that loggerheads, West Pacific and Southwest Atlantic leatherbacks,
spend more time in oceanic habitats (leatherbacks, olive ridleys, and many RMUs for hawksbills, greens, and olive ridleys).
and, to a lesser extent, loggerheads) tend to have proportionately Because individual RMUs should be considered the basic unit for
more epibiont studies in those habitats (11–17%) than the epibiont sampling, future epibiont research should focus not only
turtle species that have a truncated oceanic stage (hawksbill on understudied host species but also on understudied RMUs
and green turtles; 2 and 6%, respectively). Epibiont studies on within well-studied species.
Kemp’s ridleys and flatbacks were too few to assess biases in life Based on the geographic distribution of epibiont studies as
stage and habitat. well as their study types (All Taxa versus Taxon Specific), we have

Frontiers in Ecology and Evolution | www.frontiersin.org 14 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

identified the most prominent knowledge gaps among RMUs for (2) How do turtle behaviors (e.g., migrating, diving, mating,
each turtle species to help guide future research: nesting, etc.) and habitats influence epibiont diversity, as
well as the frequency and intensity of different epibiont
• Loggerheads: eastern portion of the South Pacific taxa?
RMU (Peru and Chile), Northeast Indian RMU (India, (3) How do surface properties and settlement cues affect the
Bangladesh, and Myanmar), and Southeast RMU (NW epibiont communities of different turtles?
Australia and southwestern Indonesia), as well as the (4) How does the diversity of meio- and micro-epibionts
globally important Northwest Indian RMU, which includes compare and/or contrast with macro-epibiont
Oman and countries surrounding the Red Sea, Persian communities?
Gulf, and gulfs of Oman and Aden. (5) What characteristics make certain “free living” taxa
• Greens: South Caribbean Atlantic RMU (northern South common sea turtle epibionts?
America and Lesser Antilles), Northeast Indian RMU (6) How do the geographic ranges of “free living” populations
(Bangladesh, Myanmar, and eastern Indonesia), Southeast of common sea turtle epibionts correspond with that of
Indian RMU (northern Australia and south-central their hosts?
Indonesia), West Central Pacific RMU (Philippines and (7) Are obligate epibiont taxa more abundant than facultative
west Micronesia), and South Central Pacific RMU (eastern taxa on certain turtle species?
Melanesia and central Polynesia). (8) Does the frequency and intensity of “pioneer” epibiont taxa
• Hawksbills: East Atlantic RMU (Mauritania to Angola facilitate higher species richness?
and into the central South Atlantic), Southeast Indian (9) Which epibiont taxa/communities are best suited to serve
RMU (Western Australia), West Central Pacific as ecological indicators of the behaviors and habitat
RMU (Micronesia), and South Central Pacific RMU preferences of their turtle hosts?
(central Polynesia). (10) How do local oceanographic and climatic factors influence
• Leatherbacks: central and eastern portions of the West epibiont diversity?
Pacific RMU, the southern portion of the East Pacific RMU (11) Can epibiont communities be used as indicators of the
(western South America), Southeast and Southwest Atlantic health status of sea turtles and/or sea turtle populations?
RMUs (eastern South America to western Africa), and (12) Are certain obligate epibiont taxa of comparable
Northeast Indian RMU (India, Bangladesh, Myanmar, and conservation concern as their endangered sea turtle
eastern Indonesia). hosts?
• Olive ridleys: excluding the East Pacific RMU, all five
remaining RMUs (West Atlantic, East Atlantic, West
Indian, Northeast Indian, and West Pacific).
• Kemp’s ridleys: northern and eastern Gulf of Mexico DATA AVAILABILITY STATEMENT
and southeast Atlantic coast of United States (North
Publicly available datasets were analyzed in this study. This data
Carolina thru Florida).
can be found here: NR and JP (2022) Data from: Sea turtle
• Flatbacks: western and northern portions of the
epibiosis: global patterns and knowledge gaps. Dryad Digital
Southeast Indian RMU (northern Western Australia and
Repository (http://doi.org/10.5061/dryad.x3ffbg7m8).
southwestern Papua) and the western and southern portion
of the Southwest Pacific RMU (Gulf of Carpentaria and
central and southern Queensland, as well as southeastern AUTHOR CONTRIBUTIONS
Papua New Guinea).
Both authors contributed to the article and approved the
submitted version.
Future Directions
We built this global database from over 100 years of sea turtle
epibiont research and made the first effort to use it by analyzing ACKNOWLEDGMENTS
broad-scale patterns and identifying knowledge gaps. However,
the capacity for the information amassed in this database to A sincere thank you to the thousands of scientists and volunteers
answer additional questions in this field is extensive. Our hope who have published on sea turtle epibiosis. A special thank you
is that this database will serve as a foundational platform on to Alex Smith and Abi Ferrazzini for their help in collating
which a novel array of hypothesis-driven questions can be tested data and Lourdes Rojas for tracking down many difficult-to-
with respect the taxonomic diversity, ecological complexity, and access articles. NR is supported by Severo Ochoa post-doctoral
evolutionary origins of sea turtle epibiosis. fellowship granted by the Institut de Ciències del Mar.
Research questions that we consider important for promoting
scientific progress in the field of sea turtle epibiosis include but
are not limited to: SUPPLEMENTARY MATERIAL
(1) How does geographic range and habitat use influence the The Supplementary Material for this article can be found
epibiont richness and diversity of different sea turtle species online at: https://www.frontiersin.org/articles/10.3389/fevo.2022.
and RMUs? 844021/full#supplementary-material

Frontiers in Ecology and Evolution | www.frontiersin.org 15 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

REFERENCES James, M. C., Davenport, J., and Hays, G. C. (2006). Expanded thermal niche for
a diving vertebrate: a leatherback turtle diving into near-freezing water. J. Exp.
Avens, L., Goshe, L. R., Pajuelo, M., Bjorndal, K. A., MacDonald, B. D., Lemons, Mar. Biol. Ecol. 335, 221–226. doi: 10.1016/j.jembe.2006.03.013
G. E., et al. (2013). Complementary skeletochronology and stable isotope Köhnk, S., Petros, C., Lomas, C., Riyad, E. M., Shameel, I., Hawlitschek, O., et al.
analyses offer new insight into juvenile loggerhead sea turtle oceanic stage (2021). Stowaways: marine leeches infecting olive Ridley sea turtles entangled
duration and growth dynamics. Mar. Ecol. Prog. Ser. 491, 235–251. doi: 10.3354/ in ghost nets in Maldivian waters. Comp. Parasitol. 88, 169–176.
meps10454 Lazo-Wasem, E. A., Pinou, T., de Niz, A. P., and Feuerstein, A. (2011). Epibionts
Bolten, A. B. (2003). “Variation in sea turtle life history patterns: neritic vs. oceanic associated with the nesting marine turtles Lepidochelys olivacea and Chelonia
developmental stages,” in The Biology of Sea Turtles, Vol. 2, eds P. L. Lutz, mydas in Jalisco, Mexico: a review and field guide. Bull. Peabody Mus. Nat. Hist.
J. A. Musick, and J. Wyneken (Boca Raton, FL: CRC Press), 243–257. doi: 52, 221–240. doi: 10.3374/014.052.0203
10.1201/9781420040807.ch9 Lazo-Wasem, E. A., Pinou, T., de Niz, A. P., Salgado, M. A., and Schenker, E.
Casale, P., Freggi, D., Basso, R., and Argano, R. (2004). Epibiotic barnacles and (2007). New records of the marine turtle epibiont Balaenophilus umigamecolus
crabs as indicators of Caretta caretta distribution and movements in the (Copepoda: Harpacticoida: Balaenophilidae): new host records and possible
Mediterranean Sea. J. Mar. Biol. Assoc. U.K. 84, 1005–1006. doi: 10.1017/ implications for marine turtle health. Bull. Peabody Mus. Nat. Hist. 48, 153–156.
S0025315404010318h doi: 10.3374/0079-032x(2007)48[153:nrotmt]2.0.co;2
Colwell, R. K., Chao, A., Gotelli, N. J., Lin, S.-Y., Mao, C. X., Chazdon, R. L., Martinez-Estevez, L., Amador, J. P. C., Amador, F. C., Zilliacus, K. M., Pacheco,
et al. (2012). Models and estimators linking individual based and sample-based A. M., Seminoff, J. A., et al. (2021). Spatial ecology of hawksbill sea turtles
rarefaction, extrapolation, and comparison of assemblages. J. Plant Ecol. 5, (Eretmochelys imbricata) in foraging habitats of the Gulf of California, Mexico.
3–21. doi: 10.1093/jpe/rtr044 Glob. Ecol. Conserv. 27:e01540. doi: 10.1016/j.gecco.2021.e01540
Corrêa, G. V., Ingels, J., Valdes, Y. V., Fonseca-Genevois, V. G., Farrapeira, C. M., Nolte, C. R., Nel, R., and Pfaff, M. C. (2020). Determining body condition of nesting
and Santos, G. A. (2014). Diversity and composition of macro-and meiofaunal loggerhead sea turtles (Caretta caretta) in the south-west Indian Ocean. J. Mar.
carapace epibionts of the hawksbill sea turtle (Eretmochelys imbricata Linnaeus, Biol. Assoc. U.K. 100, 291–299. doi: 10.1017/S0025315420000107
1822) in Atlantic waters. Mar. Biodivers. 44, 391–401. doi: 10.1007/s12526-013- Perrault, J. R., Muller, E. M., Emily, R., and Rotjan, R. D. (2015). Presence of the
0189-9 northern star coral (Astrangia poculata) as an epibiont on the carapace of a
Crespo-Picazo, J. L., García-Parraga, D., Domènech, F., Tomás, J., Aznar, F. J., nesting loggerhead turtle (Caretta caretta) in the western Gulf of Mexico. Reef
Ortega, J., et al. (2017). Parasitic outbreak of the copepod Balaenophilus Encount. 30:46.
manatorum in neonate loggerhead sea turtles (Caretta caretta) from a Pfaller, J. B., Alfaro-Shigueto, J., Balazs, G. H., Ishihara, T., Kopitsky, K., Mangel,
head-starting program. BMC Vet. Res. 13:154. doi: 10.1186/s12917-017-1 J. C., et al. (2014). Hitchhikers reveal cryptic host behavior: new insights from
074-8 the association between Planes major and sea turtles in the Pacific Ocean. Mar.
Darwin, C. (1851). A Monograph on the Sub-Class Cirripedia, with Figures of All Biol. 161, 2167–2178. doi: 10.1007/s00227-014-2498-3
Species. The Lepadidae, or, Pedunculated Cirripedes. London: The Ray Society. Pfaller, J. B., and Robinson, N. J. (2022). Data from: sea turtle symbiosis: global
Darwin, C. (1854). A Monograph on the Subclass Cirripedia, with Figures of All the patterns and knowledge gaps. Dryad Digit. Repository doi: 10.5061/dryad.
Species. The Balanidae, the Verrucidae, etc. London: The Ray Society. x3ffbg7m8
Domènech, F., Tomás, J., Crespo-Picazo, J. L., García-Párraga, D., Raga, J. A., Pikesley, S. K., Maxwell, S. M., Pendoley, K., Costa, D. P., Coyne, M. S., Formia, A.,
and Aznar, F. J. (2017). To swim or not to swim: potential transmission of et al. (2013). On the front line: integrated habitat mapping for olive ridley sea
Balaenophilus manatorum (Copepoda: Harpacticoida) in marine turtles. PLoS turtles in the southeast Atlantic. Divers. Distrib. 19, 1518–1530. doi: 10.1111/
One 12:e0170789. doi: 10.1371/journal.pone.0170789 ddi.12118
Fossette, S., Girard, C., Lopez-Mendilaharsu, M., Miller, P., Domingo, A., Evans, Pilsbry, H. A. (1916). The sessile barnacles (Cirripedia) contained in the collections
D., et al. (2010). Atlantic leatherback migratory paths and temporary residence of the U.S. National Museum; including a monograph of the American species.
areas. PLoS One 5:e13908. doi: 10.1371/journal.pone.0013908 U.S. Natl. Mus. Bull. 93, 1–366.
Frick, M. G., and Pfaller, J. B. (2013). “Sea turtle epibiosis,” in The Biology of Sea Pinou, T., Domènech, F., Lazo-Wasem, E. A., Majewska, R., Pfaller, J. B., Zardus,
Turtles, Vol. 3, eds J. Wyneken, K. J. Lohmann, and J. A. Musick (Boca Raton, J. D., et al. (2019). Standardizing sea turtle epibiont sampling: outcomes of the
FL: CRC Press), 399–426. epibiont workshop at the 37th International Sea Turtle Symposium. Mar. Turt.
Frick, M. G., Williams, K. L., and Robinson, M. (1998). Epibionts associated with Newsl. 157, 22–32.
nesting loggerhead sea turtles (Caretta caretta) in Georgia, USA. Herpetol. Rev. Pinou, T., Lazo-Wasem, E. A., Dion, K., and Zardus, J. D. (2013). Six degrees of
29, 211–214. separation in barnacles? Assessing genetic variability in the sea-turtle epibiont
Greenblatt, R. J., Work, T. M., Balazs, G. H., Sutton, C. A., Casey, R. N., and Stomatolepas elegans (Costa) among turtles, beaches and oceans. J. Nat. Hist. 47,
Casey, J. W. (2004). The Ozobranchus leech is a candidate mechanical vector for 2193–2212. doi: 10.1080/00222933.2013.798701
the fibropapilloma-associated turtle herpesvirus found latently infecting skin Rees, A. F., Al Saady, S., Broderick, A. C., Coyne, M. S., Papathanasopoulou, N.,
tumors on Hawaiian green turtles (Chelonia mydas). Virology 321, 101–110. and Godley, B. J. (2010). Behavioural polymorphism in one of the world’s largest
doi: 10.1016/j.virol.2003.12.026 populations of loggerhead sea turtles Caretta caretta. Mar. Ecol. Prog. Ser. 418,
Harder, T. (2009). “Marine epibiosis: concepts, ecological consequences and host 201–212. doi: 10.3354/meps08767
defence,” in Marine and Industrial Biofouling. Springer Series on Biofilms, Vol. Reich, K. J., Bjorndal, K. A., Frick, M. G., Witherington, B. E., Johnson, C., and
4, eds H.-C. Flemming, P. Sriyutha Murthy, R. Venkatesan, and K. E. Cooksey Bolten, A. B. (2010). Polymodal foraging in adult female loggerheads (Caretta
(Berlin: Springer), 219–231. doi: 10.1007/978-3-540-69796-1_12 caretta). Mar. Biol. 157, 113–121. doi: 10.1007/s00227-009-1300-4
Hatase, H., Omuta, K., and Tsukamoto, K. (2010). Oceanic residents, neritic Robinson, N. J., Figgener, C., Gatto, C., Lazo-Wasem, E. A., Paladino, F. V.,
migrants: a possible mechanism underlying foraging dichotomy in adult female Tomillo, P. S., et al. (2017a). Assessing potential limitations when characterising
loggerhead turtles (Caretta caretta). Mar. Biol. 157, 1337–1342. doi: 10.1007/ the epibiota of marine megafauna: effect of gender, sampling location, and inter-
s00227-010-1413-9 annual variation on the epibiont communities of olive ridley sea turtles. J. Exp.
Heppell, S. S., Snover, M. L., and Crowder, L. B. (2003). “Sea turtle population Mar. Biol. Ecol. 497, 71–77.
ecology,” in The Biology of Sea Turtles, Vol. 2, eds P. L. Lutz, J. A. Musick, and J. Robinson, N. J., Lazo-Wasem, E. A., Paladino, F. V., Zardus, J. D., and Pinou,
Wyneken (Boca Raton, FL: CRC Press), 275–305. doi: 10.1201/9781420040807. T. (2017b). Assortative epibiosis of leatherback, olive ridley and green sea
ch11 turtles in the Eastern Tropical Pacific. J. Mar. Biol. Assoc. U.K. 97, 1233–1240.
Ingels, J., Valdes, Y., Pontes, L. P., Silva, A. C., Neres, P. F., Corrêa, G. V., et al. doi: 10.1017/S0025315416000734
(2020). Meiofauna life on loggerhead sea turtles-diversely structured abundance Robinson, N. J., Majewska, R., Lazo-Wasem, E. A., Nel, R., Paladino, F. V., Rojas, L.,
and biodiversity hotspots that challenge the meiofauna paradox. Diversity et al. (2016). Epibiotic diatoms are universally present on all sea turtle species.
12:203. doi: 10.3390/d12050203 PLoS One 11:e0157011. doi: 10.1371/journal.pone.0157011

Frontiers in Ecology and Evolution | www.frontiersin.org 16 March 2022 | Volume 10 | Article 844021
Robinson and Pfaller Sea Turtle Epibiosis

Ten, S., Pascual, L., Pérez-Gabaldón, M. I., Tomás, J., Domènech, F., and Aznar, F. J. Zardus, J. D. (2021). A global synthesis of the correspondence between epizoic
(2019). Epibiotic barnacles of sea turtles as indicators of habitat use and fishery barnacles and their sea turtle hosts. Integr. Org. Biol. 3:obab002. doi: 10.1093/
interactions: an analysis of juvenile loggerhead sea turtles, Caretta caretta, in iob/obab002
the western Mediterranean. Ecol. Indic. 107:105672. doi: 10.1016/j.ecolind.2019.
105672 Conflict of Interest: The authors declare that the research was conducted in the
Vander Zanden, H. B., Arthur, K. E., Bolten, A. B., Popp, B. N., Lagueux, C. J., absence of any commercial or financial relationships that could be construed as a
Harrison, E., et al. (2013). Trophic ecology of a green turtle breeding population. potential conflict of interest.
Mar. Ecol. Prog. Ser. 476, 237–249. doi: 10.3354/meps10185
Vander Zanden, H. B., Bjorndal, K. A., Reich, K. J., and Bolten, A. B. (2010). Publisher’s Note: All claims expressed in this article are solely those of the authors
Individual specialists in a generalist population: results from a long-term stable and do not necessarily represent those of their affiliated organizations, or those of
isotope series. Biol. Lett. 6, 711–714. doi: 10.1098/rsbl.2010.0124 the publisher, the editors and the reviewers. Any product that may be evaluated in
Wahl, M. (2009). “Epibiosis,” in Marine Hard Bottom Communities, ed. M. Wahl this article, or claim that may be made by its manufacturer, is not guaranteed or
(Berlin: Springer), 61–72. doi: 10.1007/b76710_4 endorsed by the publisher.
Wallace, B. P., DiMatteo, A. D., Hurley, B. J., Finkbeiner, E. M., Bolten, A. B.,
Chaloupka, M. Y., et al. (2010). Regional management units for marine turtles: Copyright © 2022 Robinson and Pfaller. This is an open-access article distributed
a novel framework for prioritizing conservation and research across multiple under the terms of the Creative Commons Attribution License (CC BY). The
scales. PLoS One 5:e15465. doi: 10.1371/journal.pone.0015465 use, distribution or reproduction in other forums is permitted, provided the
WoRMS Editorial Board (2021). World Register of Marine Species. Available Online original author(s) and the copyright owner(s) are credited and that the original
at: https://www.marinespecies.org at VLIZ [accessed December 5, 2021]. publication in this journal is cited, in accordance with accepted academic practice.
Wyld, J. A., and Brush, A. H. (1983). Keratin diversity in the reptilian epidermis. No use, distribution or reproduction is permitted which does not comply with
J. Exp. Zool. 225, 387–396. doi: 10.1002/jez.1402250306 these terms.

Frontiers in Ecology and Evolution | www.frontiersin.org 17 March 2022 | Volume 10 | Article 844021