ONE SHELL OF A PROBLEM: CUMULATIVE THREAT ANALYSIS OF MALE SEA TURTLES INDICATES HIGH ANTHROPOGENIC THREAT FOR MIGRATORY INDIVIDUALS AND GULF OF MEXICO RESIDENTS

Human use of oceans has dramatically increased in the 21st century. Sea turtles are vulnerable to anthropogenic stressors in the marine environment because of lengthy migrations between foraging and breeding sites, often along coastal migration corridors. Little is known about how movement and threat interact specifically for male sea turtles. To better understand male sea turtle movement and the threats they encounter, we satellite-tagged 40 adult male sea turtles of four different species. We calculated movement patterns using state-space modeling (SSM), and quantified threats in seven unique categories; shipping, fishing, light pollution, oil rigs, proximity to coast, marine protected area (MPA) status, and location within or outside of the U.S. Exclusive Economic Zone (EEZ). We found significantly higher threat severity in northern and southern latitudes for green turtles (Chelonia mydas) and Kemp’s ridleys (Lepidochelys kempii) in our study area. Those threats were pervasive, with only 35.9% of SSM points encountering no high threat exposure, of which 47% belong to just two individuals. Kemp’s ridleys were most exposed to high threats among tested species. Lastly, turtles within MPA boundaries face significantly lower threat exposure, indicating MPAs could be a useful conservation tool.


Table of Contents
Human use of oceans has dramatically increased in the 21 st century, with some of the highest rates of change found within the Gulf of Mexico and Caribbean. Sea turtles are particularly vulnerable to anthropogenic stressors in the marine environment as they make lengthy migrations between foraging and breeding sites often along coastal migration corridors. Sea turtles face severe population pressure from humans, yet little is known about how movement and threats interact specifically for male sea turtles. To better understand male sea turtle movement, and the threats they encounter within their expansive ranges, we tagged 40, adult male sea turtles of four different species. We calculated movement patterns using state-space modeling (SSM) and quantified threats in seven unique categories: shipping, fishing, light pollution, oil rigs, proximity to coast, marine protected area (MPA) status, and location within or outside of the US Exclusive Economic Zone (EEZ). We found multiple clusters of male sea turtles within the Gulf of Mexico and Caribbean, with significantly higher threat severity in northern and southern latitudes for green turtles and Kemp's ridleys. These threats are pervasive, with only 35.9% of SSM points encountering no high threat exposure, of which 47% belong to just two individuals. We also found Kemp's ridleys were most exposed to high threats among tested species. Lastly, turtles within MPA boundaries face significantly lower threat exposure, supporting the use of MPAs within the United States (US) as a conservation tool.

Introduction:
Human use of oceans has dramatically increased in the 21 st century. Results from a five-year study found that by 2013, 66% of the world's oceans faced increased pressure from anthropogenic activities like fishing, pollution, global shipping, and elevated sea surface temperatures from 2008 levels [1]. Some of the highest cumulative impact scores (top 5% of scores) were in the Gulf of Mexico, Caribbean, and Central Atlantic Ocean [1]. Additional studies on human impact have found that the Caribbean scores as one of the highest regions in the world for marine-associated threats [2]. Due to high pressure from human activity, species within the Caribbean show the greatest rates of loss within neritic (coastal) habitats, where the highest species richness occurs [3]. In the Gulf of Mexico, more than 79 species are listed as at least Near Threatened by the International Union for Conservation of Nature (IUCN), which represent seven of nine marine reptiles, five of 27 marine mammals, and 19 of 44 shark species [4]. Additionally, more than 26% of endemic bony fish in the Gulf of Mexico and Caribbean are facing severe population declines [3]. High cumulative human impact on marine wildlife is disruptive and can significantly increase the chance of species extirpation or extinction [5].
One of the major ways that human activities impact marine animals is by disrupting foraging and breeding behavior. Such disruptions force animals out of optimal habitat, which can decrease an individual's overall fitness [6,7,8,9,10]. Furthermore, marine wildlife that remain near human activity often experience reduced foraging time with diminished catch because they spend time avoiding interaction with humans due to stressors such as noise from ship traffic [11,12,13]. Habituation to human activity can also be dangerous because it increases the risk of injury or death from boat strikes, capture, or aggressive behavior from conspecifics [14,15,16].
Habituation has also been documented to decrease health and increase disease risk in marine species due to improper diet [17].
Wildlife can also be disrupted by the indirect effects of human activity. For example, threats can occur along spatial gradients, as is the case with marine debris from areas with heavy shipping and fishing presence [18]. Similarly chemical disturbances, such as those from polychlorinated biphenyls (PCBs) and oil spills, are often most concentrated at the source of contamination, but with wider ranging, though more diffuse effects, through bioaccumulation or dispersal of affected individuals [19,20]. Additional pressures from threats like fishing and tourism can also be spatially autocorrelated, with high areas of pressure found closer to the shore or in clusters where activity is greatest [16,21]. However, of the disruptive activities on marine wildlife, fishing is regarded as one of the most destructive.
Fishing practices in the 21 st century have exhausted multiple fisheries across the globe, to the point that only 32% have stocks above replenishable levels [22]. Current data on fisheries suggest that even as the size of the global fishing fleet has more than doubled from 1950 to 2017, catch per unit of effort of fish during that period decreased by more than 80% in most areasevidence that global fish abundance may be rapidly declining [23,24]. Additionally, the world's fishing fleet uses methods that catch many species in non-target trophic groups, which is referred to as bycatch [25,26,27]. Despite upgrades in fishing technology to reduce bycatch, oversight to ensure newer technology is being used is lacking [26]. Fishing methods meant to catch a target species are often indiscriminate [23,26]. Undesirable fish are discarded at sea, and due to lack of a governing body or trained individuals to observe active fishing vessels, especially on artisanal fishing boats, the catch goes under-reported, or unreported [26,28]. The pressures caused by vessels from the world's fishing fleet puts fish and other marine species at risk, either directly from bycatch, or indirectly from boat strikes and pollution. Species that conduct long migrations are particularly at risk, such as elasmobranchs [29,30], pinnipeds [31,32] and chelonids [27,35,36,37]. To best protect migratory marine species, researchers rely on tracking individuals to better understand the threats they encounter within and en-route to their feeding and breeding locations. [7,29,30,31,35,36,38,39,40,41,42,43,44]. Locations where species of interest aggregate can be used by managers to design strategies that protect them, for example by establishing or expanding marine protected areas (MPAs) [7,39,42]. If placed correctly, MPAs can be beneficial at protecting migratory species in addition to residents of the locale, particularly those that hold economic value like species consumed by humans [45,46,47,48].
Because they represent a well-known, charismatic group of migratory marine megafauna, sea turtles have been used to justify the establishment of MPAs. For example, Mexico [49], Gabon [50,51], and Indonesia [52] have established MPAs in areas where tracking studies found high use by sea turtles [53]. Sea turtle tracking studies have also helped create zones that prohibit oil exploration or pipe laying in areas that intersect migration routes [53,54]. Sea turtles also benefit from previously established MPAs. Large sea turtle aggregations of multiple species can be found within US MPAs, such as the Florida Keys National Marine Sanctuary (FKNMS) and Dry Tortugas National Park (DRTO) [55].
Established in 1990 and 1992, respectively, to create critical marine habitat and protect marine resources for a number of imperiled species, the FKNMS and DRTO encompass over 2900 square nautical miles and contain a mix of restricted and prohibited human activity zones, also known as marine zoning [56,57,58]. The FKNMS and the reserves it overlaps, including 7 DRTO, have been found to have positive impacts, such as increasing population numbers for marine species, including green and loggerhead turtles and sharks, and species of economic value such as spiny lobsters (Family Palinuridae) [39,59,60,61]. Thanks to marine zoning and restricted human activities, the FKNMS and DRTO show high use from sea threatened and endangered sea turtles [39].  [62,63,64,65,66,67,68]. During nesting years, adult female sea turtles will undergo migrations of up to thousands of kilometers to their natal beaches to nest, often traveling through areas with high human activity [7,44,69,70,71]. Female turtles aggregate in the vicinity of nesting sites for the duration of the mating season, where they lay multiple clutches of eggs before returning to their foraging grounds [72,73]. There is high fidelity to nesting locations each season, often in neritic habitat, which puts sea turtles at risk of human-wildlife conflicts [7]. For example, small-scale fisheries in Greece report heavy interaction involving loggerhead turtle bycatch in nearshore waters annually, one of the primary nesting locations for Mediterranean loggerhead turtles [74].
In response to human presence, turtles may be moving to less ideal habitat where interactions are fewer. For example, in Zakynthos Island, Greece, Schofield et al. [8] determined that during the 2020 lockdown due to the Covid-19 pandemic, nesting female sea turtles moved to warmer waters closer to shore that were previously occupied by high densities of humans 8 versus staying further offshore in colder waters. Turtles moving to warmer waters when humans were absent suggests they were residing in lower quality habitat to avoid human interactionsa behavior not detected until disruption of the daily pressures posed by the tourist industry. The phenomenon of species changing their behaviors to avoid humans has been well documented.
Multiple mammalian species across the globe have shifted to nocturnal foraging patterns in response to heavy human presence, with recreation, resource harvesting, extractive activities, development, and vehicles being some of the leading causes behind this shift [75]. Species will also outright flee areas occupied by humans or other predators, even if the area is richer in resources [76]. In marine species, alteration of behavior to avoid humans has been less documented, but has been found in sperm whales [77] and killer whales [78]. The recent paper by Schofield et al. [8] also indicates that turtles may be exhibiting similar behaviors of avoiding areas with heavy human presence that warrant further study.
Although how sea turtles respond to human threat is understudied, there is strong evidence suggesting that sea turtles face multiple anthropogenic threats throughout their range [36]. However, due to differences in habitat, species may vary in the degree to which they encounter various threats [40]. Kemp's ridleys, for example, nest almost exclusively in Rancho Nuevo, Mexico, and are primarily found in the Gulf of Mexico, exposing them to a larger number of threats derived from oil pollution than hawksbill sea turtles, who are more confined to coral reefs and other tropical climates, at least within Caribbean, Southern Gulf of Mexico, and Atlantic populations [20,44,79,80,81]. Another study found that female Kemp's ridleys have much larger foraging ranges than loggerhead turtles, and minimal overlap of foraging grounds within the Gulf of Mexico, exhibiting spatial partitioning of habitat, which could indicate different exposure to certain human threats [40]. Foraging and migration timing also differs by 9 species, even in areas where population overlap occurs. One study found that although a foraging ground in Florida was shared by three species of sea turtle (loggerhead, green, and Kemp's ridley turtles), spatiotemporal partitioning existed in dive depth, duration, and the time at which diving behavior occurred [82]. Seasonal migration timing may also play a factor in threat exposure, as different populations of species will move out of shared spaces at different times. One study found that in a seagrass bed used by three species of sea turtle, Kemp's ridleys and loggerhead turtles left the area when water temperatures dropped, but green turtles remained through the winter season [83]. Therefore, due to differences in behavior, foraging location, and migration patterns, anthropogenic threats may impact the various species in different ways.
Data regarding sea turtles, although extensive, is mostly garnered from studies of females, due to the relative ease of capture on beaches when they nest [84,85,86]. Male sea turtles, however, are understudied, because they spend their entire lives in the ocean where inwater captures are more logistically difficult and financially expensive [42,86,87]. To date, the largest sample size of male sea turtles has come from Schofield et al. [7], who were able to track and record 38 adult, male loggerhead turtles, of which only five were tracked for more than one season. Because most studies have small sample sizes and cover short temporal scales, there is a data deficiency in the literature regarding male sea turtles. This is especially true for studies of migration, residency areas during non-breeding times, and the anthropogenic threats they face in those locations.
The tracking of male sea turtles is important from a conservation standpoint, as they exhibit different movement patterns than females [42,44]. While some male sea turtles reside near nesting beaches year-round, others have been found to exhibit long-distance migrations between breeding and feeding grounds with unique timing compared to females [42,44,88].
Migrations often remove species from protected areas into locations that increase an individual's risk of mortality from human interaction, such as artisanal and benthic fishing fleets in international waters or boundaries of countries where protection is weak [35,89]. Additionally, shipping lanes, and pollution from ships and oil platforms increase mortality risk [35]. There is also evidence that light pollution negatively impacts adult female sea turtles in addition to hatchlings [90,91]. Thus far, however, the risks posed by these threats have not been systematically evaluated for male sea turtles.
Male sea turtles will be increasingly impacted by human threats as climate change accelerates in the 21st century. When coupled with sea level rise and coastal development, nesting beach habitat and therefore recruitment in these already small populations will be further reduced [92,93]. Additionally, sea turtles are long lived, taking anywhere from 8 -24 years to reach sexual maturity [94,95,96]. Replacement of lost males in a population is therefore slow, and data on male mortality rates is nonexistent in most locations [87]. Migration routes and phenology may also differ in male sea turtles compared to females [7,42,44]. Lastly, operational sex ratios and the reproductive value of male sea turtles is unknown, as females are focal points of studies relating to reproduction [87]. For all these reasons, there is a need to study and examine the behaviors and movements of, and threats to, male sea turtles to better understand how and where to focus conservation efforts.
The purpose of this study was to track four species of male sea turtles to better understand their exposure to spatially and temporally variable threats. We focus specifically on threats faced by males within foraging areas and during long migrations between breeding and feeding grounds by using seven, unique threat categories: 1) within or outside of an MPA boundary; 2) within 10 km of a coastline; 3) within the Exclusive Economic Zone of the United States (US) or not; 4) fishing; 5) shipping; 6) oil rig proximity; and 7) light pollution levels. We predict that 1) tagged turtles in our dataset will conduct long-distance migrations that will put them into contact with threats; 2) threat intensity will vary along a spatial and temporal gradient with increasing distance from MPAs like the FKNMS and DRTO; 3) exposure to threats will vary by species; 4) threats will be lower within MPA boundaries than outside of them; and 5) turtle density will be lowest in areas of high threat.

Collection and Calculation of Threats/State-Space Modelling:
We collected a total of 8875 SSM points for threat analysis. Through review of scientific literature and professional consultation, we collected data for seven primary threats to male sea turtles (Fishing, Shipping, Drilling Platforms, Light Pollution, MPA boundaries, Located within or outside the US Exclusive Economic Zone (EEZ) and coastal threat (within 10 km of a coastline) [8,35,36,98,99,100,101]. Raw location data have spatial accuracy ranges that vary between 500 m to 1.5 km. Raw tracking data were therefore fit to a hierarchical, behavior-switching state-space model (SSM), which was then used to increase the accuracy of tracking data and to determine home ranges of each individual [102]. This allowed for estimation of the behavioral modes of individual turtles (unique behavioral patterns), regularize the locations in time, and account for location error in the raw data. In order to accurately depict the threats within the area of each SSM point (1.5 km), we created a two km radius buffer around each SSM turtle point using R [103], within which threats were assessed. The threat data were collected and prepared as follows:

Fishing Data:
Threats from fishing can come from a variety of sources (artisanal, longlining, commercial, nets and trawlers, etc.). Since 2016, all commercial fishing vessels within US waters over 65 feet in length are required to have an AIS (Automatic Identification System) transponder tag attached, which tracks the ships every hour via satellite GPS and ground-based receivers placed by the US Coast Guard [104]. At present, only 2% of the world's fishing vessels have AIS tags on board, but these ships account for more than 50% of total fishing effort [104]. We used a fishing density raster layer of ground-based, AIS tracked ships as a representative subsample of fishing fleet intensity from marinecadastre.gov, a joint collaborative data repository for marine-related research by the National Oceanic Atmospheric Administration (NOAA), and the Bureau of Ocean Energy Management (BOEM) [105]. This layer includes tracks of fishing vessels that leave US waters in the Gulf of Mexico and therefore provides a representative subsample for turtles that move beyond the US EEZ boundary.
Fishing intensity was measured in grid cells one hectare in size, with each cell representing the total number of fishing craft that passed through that cell with an AIS 13 transponder onboard within a given year. We added and averaged the total number of fishing vessels per cell from 2015 -2017 to get the mean fishing intensity per cell. We then created a single raster layer for analysis using R Version 4.1.0 [103]. Although only three years of fishing data were available, we assume fishing density was similar enough in previous years that the average cell values of the three years represent past fishing seasons. We averaged the fishing intensity score of all raster cells within each SSM turtle buffer and assigned that value as the fishing threat score for that point in R [103].

Shipping Data:
Shipping data were also obtained using AIS tagged ships, and were downloaded in their raw format courtesy of the US Coast Guard with certain identifiers scrubbed for privacy. We were able to obtain data for the entire study period within from the Marine Cadastre data repository [105]. These data cover the entire study area and are a suitable representative sub-sample of shipping data, clearly showing all shipping lanes within our study area.
AIS-tagged ships in the US account for 50 -60% of total shipping activity [105]. We clipped All AIS data to each 2 km turtle buffer by date and then merged the data into a single vector shapefile to create a layer of shipping points that coincide with the presence of each 2 km SSM turtle buffer. Because of the large size of the AIS data files, we ran an RStudio instance on Google's Cloud Computing Engine [103]. The total number of shipping points within each 2 km buffer were then added and assigned as the threat score for that SSM point.

Drilling Platform Data:
We downloaded drilling platforms point data, also referred to as oil rigs, oil platforms or drilling rigs to represent oil derived threats from the Marine Cadastre data repository [105]. We calculated the number of platforms within each turtle buffer by clipping the oil rig layer to the turtle layer and merging the data into a single Vector shapefile to create a layer of oil rig points within each SSM turtle buffer using ArcGIS Pro ver. 2.5 [106]. Drilling platforms were corrected by date to ensure they were in use during the date associated with the date of the 2 km turtle buffer.
Marine Cadastre, although very useful in acquiring data for the US, is missing drilling platform data for other parts of our study area, specifically Mexico and Cuba. In order to understand if turtles that left the US EEZ encountered oil threats, we used a world oil exploration shapefile, called PETRODATA, that covers oil drilling hotspots around the world [107]. Upon comparison with our existing dataset, 99.7% of oil rig points fall within the PETRODATA polygon for the US, Gulf of Mexico, oil exploration polygon. Therefore, we felt it was comparable since no public oil rig data is available for Mexico at the time of writing this manuscript. No international turtle points fell within the confines of oil polygons, so further calculations were not necessary.

Light Pollution Data:
In 2011, the SUOMI VIIRS (Visible Infrared Imaging Radiometer Suite) satellite was launched to track multiple spatial data, such as snow and sea ice cover, active wildfires, sea and ice surface temperatures, and day/night light reflectance and radiance at high resolution [108]. We created our light pollution threat layer by combining all available light radiance raster data from NOAA's Earth Observation Group public download domain and averaging the total radiance for each pixel [108]. In total, 54 raster files, ranging from January 13 th -March 8 th , 2021, were combined by taking the average light radiance of each pixel, and then recording the average value of all pixels within each 2 km turtle buffer user R [103]. We have no reason to believe that there was annual variation in light radiance during the study period, so we assume that the light data we collected are representative of all years for which we have tracking data.

MPA, EEZ, and Proximity to Coast Data:
We downloaded both the MPA Layer and EEZ Layer as vector layers from the Marine Cadastre data repository [105]. We created the coastal threat vector layer in ArcGIS Pro version 2.6 by making a 10 km buffer around all available land within the study region [106]. We assigned SSM points that were >10 km from coastline, or within the US EEZ or an MPA boundary were assigned a value of "0" to indicate threat absence, while points that were <10 km from the coast or outside of the US EEZ or an MPA boundary we assigned a value of "1" to indicate threat presence.
Because Marine Cadastre focuses on primarily US waters, we needed data on international MPAs, specifically for Cuba and Mexico. Those data were downloaded from the IUCN's World Database on Protected Areas website [109]. We followed the same format of assigning values of "1" or "0" for turtles that were within the confines of those MPAs.

Statistical Analyses:
To directly add and compare the effects of individual threats, we standardized all threat categories to a mean of zero and standard deviation of one, which allowed us to take the sum of all threats directly and create a cumulative threat score for each SSM point. Through preliminary data exploration, we discovered the data were substantially non normal, and greatly spatially autocorrelated. As a result, we removed SSM points that were within 4 km of one another. Data removal reduced the number of SSM points from 8875 to 474. Despite removal, spatial autocorrelation still existed, but the degree to which it existed was reduced. Moran's I of cumulative impact scores changed from 0.458 (p = <0.001) at distance class I to Moran's I of 0.313 (p = <0.001). Complete removal of spatial autocorrelation from our dataset would have thinned the data to too few points to be able to run an analysis. Therefore, we decided to strike a balance between reducing the data, yet also minimizing the degree to which spatial autocorrelation existed in our dataset.
To test the prediction that individual or combined threats varied with species along a latitudinal gradient we tested our thinned data using PERMANOVAs, a nonparametric test for significance as the data were still very non normal. We tested the prediction using the interactive effects of species and latitude responding to threat. We included the threat x species interaction because we expected that species may vary in their response to spatially varying threats. Because of the potential for areas with variables of high threat to be clustered, data may not show a direct, linear relationship with latitude. To better understand any latitudinal relationships present in our data, we ran breakpoint regressions between latitude and threat (individual threats vs latitude and cumulative threat scores vs latitude) to see if the breakpoint model was a better fit to the data.
Due to the large spatial gap present and low sample size, we removed turtle 14, the lone hawksbill of this study from this portion of the analysis.
In order to test the prediction that threats vary by species, we first calculated median values (due to the presence of outliers) of threat scores for each of the four numerical threats (Shipping = 0, Light = 1.3, Fishing = 1.2, Oil = 0) and used presence scores for the remaining three (Coast = 1, MPA = 1, EEZ = 1). Values above the median value for numerical threats, or categorical threats that had a score of one were categorized as "high threat" while values below the median, or had a score of zero, respectively, were considered "low threat". We then calculated the percent of days during the study period an individual turtle encountered high threats by dividing the number of days a high threat was encountered by the sum of their SSM points. We calculated average values for each species from these percentages to understand how often sea turtle species were exposed each threat during the study period. Preliminary data analysis discovered our data were very non normal. Therefore, we ran PERMANOVAs on each threat category percent by species. Turtle 14, the single hawksbill captured for our study was removed from this part of the analysis.
To test the prediction that turtles within MPA boundaries experienced lower threat, we recorded the mean time each individual turtle spent outside of an MPA using their SSM points. If individuals spent more than the mean value (23.6%) outside of an MPA, they were counted as a "non-MPA" turtle. Individuals that spent more time than the mean value within an MPA were counted as "MPA" turtles. Three threat variables of continuous data (Light, Shipping, Fishing) that were categorized as high threat were compared between non-MPA and MPA turtles for statistical significance with a Welch's T-Test. T-Test analyses were modified for unequal variance and if the equal variance assumption was violated.
To test the prediction that turtle density is lowest in areas of high threat, we created a 10x10 km grid cell fishnet over the study region and then used the 'Spatial Join' tool to merge all points within each grid of the fishnet in ArcGIS Pro version 2.5 [106]. The number of turtle points within each grid cell was treated as density for that specific area. Threat values of each grid cell were averaged if more than one turtle point was present. We then ran linear regression analyses to test for relationships between density and each individual threat, and several combinations of threat layers: All Combined threats, Oil threat (oil, shipping, coastal layers), Boat threat (fishing, shipping, coastal, EEZ layers), and Fishing threat (fishing, coastal, EEZ, MPA layers). We additionally ran breakpoint regressions on our data to determine if density responded to threat nonlinearly or in linear segments.
To better understand threat interactions through time, we added the total number of high threat categories encountered on a given day for each SSM point and then plotted the threat on a color gradient (from 0-6) by month of the year and individual turtle, which clearly displays the daily number of high threats each turtle was exposed to during the tracking period. Turtles were sorted by capture location, MPA status, and species. Plotting data in such a manner can provide a visual interpretation of threats by individual, as well as show the timing of threats by month. All data were analyzed using R version 4.1.0 [103].

Results:
Turtles were tracked for an average of 221.8 days. The tracking period ranged from 16 days  Figures 1, 3, 6-7). The fifth migratory green turtle, Turtle #16, made two migrations at similar times on two separate years. Increases in high threat exposure were found at similar times when migrations began for this individual (Figure 2; Supplemental Figure 4).
Whether it be from shipping, fishing, oil rigs, light pollution, moving within coastal boundaries or outside MPAs, or leaving on long-distance migrations outside of the US EEZ, most turtles experienced at least one threat on a daily basis. Of the 8875 SSM points, 35.9% encountered no high threat exposure, 47% of which belonging to just two individuals (turtle #25 and #4; Figure 2). However only 19 of the 8875 points, representing five individual turtles, had a cumulative threat score of 0.

Latitudinal Gradient:
Through our PERMANOVA analysis, we found a significant interaction between the effects of latitude and species on cumulative threat (Total cumulative threats F2,448 = 29.268, p = 0.001;   Figure 6). Kemp's ridleys were found to have the highest exposure of five threats (Fishing, Shipping, Light Pollution, Oil rigs and MPA threat; Figure 3). For three of these threats (Fishing Intensity, Light Pollution, and MPA threats), Kemp's ridleys encountered 100% high threat presence during their entire tracking period. They were also the only species to be found near oil rigs, with more than 50% of tracking days encountering high oil threat.
Loggerhead turtle remained significantly nearer to the coast than other species, with more than 30% of tracking days found within 10 km of coastlines (coastal threat). Green turtles were the only species to move beyond the US EEZ, with 5.2% of SSM points found in international waters. These points represent five individuals that migrated south, three to the Yucatan Peninsula and two to the northern coast of Cuba (Figure 7). Green turtles spent the most time within MPA boundaries and scored lowest in five out of the seven threat variables (Fishing Intensity, Light Pollution, MPA, Shipping and Oil threats; Figure 6). Differences between species for distance to coast and EEZ threat exposure were found to be nonsignificant (F2,36 = 0.6355, p = 0.504 and F2,36 = 0.7947, p = 0.479, respectively; Figure 3).

Threats and MPAs:
From our analysis, we found a significant relationship between high threat exposure and the  Figure   8).

Density and Threat:
There were no significant relationships between turtle density for five of the individual threats these associations were not further investigated. To further test whether or not a relationship existed between MPA boundaries predicting density, a breakpoint regression was run on all individual variables and cumulative threat, but was found to not be a significantly better fit than the initial, linear model.

Discussion:
Male sea turtles in the Gulf of Mexico, the Caribbean, and Atlantic Coast of Florida are under pressure from anthropogenic threats. Of the species we monitored, Kemp's ridleys had the highest threat exposure of all species for five out of seven threat categoriesthree on a daily 22 basis (100% of points in high exposure to fishing, light, and MPA threats). Kemp's ridleys were also the only species to be present in areas with high oil rig threat presence. Loggerhead turtles consistently experienced the next highest level of threat, with the highest coastal exposure threat among all three species. Green turtles scored the lowest threat exposure among species for five of seven threats. However, they were the only species that left the US EEZ and therefore scored highest in that category. The exposure of the aforementioned high threats varied significantly based on geographic location, MPA Status, and species. Lastly, turtles that remained within MPA boundaries at least 76.4% of the time faced significantly less exposure to high threat compared to those outside of MPA boundaries for three of our numerical threats (light, shipping, fishing).

Latitudinal Gradient:
We found significant differences between the interactive effects of species and latitude with all combined threats within our study site. Additionally, we found that the highest cumulative threat areas occurred in clusters north and south of the centrally located FKNMS and DRTO MPAs.
Cumulative impact mapping using standardized values of all combined threats found a threat gradient, with relatively high threat in southern latitudes which decreases around 24.5° N before increasing again in northern latitudes of the study area. When MPA threats are removed from our cumulative impact analysis, green turtles, and Kemp's ridleys displayed similar results, but loggerhead turtles show a reverse trend of higher threat in southern latitudes and lower threat in northern latitudes. The area with lowest scores, whether MPA threats are included or not, coincided with areas represented by green turtles within the boundaries of two MPAs, the FKNMS and DRTO. We did find an area of elevated threat within the MPA boundary however, 23 which is due to a major shipping lane that runs through the FKNMS along the southern and western side of Key West to the rest of the Gulf of Mexico and the Florida Straits. Additionally, due to zoning and restricted take zones within the FKNMS, the DRTO had lower threat exposure than the FKNMS.
Despite the cumulative trend of overall lower threat values within the confines of the FKNMS and DRTO, three individual threats (light, fishing, and shipping) had higher scores in the reserve for loggerhead and green turtles. Previous data has shown that migratory risks for female loggerhead turtles within this geographic area are high due to heavy boat traffic from fishing and ships with AIS transponders, particularly along the Florida Straits to the Atlantic coast of Florida [35]. Our data also show a similar trend for male sea turtles, with increasing threat values as turtles head northeast along the Florida Straits, and then increase again within the Gulf of Mexico. Despite an uptick in light, fishing, and shipping within the FKNMS along the Florida Straits, the lack of, or reduction in those and other threats within the FKNMS and DRTO compared to the Gulf of Mexico may indicate the effectiveness of MPAs within our study area.
Of all threats, fishing is often reported as one of the most impactful for turtles [35,36].
Previous studies within the Gulf of Mexico and Atlantic Ocean have found that fisheries bycatch represents a major source of mortality for sea turtles [35,40,41]. Hart et al. [40] also found that 77% of tracked turtles spent at least one day in high threat fishing locations. Our results also found that turtle points along the Atlantic Coast of Florida, the Florida Straits, and Gulf of Mexico had the highest fishing threat scores of all SSM points, which coincides with threats to female sea turtles [35].
Cumulative impact analysis, like the results in our study, has been used in other systems to develop ecosystem-based management practices [1,110,111]. Combined effects of all threats can help shed insight as to where the areas under highest anthropogenic influence are located.
Our study found the highest rates of cumulative impact within the Gulf of Mexico and Atlantic Coast, with the lowest rates of impact coinciding with the placement of the FKNMS and DRTO.
By restricting human activities within large aggregate areas, threats to survival can be reduced.

Threat Exposure:
We detected a significant relationship between individual threat exposure and species within our study. Overall, Kemp's ridley sea turtles faced the highest proportion threat exposure for five threats, with three high threats found at 100% of points (shipping, light pollution, MPA) and the remaining two being 85.2% (fishing) and 50.6% (oil rigs), respectively. Loggerhead turtles spent the most amount of time inside coastal waters (30.3%) and had the second highest scores of all tested threats they had values for. Green turtles were the only species that left the US EEZ (5.3%) but scored lowest for all remaining variables they had values for except for coastal threats.
There are likely several reasons for Kemp's ridley sea turtles being exposed to the highest threats in our sample. For one, within our dataset, sites to which Kemp's ridley have high fidelity (aggregate clusters of SSM points) are unprotected, exposing turtles to increased threat exposure.
Secondly, Kemp's ridley turtles are largely found within the Gulf of Mexico, which has the highest levels of fishing threat exposure from our study, as well as is documented as having the highest rates of recorded sea turtle bycatch within the US [112]. Research has found that sea turtle bycatch in the US from shrimp trawlers within the Gulf of Mexico was as high as 98% of 25 total turtle bycatch from 1990 -2007 [112]. Kemp's ridley turtles are also caught by recreational fishermen in the Gulf of Mexico. In one study, more than 12% of sampled turtles were found to have fishing hooks in their gastrointestinal tract [113]. Lastly, because they are a primarily Gulf of Mexico species, Kemp's ridleys were severely impacted by the Deepwater Horizon oil spill in 2010, along with local populations of green, loggerhead, and hawksbill turtles [20]. Notable side effects from the oil spill were deformities in developing embryos, increased mortality, reduced immune systems, movement impairments, and other symptoms associated with oil toxicity [81,114,115,116].
Despite setbacks, Kemp's ridleys have been the subject of major efforts by multiple federal and non-profit agencies to recover viable populations. In the 20 th century, Kemp's ridleys experienced a population collapse, losing more than 98% of their population between censuses in the 1940s and 1980s [44]. Despite a population rebound in recent decades, since 2010 nesting counts have plateaued in Texas and Rancho Nuevo, Tamaulipas, Mexico, where the majority of nests are found [117,118]. Part of this decline in nesting numbers has been hypothesized as a pulse event of a sudden drop in nesting females from the Deepwater Horizon oil spill, but that should be short lived [119]. Researchers have hypothesized that recent declines in nesting numbers are due to lack of available food resources. The lack of resources may stem from increases in neritic populations, competition with loggerhead turtles and fish species for food and discarded catch from fishing vessels, and increases in fishing pressure on crab and shrimp fisheries within the Gulf of Mexico [117]. Degradation of habitat could lead to further population declines in food resources for Gulf of Mexico resident sea turtles [117]. Due to the possibility that Kemp's ridleys receive food from fishing vessels and are conditioned to seek out humans for 26 food it is not surprising that we found shipping, fishing, and light threats to be the highest among tracked males of this species.
The trend in stagnant nesting success of Kemp's ridleys could also be indicative of a mismatch in placement and functionality of MPAs for sea turtles within the Gulf of Mexico.
Within the confines of our study, we found two clusters of high site fidelity for adult, male Kemp's ridleys off the coast of Louisiana (five males) and Alabama (being occupied by a single male), and one for male loggerhead turtles off of the Florida Panhandle. The three aggregates of SSM points almost entirely fall outside of any current, established MPA and are subjected to heavy exposure to anthropogenic threats which could be detrimental to the success of the species as adult turtles represent the life stage contributing to future generational growth [120].

Threats and MPAs:
At present there are mixed data on the effectiveness of MPAs. While some studies have found positive effects in establishing MPAs for threatened species, others have suggested their effects are neutral or even negative. Without proper enforcement, anthropogenic activities such as fishing can actually increase in MPAs, reducing their protective potential [101,121,122].
Revuelta et al. [121] found that without enforcement, MPAs in the Dominican Republic had significant increases in the aforementioned activities, putting resident populations within those MPAs at risk.
MPAs can also suffer from overpopulation of the species they intended to protect in the first place, which can lead to phenomena like overgrazing. Christianen et al. [123] found that MPAs within Indonesia were too small to manage the success of green turtle population increases, highlighting the need to incorporate necessary habitat into species management plans.
Lack of suitable habitat further highlights that MPAs are not a universal solution to widespread habitat degradation within a region. Despite the shortcomings of MPAs around the world, and the controversy on the actual effectiveness of MPAs to protect marine life, our data suggest that MPAs significantly reduce exposure of three threats (shipping, light, fishing) in areas that overlap with MPA boundaries.
Effective MPAs have been found to limit both extractive (fishing, oil exploration, etc.) and non-extractive threats (recreational boating, shipping, etc.) [124] as well as include active law enforcement to curb illegal activity [125]. Our dataset further supports evidence that MPAs within the US and Mexico are both effective and well-managed for male sea turtles. When placed in proper locations, and with adequate laws and enforcement, MPAs can help curb anthropogenic threats on sensitive marine species. Tracking data like those presented here have helped establish areas that reduce the severity of anthropogenic threats for resident animals [49,51,52]. Areas in Gabon, Mexico and Indonesia with satellite tracking studies have contributed to the creation of MPAs which have helped at-risk species and many tracking papers like this one have found areas that MPAs could be created or expanded [42,53,126]. If properly placed, MPAs that encompass aggregate populations of male sea turtles could help reduce mortality from anthropogenic threats [126,127].

Conclusion:
The data from this study provide us with more details on male sea turtle movements with emphasis on the Gulf of Mexico and the Caribbean, and the threats they are exposed to on a daily basis. Male sea turtles are an understudied group of their species, with most tracking studies focused on females due to being easier to tag when they nest. This study is the largest tracking study of male sea turtles to date, and reveals that although some males exhibit high site fidelity, others will conduct lengthy migrations that put them in direct overlap with multiple anthropogenic threats. Using cumulative impact analysis, we found a latitudinal gradient, with higher threat scores in northern and southern latitudes, and an area of low threat in central latitudes for all three tested species, and two of three species when MPA threats were removed.
Our data provides promising evidence to the effectiveness of MPAs when actively enforced and of a large enough size to sustain recruitment of populations of sensitive species. Lastly, we found that different species of male sea turtle face varying exposure of the same threat, with Kemp's ridleys being the most widely exposed to high threat, often on a daily basis. Kemp's ridleys are particularly subjected to high light pollution, fishing, oil, and shipping threats, as exposure is very high compared to other, tested species. We hope that these data will be able to support and encourage conservation measures for male sea turtles, especially given the continued decline in nesting numbers of these sensitive, marine reptiles.

Acknowledgments:
The data provided and all SSM calculations for this paper were given by Dr. Kristen Hart and her team from the US Geological Survey. Research funding was provided by the Eastern Shawnee Tribe of Oklahoma and the Watling Lab. I am very thankful to Dr. James Watling and Dr.
Kristen Hart for their analytical review of this thesis.  Missing data indicate lacking reliable tracking data or transmitters that terminated transmission before a given full calendar year. Turtles that have more than one calendar year of tracking data are displayed with a "-#" after their ID number. * Denotes turtles that conducted migrations, ** Denotes turtles that conducted migrations, but tracking data stopped before turtles returned to their foraging grounds. Acronyms are as follows: CC = loggerhead, CM = green, EI = hawksbill, LK = Kemp's ridley.