Sea turtles are ectothermic animals and like other reptiles have effective physiological, anatomical, and behavioral mechanisms by which they regulate their body temperature (Spotila and Standora 1985; Wallace and Jones 2008). However, the geographic and thermal ranges of nearly all species are constrained to tropical and subtropical latitudes because of an inability to persist in prolonged low temperatures (Hawkes et al. 2007; Wallace and Jones 2008).

The green turtle, Chelonia mydas, is one of the species that suffers frequent mass strandings due to coastal hypothermic stunning events (Shaver 1990). These events have also been reported for other sea turtles in single- or multiple-species stranding events, including loggerhead turtles, Caretta caretta; Kemp's Ridley turtles, Lepidochelys kempii; and hawksbill turtles, Eretmochelys imbricata (Wilcox 1896; Shoop 1980; Brongersma 1982; Meylan and Sadove 1986; Witherington and Ehrhart 1989; Schroeder et al. 1990; Shaver 1990; Morreale et al. 1992; Ogren and McVea 1995; Mitchell et al. 1997; Gerle et al. 2000; Foley et al. 2007, 2012; Bellido et al. 2008; Avens et al. 2012; Roberts et al. 2014, Shaver et al. 2017).

Juvenile green turtles behave actively when sea surface temperature (SST) is 18°–20°C or higher, but below a certain temperature threshold, they become inactive and may exhibit overwintering torpor (Felger et al. 1976; Seminoff 2000). This inactivity threshold varies among different green turtle aggregations around the globe, for example, 18°C in Florida (Mendonça 1983), 14°C in southeastern Australia (Read et al. 1996), and 15°C in the northeastern Pacific Ocean (Seminoff 2000). To avoid the negative physiological effects of cold waters, green turtles in temperate areas perform different survival strategies, such as avoiding cold coastal waters or reducing activity to cope with metabolic changes (Schwartz 1978; Witherington and Ehrhart 1989; Still et al. 2005; Roberts et al. 2014). In the latter behavior, named “winter dormancy” or “brumation”, turtles reduce their activity, movements, and metabolism (Felger et al. 1976; Schwartz 1978; Gregory 1982; Castro et al. 2007); this is similar to hibernation, but the difference is that reptiles present some specific activities, such as the need to breath regularly at the water's surface (Mayhew 1965). Only some populations of green turtles are known to practice winter dormancy, such as in the northwestern Atlantic and in the Gulf of Mexico (Felger et al. 1976; Mendoça 1983); this behavior has also been detected in loggerhead turtles in Florida (Carr et al. 1980) and in Greece and southern Italy (Ogren and McVea 1995; Hochscheid et al. 2005, 2007).

Green turtles may compensate for changes in SST with different physiological and behavioral adaptations; however, when they cannot compensate a drop in body temperature, they may suffer hypothermia. Hypothermia occurs when sea turtles are exposed to temperatures under 10°C for an extended period of time (e.g., Still et al. 2002; Shaver et al. 2017) in which they are unable to find warmer waters. This situation may occur as a consequence of unusually cold weather over a period of several days or the arrival of sudden cold fronts to their area of distribution (Schwartz 1978; Witherington and Ehrhart 1989; Milton and Lutz 2003). When SST falls to 10°C, hypothermic turtles may become lethargic and buoyant, and at 9°C, turtles start floating at the surface (Schwartz 1978; Milton and Lutz 2003) and could strand if the currents and winds take them to the coast. These floating turtles exhibit a comatose condition that, without treatment, can result in death (McMichael et al. 2008). Mortality can occur when turtles remain in waters at or below 5°–6°C (Schwartz 1978; Still et al. 2005; Roberts et al. 2014). Hypothermic stunning events are typically confined to small, distinct areas because a very specific set of atmospheric conditions is necessary to produce a significant decrease in SST (Witherington and Ehrhart 1989; Foley et al. 2007).

In Uruguay, juvenile green turtles have a continuous distribution along the Rio de la Plata estuary and the Atlantic coast, where important foraging grounds have been documented in the shallow coastal waters of the departments of Canelones, Maldonado, and Rocha (Fig. 1) (López Mendilaharsu et al. 2006; Vélez-Rubio et al. 2013). During the austral winter, a small proportion of this green turtle aggregation remains in coastal habitats, overwintering on the seafloor and undergoing winter dormancy to tolerate low temperatures (López-Mendilaharsu et al. 2006; Martinez Souza 2014; Vélez-Rubio et al. 2016) or in open slightly warmer open waters (González Carman et al. 2012). Two other sea turtle species are common in Uruguayan waters, including loggerhead turtles and leatherback turtles (Dermochelys coriacea), but their presence is scarce during the austral winter (Vélez-Rubio et al. 2013).

View Figure

• Download as Powerpoint
• Download Figure

In this article, we describe a large, unusual hypothermic stunning event that occurred in July 2012 in Punta del Este, Department of Maldonado, Uruguay. This punctual event provided the opportunity to characterize the green turtle population that overwinters in Uruguayan coastal waters. Our report may also help in predicting and responding to future mass stranding events caused by atypical oceanographic and meteorological conditions.

METHODS

The Uruguayan coast is part of a complex hydrological system that comprises the frontal zone of the Rio de la Plata estuary (RP) and the Atlantic Ocean. This is a transitional zone influenced by waters with contrasting features: warm and saline tropical waters from a branch of the Brazil Current and cold diluted sub-Antarctic waters derived from the Malvinas Current. These characteristics result in the presence of a strong alongshore salinity and temperature gradient of more than 15°C throughout the year (Acha et al. 2004; Ortega and Martínez 2007; Campos et al. 2008). Our study area is a small bay protected by the Punta del Este Peninsula and Gorriti Island, an area used by overwintering green turtles (Lezama et al. 2012; Vélez-Rubio et al. 2013).

The stranding monitoring has been coordinated by the nongovernmental organization (NGO) Karumbé, running the Marine Turtle Stranding and Rescue Network (Red de Rescate y Varamientos de Tortugas Marinas [RRVTM]) along the coast of Uruguay since 1999. This network records dead or injured sea turtles stranded on beaches. It is coordinated via 24-hr telephone hotline or e-mail, and the database is updated by Karumbé members (for more details, see Vélez-Rubio et al. 2013). Most of the Uruguayan coast is unmonitored, and the NGO gets information of strandings from reports to the RRVTM. The RRVTM records stranding events throughout the year, transporting live and comatose turtles to a rehabilitation center and releasing them after their recovery. The likely cause of stranding is determined by examining live turtles at the beach or when they are brought to the rehabilitation center. A complete external and internal examination is completed on deceased turtles through necropsy. This allows the detection of symptoms of hypothermia in turtles with good body condition, positive buoyancy, little presence of food in the esophagus and stomach, signs of pneumonia (black spots or nodules) in the lungs, and presenting no signs of other possible stranding causes (i.e., interaction with marine debris, signs of bycatch, or evidence of collision) (Thomson et al. 2009). Curved carapace length (notch to tip [CCL_n−t]) of stranded turtles is measured; green turtles with CCL < 90 cm are classified as juveniles (Almeida et al. 2011; Vélez-Rubio et al. 2013).

To analyze the oceanographic conditions during the stunning event on 13–25 July 2012, we used daily satellite data L3-mapped images of 4-km horizontal resolution from MODIS Aqua (http://oceancolor.gsfc.nasa.gov/cms). We calculated the daily SST climatology as the mean SST during 2003–2014, and the anomaly as the difference between the mean value of SST registered from 13 to 25 July 2012 and the mean SST during 2003–2014 (Fig. 2). To obtain a time-data series from the satellite images, we constructed an index box, giving the spatial mean of 210 pixels near the coast. To see how different 2012 was in relation to other years, we compared the minimum and maximum SST reached in an 11-yr period with the SST values reached in 2012. We decided to focus only on the colder months (May–September) to concentrate on the period when the event occurred. We calculated a weekly mean of SST and air temperature (AT) to smooth the daily variability, and then we found the maximum and minimum of each week except for 2012 (Fig. 3).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

To understand the meteorological conditions during the days of the stunning event, we plotted climatology and anomaly maps of the atmospheric variables, which included wind, sea level pressure, and AT. We used a Web-based plotting and analysis tool from http://www.esrl.noaa.gov/ with daily mean composites from the NCEP/NCAR reanalysis data. We also searched for historical satellite images of the enhanced infrared channel and water vapor from GOES-13 (http://satelite.cptec.inpe.br).

RESULTS

A total of 90 juvenile green turtles stranded along the Uruguayan coast from 12 to 25 July 2012. We found that strandings were a result of atypical oceanographic and meteorological conditions. Most turtles (63 turtles, or 70%) stranded over 3 d (15–17 July). Until this event, Karumbé's RRVTM had recorded a total of only 55 stranded green turtles associated with hypothermia between 1999 and 2011 (Vélez-Rubio et al. 2013; Karumbé, unpubl. data, 2012), and no sea turtle mass stranding events had been recorded in Uruguay prior to 2012. Most of the green turtles recorded in July 2012 were stranded in the external estuarine zone (departments of Canelones and Maldonado), particularly in Punta del Este (Fig. 1).

Of the 90 stranded turtles, only 35 individuals could be transported to the Karumbé Rehabilitation Center in Montevideo. The rest of the turtles were found by other organizations and released or buried; no additional information is available for these cases. Of the 35 turtles taken to the center, 20 presented buoyancy problems, pneumonia, skin infection diseases, and septicemia, and 15 appeared to be in good physical condition. All the individuals transported to the Karumbé Center were juveniles (CCL_n−t = 39.9 ± 3.96 cm SD; range = 31.5–48.5 cm, n = 35). No other sea turtle species stranded during those days.

In July 2012, record-breaking cold weather occurred throughout Uruguay with records of near-shore SST below 10°C. SST, which in winter normally reaches 11°–12°C (Defeo et al. 2009), descended over a 3-wk period, reaching an unusual low of 8°C, which is considerably lower than SST recorded during previous years (Fig. 3). A low-pressure center was registered in the southwestern Atlantic during those days in 2012, bringing persistent winds from the south and southwest. These winds reached 50 km/h, with positive anomalies of 18 km/h. The passage of a cold front was recorded with negative air temperature anomalies of −2.5°C. These winds likely had a strong influence on the mixing of the water column, which is faster in shallow coastal waters, where turtles seem to be dormant during the winter (Vélez-Rubio et al. 2016).

DISCUSSION

Our results show how the favorable conditions that allow green turtles to remain in the Uruguayan coastal waters in winter months changed abruptly in July 2012 with a sudden and prolonged decrease in AT and SST. The passage of a cold front and the persistent winds from the south likely promoted mixing of the water column, transporting cold water from the surface to the bottom and causing a decrease in the temperature of coastal waters. In that sense, sustained warm temperatures in previous winters could have increased the numbers of green turtles that decided to stay the colder months in Uruguayan coastal waters. This could have generated an ecological trap; when SST dropped in 2012, a high number of green turtles could not escape to warmer waters.

Mass strandings associated with hypothermic events are normally restricted to relatively small, well-defined areas (Witherington and Ehrhart 1989; Shaver 1990; Morreale et al. 1992), but our study area is more open and less confined (despite the presence of the Punta del Este Peninsula and Gorriti Island), so lower temperatures perhaps resulted in hypothermia even though they were not as low as the minimum SST reported for other mass stranding events (Table 1).

Table 1. Review of the available bibliography of hypothermic stunning events worldwide. For each event, we include years; location (NWA = northwestern Atlantic; NEA=northeastern Atlantic; SWA =  southwestern Atlantic; Med =  Mediterranean); sea surface temperature (SST) registered for the days of the event, including the minimum (min.) SST registered; the duration (days or weeks) of the event; and number of sea turtles affected (Cc = Caretta caretta; Lk = Lepidochelys kempii; Cm = Chelonia mydas; Ei = Eretmochelys imbricata). Size data are included for green turtles only (SCL = straight carapace length; CCL = curved carapace length). STSSN = Sea Turtle Stranding and Salvage Network.

View Fullscreen

The hypothermic stunning event described here occurred during 12 d, a similar duration in comparison with other similar events (e.g., Avens et al. 2012; Foley et al. 2012). Only green turtles were recorded during the mass stranding event reported here; other species, such as the loggerhead turtle, were probably absent because aggregations of this species are reduced in Uruguayan coastal waters during the austral winter (Vélez-Rubio et al. 2013). The size distribution of green turtles registered in the present study is similar to that of individuals reported in other hypothermic events in the northwestern Atlantic (Table 1), except in areas with a presence of larger juveniles (CCL > 70 cm), hypothermic turtles registered present a larger mean size (Witherington and Ehrhart 1989; Morreale et al. 1992).

Our study provides insights about the environmental conditions that result in unusual hypothermic stunning events, especially in temperate areas. Typically, only a small portion of the green turtle aggregation in Uruguayan coastal waters stays year-round (Martinez Souza 2014; Velez-Rubio et al. 2013, 2016). However, after successive years with positive anomalies in SST, green turtles may have been influenced to stay in Uruguay in higher numbers than usual, thus resulting in the mass stranding event reported here. The southwestern Atlantic Ocean aggregations have also been impacted by other human-related threats, such as fisheries bycatch and debris ingestion (Gallo et al. 2006; Sales et al. 2008; Lezama 2009; González Carman et al. 2011, 2013; Laporta et al. 2012; Vélez-Rubio et al. 2013; Teryda 2015; López-Mendilaharsu et al. 2016).

It is extremely necessary to monitor atmospheric and oceanographic variables (e.g., low pressure, cold fronts, AT, and SST), particularly in the extremes of the distributional range of sea turtles, to predict and effectively respond to future mass stranding events. Indeed, the combination of cold fronts and wind from the south at high speeds can cause water temperatures to decrease below 10°C and have the potential to result in additional mass stranding events. NGOs and government agencies should therefore be ready to respond to additional hypothermic stunning events when these environmental conditions are present.