Light and temperature records of the seawater associated with southern elephant seal dives during foraging trips in South Atlantic and Pacific Oceans

Elena Eder, Marcos Zarate and Mirtha Lewis

Agencia Nacional de Promoción Científica y Tecnológica PICT 01- 11749 and Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) PIP 02462 Resolución 1123/03, Australian Antarctic Division and National Research Council of Argentina Ph.D. programme (CONICET). The publication of this data paper was supported by the Belgian Science Policy Office (BELSPO, contract n°FR/36/AN1/AntaBIS) in the Framework of EU-Lifewatch as a contribution the SCAR Antarctic biodiveristy portal (biodiversity.aq).

The locations and associated data from the seawater (light level and temperature) during the feeding trips of the southern elephant seals encompassed an area between the South Atlantic and Pacific Oceans (South West [-58.75, -81.29], North East [-37.60, -26.54]). Fig. 1

Figure 1.  

Study area with the locations of the associated data from the seawater during the feeding trips of the southern elephant seals. Green points represent locations where seawater light level and temperature data were collected.

LTL devices (Platypus Engineering, Sydney, Australia) were deployed simultaneously with satellite tags (SPOT4/SPOT5; Wildlife Computers, Redmond, Washington) on 13 immature southern elephant seals from the Península Valdés colony, Patagonia, Argentina (42°45’S, 63°38’W). The setting and deployment protocols of instruments and their recovery have been described previously in Campagna et al. (2006), Campagna et al. (2007) and Eder et al. (2011). Data recorded encompass the post-resting and post-moulting feeding trips of individuals during 2005 (n = 4), 2005-2006 (n = 7) and 2007 (n = 2) seasons. Additionally, morphometric data of individuals taken during deployment and retrieval of the instruments and body mass estimates before and after feeding trips, are also provided. Standard length (SL, snout-tail length) and maximum girth (MG) were taken with a measuring tape to the nearest 0.5 cm, in ventral recumbency and body mass (EBM) was estimated using the equation in Bell et al. (1997) (EBM= 53.896 SL^1.063 * MG^1.697).

Estimation of dive durations and surface intervals

The variation of light registered by the LTLs (records were sampled at an interval of 30-s, expressed as hh/mm/ss), were used to estimate the dive durations and the surface intervals of dives, as vertical movements through the water column are reflected by changes in the light level (from maximum value at the surface = 250 arbitrary units, to total darkness in deep water = 2 arbitrary units). Simulations of dives at sea with the instruments under different conditions (different seasons, total, partial or no cloud cover and different orientations of the light sensor), determined a conservative minimum value of saturated light level of 190 units at the sea surface. The details of these simulations to estimate dive duration are described in Eder et al. (2011). According to this, dive durations and surface intervals of individuals were estimated, based on diurnal records between 06:00 and 17:00 h of the LTLs (local time; Eder et al. (2011)). Records from night or long dives during sunrise and sunset (hours of attenuated daylight) from the winter months were excluded for this analysis (Eder et al. 2011), but they were included in the dataset in order to provide the temperature variation.

Validation and correction of the temperature data

The LTLs measure seawater temperature with a resolution of approximately 0.2ºC, over a range of -12 to 31ºC. However, some devices showed temperature records from -1.1 to 31.1ºC, which may not be appropriate values for the Argentine Sea and the adjacent Oceanic Basin. For this reason, the values of the devices were validated against an autonomous thermometer (Optic StowAway Temp) activated during the diving simulations and a digital thermometer during laboratory tests, to correct the temperature values of the LTLs. Fig. 2 shows the temperature profile recorded by the devices and the autonomous thermometer during the diving simulations detailed in Eder et al. (2011). During these simulations, the recorded temperatures never fell below 10ºC and the differences between the records of the LTL and the autonomous thermometer, under different conditions, was 4.4 ± 0.8ºC and 4.4 ± 2.3ºC. Fig. 3 shows the profile of the LTL and the digital thermometer records during exposure to low and high temperatures in the laboratory. In these experiences, the temperature difference was 2.1 ± 0.6ºC when the exposure was from 0.7 to 12ºC, 7.4 ± 1.4 when the temperature was 22 to 13.7ºC and 6.3 ± 1.9 when exposed to 31.06 and 22ºC. As can be seen in Figs 2, 3, the LTL temperature records behave differently at low and relatively high values. At temperatures below 6ºC, the records tended to underestimate the temperature values, although the difference is somewhat less than at temperatures greater than 6ºC, when the records tended to overestimate the values until reaching the upper limit of the range that the device can measure. Given this irregular behaviour of the temperature records of the LTL, the temperature data obtained in the laboratory were used to obtain a general adjustment function to correct the entire range of temperatures registered by the devices (Fig. 4). Fig. 5 shows the temperatures recorded with the digital thermometer and the temperatures of the equipment once the correction was applied.

Figure 2.  

The LTLs (1-8) and the autonomous thermometer (data logger) temperature profiles during diving simulations at sea. a) From a boat (two dives), during a cloudy day, at a maximum depth of 35 m. b) From a local wharf (four dives with the devices orientated in the positions described in Eder et al. (2011)), during a clear day, at a maximum depth of 18.5 m.

Figure 3.  

LTLs (1-8) temperature profiles during the gradual rise (a) and decrease (b) in the bath temperature at lab.

Figure 4.  

Adjusted function to correct the temperatures recorded by the LTL.

Figure 5.  

Overlapped temperature profiles of the digital thermometer and the LTLs (1-8) after correction.

All records were validated. The coordinates were validated using the check_onland() function of the obistools package to verify if there are points on land. Although the dataset has only one taxon, match_taxa() was used to determine if the taxonomic name is valid. All scientific names were checked for typos and matched to the species information backbone of Worlds Register of Marine Species (http:// marinespecies.org/) and LSID were assigned to taxon as scientificNameID. The original date data columns were converted with the OpenRefine tool to ISO 8601 format, which was assigned to the eventDate field of the Dwc standard. To check the consistency of the eventID and parentEventID fields, the check_eventids() function was used.

The locations during the feeding trips of the southern elephant seals encompassed an area between the South Atlantic and Pacific Oceans (South West [-58.75, -81.29], North East [-37.60, -28.65]). Most of the records are located within the well-known foraging areas of southern elephant seals from the Península Valdés colony. These include the continental shelf, an area of less than 200 m in depth that extends 300-400 km east from the coast and is characterised by mixed coastal and stratified waters (Campagna et al. 2006), the shelf break, where the Malvinas Current carry cold sub-Antarctic waters north from the Antarctic Circumpolar Current and the encounter with low-salinity shelf water originates a shelf-break front associated with temperature and salinity gradients and increased primary productivity (McGovern et al. 2022), and the deep Argentine Basin (6000 m), where the warm-salty subtropical waters, carried southwards by the Brazil Current, meet the cold-fresh subpolar waters carried northwards by the Malvinas Current, producing the Brazil-Malvinas Confluence, characterised by increased primary productivity, large temperature gradients and intense mesoscale eddy activity (Campagna et al. 2006, McGovern et al. 2022).