Drinking by sea snakes from oceanic freshwater lenses at first rainfall ending seasonal drought

Abstract

Acquisition of fresh water (FW) is problematic for FW-dependent animals living in marine environments that are distant from sources of FW associated with land. Knowledge of how marine vertebrates respond to oceanic rainfall, and indeed the drinking responses of vertebrates generally following drought, is extremely scant. The Yellow-bellied Sea Snake (Hydrophis platurus) is the only pelagic species of squamate reptile and ranges across the Indo-Pacific oceans, having one of the largest geographic distributions of any vertebrate species. It requires FW and dehydrates at sea during periods of drought. Here we report drinking behaviors of sea snakes precisely at the transition from dry to wet season when rainfall first impacted the ocean following 6 months of seasonal drought. We show that the percentage of sea snakes that voluntarily drank FW in the laboratory when captured over eight successive days decreased from 80% to 13% before and after rainfall commenced, respectively. The percentage of snakes that drank immediately following capture exhibited a significant linear decline as the earliest rains of the wet season continued. Drinking by snakes indicates thirst related to dehydration, and thus thirsty snakes must have dehydrated during the previous six months of drought. Hence, the progressive decline in percentage of thirsty snakes indicates they were drinking from FW lenses associated with the first rainfall events of the wet season. These data reinforce the importance of accessing oceanic FW from precipitation, with implications for survival and distribution of pelagic populations that might be subjected to intensifying drought related to climate change.

Fig 1. Map of study area in Costa Rica.

A. Northwest coast of Guanacaste showing location of Golfo de Papagayo and beaches near the offshore area where sea snakes were collected (shown in gray). B. Inset showing location of Costa Rica in Central America with arrow pointing to the area of study offshore at Guanacaste.

Fig 2. Daily rainfall for the year May 2016 to May 2017.

Vertical bars quantify daily rainfall and illustrate the period of drought roughly from December 2016 to May 2017. For presentation, we calculated the mean daily rainfall (mm) from CMORPH data [19] of two grid values represented by boxes and the area shown on the inset map. The black dots represent GPS locations where sea snakes were collected. The area shown in grey represents land, and the ocean is shown in white. Note that land as well as sea is represented in part of the right-hand grid.

Fig 3. The percentage of sea snakes (Hydrophis platurus) that voluntarily drank fresh water immediately following capture on different days beginning at the end of the dry season at Golfo de Papagayo, Costa Rica.

These percentages are plotted in relation to the beginning of the very first storms of the wet season that produced heavy rainfall over the ocean where snakes were collected. The cloud symbols represent heavy rainfall that was observed directly over the ocean during the afternoon or evening beginning just before the third day of sampling. The small numbers next to each data point represent the numbers of snakes in each sample. The upper graph illustrates the daily rainfall as determined from NOAA CPC Morphing Technique (CMORPH) Global Precipitation Analyses Version 0.x [19]. The CMORPH data likely do not reflect accurately the precipitation that was observed because of the low spatial resolution relative to the area of water where we collected snakes (see Figs 1 and 2). The right-hand grid cell includes land area as well as ocean, hence the small amount of rainfall on Day 1 likely reflects precipitation we observed over land but not the ocean, and therefore not accessible to H. platurus. Following the advent of significant precipitation that we observed over the ocean (cloud symbols), the linear decline in number of snakes drinking fresh water in the laboratory was significant (dashed line in lower graph; Y = 85.819–8.7729X, R² = 0.851, P = 0.0031).

Fig 4. Three discrete points of rainfall during afternoon thunderstorms observed on 12 May 2017, Golfo de Papagayo, Guanacaste, Costa Rica.

The photograph illustrates how rainstorms over the ocean can produce patchy opportunities for pelagic snakes to drink fresh water. Photograph by H.B. Lillywhite.

Fig 5. Salinity of ocean surface waters where snakes were collected.

Data points represent salinity of water measured on each day that snakes were collected, and the vertical dashed line indicates the onset of wet-season rainfall.

Fig 6. Relative amount of fresh water that was drunk by snakes captured on different days at the transition of dry to wet season (days 2–3) at Golfo de Papagayo, Guanacaste, Costa Rica.

Data are displayed as means ± SEM. Sample sizes can be determined from Fig 3. The variation in amount of water drunk was not significant across the span of daily measurements (ANOVA, F_{6, 40}, P = 0.255).