Here the data for "Fijian sea krait behaviour relates to fine-scale environmental heterogeneity in old growth forest: the importance of integrated land-sea management for protecting amphibious animals" by Lowe, C., Keppel, G., Waqa, K., Peters, S., Fisher, R.N., Scanlon, A., Osborne-Naikatini, T, and Thomas-Moko, N  is provided. This article investigates the habitat of Yellow Lipped Sea Kraits, Laticauda colubrina, in the terrestrial realm on Leluvia Island, a small, topographically flat atoll in Fiji with coastal forest. The investigation uses concurrent microclimate measurements and behaviour surveys, as well as vegetation surveys, and the data collected for these analyses are provided here. Microclimates were significantly related to canopy cover, leaf litter depth, and distance from the high-water mark (HWM). Sea kraits were almost exclusively observed in coastal forest within 30 m of the HWM. Sloughing of skins only occurred within crevices of mature or dying trees. Resting L. colubrina were significantly more likely to occur at locations with higher mean diurnal temperatures, lower leaf litter depths, and shorter distances from the HWM. On Leleuvia, behaviour of L. colubrina therefore relates to environmental heterogeneity created by old-growth coastal forests, particularly canopy cover and crevices in mature and dead tree trunks. The importance of healthy coastal habitats, both terrestrial and marine, for L. colubrina suggests it could be a good flagship species for advocating integrated land-sea management. Furthermore, our study highlights the importance of coastal forests and topographically flat atolls for biodiversity conservation. Effective conservation management of amphibious species that utilise land- and seascapes is therefore likely to require a holistic approach that incorporates connectivity among ecosystems and environmental heterogeneity at all relevant scales.

Study Design

Six transects were established in the forested part of Leleuvia Island. Five 55 m transects were positioned perpendicular to the coast, commencing at 2 m inland of the high-water mark (HWM) – see article for details. Another 65 m transect was established in the centre of the island. Each transect was equipped with 7 (55 m transects) or 8 (65 m transect) Hygrochron DS1923 microclimate sensors (Eclo Solutions) located at 2.5 m, 7.5 m, 12.5 m, 22.5 m, 32.5 m, 42.5 m, and 52.5 m (55 m transects), and 0 m, 10 m, 20 m, 25 m, 35 m, 40 m, 50 m, and 60 m (65 m transect) from the starting point of the transect. These 43 sensors were supplemented with 17 additional sensors placed close to resting L. colubrina. Of the 60 sensors, 20 were positioned in areas with kraits present and 40 in areas without. Ten of these sensors, five each in locations with and without snakes, malfunctioned or were lost.

Microclimate Sensors

Microclimate sensors were standardised at room temperature for 24 hours, then programmed to record temperature and humidity at 20-minute intervals for two weeks. Each sensor was positioned 30 mm (approximately body height of L. colubrina) above ground on a 400 mm wooden stake of 10 mm diameter. A white paper cup, perforated with holes to facilitate airflow, was attached above the sensor and used as a radiation shield.

Snake Surveys

A survey route (Fig. 2a), which included the six transects mentioned above, was established around the island to locate sea kraits. A total of 15 surveys were completed over a 9-day period, with surveys occurring every second day (on 21, 23, 25, 27 and 29 July 2019) to minimise disturbance to the animals. Three surveys were completed on each survey day. Surveys started at 9:00, 15:00 and 21:00 hours. The direction of travel for each survey was alternated to minimise pseudoreplication.

Environmental and Vegetation Data

Data collected for each transect point and for each location that a snake was found in included the location of each snake using a Garmin etrex 20x GPS (Garmin), the behaviour exhibited by the snake (either resting, sloughing, breeding and travelling), the percentage canopy cover in a 2-m radius above the snake (or transect point), the depth of the surrounding leaf litter using a ruler, and the plant species present in a 1-m radius.

Data Processing

Relative humidity collected by the sensors was converted to absolute humidity (g.m^-3) using: AH = (6.112 x e^[(17.67 x T) / (T + 243.5)] x RH x 2.1674) / (273.15 + T), where T is temperature in °C, RH is relative humidity in %, and e is the base of natural logarithms [raised to the power of the contents of the square brackets]. the microclimate data (temperature and humidity). Temperature and humidity data was split into diurnal and nocturnal observations using the following approach in line with the Australian Bureau of Metereology (BOM): Night temperatures/humidity for a day include temperatures/humidity from one hour after sunset to one hour before sunrise in the 24 hours to 9 am, and day temperature/humidity the temperatures/humidity from one hour after sunrise to one hour before sunset on that day. The data was then used to calculate the mean and variance in temperature and humidity for both diurnal and nocturnal observations.

All data are presented in a single spreadsheet with the following sheets:

• 'Environmental Data' includes the site location (longitude and latitude), percentage canopy cover, leaf litter depth, distance to the high water mark, and snake presence or absence for the 50 locations with retrieved microclimate data
• 'Microclimate Summary - Day' includes the means and variances for temperature and absolute humidity data collected during the day
• 'Microclimate Summary - Night' includes the means and variances for temperature and absolute humidity data collected during the night
• 'Microclimate Raw' includes all humidity and temperature readings collected from the 50 sensors
• 'Snake Survey' includes all the information collected during the snake surveys
• 'Vegetation Data' includes the cover-abundance data for plant species collected for each of the 37 transect points