Sand temperatures at volcanic sand sea turtle nesting beaches

Description

Dataset of daily average sand temperatures, air temperatures, relative humidity and precipitation recorded at two volcanic sand beaches at the Pacific coast of Guatemala, Central America. Recorded from May 2018 to November 2019 using LogTag Trix-8 temperature loggers buried at 30 and 50 cm depth along the shoreline. R script for LMM analyses of dataset.

Steps to reproduce

Sand temperature data were collected from 26 May 2018 to 3 November 2019, covering two nesting seasons and one intermediate period, on the sea turtle nesting beaches of Hawaii and El Banco, Guatemala. LogTag Trix-8 temperature loggers were used for temperature recording. The loggers were programmed to record temperature simultaneously every hour. From May 2018 to November 2019, they were placed along two transects (one at Hawaii, one at El Banco) parallel to the shoreline, and on average one meter away from the high tide line, in the zone were, according to previous observations, most sea turtles place their nests. Later, the type of surrounding was defined as follows: “sand surrounding” was a plot with no vegetation cover and not close to concrete structures, “vegetation surrounding” was a plot with vegetation cover, and “concrete surrounding” was a plot with concrete structures in its proximities (< 2 m distance) either covered with vegetation or not. Each transect stretched over 1 km and consisted of ten plots 100 m apart from each other. At each plot, temperatures were measured in the sand at 30 and 50 cm below the surface. These depths were corresponding to the natural depths of Olive Ridley nests, assuring that both loggers were placed at depths where egg incubation under natural conditions would occur. Weather data of air temperature, precipitation, and relative humidity were collected from Instituto Privado de Investigación sobre Cambio Climático. The nearest weather stations to both sites were used; La Candelaria for the El Banco transect, and La Máquina for the Hawaii transect. We estimated the daily average of all sand temperature records. Then, a list of eight candidate models was created to assess the effects of local weather (air temperature, precipitation, and relative humidity), location (beach, depth, and type of surrounding), and annual cycle (year and season) variables on sand temperatures. The effects of air temperature (daily average), precipitation (daily average), relative humidity (daily average), beach (Hawaii and El Banco), depth (30 cm and 50 cm), type of surrounding (sand, vegetation, and concrete), year (2018 and 2019), and season (rainy and dry) on the sand temperatures were analyzed with linear mixed models (LMM). Julian day and the individual datalogger identity were considered as random effects with temporal autocorrelation to correct for the diurnal cycle of the weather variables and the temporal autocorrelation of these variables with Julian day [30]. Beach and year were treated as fixed effects because they consist of only two levels each, making it impossible to use them as random effects. Model selection was based on Akaike’s Information Criterion corrected for small sample size (AICc). There was no need for model averaging as ΔAICc was > 2 in all the other models compared to the most parsimonious one. Parameters that included zero within their 95% confidence interval (CI) were considered uninformative.

Institutions

Categories

Funding

Licence