We compared gut contents collected from green turtles (Chelonia mydas) foraging off the Miskito Cays, Nicaragua, in 1975 and 2020, and from females nesting at Tortuguero, Costa Rica, in 1976–1978 and 2021–2023. At the foraging ground, the seagrass Thalassia testudinum encompassed 93 % of gut contents across time. The consistency in green turtle diets over nearly five decades suggests the seagrass meadows in this area remain productive and structurally stable despite global declines, likely because this ecosystem coevolved with denser megaherbivore populations and now functions below carrying capacity. Moreover, reproductively active males at foraging grounds consumed more anthropogenic debris, possibly reflecting greater cumulative exposure to polluted neritic habitats, while females ingested more animal matter, potentially to meet the energetic demands of reproduction. Contrastingly, nesting green turtles indiscriminately ingested less and lower-quality food, suggesting minimal nutritional gain during the nesting season. Gut contents were dominated by riparian vegetation, and we observed a sharp increase in anthropogenic debris and algae in the recent sampling period. Increased debris ingestion reflects worsening riverine pollution from nearby urban areas, which is turning Tortuguero into an ecological trap: an ideal nesting habitat that naturally lacks green turtle’s main food sources and exposes nutritionally depleted individuals to high levels of pollution. The rising ingestion of algae, particularly Sargassum spp., may also reflect broader environmental changes linked to eutrophication and shifting oceanic currents. These findings underscore the need to protect resilient foraging habitats while mitigating anthropogenic threats at nesting sites through improved watershed management and pollution control

Sample collection

This study was conducted in Puerto Cabezas, northeast Nicaragua, and in Tortuguero National Park, northeast Costa Rica. The town of Puerto Cabezas serves as a major landing port for green turtles (Chelonia mydas) legally harvested for human consumption from the foraging ground near the Miskito Cays (Fig. 1). Located about 500 km south of Puerto Cabezas, Tortuguero is the primary nesting beach for green turtles in the Atlantic basin, and most green turtles nesting at Tortuguero migrate to the seagrass meadows near the Miskito Cays between reproductive seasons (Troëng et al. 2005).

The Nicaraguan green turtle fishery, which remains active, exacted a toll of over 170,000 individuals between 1991 and 2011 (Lagueux et al. 2014, 2017). In January and February 2020, one of us (RB) collected gut contents from large juvenile and adult green turtles, both male and female, that were legally harvested near the Miskito Cays.

For historical comparison, we used gut content data from 1975, collected by Mortimer (1976, 1981). Although Mortimer’s sampling encompassed gut contents from green turtles harvested at several locations along the Nicaraguan coast, we restricted our analysis to the 126 individuals reported as originating from foraging grounds near the Miskito Cays, to match the source location of the green turtles we sampled in 2020.

In Costa Rica, a legal green turtle harvest for human consumption was in place until the late 1990s, when a ban was passed (Sala Constitucional 1999; Troëng and Rankin 2005). One of us (ABM) collected gut contents from nesting green turtles that were legally harvested for human consumption by local community members of Tortuguero Village between 1976 and 1978. In recent years, the number of nesting green turtles preyed upon by jaguars (Panthera onca) at Tortuguero has increased (Troëng 2000, Arroyo-Arce & Salom-Perez 2015). Assisted by local community members and national park rangers, one of us (RB) collected gut contents of nesting green turtles depredated by jaguars between 2021 and 2023.

At both the foraging ground and the nesting beach, we collected contents from the esophagus and/or stomach, henceforth referred to as “gut contents”. While we collected all the material present in stomach and esophagus of nesting females, at the foraging ground in 2020 we subsampled gut contents. Food boli found in the esophagus and/or stomach of these individuals were homogenized in a plastic container, and material was then collected using tweezers in random grab samples—without targeting specific item types—until a 50-mL container was filled. Given the observed homogeneity of samples from the foraging ground, we are confident that these subsamples accurately represent the gut contents of green turtles from that location.

Gut contents collected at the foraging ground in 2020 were stored in 95% ethanol and gut contents collected at the nesting beach in Tortuguero were stored in 5% formalin in sea water in the first sampling period (1976-78) and 10 % formal saline solution in the second sampling period (2021-23). These methods successfully preserved the integrity of gut contents, which were analyzed in the laboratory of the Archie Carr Center for Sea Turtle Research at the University of Florida.

Gut content composition

We classified gut content components into five main categories: seagrasses, algae, animal matter, anthropogenic debris, and riparian vegetation. Seagrasses were identified to species level, whereas algae were identified to phylum level as green (Chlorophyta), red (Rhodophyta), or brown (Phaeophyta). Animal matter, riparian vegetation, and anthropogenic debris were not further subdivided for analyses.

We determined the volume of each category and the total volume of each sample by measuring water displacement in graduated cylinders to the nearest 0.2 mL. To calculate the percent relative volume of each category (Vol _i), we divided the volume of that category (V _i) by the total volume of the corresponding sample (V _total):

  Vol_i = (V_i / V_total) * 100

Additionally, we calculated the frequency of occurrence (F _i) for each category, defined as the percent of samples in which the category was present (N _i) out of the total number of samples analyzed (N _total):

 F_i = (N_i / N_total) * 100

To provide a comprehensive measure that integrates both volume and frequency of occurrence, we calculated the Index of Relative Importance (IRI _i) according to the formula modified from Horn (1966) and used for the analyses of green turtle diets by Gulick et al. (2021):

  IRI_i = 100 * (F_i * Vol_i) / sum(F_i * Vol_i)

To assess historical changes in green turtle gut contents at the foraging grounds, we relied on data collected from green turtles at foraging grounds off the Miskito Cays in 1975 (Mortimer 1976, 1981). In Mortimer’s study, the frequency of occurrence and relative dry weight were provided for each diet category, allowing us to calculate the IRI for each category in that period. While our study measured volume instead of dry weight, the comparison remains valid because both metrics are expressed as percentages of the total sample and the datasets were collected from turtles of similar demographics at the same foraging ground.

Finally, we calculated the extent of gut content overlap (Ĉλ) as per Bjorndal et al. (1997), where a value of 0 represents no overlap and a value of 1 indicates complete overlap between gut contents. In this formula, n is the number of gut content categories, x_i is the relative importance (IRI) of the category i for turtle grouping x, and y_i is the same for turtle grouping y.

  Ĉλ = 2 * sum(x_i * y_i) / (sum(x_i²) + sum(y_i²))

We compared both the Index of Relative Importance (IRI) and Ĉλ to assess variation in green turtle gut content composition. Comparisons were made across sex, maturity and reproductive status, habitat type (nesting vs. foraging grounds), and sampling period (1970s vs. 2020s). Because the data were neither normally distributed nor homoscedastic, as determined by the Shapiro-Wilk and Levene’s tests, we used Mann-Whitney U tests to evaluate whether relative gut content volumes differed among groups. All statistical analyses and figure generation were conducted in R version 4.4.0 (R Core Team 2024).

Reproductive organ analyses

In this study, maturity status refers to whether a green turtle was biologically capable of reproducing (adult vs. juvenile). Reproductive status distinguishes reproductively quiescent individuals from those that were preparing to reproduce or had recently done so (reproductively inactive vs active).

Female green turtles sampled at the foraging ground in 2020 were considered adults if oviduct width > 15 mm, ovarian stroma width > 143 mm or if their ovary presented ovulatory scars formed during ovulation (Miller and Limpus 2003, Bruno et al. 2025, Bruno et al. in review). Moreover, female green turtles at the foraging ground were considered reproductively active if they had bred in the past 12 months (indicated by the presence of ovulatory scars > 4 mm), or if they were preparing to breed in the upcoming nesting season (indicated by the presence of vitellogenic follicles > 7 mm) (Limpus et al. 2005; Pilcher et al. 2021; Bruno et al. in review). All individuals studied at Tortuguero were adult females captured on the nesting beach.

At the time male green turtles from the foraging ground were butchered in 2020, we conducted a visual analysis of the reproductive tract while it was still attached to the body. Males were considered adults if testes were cylindrical and epididymides were bulging from the body wall (Miller and Limpus 2003). Reproductive activity was confirmed histologically: males were considered reproductively active if spermatozoa were present within the lumen of the seminiferous tubules (Miller and Limpus 2003; Meylan et al. 2002), as determined via microscopic analysis following hematoxylin and eosin staining at the Universidad Nacional Autónoma de Nicaragua in León.