Estimation of the Net Nesting Effort of Olive Ridley Arribada Sea Turtles Based on Nest Densities at Ostional Beach, Costa Rica

—Ostional Beach, Costa Rica, supports a large mass nesting (arribada) aggregation of Olive Ridley Sea Turtles (Lepidochelys olivacea). A large number of egg clutches is lost to egg harvest and to nest destruction by nesting females after every arribada. Accordingly, the purpose of our study was to generate estimates of the net number of clutches left incubating from clutch densities with the use of a quadrat methodology, and to compare these data with nesting population estimates resulting from the strip transect in time methodology that is currently applied. After the conclusion of each arribada, we randomly performed 50 excavations in 1-m quadrats to count the number of clutches present. We extrapolated quadrat density data to the entire nesting area of the beach to estimate the total number of clutches remaining following each arribada and egg harvest. The mean total clutch density was 4.09 6 0.18 SE nests/m. Our results showed that quadrat and transect estimate differences ranged from 0.04 to 52.6%, with quadrat estimates typically being lower. Our results demonstrated that, in the rainy season, a large number of clutches (47.4–99.9%) was left incubating in the beach after every arribada and egg harvest. By omitting clutches that were harvested or destroyed throughout the arribada, we evaluated the reproductive potential and estimated the magnitude of clutch loss on Ostional Beach, both of which play important roles in the management of the egg harvest as a sustainable conservation strategy. RESUMEN.—Playa Ostional, Costa Rica presenta el anidamiento masivo o arribada de la tortuga marina Lora (Lepidochelys olivacea). Un número elevado de huevos se pierde durante cada arribada debido a la cosecha de huevos y a la destrucción de nidos por parte de tortugas anidadoras. Ası́ pues, el objetivo de este estudio fue el de generar estimados del número neto de nidos que permanecen incubándose a partir de la densidad de nidos utilizando una metodologı́a de cuadrantes y la comparación de estos datos con los estimados poblacionales obtenidos con la metodologı́a de transectos fijos sobre el tiempo que se está utilizando actualmente. Después de la conclusión de cada arribada, realizamos al azar 50 excavaciones en cuadrantes de 1m para contar el número de nidos presentes. Nosotros extrapolamos los datos de densidad de estos cuadrantes a la playa de anidación entera para estimar el número de nidos que quedan en la playa luego de cada arribada y de cada cosecha de huevos. La densidad promedio total de nidos fue de 4.09 6 0.18 SE nidos por m. Nuestros resultados mostraron que las diferencias entre los estimados transectos y de cuadrantes variaron de entre 0.04% a 52.6%, siendo los estimados de cuadrantes tı́picamente menores. Nuestros resultados demostraron que en la época lluviosa un gran número de nidos (entre el 47,4% y el 99,9%) permanecieron incubándose en la playa luego de cada arribada y cosecha de huevos. Al omitir los nidos que son destruidos por las tortugas durante cada arribada o cosechados, este nuevo procedimiento nos permitió hacer una evaluación del potencial reproductivo y de la magnitud de la destrucción de nidos en playa Ostional, ambos siendo factores que juegan un papel importante en el manejo de la cosecha de huevos como una estrategia de conservación sustentable. Olive Ridley Sea Turtles (Lepidochelys olivacea; Eschscholtz, 1829) are listed as a vulnerable species on the IUCN Red List and are considered to be in overall global decline (AbreuGrobois and Plotkin, 2008). Population studies indicate this species to be most abundant in the Eastern Tropical Pacific; however, quantitative estimates of abundance are uncertain because of difficulties in estimating the number of females in mass nesting events (also known as arribadas) and large interannual fluctuations in nesting abundance that characterize this species (Valverde and Gates, 1999; Cornelius et al., 2007; Eguchi et al., 2007; Valverde et al., 2012). The Ostional National Wildlife Refuge (ONWR) was established in Costa Rica in 1983 to protect the Olive Ridley Sea Turtle rookery (Campbell, 1998), which supports a nesting female subpopulation, once estimated to contain ~134,400 females (Abreu-Grobois and Plotkin, 2008). Within ONWR, the Nosara and Ostional beaches extend 7 km with variable width. The majority of sea turtle nesting activity at ONWR is concentrated at Ostional Beach, which is one of the most important Olive Ridley nesting sites in the world (Cornelius, 1981). In the early 1980s, the arribadas at Ostional contained between 35,000 and 180,000 nesting females (Cornelius and Robinson, 1981, 1983). More recent estimates indicate that this subpopulation may reach up to 470,000 over a period of up to 7 d of oviposition (Valverde et al., 2012). The difference between these estimates supports the need to test and validate alternative standardized methodologies for estimating arribada abundance. The Ostional nesting aggregation also supports a large-scale community-based egg harvest program, in which approximately 20% of the eggs laid during each arribada are harvested by the local community for commercial use (Valverde et al., 2012). The legal harvest of sea turtle eggs by the local association takes place within the first 36 h of each arribada (Campbell, 1998). The socioeconomic success of the egg harvest program so far has made it a potential model for community-based conservation at other mass nesting beaches (Ballestero et al., 2000; Campbell et al., 2007). Despite the importance of Ostional Beach as an arribada rookery, there are limited published data for this nesting beach since its discovery in 1970 and the legalization of the egg harvest in 1987 (Richard and Hughes, 1972; Campbell, 1998). This information deficiency has been attributed, in part, to methodological difficulties associated with generating robust and reliable estimates of arribada abundance (Campbell, 1998; Corresponding Author. E-mail: [email protected] DOI: 10.1670/14-152 Ballestero et al., 2000). Monitoring various aspects of the nesting population at Ostional Beach (e.g., nesting female abundance, hatching success, and clutch densities) is critical to inform the effective management of the egg harvest program so it can continue to serve as a source of income for the local community (Campbell et al., 2007; Valverde et al., 2012). Long-term monitoring of various parameters with the use of standardized and statistically robust methods is critical for providing managers with information that can be used to evaluate the status of this nesting subpopulation in comparison to historical data. A logistically practical and statistically robust methodology was developed to estimate arribada abundance at beaches around the world with the use of a ‘‘strip transect in time’’ technique (Gates et al., 1996; Valverde and Gates, 1999). This transect methodology estimates the number of egg-laying females and, therefore, the number of nests deposited during each arribada; however, this approach provides no information on the number of nests left in the sand following intraspecific clutch destruction and the collection of clutches by the egg harvest program on Ostional Beach. Previous studies at Ostional and other arribada beaches in Costa Rica report low hatching success (~8%) and have identified clutch density, rainfall, and incubation temperature as important factors affecting hatching success at arribada beaches (Cornelius et al., 1991; Honarvar et al., 2008; Valverde et al., 2012). Additionally, a recent study found a positive correlation between cumulative clutch density and clutch destruction at an arribada beach and recommended the long-term monitoring of clutch density and nest destruction at arribada beaches (Ocana et al., 2012). Several different methodologies, including quadrat sampling and simulations, have been used to estimate clutch densities at arribada beaches (Cornelius et al., 1991; Honarvar et al., 2008; Ocana et al., 2012). Therefore, we implemented an additional methodology with the use of quadrat sampling to provide another estimate of the clutch density and reproductive potential. Simultaneously implementing both methodologies provides a different perspective through a comparison of the estimated reproductive output with the estimated abundance of nesting females resulting from the current transect methodology. Our data analyses allow us to generate an alternative estimate of the reproductive output, (while providing a measure of the magnitude of clutch destruction that occurs during each arribada), and provide a comprehensive assessment of the reproductive cost of this mass nesting behavior resulting from intraspecific clutch destruction, egg harvest, and predation at the nesting beach. Estimating these factors and their potential effects on the status of the nesting aggregation supplements the data needed for the continued monitoring of the nesting beach and the egg harvest program at Ostional. MATERIALS AND METHODS Study Site.—This study was conducted between September 2010 and April 2011 at Ostional Beach within the Ostional National Wildlife Refuge (ONWR), located on the Pacific coast of the Nicoya Peninsula in Costa Rica (9.9964718N; 85.6978008W). Nesting turtles typically are more abundant during the rainy season (May through November), and the beach structure is highly dynamic, given that nearby estuaries often overflow and cause substantial erosion. The beach is marked with sector markers every 50 m from north to south. The village of Ostional is located adjacent to the main nesting beach at Ostional Beach, where arribadas tend to concentrate (Valverde et al., 2012). Arribada Census.—Nesting turtle abundance in a given arribada was estimated with the use of strip transect in time methodology following the guidelines from Gates et al. (1996) and Valverde and Gates (1999). Arribada census data through December 2010 were published in Valverde et al. (2012); data for January through April 2011, however, are presented here for the first time. Briefly, transects were set up after the first night of each arribada by placing transect markers at the high-tide line in front of each sector marker. Markers were placed where the current arribada was concentrated and along either side of the beach until nesting tapered off. We censused transects every 2 h during the peak nesting activity (>100 nesting females) of each arribada by walking from the high-tide marker to the sector marker and counting the number of nesting females with eggs present in the nest chamber that fell within the 2-m width of the transects (Fig. 1A). Censuses were conducted each night until all transects produced no counts, at which point the arribada was considered to have ended. We analyzed transect results with the use of the FIG. 1. Diagram of the (A) transect and (B) quadrat methodology as implemented in this study. 410 V. S. BÉZY ET AL.