Nest Site Preference and Hatching Success of Green (Chelonia mydas) and Loggerhead (Caretta caretta) Sea Turtles at Akyatan Beach, Turkey
Abstract
We studied nesting site preferences and hatching success of loggerhead (Caretta caretta) and green turtles (Chelonia mydas) on Akyatan Beach, Turkey. The distribution of nests varied annually, with green turtles nesting most often in vegetated areas and loggerhead turtles most commonly in nonvegetated areas. Green turtle hatching success was higher at vegetated vs. nonvegetated nest sites (p < 0.001).
There are various hypotheses regarding nest site preferences of marine turtles. Some authors suggest that it is random (Mrosovsky 1983; Hays et al. 1995), whereas others have found that texture of beach sand (Mortimer 1990, 1995), beach slope (Wood and Bjorndal 2000), artificial lights (Witherington 1992), and the distance from the nearest human settlement (Kikukawa et al. 1999) and/or beach vegetation (Karavas et al. 2005) can influence nest site selection. In general, such preferences are adaptive in nature and are correlated with survival and reproductive success (Williams and Nichols 1984).
Data on density-dependent effects on sea turtle populations are limited. Some of these studies have addressed density-dependent nest destruction as it relates to emergence success by leatherback turtle hatchlings (Dermochelys coriacea; Caut et al. 2006), effect of density-dependent factors on hatchling production of green turtles (Tiwari et al. 2006), and density effects on hatching success of olive ridley turtles (Lepidochelys olivacea; Honarvar et al. 2008).
The collection of remotely sensed data that can be used in precise habitat mapping has become more readily available thanks to the development of geographic information system (GIS) technology and the increasing use of digital technology (Rushton et al. 2004). GIS in particular is of special interest to marine turtle researchers, as complex spatial data collection and analysis can be done quickly and accurately, and is relatively easy to reproduce. GIS analysis has been successfully used in several sea turtle studies (Maktav et al. 2000; Blamires et al. 2001; Fish et al. 2005; Karavas et al. 2005; Rusenko et al. 2005).
The objectives of this study were to produce a geospatial database that would include both nest data from field seasons and spatial land cover layers, demonstrate GIS's capabilities in analyzing spatial data, and provide a foundation for future research at this site. Furthermore, green turtles have been shown to prefer nesting in vegetated areas of beaches (Bustard and Greenham 1968; Bustard 1972; Brown and Macdonald 1995; Chen et al. 2007). Using GIS, we further evaluated these findings to determine if nest success was higher in the vegetated areas than in areas closer to the tide lines, which were devoid of vegetation. We also conducted an analysis to determine if females preferred certain areas of the beach over others and attempted to identify if geographic components of the beach that are not currently recognized influence nest site selection. Finally, we assessed density-dependent hatching success.
Study Site
Akyatan Beach is one of the most important green turtle nesting grounds in the Mediterranean with almost 50% of overall nesting occurring at this site (Kasparek et al. 2001). Loggerhead turtles also nest on this beach but in many fewer numbers. The beach is located on the eastern Mediterranean coast of Turkey and is almost 22 km in length (Fig. 1).
Citation: Chelonian Conservation and Biology 10, 2; 10.2744/CCB-0861.1
Materials and Methods
Fieldwork was conducted during the 2006–2009 nesting seasons at Akyatan Beach in southeast Turkey (Fig. 1). The beach was patrolled on foot every day. Nest locations were determined by using metal sticks. Data were collected for nest location using a Garmin eTrex Vista. Land cover classes (sand area and vegetated area) as well as tide lines were mapped for the areas of the beach that were monitored daily using a Dell Axim ×51 and a GlobalSat global positioning system (GPS) receiver during the nesting season of 2007. Beach mapping was conducted after sunrise, concurrently with beach surveys for new sea turtle nests and nonnesting emergences. Three zones were identified: wet sand (the portion of the beach seaward of the high tide line), dry sand (the portion of the beach landward of the high tide line without vegetation), and vegetated (the area of the beach where vegetation was present). The vegetated zone was characterized as containing live forbs, grasses, and/or woody shrubs.
The analysis was conducted using ESRI ArcInfo version 9.2. Polylines and points were imported. Some data cleanup was performed to remove extraneous GPS points, particularly those at the start and end of data collection days. WGS84 was used as the projection for all data collection. Based on the polylines, the beach was divided into vegetated and nonvegetated zones. In addition, the nesting areas monitored daily on Akyatan were divided longitudinally into 1-km segments to assess density-dependent hatching success. These sectors were numbered from 1 to 11 (Fig. 2) starting from the easternmost portion of the beach and moving toward the westernmost portion (e.g., Area 3 is more easterly than Area 5). Nests were counted and recorded for each sector of beach. Hatching success was determined each year by excavating nests after they hatched (usually 2–4 days after the first emergence was noted). The hatching percentage was calculated by dividing the number of hatched eggs by the total number of eggs in the nest. When eggshells were fragmented, pieces were grouped together to represent one egg. This conservative methodology may result in ± 4 eggs error.
Citation: Chelonian Conservation and Biology 10, 2; 10.2744/CCB-0861.1
The hatching success among the longitudinal segments was compared by one-way analysis of variance (ANOVA). Since loggerhead turtle nesting sites were scattered on the beach, hatching success could not be compared based on nest locations. Nesting data were analyzed separately for each year. Predated nests were included in overall nest numbers. Nest success was also calculated for nonpredated nests; however, results were presented separately for predated and nonpredated nests because this would affect our results. The following tests were used to determine if results were statistically significant: χ2, difference of means, and 1-way ANOVA.
Results and Discussion
During the study period, the number of loggerhead turtle nests ranged from 3 to 31 per year. The mean clutch size was from 64 to 71 eggs. The mean hatching success was 14% to 41%, a relatively low value, presumably due to nest depredation by jackals (55% to 66% of the eggs depredated). For green turtles, a total of 170 to 562 nests were deposited each year. Mean clutch size varied between 103 and 119. The mean hatching success was higher (58% to 67%) than for the loggerhead turtles. The green turtle nests were also depredated by jackals (14% to 25% of the eggs) but percentages were lower in comparison to loggerhead turtle nests. The higher depredation by jackals of loggerhead nests may be because 1) loggerhead nests are shallower and thus easier to find and 2) the vegetation in areas where green turtles prefer to nest may provide additional protection vs. nonvegetated nest sites.
During the study period, green turtles nested more often in vegetated areas with higher hatching success (p < 0.001) whereas most of the loggerhead turtle nests were located in nonvegetated areas (Table 1). This is consistent with previous studies that also have found that green turtles nest more frequently in vegetated areas of beaches (Bustard and Greenham 1968; Bustard 1972; Brown and Macdonald 1995; Chen et al. 2007). Our study supports this finding and also showed that hatching success was higher in the vegetated areas. Wang and Cheng (1999) found that digging success by green turtles was higher in the vegetated zone than on the open beach in Wan-An Island, Peng Hu archipelago. There are some explanations for the preference of vegetated areas in marine turtles. Mortimer (1995) stated that egg chambers of the nests on the open beach are more prone to collapse. Mrosovsky (1983) reported that there is a high probability of nests on the open beach being destroyed by wave wash. On Akyatan Beach, we thought that vegetation might also provide protection against predators. This possibility is currently being investigated.
Similar to our results, Hays and Speakman (1993) found that in the absence of disturbance, loggerhead turtles prefer to nest in nonvegetated portions of the beach. However, most studies on nest site selection by loggerhead turtles have shown that patterns of nest site selection vary across rookeries (reviewed in Miller et al. 2003). For example, Karavas et al. (2005) reported that the increase of fine sand caused a reduction of nesting density in loggerhead turtles.
In terms of nest distribution, the eastern side of Akyatan Beach contained more nests than the western side except in 2007 (Table 2); however, the preferred areas as defined by the 1-km areas (i.e., sections; Fig. 2) changed from year to year (Fig. 3). Hatching success for green turtles between 1-km polygons for 2006 through 2009 were not significantly different, except in 2008 when depredated nests were included (ANOVA F10,339 = 2.914, p < 0.01). When depredated nests were not included, the hatching success between the polygons was not significantly different, except for the 2006 nesting season (ANOVA F10,311 = 1.5451, p < 0.01). During 2006, areas 2, 9, and 11 were different from areas 6, 8, and 10, and area 3 was different from area 6 (Table 2).
Citation: Chelonian Conservation and Biology 10, 2; 10.2744/CCB-0861.1
The effect of nest density on hatching success of sea turtles has been debated for some time. Bustard and Tognetti (1969) reported that nest density influenced the hatching success of green turtle nests in Australia. Cornelius et al. (1991) stated that high nest densities of olive ridley turtles caused destruction of nests and reduced hatching success. Recently, Honarvar et al. (2008) studied the effects of dense nesting on hatching success of olive ridley turtles and concluded that beaches with the highest nesting density had the lowest hatching success.
In conclusion, this study confirmed that on Akyatan Beach, most of the nesting by green turtles occurs in vegetated areas that are farther from the water. This study also showed that hatching success of green turtles was higher in the vegetated areas and is a likely function of the nests being farther from the high tide line. Loggerhead turtles nest principally in nonvegetated areas. In general, hatching success does not differ among 1-km beach polygons for green turtles.
The location of Akyatan Beach, Turkey.
Orientation of the 11 1-km areas along the Akyatan Beach face.
Density of nests in each 1-km area along Akyatan Beach.
