Thirty-Eight Years of Loggerhead Turtle Nesting in Laganas Bay, Zakynthos, Greece: A Review
Abstract
Here we present 38 yrs of nesting data of the loggerhead population in Laganas Bay, Zakynthos, Greece, adding data for 2010–2021 to existing published data for 1984–2009. This rookery, encompassing 6 discrete beaches of 5.5 km total length, features the highest nesting density in the Mediterranean. The annual nest numbers, ranging from 667 to 2018, showed no significant long-term trend. Beach nesting contributions increased on some public-accessed beaches (as did nesting success), but decreased on a remote beach that previously held > 50% of all nests. These changes might be attributed to management actions by the National Marine Park of Zakynthos that improved conditions for sea turtles (e.g., restricting nighttime access to the public), indicating that the previously recorded high nesting density on the remote beach may have been atypical. Although no significant phenological changes were detected with the available data in the 20-yr period 2002–2021, the date of the first hatched nest shifted significantly toward earlier dates in the 38-yr data set. Clutch size and body size of nesting turtles exhibited significantly decreasing trends. Hatching success, hatchling emergence success, and in-nest hatchling mortality showed significantly increasing trends. The number of viable hatchlings decreased over the seasons, albeit not significantly, possibly a result of the decreasing clutch size. Incubation durations were revealed as shortening on all beaches, a potential sign of global warming, with a consequent suggested increase of female primary sex ratio. Continuation of this long-term monitoring program is expected to provide further insights in the reproductive traits of this regionally important loggerhead population.
The loggerhead turtle Caretta caretta is a circumglobally distributed species breeding in subtropical and temperate areas. To determine population segments for the International Union of Conservation of Nature (IUCN)'s Red List assessments, Wallace et al. (2010), based on biogeographical data, identified for this species 10 regional management units (RMUs) worldwide, with one RMU containing populations breeding in the Mediterranean. Loggerheads in the Mediterranean reproduce mainly in the eastern basin, with major nesting aggregations in Greece, Turkey, Libya, and Cyprus (Margaritoulis et al. 2003) concentrating about 96% of total nests (Casale et al. 2018). Nesting areas in Greece host about 46% of all documented loggerhead nests in the Mediterranean, with Laganas Bay on Zakynthos Island holding 18% of these nests (Casale et al. 2018). Nesting in Laganas Bay has been monitored by ARCHELON since 1984, making it the oldest sea turtle program in undertaking systematic nest counts in the Mediterranean (Casale et al. 2018). Preliminary nest counts in late 1970s confirmed Laganas Bay as an important nesting area (Margaritoulis 1982). Relevant campaigns and pressures from national and international bodies resulted in the establishment of the first protective legislative acts (Warren and Antonopoulou 1990), which in 1999 led to the foundation of a marine park (National Marine Park of Zakynthos [NMPZ]), and the associated management agency in 2000 (Dimopoulos 2001). Nevertheless, it took some time for the management agency to start functioning properly due to unwillingness of local stakeholders, and lack of adequate funding in some years hindered progress (Rees 2005; Togridou et al. 2006).
Reproductive data from Laganas Bay have been published up to and including the 2009 nesting season (Margaritoulis 2005; Margaritoulis et al. 2011). Although ARCHELON annually provides all data to the NMPZ for its management needs, there is a growing necessity to publish up-to-date nesting data, preferably at regular intervals (e.g., 5–10 yrs). This would keep the broad scientific community informed and promote conservation status evaluations (e.g., Mazaris et al. 2017; Casale et al. 2018; Laloë et al. 2019), including the IUCN's Red List (Casale et al. 2018) and the European Union's (EU) Marine Strategy Framework Directive (2008/56/EC) where the loggerhead turtle is an “indicator” species for biological diversity.
Here we combine data of the recent period, 2010–2021, with published data of previous periods (1984–2009; Margaritoulis 2005; Margaritoulis et al. 2011) to examine long-term trends of reproductive traits that may assist in the conservation and improved management of this nesting population.
METHODS
Study Site. — Laganas Bay on Zakynthos Island (Ionian Sea) has a southeastern orientation with a coastline of about 20 km, and an opening of about 12 km (approximate middle of the bay: lat 37°42′N, long 20°25′E). The climate is highly seasonal, with high temperatures and low precipitation during the summer months, while autumn and winter are characterized by increased precipitation, occasional storms, and lower temperatures (for detailed climatic data see Margaritoulis 2005). The nesting habitat (total beach length: 5.5 km) consists of 6 discrete beaches (i.e., Marathonissi, East Laganas, Kalamaki, Sekania, Daphni, Gerakas) fringing the bay (Fig. 1). The beaches differ remarkably in physical characteristics, development, and human access. East Laganas, Kalamaki, and Gerakas have easy public access; Marathonissi Beach, on a small island inside the bay, is visited during daytime by many boats and people; Daphni, once a remote beach, has undergone illegal development in the form of roads, houses, and tavernas; Sekania Beach is remote, with its hinterland acquired by World Wildlife Fund Greece (Casale et al. 2018), and since the formation of the NMPZ, is an area of “absolute protection” (Dimopoulos 2001). The 6 beaches constitute a single nesting habitat with a considerable interchange of nesting turtles within and among seasons (Katselidis et al. 2005). After the creation of the NMPZ, for administrative reasons, the name of East Laganas Beach changed to Kalamaki and that of Kalamaki to Crystal Beach. For consistency with previous publications, we keep the original names.
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Field Methods. — Fieldwork was carried out by trained volunteers following standard protocols and was supervised by experienced field assistants. Observations and measurements were recorded in field notebooks and later transcribed onto standard datasheets. Subsequently, all data were entered in specifically designed electronic databases. Continuity of methods was maintained through field assistants' seminars at the start of the nesting season, and their presence in the field to provide training and guidance to volunteers. A major progress, since 2006, was the use of global positioning system (GPS) units (Garmin eTrex, Taiwan, accuracy < 3 m, 95% typical) to record location of emergences, laid nests, and hatched nests. To avoid problems with GPS accuracy, we continued the traditional method of triangulation through fixed beach markers. Another improvement was the systematic location of the nests' egg chambers through hand excavation until appearance of top eggs on all beaches except Sekania, the beach containing in general more than 50% of all nests (Margaritoulis 2005), where such digging could disrupt incubating nests. We estimated nest numbers on Sekania by the count of nesting emergences and confirmed later by the number of hatched nests. On “high-risk” sectors, wooden frames, provided by the NMPZ's management agency, were placed over nests to avoid trampling. Egg predation by mammals was negligible and hence antipredator measures were not generally used. However, a small percentage of nests (annual average, 1.4%; range, 0.3%–3.6%) were relocated within the same beach, to avoid possible inundation.
We provide below definitions of the examined reproductive parameters. Nesting success: percentage of emergences that resulted in egg-laying; clutch size: number of yolked eggs in a clutch; incubation duration: in days, from egg-laying until emergence of first hatchling; hatching success: percentage of eggs that produced a hatchling; hatchling emergence success: percentage of eggs that produced a hatchling that emerged from the nest; in-nest hatchling mortality: percentage of eggs that produced hatchlings unable to exit the nest; hatched nest: a nest that produced at least 1 hatchling on beach surface; beach nesting contribution: percentage of nests each beach contributed to the annual total. Date of first and last nest were defined, excluding extraneous early or late nests, as those of continuous nesting, and used to calculate the median date of nesting and the duration of the nesting season. We examined the trends in nesting phenology (i.e., dates of first nest, last nest, and first hatch as well as the median date) for the period 2002–2021, for which we had accurate data. Additionally, we examined the trend of the dates of first hatch, for the 38-yr period (1984–2021), as we had recorded precisely this metric in all seasons. Phenology analysis was carried out using ordinal dates.
We calculated clutch size, hatching success, hatchling emergence success and in-nest hatchling mortality from the contents of nonrelocated and nondepredated nests that were excavated posthatch. We generally opened and examined all hatched nests, ready for excavation, on all beaches until termination of our annual fieldwork. We excluded clutches with more than 199 eggs, assuming that these occurred from merging of 2 adjacent nests, not uncommon at densely nested beach sections.
We calculated incubation durations per beach for the period 2002–2021, in which period we had adequate sample sizes for all beaches. Despite recently recorded rare occurrences of incubation durations of less than 42 d, we excluded these from our results for consistency with previous publications, where such short durations were considered as errors.
For the period 1988–2021, we estimated the annual number of emerged hatchlings by using 1) the number of hatched nests, 2) the mean clutch size, and 3) the mean hatchling emergence success.
Straight carapace length, notch to tip (SCLn-t senso Bolten 1999), of nesting turtles has been recorded in Laganas Bay since 1982 (Margaritoulis et al. 2020). To determine any trend in adult size, and avoid pseudoreplication, we used measurements only for “neophyte” turtles, i.e., those without tags or tagging scars and presumably considered as first nesters. To avoid inclusion of “remigrant” turtles (those that had nested in previous seasons), which could not be differentiated in the first years of tagging, we initiated body size analysis from year 1987 when the annual percentage of neophytes was stabilized around 30% (range, 12.0%–48.4%). No body size measurements were obtained in 1999.
Statistical Analysis. — We used the Mann-Kendall trend test for investigating temporal trends per beach of the following variables: number of nests, nesting success, beach nesting contribution, and nest incubation duration, as these variables present time-series data and were not normally distributed. Additionally, the Mann-Kendall trend test was used to assess the presence of trends in nesting phenology variables, number of hatched nests, percentage of hatched nests to laid nests, clutch size, hatching success, hatchling emergence success, and in-nest hatchling mortality, as well as trends in straight carapace length of neophyte turtles. The multivariate (multisite) Mann-Kendall test was used to investigate the presence of trends in number of emergences and nests at the rookery level. The associated plots were fitted with trend lines using the Theil-Sen estimator method. Where significant trends were found, these were further investigated for single change-point detection using Pettitt's test. We used linear models in examining the relationships 1) between straight carapace length and clutch size, and 2) between the number of hatchlings and the number of laid nests, number of hatched nests, clutch size, hatching success, hatchling emergence success, and in-nest hatchling mortality since data were both linear and normally distributed (Shapiro-Wilk normality test, p > 0.05). We investigated the differences in incubation durations between beaches using the nonparametric Kruskal-Wallis test, as data did not display homogeneity of variance (Levene's test of homogeneity, p < 0.05). Further, we carried out a Wilcoxon rank sum to investigate pairwise comparisons between beaches. All analyses and plots were produced in RStudio v1.4.1106 (R Core Team 2020).
RESULTS
Overall Nesting Activity and Trends. — Nesting data for the period 2010–2021 appear in Supplemental Table S1 (all supplemental material is available at http://dx.doi.org/10.2744/CCB-1531.1.s1). For the 38-yr data set, the annual mean of 1) emergences was 4617.8 (range, 2189–8128; SD, 1303.8), 2) nests was 1209.1 (range, 667–2018; SD, 314.4), 3) nesting success was 26.2% (range, 19.2%–35.9%; SD, 3.6%), 4) nesting density was 219.8 nests/km (range, 121.3–366.9; SD, 57.2) (Table 1). High interannual variabilities were exhibited in the numbers of emergences (coefficient of variation [CV] = 0.28) and of nests (CV = 0.26) (Table 1). We found no significant temporal trends in the annual number of either emergences (z = –0.562, p = 0.574) or nests (z = –0.224, p = 0.823); however both emergences and nests displayed an initial period with a slight increase, followed by a period of decline and a final (current) period of increase (Fig. 2).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Spatial Distribution of Nesting and Trends per Beach. — Nesting was not evenly distributed among the 6 beaches with highest nesting density documented on Sekania (mean, 930.6 nests/km; range, 424.6–1609.2; SD, 294.9) and lowest on East Laganas (mean, 61.8 nests/km; range, 32.4–114.0; SD, 22.2) (Table 1). The annual number of emergences and nests varied considerably among the beaches, and this affected the contribution of each beach to total nesting. The highest nest numbers (and nesting contribution) were recorded on Sekania (mean, 604.9 nests/yr; range, 276–1046; mean nesting contribution, 50.0%), and the lowest on Marathonissi (mean, 95.3 nests/yr; range, 28–228; mean nesting contribution, 7.9%) (Table 1). Nesting success also varied considerably among beaches with highest annual values on Marathonissi (mean, 32.2%; range, 18.8%–47.5%) and lowest in Daphni (mean, 13.7%; range, 6.0%–23.3%) (Table 1).
Trends in the number of nests through time varied per beach with stable and increasing or decreasing periods (Fig. 3); however, overall there were significant trends in the number of nests on Marathonissi (z = –3.245, τ = –0.370, p = 0.001), Daphni (z = –2.945, τ = –0.337, p = 0.003), Sekania (z = –2.401, τ = –0.273, p = 0.016), East Laganas (z = 3.646, τ = 0.415, p < 0.001), and Gerakas (z = 2.894, τ = 0.330, p = 0.004), while the number of nests on Kalamaki did not show a significant trend (p = 0.213) (Fig. 3). Results of Pettitt's test detected significant single change-points (p < 0.05) for all beaches, except Kalamaki where a significant trend was not detected. Declines were observed for Daphni after the year 2000 and for Marathonissi and Sekania after 2004, whereas significant single change-points suggested increases in nest numbers after 2008 for both East Laganas and Gerakas (Fig. 3).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Nesting success over time showed 1) significant trends with mostly increasing trajectories on East Laganas (z = 3.671, τ = 0.417, p < 0.001) and Gerakas (z = 2.389, τ = 0.272, p = 0.017), 2) a significant decline on Daphni (z = –4.627, τ = –0.525, p < 0.001), and 3) no trend on Marathonissi, Kalamaki, and Sekania (p > 0.05) (see Supplemental Fig. S1). The significant trends on nesting success for East Laganas, Daphni, and Gerakas were further investigated for a single change-point detection using Pettitt's test, which revealed a significant change for Daphni after 2003 and for East Laganas after 2006; however, no single change-point was detected for Gerakas (Fig. S1).
The nesting contributions of each beach have shown significant trends across all beaches; i.e., Marathonissi (z = –3.319, τ = –0.377, p < 0.001), East Laganas (z = 4.199, τ = 0.477, p < 0.001), Kalamaki (z = 2.690, τ = 0.306, p = 0.007), Sekania (z = –3.294, τ = –0.374, p < 0.001), Daphni (z = –3.822, τ = –0.434, p < 0.001), and Gerakas (z = 3.269, τ = 0.371, p = 0.001). Pettitt's test for single change-point detection revealed significant change points (p < 0.05) for all beaches, where declines started after 1999, 2005, and 2008 in Daphni, Sekania, and Marathonissi, respectively, and increases started after 2004, 2006, and 2008 in Kalamaki, East Laganas, and Gerakas, respectively (Fig. 4).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Nesting Phenology. — In the 20-yr period 2002–2021, the dates of first nest, last nest, and first hatch as well as the median date of nesting and the duration of the nesting season did not show any significant trend (Mann-Kendall trend test, p > 0.05), while a significant relationship (r2 = 0.343, t = –4.217, p < 0.001) was found between the dates of first nest and first hatch. Nevertheless, over the 38-yr period, Mann-Kendall trend test results revealed a significant trend with date of first hatch shifting to earlier in the year (z = –3.445, τ = –0.402, p < 0.001) (Fig. 5). A single change-point was detected in the trend at the year 1998 (Pettitt's test, p = 0.002).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Hatched Nests. — The annual number of hatched nests averaged 1011.1 (range, 611–1591; SD, 229.9) without showing any temporal trend (Mann-Kendall trend test, p > 0.05). However, the annual percentage of hatched nests to laid nests (mean, 83.7%; range, 66.0%–96.3%; SD, 7.8%) exhibited a significant upward trend (z = 2.194, τ = 0.266, p = 0.028).
Clutch Size, Hatching Success, Emergence Success, and In-Nest Hatchling Mortality. — Relevant data for years 2010–2021 appear in Supplemental Table S2. Over the 38-yr period, examination of 18,463 nests provided annual mean values for 1) clutch size, 109.5 eggs (range, 92.7–130.4; SD, 9.3); 2) hatching success, 73.3% (range, 61.7%–85.2%; SD, 4.7%); 3) hatchling emergence success, 68.0% (range, 58.9%–78.9%; SD, 4.5%); and 4) in-nest hatchling mortality, 5.3% (range, 0.8%–10.4%; SD, 2.4%).
Over time, clutch size showed a significant downward trend (z = –6.451, τ = -0.733, p < 0.001), while hatching success, hatchling emergence success and in-nest hatchling mortality all displayed significant upward trends (respectively: z = 3.409, τ = 0.388, p < 0.001; z = 2.101, τ = 0.240, p = 0.036; and z = 3.672, τ = 0.419, p < 0.001) (Fig. 6).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Number of Emerged Hatchlings (Recruitment). — We estimated that over the last 34 yrs (1988–2021), about 2,531,328 hatchlings emerged from 34,379 nests (73.6 hatchlings per nest) with annual numbers ranging from 43,799 to 112,710. Over time, the annual number of hatchlings showed a nonsignificant downward trend (z = –1.305, τ = –0.159, p = 0.192). The number of laid nests had a significant positive affect on the number of hatchlings (r2 = 0.780, t = 10.656, p < 0.001); however, the relationship between the number of hatched nests and the number of hatchlings was stronger (r2 = 0.847, t = 13.325, p < 0.001). We found no significant relationships between the annual number of hatchlings and 1) clutch size (r2 = 0.069, t = 1.546, p = 0.132), 2) hatching success (r2 = 0.002, t = –0.228, p = 0.821), 3) hatchling emergence success (r2 = 0.001, t = 0.172, p = 0.865), and 4) the in-nest hatchling mortality (r2 = 0.019, t = –0.793, p = 0.434).
Incubation Durations per Beach. — Incubation durations, calculated for 14,062 nests, provided highest values on Marathonissi (overall mean, 62.2 d; range of annual means, 56.0–72.2; SD, 4.1) and East Laganas (mean, 55.9 d; range, 51.3–63.8; SD, 3.2), and lowest in Kalamaki (mean, 48.8 d; range, 46.6–52.6; SD, 1.7) (Table 2). Incubation durations were significantly different between beaches (Kruskal-Wallis χ25 = 89.652, p < 0.001). Wilcoxon Rank-Sum pair-wise test results showed that all beaches have significantly different incubation durations between them, with the exception of Kalamaki and Daphni, which indicated a similar mean incubation duration (p = 0.904) (Table 2).
The Mann-Kendall trend tests showed that over time, incubation durations displayed significant shortening trends on Marathonissi (z = –3.536, τ = -0.579, p < 0.001) and on East Laganas (z = –2.628, τ = –0.432, p = 0.009) (Fig. 7). Incubation durations on all other beaches did not display any significant long-term trend (p > 0.05); however, all seemed to have slightly shortened until around 2011 and stabilized thereafter (Fig. 7).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
Body Size of Nesting Females. — Annual body size, in terms of SCLn-t, measured on 2120 different “neophyte” turtles, averaged 76.2 cm (range of annual means, 73.6–79.5 cm; range of individual values, 60.5–95.5 cm). Over time, the annual SCLn-t exhibited a significant decreasing trend (z = –6.204, τ = –0.745, p < 0.001), with a major decrease starting at around 2003 (Fig. 8). The linear model between body size (SCLn-t) and clutch size at population level, has shown that larger individuals produce significantly larger clutches (r2 = 0.835, t = 12.730, p < 0.001).
Citation: Chelonian Conservation and Biology: Celebrating 25 Years as the World's Turtle and Tortoise Journal 21, 2; 10.2744/CCB-1531.1
DISCUSSION
Overall Nesting Activity and Trends. — The loggerhead rookery in Laganas Bay is characterized by a high nesting density, considered to be the highest in the Mediterranean (see also Casale et al. 2018). The noted increasing trend after about 2012 complies possibly with a general increase of nesting in the Mediterranean (Mazaris et al. 2017; Casale et al. 2018). The intense interannual fluctuations in the number of emergences and nests, common in Mediterranean rookeries (Margaritoulis and Rees 2001; Ilgaz et al. 2007; Türkozan and Yilmaz 2008; Sönmez et al. 2021), are caused by the turtles' complex reproductive traits as well as by food availability and environmental factors at the foraging areas (Broderick et al. 2001; Solow et al. 2002; Mazaris et al. 2009; Weishampel et al. 2010; Pike 2013).
However, other factors may be impeding population increase as the main foraging areas of Zakynthos females, in the Adriatic Sea and Gulf of Gabès (Margaritoulis et al. 2003; Zbinden et al. 2008; Schofield et al. 2013), are subject to intense fishing effort, mainly by trawlers impacting large-sized turtles (Casale 2011). In recent years, turtle bycatch has increased in the Adriatic (Lucchetti et al. 2017) and eutrophication in the Gulf of Gabès may have changed trophic conditions for resident turtles (Patel et al. 2015), hence constraining nesting numbers (Broderick et al. 2001; Mazaris et al. 2009; Patel et al. 2016). In addition, 248 dead adult females were recorded in Zakynthos in the period 2005–2019 (ARCHELON, unpubl. data, 2020), which is an alarming mortality rate since observed strandings may represent a fraction of actual mortalities (Epperly et al. 1996). Further, the average nesting population in Laganas Bay is estimated at 318 females/season (range, 176–531), considering a clutch frequency of 3.8 nests/female reported recently at the nearby rookery of Kyparissia Bay (Rees et al. 2020).
Nesting success at Zakynthos is low (26.2%; 29.4% after excluding Daphni, the beach with lowest nesting success), compared to other rookeries in Greece that may feature nesting successes of 35% or more (Margaritoulis and Rees 2001, 2003, 2006; Rees et al. 2002). Zakynthos turtles require on the average 4 emergences to deposit 1 clutch, while Kyparissia turtles require 3 (Margaritoulis and Rees 2001). The impact of increased energy expenditure from repeated nesting attempts on the anticipated reproductive output is not known and should be investigated.
Changes in Spatial Distribution of Nesting. — The significantly increasing nesting contribution of the public-accessed beaches of East Laganas, Kalamaki, and Gerakas are possibly attributed to the better enforcement of regulations by the NMPZ. The NMPZ's management agency gradually developed a successful scheme of wardening and controlled human beach use. As a result, traditional disturbances on those beaches (e.g., vehicular traffic, human presence at night, bright lights) (Arianoutsou 1988; Rees 2005) were progressively removed or reduced.
The increased nesting on those beaches caused a corresponding decrease of nesting at the remote Sekania Beach, alleviating the atypically high nesting density recorded there (annual max: 1610 nests/km).
The significant downward trend of nesting in Daphni, together with a significant decrease of nesting success, is probably a result of growing human disturbances notwithstanding the dynamic nature of this beach characterized by large percentage of pebbles deposited and eroded alternatively (Arianoutsou 1988; Margaritoulis 2005). Locals at Daphni are generally unwilling to comply with the NMPZ's guidelines and regulations (Katselidis and Dimopoulos 2000; Schofield et al. 2005; Togridou et al. 2006).
The abrupt reduction of nesting on Marathonissi was probably caused by the dramatic increase of commercial turtle-watching (and privately rented) boats operating around this small island, with many landing on the nesting beach (ARCHELON 2015). Another cause of nest reduction may be the predominance of male hatchlings produced from this beach (Zbinden et al. 2007; Katselidis et al. 2012), with low numbers of females that would return to nest on their natal beach (Heppell et al. 2003 and references therein). This deserves, however, further investigation.
Nesting Phenology. — Although we did not detect a clear phenological shift in the 20-yr period (2012–2021), the significantly decreasing trend of the date of first hatch over a longer timeframe (38 yrs; 1984–2021) provides an indication of earlier onset of nesting, although such evidence might be confounded with shortening incubation durations and nesting distribution shifting to beaches with different thermal properties. Phenological shifts were recorded in other loggerhead populations and were attributed to global warming (Weishampel et al. 2004; Pike et al. 2006; Hawkes et al. 2007; Lamont and Fujisaki 2014; Patel et al. 2016).
We also did not detect any temporal trend of the duration of the nesting season. The effect of earlier nesting on the length of nesting season has led to contradicting opinions elsewhere; 2 studies in Atlantic Florida (Pike et al. 2006; Weishampel et al. 2010) postulated that earlier nesting shortens the nesting season and 2 other studies, in North Carolina (Hawkes et al. 2007) and Gulf of Mexico (Lamont and Fujisaki 2014), suggested the opposite. Apparently, there are significant methodological differences in defining the onset of nesting, and consequently the length of the nesting season, which should be appropriately explored and standardized.
We do not know whether a possible phenological shift in Zakynthos will be sufficient to mitigate the effects of global warming on turtles. Another adaptive strategy might be that turtles will change their nesting habitats for cooler ones. In the Mediterranean, there are extensive beaches at northern latitudes; although not currently used regularly by turtles, they host a growing number of successful nestings (e.g., Bentivegna et al. 2010; Maffucci et al. 2016; Casale et al. 2018; Piroli and Haxhiu 2020). Nevertheless, such shifts in time and/or space may not be successful in the long term, mainly due to sea level rise or to other perplexities (e.g., Weishampel et al. 2004; Hawkes et al. 2009; Poloczanska et al. 2009; Maffucci et al. 2016; Patel et al. 2016; Almpanidou et al. 2018; Dimitriadis et al. 2022). This is of special concern at Zakynthos where there is generally limited resilience to changing conditions (Mazaris et al. 2009; Katselidis et al. 2013; Dimitriadis et al. 2022).
Increased Proportion of Hatched Nests to Laid Nests. — The percentage of nests that produce hatchlings in Zakynthos is generally high in comparison to other Mediterranean rookeries (Casale et al. 2018), and this is mainly due to the lack of egg predation and to the rather infrequent inundation events (Margaritoulis 2005; Patel et al. 2016). The increasing trend of hatched nests in respect to laid nests was probably a result of nest protection measures. Nest protection was gradually improved, especially after the NMPZ's formation, with 1) better targeted relocations, 2) introduction of wooden frames to prevent trampling, 3) regulation of beach furniture placements, and 4) cordoning off the nesting zone to restrict humans in areas close to the water (ARCHELON 2015).
Increase of Hatching Success, Emergence Success, and In-Nest Mortality. — The significant upward trends of these variables can be explained in general by the increasing temperatures (Mazaris et al. 2008, 2009; Pike 2014; Patel et al. 2016; Hays et al. 2017). Increased temperatures favor these metrics in temperate regions (e.g., Mediterranean) in contrast to tropical regions where current temperatures may have surpassed suitability thresholds (Pike 2014; Almpanidou et al. 2016; Hays et al. 2017; Montero et al. 2018). Temperature measurements in Zakynthos verified that the optimal maximum has not yet been reached (Katselidis et al. 2012; Patel et al. 2016), and hence hatching success should continue to rise until the thermal threshold occurs, about the year 2050 according to Pike (2014). To verify this will require continued long-term monitoring of hatching success and in-nest hatchling mortality (Hays et al. 2017; Almpanidou et al. 2018).
Decrease of Hatchling Recruitment. — Estimation of broad-scale hatchling production is a useful metric for population recruitment (Rees et al. 2016). Despite the long-term stable nest numbers and the significant increase of the percentage of hatched nests (in respect to laid nests) and of hatching success, hatchling production at Zakynthos has been declining, albeit not significantly, seemingly due to the significantly reduced clutch size.
Some important characteristics of hatchlings (e.g., body mass, growth rate, fitness) are shaped by the incubation environment, which is made up of a plethora of factors such as temperature, water content, gas exchange, and many others (Carthy et al. 2003). Changing environmental conditions, due to global warming, are expected to change hatchling characteristics. We consider it advantageous that loggerheads in Zakynthos deposit their clutches on different beaches across Laganas Bay, and hence in a variety of incubation environments, which may counter predict negative effects and ensure appropriate levels of hatchling production in the long term.
Decrease of Incubation Durations. — Incubation duration can be used as a proxy of incubation temperature and hence of hatchling sex ratio (Mrosovsky and Yntema 1980; Mrosovsky et al. 1999). All beaches, with the exception of Marathonissi and, to a lesser extent, East Laganas, displayed in the last 20 yrs mean annual incubation durations below the pivotal duration of 56.6 d (Mrosovsky et al. 2002), signaling production of predominantly female hatchlings. Male hatchlings are mainly produced on the beaches of Marathonissi and East Laganas, as reported previously (Zbinden et al. 2007; Katselidis et al. 2012). The declining trends of incubation durations, significant on Marathonissi and East Laganas, will further skew the overall female primary sex ratio, estimated during 2002 and 2003 as 68%–75% (Zbinden et al. 2007) and during 2007–2009 as 73.2%–80.6% (Katselidis et al. 2012).
It has been hypothesized that nesting fidelity may perpetuate the predominance of female hatchlings on the female-producing beaches and correspondingly decrease nesting levels at the male-producing sites (Heppell et al. 2003 and references therein). Laloë et al. (2014) explained the proliferation of loggerhead nesting in Cape Verde by the increased female recruitment due to temperature rise. In contrast to this, Zbinden et al. (2007) speculated that in view of the large thermal differences among Laganas Bay beaches, sea turtles would react to global warming by choosing the “cooler” beaches. A similar suggestion was presented for green turtles in Japan (Kobayashi et al. 2020). Although, we have not observed such a behavioral response in Laganas Bay, the hypothesis of Heppell et al. (2003) may not be the only reason explaining the decreased nesting in Marathonissi, as a plethora of other factors govern nest site selection (Miller et al. 2003; Mazaris et al. 2006), and the situation is exacerbated by the prevalence of high levels of anthropogenic disturbance.
Despite the ever-increasing number of female hatchlings, an almost 50:50 operational sex ratio was reported in Laganas Bay, currently minimizing the need of additional males (Hays et al. 2010; Schofield et al. 2017). However, the threshold number of males required for the viability of a population is not yet known (Witt et al. 2010; Laloë et al. 2017).
Reduction of Clutch Size and Body Size. — In the present study, we documented a significant downward trend of body size and clutch size at population level. These variables are strongly interrelated, with larger turtles laying larger clutches (Hays and Speakman 1991; Tiwari and Bjorndal 2000; Broderick et al. 2003; Margaritoulis et al. 2003; Zbinden et al. 2011; Casale et al. 2018; this study).
A reduction of body size (and consequently of clutch size) could be a result of greater mortality selectively affecting large-sized animals, as indicated by Kamezaki (2003). Continued loss of adults may cause early maturation in the population (Hamman et al. 2003; Le Gouvello et al. 2020). A decreased body size and clutch size of loggerheads in Turkey was explained by the declining number of females and the increased recruitment of younger turtles, which are generally smaller (Ilgaz et al. 2007). Changes in foraging areas or in the diet of the population may also cause changes to body size. This has been shown for loggerheads nesting in Laganas Bay, with those residing in the Adriatic being bigger and producing larger clutches than those residing in the Gulf of Gabès (Zbinden et al. 2011; Schofield et al. 2013); this dichotomy explained by the trophic richness in the northern areas (Schofield et al. 2013; Cardona et al. 2014; Patel et al. 2015).
Although the nesting population in Zakynthos is “stable” in the long term, the noted decline of body size, if not driven by large-scale changes in preferred foraging habitats, is potentially alarming as the current population “stability” may be attributed to an increased influx of neophyte nesters (see Hatase et al. 2002). Further, the decreasing clutch size may lead to a reduced lifetime reproductive output but this needs further investigation as it also depends on other factors, such as clutch frequency, remigration interval, and reproductive longevity (Broderick et al. 2001; Hawkes et al. 2005; Rivalan et al. 2005; Troëng and Chaloupka 2007; Zbinden et al. 2011).
Conclusions. — By analyzing long-term nesting data, collected under a systematic monitoring program on the 6 beaches fringing Laganas Bay in Zakynthos, we have highlighted some major findings, including confirmation of the regional importance of the loggerhead population and its reproductive and behavioral responses over time.
The annual nest numbers did not show any significant trend over 38 yrs and hence the nesting population is considered stable. The nesting contribution of the individual beaches has changed, with some public-access beaches receiving more nests, possibly due to improved management actions applied by the NMPZ's management agency; as a result nesting decreased in a remote beach previously featuring an atypically high nesting density. We noted the profound effect that increased temperatures, due to global warming, may have on various reproductive parameters: incubation durations showed decreasing trends on all beaches and consequently an increased female primary sex ratio, which is accentuated by a reduction of nesting activity on the coolest beach; hatching success, hatchling emergence success, and in-nest hatchling mortality showed increasing trends. We also recorded significantly decreasing trends in 1) clutch size, reducing hatchling production rates, and 2) turtles' body size, potentially suggesting early maturation. Overall, the situation for turtles nesting in Laganas Bay is in balance with reproductive rates at levels that can grow the population, but this is currently negated by other factors, likely including high adult mortality at sea and continued disturbances at some beaches. Further actions need to be taken to eliminate the unacceptable number of turtle deaths at sea, at least within the confines of the NMPZ, and to better enforce regulations at Daphni and Marathonissi.
Additionally, these results demonstrate the value of maintaining a long-term program by an NGO, dedicated in studying and protecting sea turtles in Greece. We expect that ARCHELON will continue this systematic work in the years to come and reveal further elusive traits of reproductive parameters, which can only be determined through decades of intensive work, and hence test current ideas on the effects of climate change to sea turtle populations.
Laganas Bay with National Marine Park of Zakynthos (NMPZ) borders and location of the 6 nesting beaches in analysis in the present study (MAR = Marathonissi, LAG = East Laganas, KAL = Kalamaki, SEK = Sekania, DAP = Daphni, GER = Gerakas). The marine area of the NMPZ includes 3 protection zones (A, B, C) featuring various boating regulations from 1 May through 31 October each year. Zone A: no boats allowed; Zone B: boats allowed under a speed limit of 6 knots, no anchoring; Zone C: same as B but with anchoring permitted. (Color version is available online.)
Number of emergences (A) and nests (B) over 38 yrs (1984–2021) in Laganas Bay, where no significant trends were found (multivariate Mann-Kendall, p > 0.05). Plots fitted with trend lines using the Theil-Sen estimator method. Grey band represents the 95% confidence interval (CI) for the fitted model.
Trends in annual nest numbers at the six beaches of Laganas Bay, over the 38-yr period 1984–2021. Significant trends and single change-points detected on all beaches except Kalamaki. Vertical dashed lines represent the year detected by Pettitt's test as a single change-point with respective significant levels. Plots fitted with trend lines using the Theil-Sen estimator method. Grey band represents the 95% CI for the fitted model (see Fig. 1 legend for definition of beach name codes).
Trends of the annual beach contribution to total nesting at the 6 nesting beaches of Laganas Bay, over the 38-yr period 1984–2021. Significant trends and single change-points detected on all beaches. Vertical dashed lines represent the year detected by Pettitt's test as a single change-point with respective significant levels. Plots fitted with trend lines using the Theil-Sen estimator method. Grey band represents the 95% CI for the fitted model (see Fig. 1 legend for definition of beach name codes).
Trend of year day for the first hatched nest for the 38-yr period (1984–2021). Plot fitted with a trend line using the Theil-Sen estimator method. Vertical dashed line represents the year detected by Pettitt's test as a single change-point with respective significant level. Bands around the fitted line represent the 95% CI for the fitted model.
Trends of the annual clutch size (A), hatching success and hatchling emergence success (B), and in-nest hatchling mortality (C) over the 38-yr period (1984–2021). Mann-Kendall tests found significant trends on all 4 variables. Plot fitted with a trend line using the Theil-Sen estimator method. Bands around the fitted lines represent the 95% CI of the fitted models. (Color version is available online.)
Trends of incubation durations per beach in Laganas Bay for the 20-yr period 2002–2021, showing decreasing trend on all beaches. Plots fitted with trend lines using the Theil-Sen estimator method. Trends in Marathonissi and East Laganas are significant (see text). Grey band around the fitted line represents the 95% CI of the model. Dotted horizontal line represents the pivotal incubation duration of 56.6 d at the nearby rookery of Kyparissia Bay (Mrosovsky et al. 2002). Beach name codes as per Fig. 1.
Trend of mean annual body size (SCLn-t) of neophyte nesting females in Laganas Bay over 34 yrs (1987–2021; no measurements in 1999). The Mann-Kendall trend test showed a significant trend (p < 0.05). Plot fitted with trend line using the Theil-Sen estimator method. Grey band represents the 95% CI for the fitted model.
Contributor Notes
Handling Editor: Sandra Hochscheid
