Loggerhead Strandings and Captures Along the Southern Spanish Coast: Body Size–Based Differences in Natural Versus Anthropogenic Injury
Abstract
Human and natural factors threaten loggerhead turtle (Caretta caretta, Linnaeus 1758) survival in the western Mediterranean basin. Loggerhead stranding data in the south of Spain provide evidence that human activities are having a more detrimental effect on larger-sized loggerhead turtles. The straight carapace length of 162 individuals affected by natural causes averaged 35.15 cm; whereas, that of 179 individuals affected by human-related causes averaged 46.23 cm. The difference was highly significant.
Immature loggerhead sea turtles (Caretta caretta, Linnaeus 1758) that hatch on western Atlantic nesting beaches arrive at eastern Atlantic foraging habitats when they are approximately 20 cm long and 1 to 2 years old (Hays and Marsh 1997; Bjorndal et al. 2000). Some of these juvenile turtles enter the western Mediterranean Sea through the Strait of Gibraltar (between southern Spain and northern Morocco) and the Alboran Sea (Camiñas and De la Serna 1995). These juvenile individuals seem unable to leave through the strait until they reach a carapace size greater than 55 cm due to strong prevailing currents (Camiñas 2006; Revelles et al. 2007a; Eckert et al. 2008), and it is possible that many Atlantic loggerheads remain in the Mediterranean for much longer (Casale et al. 2009). Thus, the western Mediterranean Sea represents an important foraging area for western Atlantic loggerhead nesting stocks (Revelles et al. 2007b). Genetic studies in turtles stranded on the Andalusian coast (South of Spain) and captured in adjacent areas show a clear predominance of individuals from the Atlantic nesting beaches with little contribution of Mediterranean loggerheads (Carreras et al. 2006).
Several human factors, such as pollution with heavy metals or the use of drifting longline and trammel nets, have been described as threats to sea turtles in the western Mediterranean basin (Carreras et al. 2003; Báez et al. 2007a, 2007b; Tomás et al. 2008; García-Fernández et al. 2009), and some of these lead to turtles stranding (Bellido et al. 2009). Loggerheads may also be damaged due to natural causes, such as cold-stunning, debilitated turtle syndrome (DTS), or pathologies that marine turtles suffer in the wild (Norton et al. 2004; Alava et al. 2005; Bellido et al. 2007). In addition, some causes of loggerhead damage on the Mediterranean Spanish coast are related to body size. This is the case of turtles stranded showing signs of interaction with longlines, which are larger than others stranded due to other causes along the eastern Spanish coast (Tomás et al. 2008).
Large juveniles and adult loggerheads are more valuable for maintaining this species' populations than are eggs or small juveniles (Crouse et al. 1987; Congdon et al. 1993). In the western Mediterranean basin Casale et al. (2007) estimated through experimental studies that annual mortality is 27% for loggerheads coming from Atlantic nesting beaches. In this context, it is important to check if the negative impact on loggerheads on the southern Spanish coast is biased for size (or for the age of the sea turtles).
In this study, we compared the body size of loggerheads stranded and incidentally captured along the Spanish coast around the Strait of Gibraltar. We also examined the relationship between turtle body size and the prevalence of natural and anthropogenic sources of mortality.
Methods
The Strait of Gibraltar divides the Andalusian coast (1100 km; lat 37°10′N, long 7°23′W to lat 37°22′N, long 1°37′W) into 2 parts, Atlantic and Mediterranean (Fig. 1). Loggerhead stranding and bycatch data were recorded systematically from 1997 to 2006 by the Threatened Marine Species Recovery Centre of Andalusia (CREMA). CREMA has a volunteer stranding network with a strong presence along the entire Andalusian coast, complemented by local police and scientific groups, which collaborate on finding stranded turtles. This network records dead, sick, or injured turtles stranded on beaches, captured incidentally by fishermen, or found floating dead or in a weakened condition near the coast. It is coordinated via a 24-hour telephone hotline. From 1997 to 2006, a total of 1045 loggerhead turtles were recorded on the Andalusian coast, 910 of them stranded alive or dead on beaches and 135 incidentally captured or collected adrift by fishermen and sailors. We were not able to identify the cause of death or injury in all cases due to the state of decomposition of the carcasses or to the lack of access to animals (Kopsida et al. 2000). Causes of stranding or incidental capture were identified for 341 loggerheads (33%).
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0802.1
Turtles found alive were moved to the recovery center for veterinary examination and rehabilitation. Turtles were clinically evaluated by a team of 2 veterinarians and 2 biologists. The cause of injury was determined by external inspection, radiography, and if the animal died by necropsy. Fresh carcasses were brought to the recovery center for necropsy when possible. A trained team performed necropsies of fresh carcasses following a detailed protocol. Dead turtles in an advanced state of decomposition were examined by trained volunteers. The primary cause of death was determined based on complete external and internal examination and gross necropsy.
Turtles showing evidence of anthropogenic impacts were classified into 4 categories: 1) longline fishery (branch lines and/or hooks were found upon examination), 2) net fisheries (nets or marks of nets around flippers or neck), 3) trauma (external wounds caused by boat or propeller strikes), and 4) intoxication (turtles poisoned by ingestion of plastics, oil, and other water-borne contaminants). When anthropogenic causes were not apparent, live turtles were assigned to 1 of 3 natural causes: 1) buoyancy problems, 2) cold stunning, and 3) DTS (emaciated turtles with small barnacles on the skin).
Straight carapace length (SCL) measurements were taken from the 341 loggerheads examined for this study. The SCL data were tested for normality using the Kolmogorov-Smirnov test. The mean SCLs of turtles grouped in different categories were compared using analysis of variance (ANOVA; Sokal and Rohlf 1981) and Fisher's least significant difference (LSD) tests (Meier 2006) when the SCL distribution was normal. When SCL data were not normal, we used a nonparametric Wilcoxon test (Sokal and Rohlf 1981).
Because buoyancy problems may be due to natural (e.g., cold stunning; Bellido et al. 2008) or human (e.g., ingestion of oil or plastics; Lutcavage et al. 1997) causes, we explored the possibility of these problems being related to other causes by comparing the SCL frequency distribution in turtles affected by buoyancy problems to those of turtles with a similar mean SCL affected by other causes using the Kolmogorov-Smirnov test for 2 samples. Means are followed by standard deviation unless otherwise noted.
Results
The studies were focused on the 341 specimens for which causes of stranding or incidental capture were identified, 294 of them alive and 47 dead. The mean SCL for these turtles was 40.1 ± 14.07 cm. Longlines, net fisheries, and DTS were the main causes in records between 1997 and 2006 (Fig. 2). A total of 162 turtles presented signs of natural causes (buoyancy problems, cold stunning, and DTS combined), with a mean SCL of 35.2 ± 14.01 cm; whereas, 179 turtles presented signs of human causes, with a mean SCL of 46.23 ± 13.31 cm. The SCL data for turtles with natural causes did not have a normal distribution. However, the distribution of all turtles impacted by human causes as well as the SCL distributions for turtles affected by each of the 7 human causes considered separately were all normal. There was a low incidence of specimens affected by intoxication and trauma among loggerheads on the Andalusian coast, with a nonspecific size distribution (CV for intoxication = 0.38; CV for trauma = 0.30).
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0802.1
Turtles with natural pathologies presented a significantly lower mean SCL than turtles with human-caused pathologies (Wilcoxon W = 20,749.000; p < 0.001). Differences between the mean SCL for all 7 causes considered separately were significant (Table 1). Three size groups were found in relation to the cause diagnosed (Fig. 3). Size pattern A (mean SCL = 46.5 ± 14.89 cm) included the largest loggerheads; these were affected by buoyancy problems, longlines, net fisheries, trauma, and intoxication. All the human causes were within this group, as well as natural-cause buoyancy problems. Thus, human activities mainly affect subadults and larger juveniles. Comparison of the SCL frequency distribution of turtles affected by buoyancy problems with those affected due to human causes (Table 2; Fig. 4) showed that the differences were not significant, although they were greater in relation to longlines and practically nonexistent in relation to net fisheries. Size pattern B (mean SCL = 33.5 ± 9.97 cm) included the loggerheads affected by DTS. Finally, pattern C (mean SCL = 21.1 ± 6.46 cm) only included the youngest loggerheads affected by cold stunning symptoms (Fig. 4).
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0802.1
Citation: Chelonian Conservation and Biology 9, 2; 10.2744/CCB-0802.1
Discussion
The stranding records do not constitute a scientific sampling because strandings are random phenomena, dependent on parameters such as mortality, interaction with human activities, currents, tides, or prevailing winds (Epperly et al. 1996; Hart et al. 2006). In addition, the relation between the number of turtles dead offshore and the numbers of turtles stranded is unknown (Peckham et al. 2008). Nevertheless, the information provided by the strandings deserves a deep study. Live turtles incidentally captured and dead carcasses provided information about age, size, abundance, grow rates, main sources of threat, geographical patterns, or feeding (MacLeod et al. 2004; Chaloupka et al. 2008). It is a responsibility of the scientific community to ensure that this information is properly processed and filed for a better understanding of the species.
The average SCL values of the loggerheads recorded on the southern Spanish coast indicate that immature turtles predominate in this area (Bolten 2003; Kamezaki 2003). Nevertheless, we recorded specimens with an SCL greater than 70 cm that could have been mature loggerheads returning from the Atlantic coast for the second time or more (Eckert et al. 2008).
Turtles affected by net fisheries had a wider SCL range than those affected by longline fisheries. In the southern Spanish coast, a large and heterogeneous fleet uses different sorts of techniques, such as bottom trawling or other artisanal gear (Camiñas 2005). This fleet makes use of both the neritic and pelagic ecosystems and affects smaller turtles living in deep waters and larger individuals living in shallow waters.
On the Spanish coast of the Alboran Sea, which is part of our study area, stranded loggerheads were shown to have a smaller mean SCL than turtles captured by longline fisheries in nearby fishing grounds (Bellido et al. 2007). Laurent et al. (2001) found that loggerheads captured by longline fisheries differed in size as a function of the gear used and target species. Two longline fisheries operate in our study area with 2 different target species: longlines targeting swordfish (Xiphias gladius, Linnaeus 1758) and longlines targeting bluefin tuna (Thunnus thynnus, Linnaeus 1758). The mean SCL of loggerheads by-caught by longlines targeting swordfish was 49 cm (range = 28 to 71 cm), whereas the mean SCL of loggerheads by-caught in longlines targeting bluefin tuna was 54 cm (range = 36 to 72 cm; Laurent et al. 2001). The hooks used by these fisheries are large, and it is possible that they could only be swallowed by larger turtles; whereas, smaller specimens would not be affected by these fisheries.
Two natural causes, DTS and cold stunning, are more frequent among smaller loggerheads (i.e., those with SCL < 40 cm). Atlantic loggerheads are frequent in shallow waters over the Mediterranean continental shelf, and occupy neritic and pelagic habitats (Casale et al. 2008). In these waters, a pure oceanic/epipelagic stage is restricted to very small loggerhead turtles, which are probably unable to dive to the sea bottom (Casale et al. 2008). According to our results, these smaller turtles suffer less human pressure than larger specimens, and the only significant contributor to their stranding is cold stunning. Bellido et al. (2008) suggested that when loggerheads are cold stunned they may make eco-physiologic use of beaches to avoid cold waters and get warm. We found no significant differences between the size ranges of turtles affected by buoyancy problems and those affected by human causes. The differences were greater regarding longlines and intoxication and slight in relation to net fisheries and trauma. Thus, it is possible that buoyancy problems in the loggerheads stranded on these coasts may mainly be of human origin, the most probable human causes being net fisheries and trauma, although in many cases we cannot establish with certitude the human origin of the buoyancy.
Therefore, it is possible that human activities are having a negative effect on larger specimens, but have little impact on smaller sea turtles. This suggests that when younger loggerheads arrive at the Mediterranean basin they are somewhat vulnerable to human threats, but as they become older they become more exposed to human activities, especially to fisheries. In this way, the survival rate of loggerheads in the Mediterranean basin decreases as their size increases.
Although the western Mediterranean basin has been considered an important pelagic habitat for juvenile and subadult loggerhead sea turtles (Carreras et al. 2006), both pelagic and neritic habitats are used by these turtles in the Mediterranean (Casale et al. 2008). Small loggerheads are more frequent in pelagic waters, while larger specimens prefer neritic waters (Casale et al. 2008). These larger loggerheads in neritic or transitional stages can swim close to the coast in benthic feeding grounds (Eckert et al. 2008), where human activities are more pronounced. Thus, specimens returning to the nesting beaches on the American coast are more vulnerable than young loggerheads arriving at the Mediterranean basin.
This is an important issue since, for long-lived species such as sea turtles, the natural trend is to present high mortality rates in the first period of life and higher survival rates in larger juvenile and adult phases. Thus, the survival rate of larger juveniles has a strong effect on sea turtle population dynamics (Crouse et al. 1987; Heppell et al. 2003; Casale et al. 2007). Thus, it is possible that the damage suffered by loggerheads in the Mediterranean, affecting larger juveniles or adults, may contribute to the drop in the nesting population in areas such as Florida. Understanding the relationship between body size and stranding dynamics may be useful in sea turtle conservation programs, especially in an underrepresented region of the Mediterranean Sea.
Andalusian coast (map using seaturtle.org/maptool).
Distribution of loggerhead records on the Andalusian coast during 1997–2006 grouped by cause of damage (n = 341).
Mean straight carapace length (SCL) taken from all loggerhead turtles (Caretta caretta) recorded, grouped by cause of damage. Causes without significant differences in SCL according to the LSD test share the same superscript letter. n = 341. Standard deviation is shown as an error bar.
Straight carapace length frequency distribution in turtles stranded on the Andalusian coast or by-caught in marine fisheries.
