Body condition indices have been used to describe the health and well-being of several species of wildlife (Stevenson and Woods 2006). A body condition index is a metric derived most often from the relationship between an animal's length and its mass. It is used as a measure for individual condition, usually with reference to nutritional status and the energy reserves (Harder and Kirkpatrick 1996 in Copeland 2004). Changes in the body condition, resulting from changes in tissue biochemical composition and mobilization of energy reserves, may be related to season, life-history stages, health status, or exposure to stressors (Barton et al. 2002). The reproductive success, survival, and thus, population dynamics depend on the body condition of the organisms (Jones et al. 1999; Stevenson and Woods 2006).

Studies focused on determining and evaluating the physiological state and the health of wild sea turtles, have generated reference values for the clinical parameters of individuals, considering diverse factors such as seasonality, location, sex, size, age group, levels of contaminants, disease, physical state, and body condition (Bolten and Bjorndal 1992; Aguirre et al. 1995; Hasbún et al. 1998; Aguirre and Balazs 2000; Keller et al. 2004; Hamann et al. 2006; Whiting et al. 2007). Different approaches have been employed to estimate and evaluate individual or group body condition. The methods most commonly used for sea turtles include evaluation of slopes and residuals of the length–weight relationship, comparison of the weight/length ratio, and estimation of diverse indices of body condition (Bjorndal et al. 2000; Work et al. 2001; Jessop et al. 2004; Hamann et al. 2005). Some of these estimates are affected by the (small) sampling size (Work et al. 2001). However, these estimations have been considered unsatisfactory for the evaluation of the body condition of reptiles (Jacobson et al. 1993) and, specifically for sea turtles, the need to develop quantitative criteria to assess body condition has been suggested (Work et al. 2001). The body condition index that is most cited and most widely used in sea turtle health-assessment studies is Fulton's condition factor, the use of which was proposed by Bjorndal et al. (2000) to determine the body condition of the green turtle (Chelonia mydas, Linnaeus 1758). Fulton's index has been estimated for diverse sea turtle species and is calculated from the length–weight relationship (Beverton and Holt 1957).

The distribution of the East Pacific green turtle (known regionally as the ‘black turtle’), ranges through the occidental coast of Baja California, Gulf of California, Islas Revillagigedo, Islas Galápagos to south Peru (Márquez 1990). Currently no subspecies are recognized, and no evidence of genetic differences has been found between populations of green turtles (Karl and Bowen 1999); however, this group is considered as a population confined to the East Pacific, semi-isolated from populations of green turtles in Asia, Indo-Pacific, and Hawaii (Carr 1978; Pritchard 1997, 1999). The bays and lagoons of the occidental coast of Baja California and Gulf of California are the most important feeding areas for developing green turtles in their northern distribution (Seminoff et al. 2002a, 2003; Nichols 2003; Brooks et al. 2004). Subadults and adults predominate along the Gulf of California (Seminoff et al. 2002a), whereas smaller juveniles are most common along the Pacific coast (Koch et al. 2007). Juveniles are found throughout the year (Koch et al. 2007) and they are estimated to remain in these zones for decades (Seminoff et al. 2002b) prior to reaching sexual maturity.

In Mexico, sea turtles have been protected since 1990 when a complete ban on sea turtle catch was established and their protection became mandatory throughout the country (Diario Oficial de la Federación 2002). Nevertheless, the threats that mainly affect populations of green sea turtles in the Baja California peninsula are the incidental catch by commercial and artisanal fisheries (Gardner and Nichols 2001; Nichols 2003; Koch et al. 2006), being particularly elevated in Bahía Magdalena (BMA) and Laguna San Ignacio (LSI; Koch et al. 2006; Mancini 2009), and the illegal hunting for human consumption or for commerce of the meat, where the green turtles represent 77% of the consumption of sea turtles species (Mancini and Koch 2009). Punta Abreojos (PAO) is a fishing ground where the consumption of turtle meat is a common or traditional practice, and BMA is one of the 3 most important centers for regional sea turtle traffic (Mancini and Koch 2009).

Additionally, the degradation or destruction of their habitat by human activities, whether through contaminants, garbage, or urban development, is another threat for green sea turtles in the Peninsula (Gardner et al. 2003, 2006). The growing fishery industry, urban development, and increase of tourism activities are potential sources of pollution that have been generating health concerns (Castellanos and Mendoza 1991 in Maya and Guzmán 1998; Méndez et al. 1998; Shumilin et al. 2001). There is little information available on the health status of the East Pacific green turtle. Biochemical evidence indicated that injured green sea turtles from PAO (e.g., emaciated, open wounds, barnacles) had minor concentration of calcium, potassium, phosphorous, and a higher activity of cholinesterase than healthy ones (Labrada-Martagón et al. 2010). Those differences have not been found in BMA (Labrada-Martagón et al. 2010), where the body condition of green turtles has been increasing in recent years (Koch et al. 2007). Additionally, seasonal, annual, and regional differences in the clinical parameters (e.g., glucose, lipids, proteins) of the green turtles have been also found, suggesting variations in the quality of the habitat, in food availability, and/or of major nutritional quality for the sea turtles (Labrada-Martagón et al. 2010).

The goal of this study was to evaluate the body condition of the black turtle by using weight and length data. An annual and seasonal body condition factor (a) by site, and the index of relative body condition (K_rel) grouping individuals by study site and by physical appearance (healthy, injured), were calculated. Differences in body condition between study groups were evaluated. The correlation between the relative body condition index and the concentration of calcium, uric acid, lipids, albumins, and total proteins were evaluated.

METHODS

Study Area

During the years 2005 to 2007, live green turtles were captured in 3 zones located on the western coast of Baja California Sur: Punta Abreojos (PAO; 26°43′57″N and 113°35′44″W), Laguna San Ignacio (LSI; 26°43′ to 26°58′N and 113°08′ to 113°16′W), and Bahía Magdalena (BMA; 24°20′ to 25°43′N and 111°20′ to 112°15′W; Fig. 1).

View Figure

• Download as Powerpoint
• Download Figure

In PAO the sea turtles were captured in Estero La Bocana, which lies between 26°44′ and 26°49′N and between 113°37′ and 113°42′W (Guzmán 1998; Fig. 1). It is a mangrove channel, parallel to the coast, which is bordered by a sandbar that extends from the community of Punta Abreojos in the south to the extreme north where its mouth opens into the Pacific Ocean (INEGI 1995; Guzmán 1998). The direct contact with the Pacific Ocean exposes PAO to the continual oceanographic variations (e.g., upwellings, eddies, the California current system), which affect the primary production and the food web (Peterson et al. 2006). PAO is one of the most important fishery centers of the peninsula; this industry has been generating environmental health concerns in the zone, mainly due to the lack of necessary infrastructure for the waste elimination (Castellanos and Mendoza 1991 in Maya and Guzmán 1998).

Laguna San Ignacio is a coastal lagoon (17,500 ha) protected from the Pacific Ocean by a sandbar, is a shallow lagoon (2–4 m deep) with mangrove channels < 20 m deep. LSI is known as a tourism attraction for whale- and bird-watching. Additionally to the artisanal fishing fleet, the tourism industry generates boat traffic in the lagoon, which affects the sites with the highest density of whales and birds (Maya and Guzmán 1998).

In BMA the sea turtles were captured within Estero Banderitas, located in the northeastern zone of BMA (24°47′ to 25°01′N, 112°04′ to 112°10′W; Fig. 1); it is the channel with the greatest growth of mangroves in the lagoon (42 km²; Koch et al. 2007). BMA is protected from the Pacific Ocean by 2 islands (Margarita and Magdalena) that form a barrier (175 km in length; 1000 km² in area) and is linked to the ocean by 5 channels (Guzmán 1998; López-Mendilaharsu et al. 2005; Kampalath et al. 2006). BMA has great importance for the artisanal and industrial fishery; both fisheries generate the highest volume of fish species captured in the peninsula (Martínez 1998). Industries like a thermoelectric power plant, a sardine cannery, and whale-watching are also established in BMA (Guzmán 1998).

Turtle Capture

Field work was conducted in the following seasons defined by Koch et al. (2006); summer: May through October, winter: November through April. Sea turtles were captured with fishing nets (100 m long; 20-cm mesh), suspended in the water and monitored every 2 hours throughout the sampling period (up to 24 hours) to search for entangled turtles. For each sea turtle captured, straight carapace length (SCL, from the nuchal notch to the tip of the distal marginal scute) was measured with calipers (± 0.1 cm) and weight with a 150-kg spring scale (± 0.1 kg). Metallic tags were attached to both posterior flippers of each turtle before its release (Balazs 2000).

The physical state of the turtles, determined by visual inspection, was defined according to Labrada-Martagón et al. (2010) as (a) healthy, absence of apparent disease; (b) injured, if an epibiotic species was present and/or when at least one of the following conditions were present: barnacles or ectoparasites, any external skin or carapace lesion, recent traumatic injuries (scars were discounted), flipper amputations, obvious signs of illness (e.g., emaciation); and (c) no physical data, the turtle had not been visually inspected.

Estimation of Cody Condition

The body condition of the green turtle was estimated from the length and weight data of the captured individuals. Within each study zone data were grouped to obtain the relationships between weight (kg) and SCL (cm) of the sea turtles for each season and year. The weight–length relationship

was expressed in its logarithmic form, and the parameters a and b were estimated for each study group

The parameter, a, corresponds to the body condition factor expressed in grams (a × 1000; Froese 2006) of the sea turtles. The seasonal and the annual body condition factors (a) for each zone of study are presented as log a ± the 95% confidence interval (CI), with their respective equations, in Table 1. Differences in the body condition factors between seasons and years were evaluated by using the comparison of two intercepts (Kleinbaum and Kupper 1978).

Table 1 Seasonal and annual body condition factor in grams (a × 1000) of the green turtle (Chelonia mydas) obtained from the weight (kg) and length (cm) relationship by zone of capture. PAO  =  Punta Abreojos; LSI  =  Laguna San Ignacio; BMA  =  Bahía Magdalena; CI  =  confidence interval; W  =  weight; L  =  length; n.a.  =  not applicable.

View Fullscreen

From equation 1, the weight–length relationships were estimated grouping the individuals by zone, and by physical appearance (healthy and injured) for each zone. The relative body condition index (K_rel) was calculated according to Le Cren (1951) as described in Froese (2006):

where a and b are the parameters obtained from the corresponding length–weight relationship of the study groups. Differences in the estimated index between study groups were evaluated by comparisons of slopes and elevations using analysis of variance (zones) and Student's t-tests (physical appearance) according to Zar (1996). The average ± standard deviation (SD) of the relative body condition index are presented by zone and physical appearance of the individuals (Table 2).

Table 2 Relative body condition index (K_rel) and Fulton's body condition index for the green turtle (Chelonia mydas) by zone and physical appearance. PAO  =  Punta Abreojos; LSI  =  Laguna San Ignacio; BMA  =  Bahía Magdalena; SD  =  standard deviation.

View Fullscreen

Fulton's body condition index was calculated only as a reference and in order to compare with previous studies published by other groups. The index was calculated by dividing the mass (kg) by the cube of the length, SCL (cm), and multiplying by 10,000 (Ricker 1975 in Nash et al. 2006). Fulton's body condition index (mean ± SD) by zone and by physical appearance of the green turtles is presented in Table 2.

The correlation between the relative body condition index and the concentration of calcium, uric acid, cholesterol, triglycerides, albumin and total proteins of the sea turtles reported by Labrada-Martagón et al. (2010) was evaluated, polling data of individuals from BMA and PAO between 40 and 55 cm SCL, in order to eliminate the SCL effect. Prior to the analysis, the correlation between the body condition index and the SCL was evaluated until the size interval with no significant correlation was found (PAO, r  =  0.38, p  =  0.30; BMA, r  =  0.60, p  =  0.06). All results with p < 0.05 were considered statistically significant.

RESULTS

Turtle Capture

From 2005 to 2007, 52 green turtles were captured in PAO, 12 green turtles in LSI, and 49 green turtles in BMA. Injured sea turtles accounted for 35% of those captured in BMA, 44% of the individuals captured in PAO, and 25% of the sea turtles captured in LSI. For all zones and all years of study, the predominant age group was juveniles (PAO, 70%; LSI, 100%; BMA, 88%), followed by subadults (PAO, 25%; BMA, 12%). Throughout the course of the study, only 3 adults (> 85 cm SCL) were captured (PAO). For turtles captured in PAO, the SCL ranged between 43.9 and 92.4 cm (mean: 62.8 ± 11.4 cm) and weight between 11.3 and 107.9 kg (mean: 38.2 ± 23.4 kg); in turtles from LSI the SCL ranged from 39.7 to 62.2 cm (mean: 49.3 ± 6.9 cm) and weight from 10 to 32 kg (mean: 18.2 ± 7 kg); in green turtles from BMA the SCL range was 40.4–80.3 cm (mean: 56.2 ± 8.96 cm) and weight ranged between 4.7 and 68.1 kg (mean: 28.2 ± 1 5.5 kg).

Seasonal and Annual Body Condition Factor

The estimated values of the condition factor of the green turtles, as well as the corresponding models of the length–weight relationship, grouped by season, year, and capture zone are presented in Table 1. In PAO the green turtles showed a greater a in 2006 than did those in 2005 (p  =  0.02; Fig. 2). The sea turtles from LSI were captured in summer of 2006; therefore, annual and seasonal differences could not be evaluated. The great variability in the weight of turtles from BMA is reflected in the CI of the estimated condition factors a (Fig. 2). In BMA, the a estimated from data of green turtles captured during 2005 was 86-fold higher than that observed in 2006, and 2-fold greater than that of 2007 (Table 1). In PAO and BMA an increase from winter to summer in a of the green turtles was observed: 4-fold higher in BMA and 2-fold higher in PAO (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Relative Body Condition Index

Average (± SE) Fulton's body condition index estimated for the green turtles captured in PAO was 1.25 ± 0.41, in LSI 1.46 ± 0.11, and in BMA was 1.5 ± 0.6. The K_rel for green turtles grouped by capture zone and physical status are shown in Table 2. The green turtles from LSI presented a lower K_rel than the turtles from the other zones due to their smaller size (Table 2), although these differences were not statistically significant (F_{(2, 109)}  =  2.4, p  =  0.09; Fig. 3A). When K_rel was expressed relative to the SCL of the individuals, no differences in the increase of K_rel per cm of SCL were found between zones (slopes: F_{(2, 16)}  =  2.87, p  =  0.06); nevertheless, the green turtles from PAO had a lower K_rel than those from LSI (elevations: F_{(2, 108)}  =  96.95, p < 0.001; Fig. 3B).

View Figure

• Download as Powerpoint
• Download Figure

Injured green turtles showed a lower body condition than healthy turtles in all 3 studied zones (Table 2). Those differences were statistically significant for turtles captured in PAO and BMA (elevations: p < 0.01; Fig. 4); data of LSI were not evaluated due to the small sample size. Additionally, in PAO the value of the parameter b for the group of injured green turtles was significantly lower than in healthy turtles (t  =  4.12, p < 0.001); that is, the rate of increase of K_rel per cm of SCL is lower for injured than for healthy turtles (Fig. 4a). The K_rel for the sea turtles was not correlated to any of the biochemical parameters evaluated when SCL effect was removed.

View Figure

• Download as Powerpoint
• Download Figure

DISCUSSION

The body condition of an animal refers to its energetic state; an animal in good condition is assumed to have more energy reserves than an animal in poor condition (Harder and Kirkpatrick 1996 in Copeland 2004). Traditionally, conservation biologists have approached species health at the population level. A population flourishes or wanes based on the health of its individuals (Stevenson and Woods 2006). In studies involving evaluation of the physiological and clinical health of sea turtles, the creation of adequate and specific criteria for the determination of the body condition is a priority. This would allow a better interpretation of the basal values and the health parameters evaluated in East Pacific green turtles and the adequate comparison among species of sea turtles around the world.

To our knowledge, this is the first estimation of the body condition of green turtles in PAO and LSI, Baja California Sur. Currently, there is little information on the general health of the sea turtles that inhabit these feeding areas, probably due to their inaccessibility because they are far from urban areas. Despite their small size (< 62 cm SCL), the body condition of the green turtles in LSI was higher than that estimated for juveniles from the Gulf of California of up to 77.3 cm SCL (Seminoff et al. 2003). Juveniles from BMA also had higher body condition (1.56 ± 0.67) than green turtles from the Gulf of California (1.42 ± 0.015; Seminoff et al. 2003). Koch et al. (2007) suggested that, in recent years, there has been a trend toward better body condition of the green turtles that inhabit BMA. Our results support this idea because estimated average body condition for BMA was slightly higher (1.50 ± 0.6; 2005–2007) than the value (1.35 ± 0.12) reported for the years 2000–2003 by Koch et al. (2007). Given that juveniles were the most abundant age group (70%–100%), it can be said that the body condition estimated in this work is represented by juveniles in the three most important foraging areas of the western coast of Baja California Sur.

The average body condition estimated for green turtles from PAO was lower than those from BMA and from LSI when size is considered. PAO was also the site where the larger proportion of injured sea turtles was captured, and the only zone that showed annual differences in the body condition factor. Annual differences in the body condition of green turtles have not been reported before in BMA or the Gulf of California (Seminoff et al. 2003; Koch et al. 2007). Probably the higher population density in PAO, relative to BMA and LSI (Labrada-Martagón et al. 2010), influenced the body condition of the individuals. Juvenile green turtles, captured in a feeding zone in the Bahamas, had a lower body condition during the years in which a higher relative density and a smaller growth rate were found, implying that these two variables are correlated with body condition (Bjorndal et al. 2000). An environment with a population density lower than its carrying capacity and, therefore, with greater abundance of food per sea turtle, may lead to a better body condition of the individuals (Seminoff et al. 2003).

The body condition of the injured green turtles from PAO and BMA was lower than in healthy turtles. Additionally, injured sea turtles from PAO showed a slower rate of increase in body condition per cm of SCL. In the Gulf of California differences in body condition between healthy and injured individuals have not been found (Seminoff et al. 2003). Differences in blood biochemical parameters (e.g., potassium, calcium, cholinesterase) between injured and healthy green turtles have also been reported in PAO but have not been found in green turtles from BMA nor LSI (Labrada-Martagón et al. 2010). The regional differences in the body condition and biochemical parameters reflected differences in habitat, availability, and composition of food. The Pacific coast of the Peninsula of Baja California has many oceanographic variations (e.g., upwellings, sea surface temperature, water masses) that modify the primary production and, thus, affect the food chain (Peterson et al. 2006). It is possible that green turtles from PAO are more susceptible to environmental changes and food availability due to the close proximity of this coastal lagoon to the Pacific Ocean, as reflected in differences of their body condition. Future studies are needed to monitor the health status of green turtles of PAO, a feeding area with the highest density of green turtles in the Pacific coast (Labrada-Martagón et al. 2010) where sea turtles develop prior to reaching sexual maturity, and their migratory and breeding periods.

The estimated seasonal and annual body condition factor for BMA showed wide CIs, due to the great variability in the weight of the individuals. This could be due, on one hand, to individual variations in measuring techniques and, on the other, to a relationship between foraging behavior and the greater area available in the mangrove channels of BMA in comparison to that of PAO. Although green turtles show a strong fidelity to their feeding areas on the western coast of Baja California Sur, individuals do move between diverse patches of vegetation within the lagoon complex of BMA, which are defined by the marked changes in tides and currents (Nichols 2003; Talavera-Saenz et al. 2007). It is possible that individual differences in diet, whether in quantity or quality, are reflected in selectivity (López-Mendilaharsu et al. 2005), nutritional status (Bjorndal 1985), and, therefore, in the weight and body condition of the individual sea turtles.

The sea turtles captured in PAO and BMA presented a body condition index that was 2- and 4-fold higher in summer than in winter, respectively. The seasonal difference in the body condition of the green turtles that inhabit BMA has been reported and it has been suggested that, specifically for the coastal lagoons, it may be due to changes in the temperature of the water and the availability of food, which leads to a higher body condition (Koch et al. 2007). Additionally, the composition of the marine flora and the composition of the diet of the turtles that inhabit the BMA lagoon complex vary by season (López-Mendilaharsu et al. 2003; Hernández-Carmona et al. 2007).

During summer and in the years (2005 in BMA, 2006 in PAO) in which the green turtles demonstrated a higher body condition, it coincided with higher concentrations of lipids (triglycerides and cholesterol), glucose, total proteins, uric acid, potassium, and the activity of the enzyme GGT, as well as with a lower activity of the transferases alanine aminotransferase (ALT) and aspartate aminotransferase (AST) (Labrada-Martagón et al. 2010). Nevertheless, those parameters were not correlated to the relative body condition index when the size effect was removed. An increased level of glucose in the blood is indicative of a greater consumption of carbohydrates in the diet. For sea turtles, uric acid has been considered as a signal of increased ingestion of proteins resulting from an omnivorous diet (Whiting et al. 2007). The concomitant elevation in the levels of total proteins, globulins, and albumin reported by Labrada-Martagón et al. (2010) supports this scenario. Triglyceride levels are a useful tool in determining the general body condition and nutritional status of the green sea turtle (Aguirre and Balazs 2000; Swimmer 2000; Hamann et al. 2005); it has been suggested that, together with other metabolites, analyses of triglycerides should be incorporated and used in studies of diet of sea turtles (Bolten and Bjorndal 1992; Hamann et al. 2005; Whiting et al. 2007).

The body condition indices, evaluated together with blood biochemistry parameters, can be a useful tool as biomarkers of the physical and nutritional status of the sea turtles. With these tools we showed that green turtles had a better body condition during summer and during the years 2005 (BMA) and 2006 (PAO), which complements the annual and seasonal information found in clinical parameters, suggesting a greater availability and/or nutritional quality of food during said periods.

The indices proposed in this work are useful and practical estimations for the evaluation of the body condition of green turtles, based on individual weight and size data. The relative body condition index, compared by slopes, allows evaluating differences not just between years, seasons (evaluated with body condition factor a in this study), or zones, but to identify healthy and injured individuals, and alterations in the body condition relative to the increase in size of the green turtles. When the objective is to estimate the relationship between some measure (e.g., body condition) and one or more independent variables (e.g., environmental, health status), typically the measurements are obtained at a number of stressful conditions, and then the resulting data are used to estimate the relationship between the measurements of performance and the stress variables (Cattet et al. 2002). In this sense, the use of the slope provides an overall trend of the data, evidencing the performance between the measure and the independent variables in different situations, instead of using the measures of central tendency (e.g., arithmetic mean) that just provide a description of the preponderance of the values somewhere around the middle of the range of observed values (Zar 1996).

We recommend the use of the relative body condition index instead of the body condition factor a, because it generates a larger CI in response to individual variations in weight of green turtles. The use of a single statistic (i.e., “condition factor a”) has been questioned when assessing fish condition because it results in a loss of information and in an inaccurate representation of the length–weight relationship (Cone 1989 in Jones et al. 1999). Within a given species, interpretation of the body condition is difficult and prone to errors due to differences in the particular dimensions measured, such as body shape, average length, seasons, gender, age, and/or gonadal development (Le Cren 1951 in Froese 2006). We recommend estimating the body condition index considering, as was done in the present study, abiotic (year, season, geographical region) and biotic (age, size, weight, sex) factors. Care must be taken when these tools are used, and conclusions made regarding the health of sea turtle populations, because an overestimation of the body condition could lead to inadequate decision-making when the goal is conservation of populations and species.