Size Structure of Hawksbill Turtles (Eretmochelys imbricata) from Taxidermied Specimens in Private Collections Captured Along the Western Coast of the Gulf of California

Abstract

Human exploitation has been a major driver of marine turtle population declines, particularly affecting naturally scarce species such as the pantropical hawksbill turtle. Although hawksbill sea turtles have been documented in the Gulf of California since the early 20th century, data on their historical demography during periods of high exploitation in this region are nonexistent. We investigated the size structure of hawksbill turtles from the Western Central Gulf of California by examining a unique sample of decorative taxidermies, corresponding to 31 specimens captured during fishing operations near Santa Rosalía, Baja California Sur, Mexico, between 1980 and 1990. An analysis of the curved carapace measures revealed a length range (nuchal notch to posterior of supracaudals) of 29.5–59.5 cm (mean = 38.75 ± 6.67 cm) and a width range of 25.0–51.5 cm (mean = 33.63 ± 5.66 cm), with 87% of specimens having lengths between 30 and 45 cm. Based on the carapace length measurements, we estimated the ages to be between 7 and 20 years, indicating that the population included juveniles. Our findings provide baseline data for an understudied period and region, suggesting that this area previously served as an important juvenile habitat. These results contribute essential historical demographic information for conservation planning.

1. Introduction

The hawksbill turtle (Eretmochelys imbricata Linnaeus) is a pantropical species, distributed in the Atlantic, Indian, and Pacific Oceans. Indiscriminate hunting for their highly valued meat and shell, or to preserve them as valuable decorative taxidermies [1,2,3], has been a major driver of the decline of population declines worldwide. Other negative factors include loss of nesting and foraging habitat due to coastal development, mortality from incidental capture in fishing gear, and climate change effects that alter nesting patterns, food availability, and temperature-dependent sex determination, which can have long-term consequences for the population [2,4]. These threats have caused a population size reduction of more than 80%, leading to its classification as “Critically Endangered” on the IUCN Red List [5] and as “Endangered” in Appendix I of the Convention on International Trade in Endangered Species of Wild Fauna and Flora [6], and its inclusion in the Norma Oficial Mexicana-059-SEMARNAT-2010 [7]. Despite protections under national and international law, hawksbill turtles continue to face many of these pressures throughout their life cycle [8].

In the Eastern Pacific (EP), the distribution of the hawksbill turtle extends from the Baja California Peninsula (BCP) in the north to Peru in the south [1]. The presence of hawksbills in the Gulf of California (GC) has prompted significant research, particularly on their foraging ecology and recruitment patterns. Notably, no nesting activity has been recorded along the GC coastlines. Genetic analyses have shown that most hawksbills in the region originate from the nesting aggregation at Costa Careyes, Jalisco, Mexico, and form a genetically distinct population [9,10,11]. This genetic distinction, combined with the observed patterns of foraging site fidelity and natal philopatry, underscores the need for tailored conservation measures at both local and regional scales [12].

The GC population is especially intriguing due to its dependence on the region as a primary foraging ground, despite the absence of local nesting sites [13]. Recent studies have started to uncover the ecological dynamics of these aggregations, which appear to be dominated by juvenile turtles exhibiting complex habitat-use patterns [14,15]. Most individuals in the GC measure below 60 cm curved carapace length (CCL) and are presumed juveniles. An analysis of size-class distribution in known aggregation sites has revealed that juvenile and subadult individuals constitute the majority, with adults being less common [14,15,16]. However, given the vast area of the GC, demographic knowledge remains uneven-particularly in the central-western region where data are scarce.

This knowledge gap is especially concerning considering the central-western coast’s historical role as a hawksbill hunting ground, where turtles were captured for decorative purposes between 1980 and 1990 (pers. comm. local fishermen). The area, located between established foraging zones such as Bahía de Los Ángeles and Ligüi [8], is likely a significant developmental habitat. Yet, unlike other better-studied areas [16], it remains largely overlooked despite its proximity to active fishing grounds and its possible function as a migratory corridor [10].

Reliable demographic data, such as size-class distribution, are fundamental to sea turtle conservation, as they inform population health, recruitment success, and habitat quality [17,18]. For hawksbills in particular, these data help identify spatial or temporal shifts in population structure, whether due to historical exploitation or contemporary conservation efforts. Additionally, correlating size data with abundance, fecundity, and genetic patterns supports informed management decisions [18].

Collecting accurate morphometric data from hawksbill turtles is inherently challenging due to their low population density, mobility, and the remoteness or inaccessibility of key habitats—often within Protected Natural Areas—along with limited specimen availability and logistical constraints [1,19]. Nevertheless, past efforts have produced valuable insights into regional population structure along both coasts of the BCP.

Therefore, the objective of this current study was to conduct the first comprehensive assessment of hawksbill turtle size structure in the central-western GC using a historical sample (1980–1990) of taxidermied specimens from private collections. Our goals were to provide previously unavailable data from a strategically important but understudied region, to place these findings within the broader ecological context of hawksbills in the GC, and to demonstrate the value of historical collections in addressing crucial gaps in conservation science. By analyzing specimens captured during the peak years of exploitation, we offer insights into pre-conservation population characteristics that can inform current recovery assessments.

2. Materials and Methods

2.1. Specimen Collection and Measurements

Between December 2019 and January 2020, we were allowed to examine thirty preserved (stuffed) turtle specimens and one carapace kept as decorative items in 31 private fishermen households in Santa Rosalía, Baja California Sur, Mexico. Species identification was conducted following diagnostic characteristics such as the number of prefrontal scales, the number and arrangement of carapace scutes, and head size and shape [20]. Based on these characteristics, we identified 31 hawksbill turtles. According to the owners, the specimens were originally captured by hand during nocturnal spearfishing using Hookah diving equipment between 1980 and 1990, in the vicinity of the fishing harbor of Santa Rosalia, from La Trinidad fishing camp (27°49′23″ N, 112°42′46″ W) to Punta Chivato (27°04′58″ N, 111°57′29″ W), including the islands of Tortuga and San Marcos (Figure 1).

Six morphometric measurements were recorded for each specimen, namely curved carapace length (CCL), curved carapace width (CCW), straight carapace length (SCL), straight carapace width (SCW), plastron length (PL), and plastron width (PW), following [21]. The CCL and CCW were recorded using a flexible plastic measuring tape (±1 mm precision) by a single observer. The CCL was measured from the nuchal notch to the posterior margin of the supracaudal scutes, while the CCW was measured at the widest point, using the first projections of the axillary keels as reference points. Straight measurements of both the carapace and plastron were obtained through photographic analysis using the FIJI/ImageJ software program version 2.16 (±0.1 mm precision) [22]. In the present study, morphological criteria could not be applied to determine the sexes of the taxidermized specimens, as both the tail length and cloaca position were retracted and deformed during the taxidermy process.

2.2. Statistical Description

The curved carapace length (CCL) and curved carapace width (CCW) are the preferred metrics for statistical analyses in sea turtle research due to their practicality under field conditions, non-invasive nature, compatibility with historical datasets, accuracy in representing the true shell size, and consistent anatomical reference points that ensure reliable data collection across studies [21].

For the statistical analysis, we only considered the measurements of the 31 hawksbill turtles. Since our study focused primarily on these variables, to validate the use of classical descriptive statistics (mean and standard deviation: SD), chi-squared normality tests (X2; α = 0.05) [23] were conducted on all six measurement groups. For normally distributed variables, the means and standard deviations were calculated; for non-normal distributions, medians and ranges are presented [23]. Additionally, Pearson correlations were computed between all pairs of normally distributed variables, using simple linear models (α = 0.05) [23]. To determine the age range of the measured specimens, we used the two models indicated in the literature [24]: the generalized additive mixed model (GAMM) and the von Bertalanffy model.

3. Results

The taxidermied hawksbill specimens exhibited curved carapace length (CCL) measurements ranging from 29.5 to 59.5 cm (mean = 38.75 SD = 6.67 cm) and curved carapace widths (CCW) from 25 to 51.5 cm (mean = 33.67 SD = 5.66 cm;

Table 1. Normality tests for six morphometric variables recorded from preserved hawksbill turtles kept as decorative items. X² values and significance (p). Average ($\overline{\mathrm{x}}$) and standard deviation (SD) values are presented for normal variables. Mean and range values are presented for non-normal variables. n: sample size; CCL: curved carapace length; CCW: curved carapace width; SCL: straight carapace length; SCW: straight carapace width; PL: plastron length; PW: width.

	CCL	CCW	SCL	SCW	PL	PW
X2	1.69	0.65	6.94	2.07	4.23	1.76
p	0.19	0.42	<0.01	0.15	0.04	0.18
$\overline{\mathrm{x}}$	38.75	33.67	NA	28.13	NA	26.19
SD	6.67	5.66	NA	4.39	NA	5.35
Median	NA	NA	35.8	NA	26.8	NA
Range	NA	NA	27.3–58.7	NA	22.3–47.6	NA
N	31	31	31	31	30	30

). Among the six morphometric variables, the straight carapace length (SCL) and plastron length (PL) did not satisfy the normality assumption, while the remaining four variables were normally distributed (α > 0.05, Table 1). Most specimens (87%) displayed a CCL between 30 and 45 cm, with only one specimen showing a CCL < 30 cm and three with a CCL of >45 cm. Similarly, 74% of the CCW measurements ranged between 25 and 44.3 cm (Figure 2), with one specimen having a CCW of >45 cm and seven with values < 30 cm.

The linear correlations between the four normally distributed variables were all significant (p < 0.01), with correlation coefficients >0.94 (

Table 2. Pairwise linear equations between four normally distributed morphometric variables. The ordinate to the origin (a), the value of the slope (b), the correlation coefficient (r), and the value of the test statistic (F1,28 l.d.) are indicated for each case. In all cases, the significance level was less than 0.01 (CCL: curved carapace length; CCW: curved carapace width; SCW: straight carapace width; PW: plastron width).

		CCW	SCW	PW
CCL	a	1.41	3.59	−2.70
	b	0.83	0.63	0.75
	r	0.98	0.96	0.94
	F	667.48	364.00	225.80
CCW	a		2.63	−3.38
	b		0.76	0.88
	r		0.97	0.94
	F		661.00	228.98
SCW	a			−5.94
	b			1.14
	r			0.94
	F			252.33

In all cases, the significance level (p) was less than 0.01.

). The CCL-CCW relationship was nearly perfect within the studied size range (Figure 3), with the CCW increasing by 0.83 cm for each centimeter increase in the CCL. The raw data used in the correlation analyses are provided in Appendix A.

4. Discussion

While private collections present accessibility challenges, multiple studies demonstrate their unique value for conservation biology. These collections often contain critical specimens unavailable in formal institutions, providing irreplaceable biogeographic and morphometric data [25,26]. In our case, the 31 analyzed specimens—from a key historical period (1980–1990)—offer baseline information impossible to obtain otherwise, as documented in analogous studies with fish taxidermy [27] and arthropods [28]. As noted by [29], these materials, while imperfect, constitute a unique information channel when properly contextualized, particularly for reconstructing historical population patterns and prioritizing conservation efforts.

Historical accounts of sea turtle exploitation in the GC suggest that the predominance of juvenile turtles may result from prolonged adult extirpation [30] in their developmental grounds. While this hypothesis is plausible for green turtles (Chelonia mydas), which were historically the most widely captured species in the western GC, the situation differs for hawksbills. Near Santa Rosalía, according to anecdotes from fishermen, hawksbills were specifically targeted at smaller sizes for taxidermy, as these individuals were easier to display as curiosities in fishermen’s homes or given as gifts for social occasions. This contrasts with practices in other regions, such as the Caribbean, where adult hawksbills were primarily hunted for their meat and carapace. Despite this size selectivity, our results align with the general size structure recorded for the GC, where juvenile and subadult stages predominate. Further studies are needed to determine whether the size distribution observed in these taxidermized specimens reflects the current hawksbill aggregation in the central-western GC.

The size ranges for taxidermied specimens in the current study indicated that juvenile hawksbill turtles use the west-central coast of the GC as foraging and developmental grounds [10,11]. In this regard, Martinez-Estevez et al. [16] mention that in some regions of the Pacific Ocean, hawksbill recruitment sizes are generally larger than in the Caribbean, where individuals as small as 20 cm of straight carapace length arrive in neritic habitats. Further studies are required to determine if hawksbill turtles continue to use this area, which underscores the importance of understanding the biology, ecology, and demography of this foraging site, and its connectivity with Eastern Pacific nesting areas, particularly Costa Careyes.

The curved carapace length (CCL) serves as a critical metric in sea turtle studies; it is a key measure for the generation of essential information for management, conservation, monitoring, and inter/intraspecific comparisons, particularly for the determination of growth rates and reproductive sizes and for age group categorization (i.e., juveniles, subadults and adults). Following refs. [17,24], hawksbills are classified as juveniles (CCL < 60 cm), subadults (60–80 cm), and adults (>80 cm), although these thresholds may vary by population. Under these criteria, all specimens in our study were juveniles. Martínez-Estevez et al. [14] found similar results with 95% (92 of 97) of the hawksbill turtles captured for their study in the GC being juveniles, similar to those reported by other authors [1,13,31]. An analysis of the relationship between CCL and CCW revealed an almost perfect linear correlation, indicating isometric growth. This finding is consistent with observations in other Cheloniid species [32].

Hypothetically, sea turtles experience indefinite growth [33]; however, most species show an accelerated growth rate in their earlier stages and do not grow significantly after sexual maturity [34,35,36], although maturity cannot be strictly linked to size alone. Sexual maturity in hawksbill turtles varies between populations: in the South Pacific region in Australia, individuals reach sexual maturity between 20 and 30 years of age and at a CCL of 80–90 cm [17,30]; in the Eastern Pacific Ocean, sexual maturity was calculated to be between 16 and 26 years at a CCL of 72.2–89.5 cm [31]. Given the minimum (29.5 cm) and maximum (59.5 cm) CCLs recorded in the current study, the ages of these preserved specimens ranged from 4.1 to 12.0 and 3.2 to 9.4 years (GAMM and von Bertalanffy, respectively; see [31]).

As observed in other sea turtle species, the sex of a hawksbill turtle can be determined using morphological characteristics, such as the tail length and cloaca position, as well as behavioral traits, including the chasing and courting of females by males. However, these traits are evident and discernible only in adults and not young individuals [17]. In the present study, morphological criteria could not be applied to determine the sexes of the taxidermized specimens, as both the tail length and cloaca position were retracted and deformed during the taxidermy process, and the sample was composed of immature turtles that do not exhibit the secondary sexual characteristics in adults (tail length).

Despite potential biases from the selective capture of small individuals near Santa Rosalía, the average CCL and CCW values align with those recorded in other Pacific Ocean populations, supporting the hypothesis that the central-western GC serves as a recruitment and development area for hawksbill turtles. Given the current conservation status of this species, our study highlights the importance of monitoring this critical habitat and provides baseline data that can inform local and regional conservation efforts. In particular, the identification of this area as a developmental habitat underscores the need to assess and mitigate potential threats such as turtle bycatch, illegal take, and degradation of foraging habitats. Engaging local communities in these efforts will be key to ensuring effective management. The recovery of hawksbill turtles in the Eastern Pacific depends on coordinated conservation actions throughout their foraging and nesting habitats [37], and this study contributes to those efforts by drawing attention to a previously overlooked region.