INTRODUCTION

The American horseshoe crab, Limulus polyphemus (Linnaeus, 1758), is an arthropod (class Merostomata, order Xiphosura, family Limulidae) that inhabits 2 defined areas of the American continent, the Atlantic coast of the United States of America (USA) and the Yucatán Peninsula (Mexico), of which in the second it is known as mex (Mayan for spider) (^{Zaldívar-Rae et al. 2009}). The history of life on Earth for this arachnid congener dates back to the Ordovician period, with Lunataspis aurora as the earliest horseshoe crab fossil record and trilobites as their closest evolutionary ancestors (^{Rudkin and Young 2009}). The American horseshoe crab has undergone minimal modifications in its physiognomy since its appearance around 450 million years ago, and it exhibits a solid adaptation to mangroves or beaches with moderate coastal dynamics (^{Jackson et al. 2002}). In these habitats, L. polyphemus feeds on a variety of invertebrates and, because it obtains nutrients through the separation of organic matter, it can be classified as a decomposer; furthermore, in different life stages it is prey to various species of birds, crocodiles, tiger sharks, turtles, raccoons, among others (^{Walls et al. 2002}). Over the past decades, the American horseshoe crab has become relevant in the field of biotechnology because its hemolymph contains limulus amebocyte lysate (LAL), a molecular complex with high economic value used for detecting endotoxin bacterial contamination in the biomedical and pharmacological industry worldwide (^{Novitsky 1991}).

In addition to its evolutionary, ecological, and biotechnological importance, L. polyphemus has been used in the USA for several decades as bait to catch other species of greater importance to fisheries, such as the eel Anguilla rostrata and the sea snails Busycon spp. (^{Bianchini et al. 1981}, ^{Botton and Ropes 1987}, ^{Kreamer and Michels 2009}). A different situation occurs in Mexico, as L. polyphemus has been documented to be used as bait for Octopus spp. in some localities of the Yucatán Peninsula (^{Zaldívar-Rae et al. 2009}, ^{Munguía-Gil 2010}, ^{Salas et al. 2011}, ^{Sandoval-Gío et al. 2016}, ^{Smith et al. 2017}); however, this is an illicit activity because the Official Mexican Norm NOM-059-SEMARNAT-2010 categorizes this species as “endangered” and prohibits its capture (^{DOF 2010}). Furthermore, the International Union for Conservation of Nature has classified L. polyphemus as a “vulnerable species” on its Red List since 2016 and it has emphasized that its populations are gradually declining (^{Smith et al. 2016}).

The octopus fishery is a prominent activity in the Yucatán Peninsula accounting for 80% of total national production, and Yucatán, the northernmost state in this region, has reported catches of up to 18,000 t per year of 2 species, Octopus vulgaris and Octopus maya (^{Velázquez-Abunader et al. 2013}, ^{Markaida et al. 2019}, ^{Sauer et al. 2019}). The O. maya fishery produces 15,000 jobs and over 27 million dollars per year in commercial value, with approximately 80% of catches being exported to Europe and Asia (^{Jurado-Molina 2010}, ^{Rosas et al. 2014}). The octopus fishing season spans from August 1 to December 15 each year in the Yucatán Peninsula, and significant catches have been recorded in the continental marine zone corresponding to Río Lagartos and San Felipe, the 2 most important ports in the Ría Lagartos Biosphere Reserve (RLBR), a protected natural area in eastern Yucatán (Fig. 1) (^{Munguía-Gil 2010}, ^{Velázquez-Abunader et al. 2013}).

Figure 1 San Felipe (SF) and Río Lagartos (RL) in the Ría Lagartos Biosphere Reserve (solid outline), Yucatán, Mexico. 

Population studies for L. polyphemus in Mexico date back at least 30 years. However, studies identifying illegal catches as one of the main risk factors for this arthropod in the Yucatán Peninsula are new and mention only how fishermen catch American horseshoe crabs, with no details on the number of caught individuals or catch volumes (^{Salas et al. 2011}, ^{Smith et al. 2017}). Although the issue is seldom addressed by RLBR residents because they appear to be afraid of identifying those who could be involved in illegal fishing, a recent qualitative study revealed that a significant percentage of the surveyed inhabitants from Río Lagartos and San Felipe (71%) believe that mex poaching for use as bait in the octopus fishery could threaten the sustainability of L. polyphemus populations in the RLBR area (^{Sandoval-Gío et al. 2016}). In the ecological context, a limitation to knowing whether this illegal extraction is quantitatively affecting L. polyphemus populations in the Yucatán Peninsula is the fact that studies addressing seasonal population trends have been discontinuous and have therefore not offered precise data on the demographic behavior of this species (^{Gómez-Aguirre 1993}, ^{Gómez-Aguirre and Yáñez-Martínez 1995}, ^{Ortiz-León and Rosas-Correa 2011}, ^{Rosas-Correa and Ortiz-León 2012}).

Regarding the ethology of the species, the American horseshoe crab in the Yucatán Peninsula is known to reproduce year round (^{Smith et al. 2017}), with presence of nests in the RLBR in spring and summer (^{Sandoval-Gío et al. 2018}), partly coinciding with the start of the octopus fishing season on the first day of August. Governed by the lunar cycles, L. polyphemus migrates from deep waters to estuaries to spawn. During the high tide, on nights with a full moon, the male and female, in amplexus, head to the high water line, where the female digs nests and lays numerous eggs that the male then fertilizes; the eggs are covered in substrate by wave action and become exposed the following lunar cycle, when swimming trilobite larvae hatch (^{Ortiz-León and Rosas-Correa 2011}).

The Guiding Plan for the Sustainable Development of Fisheries and Aquaculture in Yucatán states that, in order to guarantee the sustainability of ecosystems and fishery resources, appropriate bioecological data that help determine the anthropogenic influence on those resources must be provided, especially if those resources are potentially threatened (^{Flores-Nava et al. 2016}). In this context, this study was carried out in the 2 most important communities of the RLBR, Río Lagartos and San Felipe, where 2 conditions of interest converge: (1) the reported abundance of L. polyphemus is among the most important in Yucatán and (2) the octopus season is very active, with indications of illegal use of American horseshoe crab as bait. Thus, to know what effect the octopus fishery has on the L. polyphemus population, samplings were carried out over 4 consecutive years (2015, 2016, 2017, and 2018) at 2 specific times each year: the first a few weeks before the octopus season began and the second a few weeks after this season began. Our hypothesis states that significant decreases in parameters such as the number of individuals, number of females, and average length of L. polyphemus occurring after the start of the octopus season could be indicators of this fishery representing a risk for the American horseshoe crab populations in these 2 localities of the RLBR. The obtained results will contribute to establish sustainable fishery management plans and conservation programs for L. polyphemus in Mexico.

MATERIALS AND METHODS

The study was carried out in the localities of Chiquilá, in Río Lagartos (21°35.805′ N, 88°87.410′ W), and Punta Morena, in San Felipe (21°34.102′ N, 88°14.224′ W), both in the RLBR, Yucatán, Mexico (Fig. 1), during the 2015, 2016, 2017, and 2018 rainy seasons (May to September). The predominant weather in the RLBR is Awo(x′) (Köppen, modified by ^{García 2004}), with annual rainfall averaging 669.9 mm (^{CONAGUA 2017}).

Two samplings were carried out each year. The first sampling was done before the octopus season started and the second after the start of this season (Table 1). Considering the ethology of the species, samplings were done in a single night with a single replica during nights with full moon or nights closest to this lunar phase to allow sighting the largest number of individuals upon their arrival onshore en masse.

Sampling stations were chosen at random in 2015, and the geographic location of each station was recorded with a Garmin Etrex 30 GPS. Samplings in 2016, 2017, and 2018 were conducted at the same locations. A 10 × 100 m (1,000 m²) transect was set at each of the sampling stations, which were all located in coastline areas with mangrove vegetation. At each sampling point, triplicate measurements were taken of depth, using a lead line graduated in centimeters, and the physicochemical parameters (namely, temperature, dissolved oxygen, pH, and conductivity), using a HACH multiparameter sonde. Marine sediment was obtained using a corer made out of a PVC tube measuring 40 cm long and 5 cm in diameter, and samples were placed in labeled resealable 18 × 20-cm Ziplock plastic bags. Sediment samples were taken to the laboratory at the Technological Institute of Chiná, Campeche, Mexico, for a granulometric analysis using the ^{Folk mechanical sieving method (Folk 1966)}.

Of the live specimens, morphometric data were collected using a vernier caliper (mm) along the longitudinal section, that is, measurements of the length of the prosoma, which is the anterior region of chelicerates; the opisthosoma, which is the middle part; and the telson, which is the posterior region. Total length (TL) was determined by the sum of these 3 measurements: TL = prosoma + opisthosoma + telson. Each sampled individual was also weighed (g) with an OHAUS digital balance with 0.1 g precision. The sex of live specimens was determined by the shape of the pedipalps, and sexual maturity was estimated from the width of the prosoma. After this, the specimens were released. The General Directorate of Wildlife authorized sampling through licenses SGPA/DGVS/10056/15 and SGPA/DGVS/002163/18.

The data for the physicochemical parameters of water and the average lengths of organisms were subjected to a one-way analysis of variance (ANOVA), with the Tukey post-hoc test, to assess the difference between means (P > 0.05) using GraphPad Prism v.7.0 (GraphPad Software, USA). GRADISTAT (Kenneth Pye Associates, United Kingdom) was used to analyze the main sedimentological characteristics (phi, classification, symmetry, and type of curve). The abundance of organisms found during each sampling at both locations was analyzed using a chi-square test with a 95% confidence level to distinguish differences between the number of organisms observed before and after the start of the octopus season for each year (2015-2018).

RESULTS

No significant differences in the physicochemical parameters of water were found between sampling years at each locality or between localities (P > 0.05) (see global results in Table 2). The granulometric analysis showed no significant differences either. Specifically, sediment in Río Lagartos consisted of 94.66% (±2.08) sand and 5.33% (±2.08) silt. In addition, sedimentology showed poorly graded medium sand, with an almost symmetrical and highly platykurtic curve. In San Felipe, the percentage of sand was also the most prominent (95.00% ± 2.00) and that of silt was low (5.00% ± 2.00). Sedimentology there showed poorly graded medium sand, with an asymmetric and leptokurtic curve.

During the 2015-2018 period, the abundance of live L. polyphemus individuals varied from 10 to 21 specimens in Río Lagartos and from 9 to 24 in San Felipe before the octopus season (Fig. 2). The number of live L. polyphemus specimens decreased in both localities after the beginning of the octopus season each year. For 2015, the number of organisms found before and after the octopus season in both localities did not differ (χ²= 3.000 ˂ 3.845; d.f. = 1, P ˂ 0.05). For 2016, 2017, and 2018, the number of organisms found at Río Lagartos and San Felipe before and after the octopus season did change significantly (χ² = 6.916 > 3.845, χ² = 30.175 > 3.845, χ² = 22.000 > 3.845, respectively) (d.f. = 1, P ˂ 0.05).

Figure 2 Number of live Limulus polyphemus specimens found at Río Lagartos (a) and San Felipe (b), Yucatán, Mexico, in the period from 2015 to 2018. White bars: before the octopus fishing season. Black bars: during the octopus fishing season. 

The analysis of average TL for L. polyphemus individuals by locality showed significant differences, with larger specimens being observed at both localities before the octopus fishing season (P < 0.05) (Fig. 3a, b ). For the analysis of average L. polyphemus lengths by year, ANOVA also showed significant differences, with the highest values before the octopus fishing season (P ˂ 0.05). The highest average length (306 mm) was recorded during the sampling done before the start of the 2017 octopus season (Fig. 3a, b).

Figure 3 Mean length (cm) of Limulus polyphemus by location (a) and by year (b) in the 2015-2018 period. Homogeneous groups (P > 0.05) are indicated with the same letters for each treatment. A: before the octopus fishing season. B: during the octopus fishing season. 

DISCUSSION

Río Lagartos and San Felipe, in the RLBR, have been reported as aggregation and reproduction sites for L. polyphemus in Mexico (^{Zaldívar-Rae et al. 2009}, ^{Sandoval-Gío et al. 2018}), but quantitative records of populations of this species have not been continuously documented. The present study represents the first attempt in Mexican territory to assess L. polyphemus in a specific area and season for 4 consecutive years (2015-2018), and the obtained results can therefore provide a helpful overview of the populations. Recording data before and after the start of the octopus season in the 2 most important localities of the RLBR can particularly reveal the local impact of this fishery on the demographic dynamics of the American horseshoe crab inhabiting this protected natural area.

The present investigation showed that in the 4 consecutive sampling years at Río Lagartos and San Felipe, L. polyphemus counts markedly decreased after the start of the octopus season at both study sites, suggesting that this mollusk fishery has a negative impact on limulus populations. The results of the chi square analysis indicated population abundances were different between 2016, 2017, and 2018, that is, that the number of organisms decreased significantly due to the illegal extraction of organisms for use as octopus bait.

Different risk factors have been posed as causes of L. polyphemus population declines in the Yucatán Peninsula, notably habitat loss, pollution, and poaching (^{Zaldívar-Rae et al. 2009}, ^{Smith et al. 2017}). In the RLBR, even under the environmental protection status granted to L. polyphemus, as it is an endangered species, illegal catches for use as bait in the octopus fishery have been disclosed (^{Salas et al. 2011}, ^{Sandoval-Gío et al. 2016}, ^{Smith et al. 2017}). A recent study determined that a significant percentage (70.67%) of Río Lagartos and San Felipe inhabitants maintain that L. polyphemus poaching could be the main threat driving population declines for this species, outweighing habitat alteration on the coastline or pollution (^{Sandoval-Gio et al. 2016}). Other studies have reported that this species is used as octopus bait when the traditional baits, namely the Florida stone crab (Menippe mercenaria) and the longnose spider crab (Libinia dubia), go scarce (^{Salas et al. 2011}, ^{Smith et al. 2017}, ^{Markaida et al. 2019}).

The first studies on the population dynamics of the American horseshoe crab in coastal areas on the Yucatán Peninsula were carried out between the mid-1980s and mid-1990s, reporting a declining trend in the number of individuals in the areas with greatest anthropogenic disturbance (^{Gómez-Aguirre 1993}, ^{Gómez-Aguirre and Yáñez-Martínez 1995}). ^{Gómez-Aguirre (1994)} found 63L. polyphemus individuals, including adults and juveniles, at Río Lagartos in 1985 but found less than 20 live specimens in this same locality 9 years later. Two studies on Holbox Island, an area adjacent to the RLBR, covering the 3 climatic seasons in the region (dry, rainy, and northerly winds) recorded a total of 331 (^{Ortiz-León and Rosas-Correa 2011}) and 695 (^{Rosas- Correa and Ortiz-León 2012}) L. polyphemus individuals, with a maximum of 30.5% of individuals being live organisms, revealing that the American horseshoe crab population was scarce. ^{García-Bielma (2011)}) monitored the population in Términos Lagoon, Campeche, and found a stable trend in the number of organisms, however, noting negative aspects that could threaten this stability, such as changes to the coastline, habitat pollution, and trawl net bycatch; the study did not identify the octopus fishery as a risk factor. The present study, carried out during the rainy season, revealed a low number of organisms per study site (maximum of 21 in Río Lagartos and 24 in San Felipe, both prior to the 2017 octopus season). In the samplings after the start of the octopus season, the counts drastically decreased: only 2 organisms were observed in 2018 in Río Lagartos and none in San Felipe.

It should be noted that the physicochemical parameters of water and sediment texture did not vary significantly between localities during the present study. Data for these parameters were similar to those in previous reports (^{Sandoval-Gío et al. 2018}) and so the physicochemical composition of water and the composition of sediments can be discarded as factors having a negative influence on the low number of American horseshoe crabs found in the study.

In the present study, in addition to observing a decrease in the number of L. polyphemus individuals, the mean TL of sampled individuals decreased after the octopus fishing season started. Maximum values reached a global average (2015-2018) of ~280 mm in San Felipe (before the octopus season) (Fig. 3a); the maximum average length was ~300 mm before the 2017 octopus season (Fig. 3b). In contrast, TL values during the octopus season did not exceed 260 mm. In the individual count, the 2 specimens with the largest TL measured 400 and 429 mm, and both were collected during the samplings conducted before the octopus season in Río Lagartos in 2015 and 2017, respectively (data not shown). Regarding these records, other authors have reported similar sizes in previous studies carried out in the Yucatán Peninsula (^{Ortiz-León and Rosas-Correa 2011}, ^{Rosas-Correa and Ortiz-León 2012}).

In L. polyphemus, as in other species, size and weight increase with age (^{Vijayakumar et al. 2000}). The results of the present study show a decrease in average TL after the start of the octopus season, suggesting that in the context of the illegal supply of L. polyphemus to the octopus fishery, poachers could be extracting the largest and therefore oldest individuals to profit from more flesh. The predilection of poachers for larger organisms could have several negative implications for the American horseshoe crab, especially in terms of reproduction and spawning, since the largest American horseshoe crabs are normally females (^{Smith et al. al. 2017}). During most of our samplings, we found fewer females during the octopus season than before it started; however, this trend was more evident in Río Lagartos but not in San Felipe (data not shown).

On another note, it is important to emphasize that we cannot dismiss the fact that other factors from the life history of the species, for example, migration, recruitment, reproductive season, or natural mortality, could be causing the observed changes in the size of the American horseshoe crabs. However, biological studies in the study area to date are insufficient. ^{Swan (2005)} studied the migration of L. polyphemus in the North Atlantic coast of the USA and found that a high percentage (75%) of organisms tagged across 17 years migrated short distances (from 0 to 20 km), compared to the 1.57% that migrated more than 100 km. American horseshoe crabs migrate from shallow to deep waters, especially when they reach sexual maturity, but they return to the intertidal zone to spawn during high tide (^{Barlow et al. 1986}, ^{Smith et al. 2017}). Because the RLBR is a spawning site for American horseshoe crabs (^{Sandoval-Gío et al. 2018}), they would be expected to return to the intertidal zone to reproduce and spawn during full moons, assuming they did migrate to deep waters, but the numbers presented here show the opposite, at least for the seasons during which the present study was carried out. Hence, its use as octopus bait is concerning, since the number of American horseshoe crabs in a population will determine their distribution range; the more abundant the population, the greater the migration distances. The L. polyphemus population in the Yucatán Peninsula has been described, however, as not very abundant (^{Zaldívar-Rae et al. 2009}, ^{Smith et al. 2017}).

In Yucatán, Octopus (O. maya and O. vulgaris) fishing is considered to be the most profitable fishing activity between August and December and, together with the red grouper Epinephelus morio, it represents 70% of total catches in this Mexican province (^{Munguía-Gil 2010}). The commercial and exporta value for this fishery has increased in recent years, creating direct employment and substantial foreign currency incomes in southeast Mexico, but the activity could be unsustainable if it depends on affecting species with high-risk status such as L. polyphemus.

Currently, the reference framework known as Driver- Pressure-State-Impact-Response is useful to better under stand the cause-effect relationships between anthropogenic impact and the environmental components of an ecosystem (^{Gebremedhin et al. 2018}). Therefore, fishing for O. maya and O. vulgaris could signify strong pressure derived from economic demand, which could change the status of L. polyphemus populations in the RLBR (^{Smith et al. 2017}). We recommend that decision makers consider the data presented here for an analysis seeking efficient responses regarding the establishment of sustainable fishery management plans and conservation programs for the affected species in southeast Mexico.

In conclusion, during the study period, the relative abundance and average length of American horseshoe crabs sampled in 2 localities of the RLBR decreased after the start of the octopus season, in comparison with data recorded before the season began. Nevertheless, the analysis does not allow estimating the magnitude of the effects at the population level. This confirms our hypothesis: the significant decrease in the population and morphometry of L. polyphemus at the sampling sites suggests a possible negative influence related to its use as bait in the octopus fishery. However, we recommend comparing these data with population studies in other localities inhabited by the American horseshoe crab with no impact of illegal fishing.

The results presented here should be taken with caution, given the low methodological replication. Therefore, more extensive research that considers other biological factors, such as migration, life cycle, and mortality, of the mex in the RLBR, coupled with socioeconomic studies, could more solidly elucidate the potential risk of the octopus fishery on the limulus populations in this protected natural area. Unlike the Atlantic coast of the USA, where environmental education programs for the preservation of American horseshoe crabs have been historically more frequent (^{Walls et al. 2002}), in Mexico events of this nature are inexistent. Transdisciplinary research is clearly needed to channel efforts leading to the conservation of this species in risk areas.