INTRODUCTION

The blue land crab Cardisoma guanhumi Latreille in Latreille, Le Peletier, Serville and Guérin, 1828, is a widely distributed species spanning from southeastern Florida, United States, to Santa Catarina, Brazil, including Bermuda and the West Indies in the Caribbean Sea (^{Taissoun, 1974}; ^{Branco, 1990}). In Colombia, it is distributed in all the departments located on the Colombian continental Caribbean and on the island of San Andrés.

C. guanhumi populations currently face two main threats. Given that the species constitutes an important economic resource in places such as Venezuela (^{Taissoun, 1974}; ^{Carmona-Suárez, 2011}), Brazil (^{Oshiro et al., 1999}; ^{Botelho et al., 2001}), Mexico (^{Rivera, 2005}), Cuba (^{Álvarez et al., 1993}) and other Caribbean islands, in which it has even been overexploited, one of these threats is extraction. The other is the disruption or loss of their habitat (^{Rodríguez-Fourquet and Sabat, 2009}). In Colombia, the species is classified as Vulnerable (VU) due to the pressures it faces through habitat deterioration and indiscriminate capture (^{Bermúdez et al., 2002}).

Knowing the growth parameters and mortality rates of the populations of a species is of great importance to generating adequate management and conservation strategies, as these allow us to evaluate how extraction or other types of anthropogenic activities can affect the populations. Unfortunately there are few studies on these aspects for C. guanhumi populations. Under laboratory conditions, ^{Henning (1975)} estimated that an individual with 98 mm carapace width can be 13 years old and only two studies, both conducted in Brazil, have been carried out to determine the population growth parameters of this species. The first was carried out in an area where this crab is exploited, where a K= 0.594 years^-1 was estimated (^{Botelho et al., 2001}). The second study was carried out in an area where commercial catches do not take place, and the estimated K was 0.145 years^-1 (^{Costa, 2016}), which, in both cases, provides evidence of the species'' slow growth rates. Bearing this in mind, the purpose of this study is to estimate the growth parameters of C. guanhumi populations in protected areas in the department of Magdalena, in the Colombian Caribbean, based on the size frequencies in order to obtain relevant information to help manage the species.

MATERIALS AND METHODS

Study areas

Five locations in the Colombian Caribbean belonging to two natural parks in the department of Magdalena were chosen: The Centro de Visitantes in the Salamanca Island Road Park (SIRP), and Bahía Concha, Bahía Nenguange, Cabo San Juan del Guía and Los Naranjos in the Tayrona National Natural Park (TNNP) (Figure 1). Despite being protected areas, these places are tourist attractions in which environmental modifications have been made, related to housing, transport or tourism, and where, despite the efforts of the relevant entities, illegal extraction of this crab may take place. However, there is no information on the intensity with which this occurs.

Figure 1 Location of sampling areas. 

RESULTS

A total of 816 individuals were collected during the six-month sampling period. Of these, 181 in Bahía Concha, 154 in Cabo San Juan del the Salamanca Island Road Park, 186 in Nenguange, 183 in Los Naranjos and 112 in the Salamanca Island Road Park (SIRP). Except for the VIPIS location, in all other locations, the individuals were most active during the night, making SIRP the smallest sample size.

A Spearman correlation (^{Sokal and Rohlf, 1979}) of the carapace width-length ratio was calculated, in which it was evidenced that the measurements are strongly associated (correlation coefficient: 0.99, p < 0.001), thus, only the carapace width (AC) was used for the other analyses.

The size frequency distribution (Figure 2) shows that, in general, the males present a wider distribution, between 20.2 and 103.7 mm (61.35 ± 0.87 mm) (average ± standard error), while in females it was between 25.9 and 91.9 mm (57.66 ± 0.50 mm). In females, a modal distribution of 62-64 mm was observed, while in males it was not possible to observe a clear modal distribution. Male weights were between 4 and 516 g (133.49 ± 5.21 g), and female weights were between 9 and 266 g (95.90 ± 2.12 g). Sexual ratio differed significantly from 1:1 (x²= 20.08, p < 0.001), with a predominance of males.

Figure 2 Histogram of Cardisoma guanhumi size frequency for males and females collected between May and October 2016 in protected areas within the department of Magdalena. 

The width-weight relationship for both males and females showed negative allometric growth (b = 2.93, IC = 2.89 - 2.97, t-student = 3.186, p = 0.002 for males and b = 2.80, IC = 2.74 - 2.86, t-student = 6.397, p < 0.001 for females) (Figure 3).

Figure 3 Carapace width-weight relationship for C. guanhumi males (a) and females (b) collected between May and October 2016 in protected areas within the department of Magdalena. 

For the growth parameters of C. guanhumi in the department of Magdalena, a higher AC_max was obtained in males than in females, making the estimated AC_∞ higher for males. When entering the respective values of AC°° in the ELEFAN K scan subroutine, the males had a lower K value than the females. This indicates that males are slower growing and have a longer lifespan than females (Table 1 and Figure 4). With regard to the parameters estimated for the PNNT, a higher K value and a lower longevity were found than those obtained for the department (taking into account both PNNT and VIPIS individuals) (Table 2).

Table 1 Growth parameters for Cardisoma guanhumi males and females in the department of Magdalena. n= number of individuals sampled, AC_max = maximum carapace width obtained in the study, AC_∞=asymptotic carapace width, K= curvature parameter, t0= the theoretical age at which the individual is size 0, t_max= maximum age, Ø′=growth performance index. 

Table 2 Growth parameters for Cardisoma guanhumi males and females in the department of Magdalena. n= number of individuals sampled, AC_max = maximum carapace width obtained in the study, AC_∞ = asymptotic carapace width, K= curvature parameter, t₀= the theoretical age at which the individual is size 0, t = maximum age, Ø’=growth performance index. 

Figure 4 von Bertalanffy growth curve for Cardisoma guanhumi males and females captured in different localities in the Magdalena department. The growth curves were plotted up to the AC_max age obtained in the study. 

Using this same routine, the monthly frequency distributions of carapace width for the Magdalena department population were plotted and it was found that there is no modal progression, as it is not possible to visualize cohorts (Figure 5). The growth performance index (Ø’) was similar in males and females. The estimated total mortality rate for the department's population was Z=0.41 year^-1, the natural mortality rate was M=0.36 year^-1 and the fishing mortality rate was F=0.05 year^-1.

Figure 5 Monthly carapace-width frequency distributions of Cardisoma guanhumi individuals captured between May and October 2016. 

DISCUSSION

The distribution of the monthly frequencies was multimodal, since it was not possible to observe the presence of cohorts throughout the months. This has been previously observed for this species and has been attributed to the fact that the recruitment of juveniles is continuous, and to the species' characteristic slow growth (^{Silva et al., 2014}; ^{Costa, 2016}).

It was possible to show that the males reached larger sizes, grew more slowly and had greater longevity than the females. ^{Costa (2016)}, using a mark-recapture method, obtained K = 0.159 year^-1 for males, which is very similar to that obtained in this study (K = 0.150 year^-1). The author found, however, that females grow slower than males (Table 3). This may be related to the AC_∞ estimated in that study, which was higher for females than for males (Table 3), which is unusual since it has been widely reported that Gecarcinidae family males reach larger sizes and higher weights than females, as the latter invest more energy in reproductive processes (^{Hartnoll, 1982}, 1988).

Table 3 Comparison of the estimated growth parameters for Cardisoma guanhumi with respect to the literature. M= males, F= females, T= total. AC_∞= asymptotic carapace width, K= growth coefficient, t = theoretical lifetime, phi prima index. *Data calculated from the information reported by the authors. To estimate t_max we used: t_max= t₀ + (3/K) (^{Taylor, 1958}), if t₀ was not reported, 0 was used in the equation. For phi prima the following was used: ^{Pauly y Munro (1984)}. **Carapace length measurements (LC). For ^{Botelho et al. (2001)} t_max was not estimated because the authors calculated K from LC values. *** Data not reported by the author. 

As previously mentioned, there are three approaches to growing this species: one in laboratory conditions (^{Henning, 1975}), and the other two in natural environments. In the first of these, ^{Botelho et al. (2001)} used the Bhattacharya method (which is also based on size frequencies), and in the second, ^{Costa (2016)} compared two methodologies: ELEFAN and mark-recapture (Table 3). In both studies, the K estimates were higher than those in this paper, possibly because of the maximum reported sizes and the length measurements used, which are important factors in the estimation of growth parameters in the size frequency analyses. For example, Botelho et al. (2001), whose samples were taken in a locality in which C. guanhumi was exploited, estimated the growth parameters based on LC measurements and not AC ones. They reported an LC_max=62 mm, which could lead to a high K being calculated.

70 mm, estimated two values for K, the first result from an ELEFAN analysis (n=1078) and the second using marking and recapture data (n=291). Thus, ^{Costa (2016)} obtained a higher K with ELEFAN than the K estimated from tagging, catch and recapture data. This can be explained because to estimate the K parameter, the ELEFAN routine is based on AC_∞ (which is calculated from the AC_max acquired in the sample). It may be that during sampling, Costa (2016) was not able to capture an individual with an AC_max representative of the population (it is likely that there were larger sizes considering that the sampling was carried out in a surveillance area where no commercial catches of C. guanhumi occur). Thus, when calculating a smaller AC_∞ the value of K is overestimated.

While with the tagging marking and recapture data, ^{Costa (2016)} was able to estimate the increase in length from direct changes in crab growth, and therefore the K parameter was not affected by the AC_max obtained in the sampling.

Based on the growth parameters for the PNNT, it was observed that. C. guanhumi showed faster growth (K= 0.29 year^-1) and shorter longevity (t_max = 10.42 years) than at departmental level (K= 0.15 year^-1 and t_max = 20.06 years). Even though both PNNT and VIPIS are part of the Magdalena department, it is possible that the differences obtained when analyzing the PNNT data separately are related to the proximity between the four localities of the PNNT, since, by being closer, these populations are more closely related and therefore have more common characteristics, such as growth. Additionally, it was evident that C. guanhumi individuals from VIPIS showed a different behavior from other localities, as during most months, the individuals showed little activity outside the burrows, restricting the sample size in this locality.

The growth performance indices (Ø’) for the department were similar to those calculated for the ^{Costa tagging and recapture data (2016)}: 3,246 and 3,191 for males and females respectively (Table 3). On the other hand, the Ø’ calculated for ^{Botelho et al. data (2001)} was 3,621, slightly higher than the others (Table 3). However, as previously mentioned, these differences may be related to the fact that the study used the LC and not AC for size frequency analysis.

The total mortality rate (Z = 0.41 year^-1), natural mortality (M = 0.36 year^-1) and fishing mortality (F = 0.05 year^-1) in the department were low, which is consistent with the fact that the surveys were conducted in protected areas. It should be noted that for Bahía Concha there have been reported events of extraction of C. guanhumi, so a study focused on this place is required to evaluate whether the mortality caused by extraction is high here. ^{Costa (2016)} estimated a total mortality Z = 2.31 year^-1, based on monthly catch data and values of AC_∞ and K obtained with tagging and recapture data (108.03 mm and 0.145 year^-1). This high mortality in Costa (2016) could be related to the difference between AC_∞ (108.03 mm) and AC_max (70 mm), which would be interpreted as high mortality in adult individuals. This author did not determine fishing mortality, as she worked in an area where they are not caught for commercial purposes and therefore, assumed Z=M.

From the growth parameters estimated in the present study, and those obtained by ^{Costa (2016)} considering that the marking and recapture data provide more accurate information on growth, it is possible to infer that C. guanhumi is a slow growing and longevous species, with a life expectancy of more than 10 years. These characteristics have also been reported in other large terrestrial crustaceans such as Ucides cordatus, which has been estimated to live up to 17 years (^{Diele and Koch, 2010}).

It should also be noted that differences in growth parameters might not be related only to the methodology used, but also to other factors that may influence growth rate and longevity. These include food availability, habitat quality, and extraction, among others, which may generate differences between populations of the same species. This may explain the differences obtained in the value of K obtained at departmental level with respect to the K obtained for the PNNT. It is also possible that if the localities within the PNNT were evaluated separately, differences in crab growth could be observed between these.

On the other hand, the implementation of capture and recapture methods should be considered in order to obtain more robust data in growth analyses in species such as C. guanhumi characterized by slow growth, longevity and constant recruitment of juveniles (^{Silva et al, 2014}; ^{Costa, 2016}), which may present difficulties when using methods based on size frequencies.

CONCLUSIONS

In general, the C. guanhumi population studied showed negative allometric growth, is longevous, and slow growing, which coincides with what has been previously reported for this species. This is of great importance because when a slow-growing species is subjected to intensive extraction, it requires more time to recover its stocks (^{Silva et al, 2014}). This should be taken into account when devising management plans for the responsible use of this crab. In addition, low fishing mortality (extraction) was obtained, indicating that, from a general perspective for the sampled area, the species is not being overexploited.