Comparative morphology of the first zoea of twelve brachyuran species ( Crustacea : Decapoda ) from the Amazon region

The laboratory-hatched first zoeal stage of twelve brachyuran species collected in the estuarine area of the Caete River in the Amazonian region are described and illustrated in the present study: P. americanus Saussure, 1857, Eurytium limosum (Say, 1818), Sesarma curacaoense De Man, 1892, S. rectum Randall, 1840, Armases rubripes (Rathbun, 1897), Aratus pisonii (H. Milne Edwards, 1837), Ocypode quadrata (Fabricius, 1787), Uca rapax (Smith, 1870), U. maracoani (Latreille, 1802), U. thayeri Rathbun, 1900, Ucides cordatus (Linnaeus, 1763) and Pachygrapsus gracilis (Saussure, 1858). Through intraspecific comparisons of the respective larval stage, an identification key was generated and provided. Most of the studied species presented morphological differences (e.g. type and presence or absence of setae) when compared to the same species previously described in the literature.

Brachyuran crabs are among the most important groups of crustaceans in coastal ecosystems, and one of the most abundant macroinvertebrate groups that inhabit tropical mangroves and tidal flats (CLARK & PAULA 2003).The taxonomic classification of brachyuran crabs is still mostly based on the morphological characteristics of the adults.However, it has been hypothesized that adults share many convergent adaptations to their specific benthic environment (FLORES et al. 2003, CLARK 2009).A valuable, but frequently overlooked alternative to adult characters is larval morphology (CLARK 2009).Taxonomically useful characters found in larvae are believed to be less problematic, because immature forms inhabit a planktonic environment with relatively uniform characteristics (RICE 1980).Phylogenetic studies on brachyuran crabs show that larval morphology is more congruent with molecular data than with adult morphology (HULTGREN et al. 2009).
According to CLARK et al. (1998), detailed, standardized descriptions of larvae obtained in the laboratory are necessary in many kinds of studies.As an example, the patterns of setae on the appendages can be used in systematic analyses because they are very conservative (FLORES et al. 2003, VIEIRA & CALAZANS 2010).

MATERIAL AND METHODS
Field-collecting was conducted in different localities along the mangrove estuary of the Caeté River (Pará, Brazil).Ovigerous females were captured manually and later taken to the laboratory.Next, two females of each species were carefully washed and identified according to RODRIGUEZ (1980) and MELO (1996).Species and their locations, as well as their time of hatching, are shown in Table I.
Morphometric data and illustrations are based on about 10 larvae of each species.Carapace length (CL) was obtained by measuring from the base of the rostral spine to the posterior margin of the carapace; dorsorostral length (DRL) was obtained by measuring from the tip of the rostral spine to the tip of the dorsal spine.Mean values and standard deviation were calculated for each species (Table II).
Table I.Collecting locality and hatching date of species obtained at the Caeté River estuary, Pará, Brazil.

Species
Collecting locality Hatching date

Dec. 2010
The females were isolated and stored in 5 L-tanks containing seawater (salinity = 30) with constant aeration, and subjected to an artificial photoperiod of 12/12 hours until the larvae hatched.
After hatching, individuals of the first larval stage displaying active swimming were preserved in glycerine + 70% ethanol (1:1).Larvae were dissected with fine needles, measured, and then illustrated using a Coleman microscope equipped with camera lucida and a Zeiss Axioskop 40 compound microscope, both with a micrometer eyepiece.For improved visualization some structures were stained using methylene blue (0.5%).The illustrations and morphological characteristics were used to construct a dichotomous key.Characters in the key are based mainly on external characters that can be viewed without dissection.The terminology used in the descriptions follow FACTOR (1978), POHLE & TELFORD (1981), CLARK et al. (1998), RIEGER & SANTOS (2001), GARM (2004), HORN & BUCKUP (2004).Samples of larvae and females were deposited in the Museum Emilio Goeldi (Pará, Brazil)

Aratus pisonii (H. Milne Edwards, 1837)
Carapace (Fig. 42).Globose, smooth with 1 pair of posterodorsal simple setae and 2 pairs of anterodorsal simple setae.Dorsal spine present and curved, with some minute spines on the anterior surface.Mediolateral inferior region with a distinct protuberance.Lateral spines absent.Rostral spine length equal to antennal protopod.Eyes sessile.Antennule (Fig. 43).Uniramous, conic-shaped and unsegmented with 3 aesthetascs and 2 simple setae of different sizes.Antenna (Fig. 44).Protopod well-developed armed with 2 rows of unequal spines.Exopod with 2 unequal simple setae and 2-minute terminal spines.Mandible.Symmetric, palp absent, incisor process with a prominent tooth associated with four small and two teeth arranged separately on the inner margin.Molar process with a large tooth in base and circular region provided of small and irregular teeth.

DISCUSSION
The morphological characteristics of brachyuran larvae, especially in the early stages, are essential to systematic and phylogenetic studies (CLARK et al. 1998), because they help to establish character states as primitive or derived (RICE 1983).
According to RICE (1983), the evolutionary trends of Eubrachyuran larvae (zoea) indicate that more derived taxa show a reduction in spines, setae and segmentation, compared with the more primitive taxa.However, such trends, which are presumably associated with a more efficient exploration of the pelagic environment, may evolve independently.Characters of the larval stages, as well as characters of the adults, are liable to convergence (RIEGER 1998).
Nevertheless, the use of larval characteristics in systematics has been widely accepted and applied by taxonomists (CUESTA & ANGER 2001).Furthermore, as more larvae are being described, morphological comparisons become more feasible (RIEGER 1998).
However, it is important to emphasize that it is difficult to differentiate among the larvae of some species, especially congeners, since their distinctions are based only on minor morphological differences.In such cases, the reliability of specific descriptions may be feasible only when the specimens described are born in the laboratory.Descriptions of larvae collected in the field are often generic due to the difficulty in identifying larvae at the level of species (KORNIENKO & KORN 2009).One example of this problem are Hemigrapsus sanguineus, H. penicillatus and H. longitarsis, described by HWANG et al. (1993), HWANG & KIM (1995) and PARK & KO (2001), respectively.These species have almost identical initial stages, and can only be distinguished when they develop into more advanced stages.
The close similarity observed among species of Hemigrapsus was also noted in the present study for species in the Panopeidae, P. americanus and E. limosum.Even though our description of P. americanus agrees with that by NEGREIROS-FRANSOZO (1986), it differs from it in various aspects.NEGREIROS-FRANSOZO (1986) described some different features, such as a pair of posterodorsal simple setae on the carapace, a small protuberance on the anterior and posterior carapace regions, and an additional small spine on each side of the telson furca; there is also a difference in the aesthetasc, and seta numbers on the antennules compared to those described by NEGREIROS-FRANSOZO (1986) (Table III).There are also discrepancies in number of setae of the coxal endite of the maxillule, and the basial and coxal endites of the maxilla.Other differences found in the description of P. americanus refer to the types of setae originally described.NEGREIROS-FRANSOZO (1986) reported the presence of feathery, i.e. plumose setae on most appendages for which we described plumodenticulate or sparsely plumose setae.This was also observed for other species described in this paper.
When describing Eurytium limosum, KURATA et al. (1981) did not consider morphological and meristic features of several appendages, including the number of aesthetascs and setae on Comparative morphology of the first zoea of twelve brachyuran species ZOOLOGIA 30 (3): 273-290, June, 2013 the antennules, the type and quantity of setae on the maxillule, maxilla, as well as maxillipeds, among other characteristics (Table III).While CLARK et al. (1998) argue that taxonomists should be discouraged to establish relationships based on incomplete descriptions, the description of KURATA et al. (1981) has much in common with the present one, for instance in the abdomen, antenna, telson, and carapace spines.However, E. limosum is very similar to other species of Panopeidae, and larvae can only be distinguished from the second stage of the zoea on, which has the antenna without any spinous process.
A great similarity among larvae of species in the same genus has also been demonstrated for Sesarmidae (GUERAO et al. 2004, CUESTA et al. 2006b, GUERAO et al. 2007).Although the morphology of species of Aratus, Armases and Sesarma is quite similar in the first zoea, these species can be distinguished, in most cases, by minor differences in the pattern of their setae (SCHUBART & CUESTA 1998).Some species, such as S. curacaoense (ANGER et. al. 1995, SCHUBART & CUESTA 1998), are more clearly differentiated, mainly by the presence of pleopods on the abdominal somites and partially differentiated pereiopods (observed but not described in this paper) that are not present in other species.These characteristics, together with the number and pattern of setae, maxillule and maxilla (SCHUBART & CUESTA 1998) (Table III) characterize S. curacaoense morphologically as more derived compared to other species of the genus, such as S. reticulatum and S. rectum (ANGER et al. 1995). Both ANGER et al. (1995) and SCHUBART & CUESTA (1998) described a very conservative pattern of appendages and setae for S. curacaoense, which is consistent with the results of the present study.However, they differentiated the setae only as simple, plumose and plumodenticulate.
Unlike the description of S. curacaoense, our description of S. rectum differs in several respects from the one previously given by FRANSOZO & HEBLING (1986).The latter did not report the presence of a pair of posterodorsal setae, and two pairs of simple setae on the anterodorsal portion of the carapace, as well as the following features: two rows of unequal spines on the antennal protopod, cuspidate and plumodenticulate setae on the basial endite, and plumodenticulate and simple setae on the coxal endite of the maxillule.Further differences are present in various structures, particularly in the maxilla, where the type or the number of setae arranged on the basial and coxal endites vary (Table IV).Discrepancies in the description of those characters, according SCHUBART & CUESTA (1998), are generally due to subject interpretations of various taxonomists, regional, or intraspecific variation.However, caution is needed when making observations on those variations, because descriptions are often based on features commonly regarded as highly conservative, as the pattern of setae on the appendages (FLORES et al. 2003).
According to CLARK et al. 1998, many recent studies have failed to provide detailed descriptions of some features, mainly involving the pattern of setae of the appendages.This is more common in older contributions, including the description of A. rubripes by DÍAZ & EWALD (1968).The morphological differences between that description and the present study are likely due to DÍAZ & EWALD's failure to observe the two unequal rows of spines on the antennal protopod, an additional seta on the antennule and coxal endites of the maxillule and maxilla, among other features (Table IV).Possibly the lack of criteria for publishing a description has resulted in such discrepancies.
ZOOLOGIA 30 (3): 273-290, June, 2013 Although the descriptions by FRANSOZO et al. (1998) andCUESTA et al. (2006a) are relatively recent, both differ in several aspects from the present description (Table IV), particularly when it comes to setal types.This may be the reason why CUESTA et al. (2006a) have not considered it.CUESTA et al. (2006b) re-examined the samples described by FRANSOZO et al. (1998), and found several differences between the two descriptions.These data, together with those presented in this study, point to the existence of interspecific variability (CUESTA et al. 2006b).Thus, morphological consistency among populations of A. pisonii may be questioned, and highlights the need for studies that address phylogenetic aspects to assist the resolution of disagreements regarding the morphology of this species.
Although not using current descriptive standards, DIAZ & COSTLOW (1972) described in detail the morphological features of O. quadrata larvae.Their description did not differ significantly from the present one, but lacked some information, such as the presence of a pair of posterodorsal simple setae on the carapace, an additional simple seta on the coxal endite of the maxillule and two minute spines on the antennal exopod (Table V).These differences may have to do with methodological limitations in the preparation of the larvae for microscopy, or result from optical limitations, possibly impairing the description of the appendages.However, other plausible reasons such as population and interspecific variation cannot be ruled out.(CE) coxal endite, (BE) basial endite, (s) simple seta, (pls) plumose seta, (pds) plumodenticulate seta, (cps) cuspidate seta, (aes) aesthetascs, pl sp (plumose spine), (sp) spine, (spd) plumodenticulate seta, sers (serrate seta), (ser sp) serrated spine, (sps) sparsely plumose seta, (hs) seta similar to the hamate, (nd) not described, (?) not defined.
The recent description of U. maracoani by NEGREIROS-FRANSOZO et al. (2009) differs from ours in a number of ways, often related to the omission or classification of setal types, such as coxal and basial endites of the maxillule and maxilla, carapace and abdominal somites (Table V).Such inconsistencies have raised questions about the stability of the morphological pattern of this species, establishing the need to investigate possible population variations.
In describing the first larval stage of U. rapax, SERBINO (2008) mentioned more details than NEGREIROS-FRANSOZO et al. (2009) in their description of U. maracoani, and description of U. thayeri by ANGER et al. (1990).However, her description is somewhat different from that in the present study, particularly regarding the number of setae and aesthetascs on the antennule, the types of setae on the basial endite of the maxilla, the coxal and basial endites of the maxillule (Table V), and the type of setae on the basis and endite of the first and second maxilliped.
According ANGER et al. (1990), there is great need for further laboratory investigations on the larval stages of species of Uca.Such investigations would contribute to the clarification of the taxonomic relationships of the group, and allow the identification of larvae in ecological studies.SERBINO (2008) noted that many species of this genus have unique characteristics, while others species have variable characters, and some do not have any features that can be used to distinguish them.
We also found several differences with respect to the description of Uca thayeri by ANGER et al. (1990).However, these were not as pronounced as for U. maracoani.The disagreements involve mainly the terminology used for setae of the maxillule and maxilla, and the number of setae on the antennules, coxal endite of the maxillule and coxal and basial endites of the maxilla (Table VI).
The description of U. cordatus by RODRIGUES & HEBLING (1989) does not differ much from that in the present study.The former did not report the presence of a single seta on the antennule and a small lateral spine on each furca, which also have serrated margins (Table VI).Although few, these differences may interfere with the identification of specimens collected in the plankton, which may hinder or even jeopardize comparative morphological or ecological studies of zooplankton.
Among all the larval descriptions analysed in the present study, we found that the one of P. gracilis by BROSSI-GARCIA & RODRIGUES (1993) differed the most (Table VI).This relates not only to inadequate descriptions of some appendages, but also to discrepancies in the patterns of setae.
According to CLARK et al. (1998), the format of descriptions of brachyuran larvae may vary considerably among authors.But in some cases this variation is clear, establishing a morphological pattern, which is not very satisfactory for comparative studies.Therefore, in addition to more detailed studies on the larval morphology relating to variability, we also need the help of molecular analyses to clarify certain taxonomic relationships.
Despite these shortcomings, the present study provides additional information on the morphology of brachyuran larvae to assist in the identification of plankton samples from different regions, and to update knowledge on comparative morphology of larvae among the groups investigated here.

Table II .
Measurements of zoea I of the species described in this study.

Table III .
Morphological comparison between the present study and previous descriptions of Panopeus americanus., Eurytium limosum and Sesarma curacaoense.

Table IV .
Morphological comparison between the present study and previous descriptions of Sesarma rectum, Armases rubripes and Aratus pisonii.

Table V .
Morphological comparison between the present study and previous descriptions of Ocypode quadrata, Uca maracoani and Uca rapax.Comparative morphology of the first zoea of twelve brachyuran species ZOOLOGIA 30 (3): 273-290, June, 2013Table VI.Morphological comparison between the present study and previous descriptions of Uca thayeri, Ucides cordatus and Pachygrapsus gracilis.