An integrative approach : using fi eld and laboratory data to characterize shell utilization and selection patt ern by the hermit crab Pagurus criniticornis ( Paguridae ) from Anchieta Island , Brazil

Th e aim of this study was to characterize the patt ern of gastropod shell occupation in the fi eld and selection of shell size and type under laboratory conditions by the hermit crab Pagurus criniticornis (Dana, 1852), inhabiting the infralitt oral area of Anchieta Island, São Paulo, Brazil. Hermit crabs were obtained monthly during 1999 by SCUBA diving. For experiments under laboratory conditions, samplings were performed in 2002. Th e hermit crabs occupied 16 species of gastropods shells. Cerithium atratum (Born, 1778) was the most occupied shell (89.31%), followed by Morula nodulosa (4.73%) (Adams, 1845). No diff erence was observed in the patt ern of occupation between males and females. Th e equations that best demonstrated the relationship between hermit crabs and their shells were those that involved Shell Wet Weight (SWW) and Shell Internal Volume (SIV). Th e laboratory experiments were in accordance to the patt ern of occupation observed in the fi eld; the mean value of SAI (Shell Adequacy Index) recorded to the population studied was 1.13 with a trend to increase this value in the last size classes. Th e results obtained corroborate with the hypothesis of the CORRESPONDING AUTHOR Fernando L. Mantelatto mantel@usp.br SUBMITTED 13 January 2016 ACCEPTED 11 May 2016 PUBLISHED 9 September 2016 Guest Editors Maria Lúcia Negreiros-Fransozo and Adilson Fransozo DOI 10.1590/2358-2936e2016011 orIgInal artIcle This article is part of the special series offered by the Brazilian Crustacean Society in honor to Nilton José Hebling in recognition of his dedication and contributions to the development of carcinology in Brazil.

In natural populations, the pattern of occupation of gastropod shells observed by hermit crabs is dependent of many factors such as the local availability of different species of gastropods, the presence of sympatric species or by the hermit crabs' preference to certain species of gastropods shells (Bertness, 1981a;Dominciano et al., 2009).However, in complex communities compounded by more than two coexistent species, the explanations regarding the mechanisms responsible by the shell choice and/or shell preference remain unclear and are not satisfactory.The reasons, in most cases, are due to the absence of complete methods of evaluation, including natural and laboratory observations and the absence of knowledge regarding the resource (live gastropods and empty shells).
The objective of the present study was to determine the pattern of occupation and selection of gastropod shells by P. criniticornis population inhabiting the infralittoral of Anchieta Island, São Paulo, Brazil.We characterized morphometric relationships and the pattern of selection for shell size and type under laboratory conditions, in order to provide subsidy to elucidate questions about the dynamics of the hermit crabs' community on this island and investigate the hypothesis of occupation conducted by availability resource (shell) associated to choice mechanism of gastropod shells by P. criniticornis.

Study area
Located in Ubatuba city, Anchieta Island (23°33'S 45°05'W) is the second largest island of the north coast of São Paulo State, with a total area of about 10 km 2 .This island has been affected by anthropogenic activities such as tourism and fishery exploitation until 1977, when it was declared an ecological reserve of São Paulo State (see Mantelatto and Garcia, 2002 for details).

Shell occupation pattern
The samples were obtained in the infralittoral rocky/ sandy area from January to December 1999.The hermit crabs were captured by two persons using SCUBA diving methods during 30 min over the same area of about 800 m 2 , at 9-10 m deep.After collection, the animals were frozen and transported to the laboratory where they were fixed in 80% alcohol for later analysis.Animals were removed from their shells, counted, measured (CSL = Cephalothoracic Shield Length), and weighed (WW = Wet Weight).Shell species occupied by these individuals were identified according to Rios (1994) and the taxonomic nomenclature revised by a specialist, measured (SAW = Shell Aperture Width and SAL = Shell Aperture Length) and their wet weight (SWW) was determined.Shell internal volume (SIV) was measured by the method of Bertness (1981 b) modified by Mantelatto and Garcia (2000) by filling the empty shells with fine sand (Ø = 0.25 mm 3 ) with a pipette (0.01 mm 3 ).The voucher crabs were deposited in the Crustacean Collection of the Biology Department (CCDB) of Faculty of Philosophy, Science and Letters at Ribeirão Preto (FFCLRP), University University of São Paulo (USP) (catalogue numbers: from CCDB 1476 to 1484 and 1741 to 1745).
The mean size of individuals of both sexes was compared by the Mann-Whitney test (Zar, 1996).The level of significance was 0.05.
The relations between hermit crabs size and shell variables were determined by regression analysis and by correlation coefficients.The chi-square test (χ 2 ) was used to compare occupancy percentage of shell species between males, ovigerous females and nonovigerous females.

Shell selection under laboratory conditions
After the knowledge of the population profile (Mantelatto and Garcia, 2002;Mantelatto et al., 2007) and of the pattern of shell utilization by this population in the field, two types of laboratory experiments were performed: shell species and size preferences, which were conducted for the most occupied shell types in the field.For this, samples were additionally carried out during 2002 in the same area, following the above described method of collection.
Three replicates were conducted for each type of experiment, which occurred in a glass aquarium (30x30x40 cm) with flowing and oxygen-saturated seawater, where the animals chosen independently of sex or size (N = 15) were placed naked with a large number of shells (N = 150), during 72 hours (time previously established).After that, the hermit crabs were removed from the preferred shell, and both were measured as described above for shell occupation.Each animal was used only once in order to avoid any acquired behavior.
The shell species experiments were conducted in pair-wise fashion, i.e., the hermit crabs were placed in the aquarium with empty shells of two different species and similar size (75 of each species) than those found in the field.Shell size preference was tested separately for most occupied shell species using shells of the same species but of varied sizes.All experiments were performed according to method described by Garcia and Mantelatto (2001).
Shell species preference was estimated based on the frequency of which species was chosen by the individuals.The chi-square test (χ 2 ) was used to compare occupancy percentage of shell species.Morphometric relationships were established by regression analysis and Spearman correlation coefficients (r s ) (p < 0.05).
Shell fitness of hermit crabs collected in the field was assessed using a shell adequacy index (SAI) (Vance, 1972a).This index is defined as: SAI = suitable crab size/actual crab size, where SAI = 1 indicates that, in field, the animal occupies a shell of adequate size while SAI ≠ 1 indicates a crab in a shell larger (SAI > 1) or smaller (SAI < 1) than adequate size.
Shell species occupation as a function of hermit crab size is illustrated (Fig. 1).The diversity of shells utilized decreased with the increase of individuals' size.Cerithium atratum was the most occupied shell in all size classes, except in the first class, where this corresponded to 12.5% of occupied shells.There was no difference between species of shell occupied by males, ovigerous females and non-ovigerous females.
The relationship between the shells and the hermit crabs was tested only to C. atratum and M. nodulosa, due to a low percentage of occupation for the other species of gastropod shells.The equations that best demonstrated this relationship were those that involved shell wet weight (SWW) and shell internal volume (SIV) (Tab.2), mainly to the most occupied shell species, C. atratum.

Experiments
The experiments were conducted only for most the occupied species of shell, i.e., C. atratum and M. nodulosa.

P. criniticornis, shell utilization and selection pattern
Nauplius, 24: e2016011  All relations between dimensions of hermit crabs and selected shells were significant (p < 0.05), but P. criniticornis showed a preference strongly associated with SIV and SWW.The higher correlation coefficients demonstrated that C. atratum shells are most adequate to the various hermit crab sizes (Tab.4).The pattern of shell choice of males, non-ovigerous females and ovigerous females was not different for both species of shells (Tabs.5, 6).

Shell Adequacy Index
The mean SAI value found to P. criniticornis population was 1.13 ± 0.25.Shells occupied by hermit crabs in the field were larger than adequate size (SAI > 1) in the most size classes; however, larger individuals showed a trend to occupy shells of adequate size (SAI = 1).
There were no differences in this pattern observed between males, non-ovigerous females and ovigerous females (Fig. 2).For the two most occupied shell species tested the pattern of adequacy was similar,    i.e., SAI > 1 in the first size classes and this adequacy increased as the hermit crabs reached higher size classes (Fig. 3).

dIscussIon
Pagurus criniticornis was found occupying a large set of 16 gastropod shell species.Despite this diversity of occupied resource, there was not plasticity in shell type utilization, since C. atratum was the supranumerary occupied shell, with 89.31% of rate of occupation.
Cerithium atratum was also one of the most frequently occupied shells by other small hermit crabs in this area, such as Paguristes tortugae (studied by Mantelatto and Dominciano, 2002), Pagurus brevidactylus (Mantelatto and Meireles, 2004), Calcinus tibicen (Mantelatto and Garcia, 2002) and Pseudopaguristes calliopsis (Biagi et al., 2006) in Anchieta Island area.This clear coexistence and the low percentage of damaged shells occupied by P. criniticornis demonstrated the good shell stock potential in this area, as accounted by Mantelatto and Garcia (2002), Meireles et al. (2003) and Mantelatto and Meireles (2004).This resource stock is one of the most

P. criniticornis, shell utilization and selection pattern
Nauplius, 24: e2016011 important condition that allows the stable coexistence between species and do not cause a collapse in the use of resource and effects on population characteristics.The pattern of shell utilization by hermit crabs is a result of a complex interaction among availability of shells (Reese, 1969;Bertness, 1980;Mantelatto and Garcia, 2000;Mantelatto and Dominciano, 2002;Mantelatto and Meireles, 2004) and selection of this resource, considering that hermit crabs can discriminate between different species and size of shells (Reese, 1962;1963;Markham, 1968;Young, 1979;Hazlett, 1996;Mantelatto and Meireles, 2004).
Shell occupation pattern was more diverse in the initial size classes of P. criniticornis.According to Bollay (1964), individuals of smallest size tend to occur in most variety of small shells, since these are not heavily dependent on species of shells to survive and reproduce.
Patterns of shell utilization vary between hermit crab populations and are influenced by the type and size of shells available in the survey, the locality and hermit crabs' shell preference (Mantelatto and Garcia, 2002).Lively (1988) and Osorno et al. (1998) affirmed that hermit crabs prefer shells which the architecture maximizes their internal space.The present study corroborated this hypothesis, considering the good correlation values found between hermit crabs and C. atratum shells, mainly those related with weight and internal volume.
The availability of empty shells is a limiting factor to hermit crab populations, which is greatly influenced by the relative abundance of different live gastropods, their mortality rates (Bertness, 1980;Ohmori et al., 1995) and by inter/intraspecific competition (Reese, 1969).Meireles et al. (2003) studied the availability and diversity of local resources in Anchieta Island areas, where C. atratum was the most abundant shell, followed by Pisania auritula (Link, 1807), Astraea olfersii (Philippi, 1846) and M. nodulosa.
The shell occupation pattern of P. criniticornis showed similarity with availability of shells in the area; however, the rate of occupation of C. atratum and M. nodulosa by P. criniticornis was not proportional to the rate of availability found by Meireles et al. (2003), suggesting the preference for C. atratum as commented by Meireles et al. (2008).The experimental results confirmed the strong preference of individuals of this population for C. atratum shells, independently of sexes.This allows us to infer that this population's dynamics is directly related with C. atratum occupation.
The mean SAI value found for the population of P. criniticornis was 1.13, which means that, in general, these animals utilize shells larger than the suitable size.However, value of shell adequacy index decreased with increase of crab size showing that the larger individuals were occupying relatively adequate shells.Meireles and Mantelatto (2005) observed similar pattern for the coexistent population of Pagurus brevidactylus from Anchieta Island.
The occupancy of larger shells than the adequate size by small individuals can be caused by an insufficient availability of small shells (Gherardi et al., 1994).According to Mantelatto and Meireles (2004), small shells represent a limited resource to the hermit crabs from Anchieta Island.Moreover, C. atratum was occupied by five of nine hermit crabs species that live in this area (Mantelatto and Garcia, 2002).In this case, the acquisition of shells larger than the adequate size is less strenuous than a continuous search for ideal sized shells.
In this way, the occurrence of individuals occupying larger shells than adequate ones is advantageous, since the interaction between specimens of the same population is more frequent than the probability of finding new and empty shells (Spight, 1977;1985).These interactions allow changes that lead to better adequacy of individuals to their shells (Vance, 1972b) and consequently increase the potential of population growth.Thus, we are convinced that the pattern of adequacy observed in the studied population would be a satisfactory way to minimize the interspecific competition that it is exposed to.
When observed in an intertidal habitat, P. criniticornis showed abilities as explorer competitor and to acquire resources more rapidly than other species (Turra and Denadai, 2004).In this intertidal area, despite showing a preference for some species of gastropods, P. criniticornis does not seem to present such a rigorous pattern for selecting shells of higher quality or size, using mostly the shells that are available in the area, even if eroded or damaged (Turra and Denadai, 2004).Comparing these results with the obtained here, we can infer that differences are due to the resource availability and competition by coexistence species in the island P. criniticornis, shell utilization and selection pattern Nauplius, 24: e2016011 assemblage, which make us to argue in favor of a future comparative study between both populations in order to elucidate this scenario.
The results obtained here indicate that the pattern of occupation found to P. criniticornis population is probably conducted by the availability of shells on the locality and by hermit crabs' choice of this resource, considering the relatively abundance of C. atratum on studied area and, additionally, the good adequacy of shell and hermit crab morphological characteristics, which is evidenced by high correlation coefficients, being, however, the second factor more influent under the determination of this pattern.
In recent studies, it was corroborated that in P. criniticornis population, there is a range of factors that influences shell exchange, and that some of them could represent trade-offs between immediate and long-term survival (Buranelli et al., 2015).In other words, these authors postulated that hermit crabs face different trade-offs related to the behavior of choice, selection, and exchange of shells, indicating that different factors may affect behavior, depending on many aspects as environmental features, sex, individual size, shell size, and shell quality.

acKnowledgeMents
The present study is part of the research project on hermit crab fauna from Anchieta Island and the major funding was provided by a grant to FLM by Fundação de Amparo à Pesquisa do Estado de São Paulo -FAPESP (Proc. 1998/07454-5).This manuscript was also part of Master Dissertation project by FCRF, which is indebted to Conselho Nacional de Desenvolvimento Científico e Tecnológico -CNPq for a fellowship.RB and ALM are indebted to FAPESP (Procs. 1997/14245-0 and2000/02554-3;1999/11679-5 and2002/01646-7) for Master and Doctoral fellowships, respectively.FLM is grateful to CNPq for Research on-going fellowship (304968/2014-5).We thank Secretaria do Meio Ambiente do Estado de São Paulo, IBAMA and Parque Estadual da Ilha Anchieta for permission (Proc.42358/98) and facilities during the sampling work.Special thanks are due to members of the Laboratory of Bioecology and Crustacean Systematics of FFCLRP/USP for help during the field and laboratory work, to the Programa de Pós Graduação em Biologia Comparada -FFCLRP/USP for support and assistance, and to Helena Matthews-Cascon (Universidade Federal do Ceará) for help with gastropod shell taxonomic revision.All experiments conducted in this study comply with current applicable state and federal laws.The authors are honored to recognize the many achievements of our friend and colleague, Nilton J. Hebling, by contributing to this commemorative issue.

Figure 2 .
Figure 2. Pagurus criniticornis.Variation of the mean shell adequacy index (SAI) values for total of population, males, non-ovigerous females and ovigerous females.(Data are expressed as mean ± SD)

Figure 3 .
Figure 3. Pagurus criniticornis.Variation of the shell adequacy index (SAI) values for each most occupied shell species in relation to the size classes.(Data are expressed as mean ± SD).

Table 2 .
Pagurus criniticornis.Regression equations for measures between hermit crab (CSL = Cephalothoracic Shield Length and WW = hermit crab Wet Weight) and the two most occupied species of shells (SAW = Shell Aperture Width; SAL = Shell Aperture Length; SWW = Shell Wet Weight; SIV = Shell Internal Volume).[N = number of individuals; r s = Spearman correlation coefficient].

Table 3 .
Pagurus criniticornis.Laboratory shell species preference observed among the two most occupied in the field.

Table 5 .
Pagurus criniticornis.Regression equations for measures between hermit crab and the shells of Cerithium atratum chosen under laboratory conditions, according to sexes.(CSL = Cephalothoracic Shield Length; WW = hermit crab Wet Weight; SAW = Shell Aperture Width; SWW = Shell Wet Weight; SIV = Shell Internal Volume; N = number of individuals; r s = Spearman correlation coefficient).

Table 6 .
Pagurus criniticornis.Regression equations for measures between hermit crab and the shells of Morula nodulosa chosen under laboratory conditions, according to sexes.(CSL = Cephalothoracic Shield Length; WW = hermit crab Wet Weight; SAW = Shell Aperture Width; SWW = Shell Wet Weight; SIV = Shell Internal Volume; N = number of individuals; rs = Spearman correlation coefficient).