Processes regulating early post-settlement habitat use in a subtidal assemblage of brachyuran decapods

gy and Ecology 344 (2007) 10–22www.elsevier.com/locate/jembe

Journal of Experimental Marine Biolo

Processes regulating early post-settlement habitat use in a subtidalassemblage of brachyuran decapods

L. Miguel Pardo a,b,⁎, Alvaro T. Palma b,c, Catalina Prieto c, Patricia Sepulveda c,Isabel Valdivia c, F. Patricio Ojeda b,c

a Universidad Austral de Chile, Instituto de Biología Marina, Laboratorio Costero Calfuco, Casilla 567, Valdivia, Chileb Center for Advanced Studies in Ecology and Biodiversity, Pontificia Universidad Católica de Chile. Alameda 340 Casilla 114-D, Santiago, Chile

c Departamento de Ecología, Pontificia Universidad Católica de Chile, Alameda 340 Casilla 114-D, Santiago, Chile

Received 30 June 2006; received in revised form 10 November 2006; accepted 15 December 2006

Abstract

In highly mobile animals post-settlement dispersion of juveniles can strongly influence the observed patterns of abundance anddistribution. To explore the relative importance of factors regulating the use of habitat by crabs we performed a multi-speciesmanipulative experiment in a subtidal environment of the central Chilean coast. First, demographic patterns were established byperforming a year-round crab survey in three discrete and well-known subtidal crab habitats: (1) algal turf, (2) cobbles and (3) shellhash. Second, habitat preferences were experimentally evaluated using concrete trays that were filled with different substrate typesthat simulate natural habitats. Settlement and recruitment rates were estimated from experimental trays that were left in the field andsurveyed after 2 weeks (complete experiment was repeated 7 times throughout 1 year). Third, mortality, due to predation, wasassessed by covering 50% of the trays with a 4-mm mesh-size screen that excluded large predators (i.e., fishes, shrimps). Fourth,habitat colonization rates were evaluated by quantifying the arrival, into open trays, of large juveniles (secondary dispersal). Themost abundant species in this system (Paraxhantus barbiger, Cancer setosus, Taliepus dentatus and Pilumnoides perlatus)displayed clear habitat preferences at the time of settlement, evidenced by differences in density of recruits among habitats.Recruitment regulation by predation seemed to explain the observed patterns in only one case. For most species, however, evidenceof ontogenetic change in the use of habitat, through active habitat redistribution by large juveniles, was detected. Thus, secondarydispersal among habitats seems to outweigh the influence of megalopae's habitat selection and post-settlement mortality asresponsible for the observed demographic patterns.© 2007 Elsevier B.V. All rights reserved.

Keywords: Exclusion experiment; Habitat selection; Migration; Recruitment

⁎ Corresponding author. Universidad Austral de Chile, Instituto deBiología Marina, Laboratorio Costero Calfuco, Casilla 567, Valdivia,Chile. Tel./fax: +56 63 22 14 55.

E-mail address: [email protected] (L.M. Pardo).

0022-0981/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.jembe.2006.12.024

1. Introduction

For marine benthic invertebrates undergoing mer-oplanktonic development, recruitment (i.e., the transi-tion between pelagic and benthic stages) has beenrecognized as an influential driving factor in the spatialorganization of populations (Roughgarden et al., 1988).Recruitment can be limited by larval supply if settlement

mailto:[email protected]

http://dx.doi.org/10.1016/j.jembe.2006.12.024

11L.M. Pardo et al. / Journal of Experimental Marine Biology and Ecology 344 (2007) 10–22

rates correlate with the abundance of early post-settle-ment stages or regulated by density-dependent factors ifhigh mortality occurs soon after settlement (Underwoodand Keough, 2001). When populations are recruitment-limited, three interacting factors are frequently invoked:(1) selection of settlement habitat by competent larvae,(2) habitat-specific post-settlement mortality, and (3) ac-tive secondary dispersion among habitats by juveniles.The relative importance of these factors, likely to affectthe abundance and habitat distribution patterns, can varydepending on the ecological and life history traits oftarget species in a given environment.

Substrate selection represents a critical behaviorduring settlement, particularly for many sessile organ-isms because the “right choice” can have a superlativeimpact on their fitness (Strathmann et al., 1981; Herbertand Hawkins, 2006). In fact, much of the early literaturesupporting recruitment–regulation considered sedentaryadults such as polychaetes,mytilids and barnacles (KeoughandDownes, 1982; Connell, 1985;Menge and Sutherland,1987). In contrast,mobile species have the option of furtherhabitat exploration after settlement; in consequence, ha-bitat selection for settlement should be less critical. How-ever, even for highly mobile species, post-larval stages canalso exhibit strong habitat selection (e.g. Underwood,2004; Forward et al., 2005). Thus, habitat selection at thetime of settlement could have an even deeper influence onthe patterns of distribution and abundance of early benthicstages (Wahle and Steneck, 1992).

On the other hand, during and after settlement, mero-planktonic organisms are subject to high predation pres-sures from large benthic invertebrates and demersalfishes (Wahle and Steneck, 1992; Stachowicz and Hay,1996; Tupper and Boutilier, 1996; Jones, 1997; Palmaet al., 1998). Such strong interactions result in high mor-tality rates of settlers and young juveniles (Gosselin andQian, 1997). If different habitats used by post-settlementindividuals provide differential protection against pre-dation, then a discrete spatial distribution would be areflection of habitat-specific mortality. In this case, pre-dation would exert a strong direct influence on the prey'sdemography.

Mobile animals can respond to predation pressuresby selecting habitats with lower risk, typically structur-ally complex ones (i.e., Palma et al., 1998). Post-settle-ment migration might occur due to varied predationvulnerability during ontogeny, which also implicateschanges in the value of habitats as refuge (Richards,1992; Pardieck et al., 1999). Thus, differential habitatutilization due to juvenile movements appears as a strongfactor explaining demography at both local and regionalscales (Palmer et al., 1996).

In the case of crabs, the demographic influence ofthese three factors (habitat selection by megalopae,habitat-specific variation in mortality rates of settlersand juvenile dispersal) have been studied under labo-ratory conditions (Hedvall et al., 1998) as well as in thefield (Fernández et al., 1993; Eggleston and Armstrong,1995; Moksnes, 2002; Moksnes and Heck, 2006). Thesehave been mainly single-species studies, thus makinggeneralizations difficult, especially when even relatedspecies can exhibit contrasting life history traits. Forexample, post-settlement movement is expected to berelatively more important in active predators comparedto detritivorous, less active species, with reduced foragingbehavior.

Around 60 species of brachyuran decapods can befound along the central coast of Chile (Lancellotti andVásquez, 2000). Many of them are of commercial valueand represent important components of the shore foodwebs (Morales and Antezana, 1983; Ojeda and Fariña,1996; Fernández and Castilla, 1997; Palma and Ojeda,2002; Fariña et al., in press). In spite of their diversityand abundance, information on processes affecting theirpopulation dynamics (i.e., recruitment, distribution pat-terns) is very limited (Jesse and Stotz, 2003), especiallyfor early stages (but see Fernández and Castilla, 2000;Palma et al., 2003, 2006).

The species Paraxhantus barbiger, Cancer setosus,Taliepus dentatus and Pilumnoides perlatus are amongthe most conspicuous and abundant brachyuran speciesinhabiting the shallow rocky subtidal of central Chile(Table 1). In this study, we first assessed the recruitmentpatterns of these species at both temporal (monthly) andspatial (three different habitats) scales then, using a ma-nipulative field experimental approach, we explored thelikely processes responsible for the patterns of abun-dance and distribution of megalopae, recruits (crab I andII instars) and juveniles of these species.

2. Methods

2.1. Study area

The study was conducted in the rocky subtidal ofPunta de Tralca (33°35′S, 71°42′W), a small bay lo-cated in the central coast of Chile, protected from theprevailing strong southerly winds. The nearshore bot-tom of the bay (between 5 and 20m in depth) is dominatedby three main microhabitats: (1) algal turf, principallycomposed of filamentous red algae nomore that 5 cm highgrowing on flat rocky surfaces; (2) boulders and cobbleson top of shell hash; and (3) shell hash (Fig. 1). Therelative proportion of these different types of habitats was

Table 1Life history traits of four crabs species present along the central coast of Chile

Family Bathymetricrange (m)

Adult sizeCW (mm)

Megalopa sizeCL (mm)

Adult habitat Diet References

Paraxhantus barbiger(Poepping, 1836)

Xhantidae 0–28 70 3.0 Rock and shell hash Detritus 1, 2, 3, U.D.

Cancer setosusMolina, 1782

Cancridae 0–25 107 3.3 Sand and rocks Invertebratesand carrion

1, 2, 4, 5, 6, 7

Pilumnoides perlatus(Poepping, 1836)

Xhantidae 0–54 24 1.4 Secondary substrate Detritus 1, 2, 8, U.D.

Taliepus dentatus(H. Milne Edwards, 1834)

Majidae 0–22 93 0.8 Secondary substrate Algae and epiphytes 1, 2, 3, 9, 10

References: (1) Lancellotti and Vásquez (2000); (2) Retamal (1999); (3) Palma et al. (2006); (4) Cerda and Wolf (1993); (5) Leon and Stotz (2004);(6) Wolff and Cerda (1992); (7) Quintana and Saelzer (1986); (8) Fagetti and Campodonico (1973); (9) Fagetti and Campodonico (1971);(10) Manriquez and Cancino (1991); U.D.=unpublished data.CW=carapace width; CL=carapace length.

12 L.M. Pardo et al. / Journal of Experimental Marine Biology and Ecology 344 (2007) 10–22

estimated by recording their presence every 1 m depth,along four parallel 100-m-long transects 10 m apart.

2.2. Field crab recruitment estimation

To evaluate the habitat-specific distribution and abun-dance of different ontogenetic stages of four crab speciesthe three microhabitats mentioned above were system-atically surveyed from May 2004 to May 2005 (seventrials 60–70 days apart). Each time every microhabitatwas surveyed by haphazardly placing 4 to 5 squareframes (0.25 m2) and all the material and individualscontained within were suctioned with an airlift deviceinto a 1-mmmesh-size catch bag (based on methodology

Fig. 1. Location of the study site and percent cover of th

by Wahle and Steneck, 1991). This methodology haspreviously demonstrated to be suitable for the estimationof the abundance of post-larvae, settlers and juveniles ofbrachyuran decapod crustaceans, particularly for thesesame species (Palma et al., 2003, 2006) which do notdisplay swimming behaviour as other species do whendisturbed (e.g. Callinectes sapidus). In the laboratory,samples were analyzed and individuals identified andmeasured. Only the four most abundant and persistentbrachyuran species (P. barbiger, C. setosus, T. dentatusand P. perlatus) were considered for this study.

To determine ontogenetic changes in habitat use bythese species, a multiple logistic regression analysis wasperformed using the CATMOD procedure in the SAS

e different microhabitats considered in this study.


statistical package (Stokes et al., 1995). Logistic re-gressions represent the most suitable approach since itmodels a polynomial categorical dependent variable(e.g. three levels of habitat use) with the independentvariables (i.e., size classes). When the logistic regressionwas significant, a pair-wise comparison between sizeclasses was performed using the PROCFREQ procedureimplemented in the SAS statistical package. Given theelevated number of comparisons, we obtained the cri-tical alpha values for significances using the Bonferronicorrection (Sokal and Rohlf, 1995). Data from all trials(7) were pooled to obtain an integral size frequency foreach microhabitat and species using size–frequencyhistograms.

2.3. Experimental crab recruitment estimation

Habitat preferences of megalopae and recruitswere experimentally evaluated using concrete trays(0.5×0.5×0.2 m). These trays represent discrete andindependent experimental units, which were filled withdifferent substrates that simulated natural microhabitats.Algal turf was simulated using 1-cm-long artificialgrass. Cobble microhabitat was simulated using naturalcobbles (0.1 to 0.2 mmaximum diameter) obtained froma quarry. The shell hash utilized (2 kg per tray of 0.1- to0.2-cm-sized shell) was obtained from the same area butthermally sterilized prior its use. In total, 30 trays (10units assigned to each microhabitat) were deployedwithin 0.25 ha between 10 and 12 m in depth. Trayswere haphazardly placed on a shell hash bottom at least5 m apart.

To assess the importance of predation as a factoraffecting the abundance and distribution of early benthiccrab stages, we used an exclusion treatment with twolevels (open and closed trays). Half of the trays for eachsimulated microhabitat type were closed with a stainlesssteel screen of 4-mm mesh opening. This screen ex-cluded large predators, such as fish or shrimps, but re-presented no barrier for newly settled crabs (megalopaeand recruits). Other potential small predators, like juvenilesof rock shrimp Rhychocinetes typus were able to enter thetrays as well. Although we did not directly evaluate theability of newly settled crabs to pass across the screen, thepreliminary results of this experiment showed higherabundance inside the closed trays compared to the openones, indicating that a screen effect, if existing, is ne-gligible. In addition, other similar studies on crab settle-ment that used exclusion cages found no significant screeneffects (Palma et al., 1998; Moksnes and Heck, 2006).

Since trays could represent shelters for crabs as wellas for their potential predators, additionally we deployed

10 empty concrete trays that served as controls. Five ofthem were covered with the same stainless steel screento evaluate the possible influence of trays as attractorsfor individuals without the influence of predators.

Seven trials were performed during the year, appro-ximately one every other month. Trays with artificialsubstrata were first surveyed approximately 2 weeksafter they were installed using the same airlift suctioningtechnique as described above. The timing of surveysvaried depending on sea conditions.

From the two different surveys (natural substrataand experimental trays) we considered three ontogeneticstages for each species: (1) megalopae, (2) recruits and(3) juveniles. Megalopae were identified from descrip-tions in the literature, when available (i.e., Fagetti andCampodonico, 1971, 1973; Quintana and Saelzer, 1986).Unknown megalopae were identified by allowing them tomolt into juvenile stages under laboratory conditions. Weoperationally defined recruits as those individuals thatsurvived the pelagic–benthic transition and became firstand second crab instars (CI and CII). These instars weremeasured (carapace width) and identified after rearingindividuals in the laboratory. Juveniles were those re-maining from the samples obtained from the 4-mm meshscreen-covered trays. Only P. perlatus displayed sexualmaturity at a smaller size (Pardo, unpublished data), how-ever, we did not separate these stages from our analysesin order to keep a suitable sample size for statisticalpurposes.

Differences in crab abundance from experimental trayswere tested separately for each species and ontogeneticstage. Mixed-model three-factor ANOVAs were per-formed in each case. Independent variableswere exclusion(open or closed) and microhabitat (artificial grass, cobbleand shell hash), which were considered fixed factors. Trialwas considered as a random factor. Before running theANOVAs, Cochrane's test was used to evaluate assump-tion of homogeneity, when this was not reached; data were(lnX+1) transformed. The Student–Newman–Keuls(SNK) test was used for post-hoc comparisons when sig-nificant differences were found in themean factors or theirinteractions. To include possible predator (i.e., fishes andshrimps) aggregation effect on the abundance of mega-lopae and early benthic crab stages, we recorded the abun-dance of predators on each tray and they were used ascovariate in our analyses. However, the covariates werenever significant (PN0.1), thus they were eliminated fromthe analyses. Analyses for megalopae and recruits wererestricted to the settlement season, which was determinedusing year-round data from experimental closed trays. Inthe case of juveniles all trials were considered for theanalyses.


3. Results

3.1. Field patterns of abundance and distribution

All species examined displayed a non-random dis-tribution on the natural microhabitats surveyed throughouttheir size ranges, including megalopae stage (logisticregression: P. barbiger: χ2=201.9, df=16, pb0.001;C. setosus; χ2=126.5, df=16, pb0.001; T. dentatus;χ2 =173.8, df=12, pb0.001; P. perlatus χ2 =625.3,df=16, pb0.001) (Fig. 2). Since for all species, at leastwithin one size-class, individuals were distinctivelydistributed in a discrete fashion among substrates.

Specifically in the case of P. barbiger, megalopaeand the first size class (4 mm CW) did not show dif-ferences in the proportion of individuals that used eachmicrohabitat. However, the second size class (6 mmCW) used cobbles in greater proportion compared toother substrates. The remaining size classes used cobblesalmost exclusively (Fig. 2). Megalopae of C. setosuswere found associated with algal turf and shell hash in

Fig. 2. Distribution–frequency pattern of individuals of the four main speciesare structured by size and habitat. Values correspond to the annual grouping othe proportional use of microhabitats between size classes.

equal proportions, while they were absent from cobbles.Individuals belonging to the first size class werepreferentially found on algal turf, while shell hash andcobbles were utilized in lesser proportion. From thesecond size class on (6–11 mm CW) individuals exhi-bited strong preference for algal turf with a marginal useof cobble. However, larger C. setosus (N12 mm CW)exhibited a different pattern, whereas a high proportionof them was found on cobble substratum (Fig. 2). Mega-lopae of T. dentatus were found on the three microhab-itats but most abundantly on algal turf. The preferencefor algae turf was even stronger for the first size classes(3–5 mm CW), while larger individuals (N6 mm CW)were only found on cobbles (Fig. 2). Megalopae ofP. perlatus were found in greater proportion in algal turfcompared to shell hash, while they were absent fromcobbles. First size class individuals of this species oc-curred preferentially on algal turf and secondarily oncobbles. Larger individuals were comparatively less abun-dant and were found on the three different microhabitats(Fig. 2).

of brachyuran decapods obtained from natural substrata surveys. Dataf the data (7 surveys). Different letters indicate significant differences in

Table 2Probabilities for significant differences from pair-wise comparisons(Tukey's unequal n HDS) between time periods for settlement rates ofmegalopae

Pair-wisecomparison

P. barbiger C. setosus T. dentatus P. perlatus

May-04 Jul-04 0.68 0.03 1.00 1.00May-04 Sep-04 1.00 0.02 0.05 1.00May-04 Nov-04 0.00 0.75 0.07 0.00May-04 Jan-05 0.00 0.72 1.00 0.13May-04 Mar-05 1.00 0.07 1.00 1.00May-04 May-05 1.00 0.02 1.00 1.00Jul-04 Sep-04 0.72 1.00 0.08 1.00Jul-04 Nov-04 0.00 0.68 0.08 0.00Jul-04 Jan-05 0.00 0.72 1.00 0.12Jul-04 Mar-05 0.99 1.00 1.00 1.00Jul-04 May-05 1.00 1.00 1.00 1.00Sep-04 Nov-04 0.00 0.03 0.99 0.00Sep-04 Jan-05 0.00 0.04 0.03 0.12Sep-04 Mar-05 1.00 1.00 0.03 0.94Sep-04 May-05 0.80 1.00 0.03 1.00Nov-04 Jan-05 0.75 1.00 0.05 0.00Nov-04 Mar-05 0.00 0.79 0.06 0.00Nov-04 May-05 0.00 0.05 0.05 0.00Jan-05 Mar-05 0.01 0.82 1.00 0.95Jan-05 May-05 0.00 0.09 1.00 0.91Mar-05 May-05 0.92 1.00 1.00 1.00

Bold: significant p-values.


3.2. Seasonal settlement rates in experimentalmicrohabitats

Although all species exhibited a seasonal settlementpattern, only P. barbiger megalopae occurred through-out the year. For this species, however, settlement signalwas weak during winter (July) and strong during springand summer months (November through January). Set-tlement of C. setosus was similarly high during May2004, November 2004 and January 2005, lower inMarch 2005 and absent in May 2005. Settlement ofT. dentatus was restricted to spring months. P. perlatusalso exhibited a comparatively short settlement period,with megalopae only occurring in late spring and earlysummer (Fig. 3 and Table 2). Settlement schedule for thetarget species in our study area was established based onperiods with records N10% of annual settlement rates.Thus, we established a settlement period for P. barbigerfrom September to May, for C. setosus from Novemberto May, for T. dentatus from September to Novemberand for P. perlatus from November to January.

3.3. Exclusion experiment by microhabitat

In general megalopae of all crab species, with theexception of P. barbiger, were more numerous on thesimulated algal turf microhabitat (Figs. 4–7). ForP. barbiger and P. perlatus, the ANOVAs exhibitedsignificant triple interactions among trial, microhabitattype and exclusion treatment (Table 3). Post-hoccomparisons (SNK) indicated that in those trials withhigher settlement rates there was a selective substratechoice. P. barbiger megalopae were significantly moreabundant on shell hash and the abundance on othermicrohabitats did not differ from the control (Fig. 4).

Fig. 3. Larval supply throughout the year for the four brachyuranspecies considered. Data are grouped for individuals collected in fourdifferent artificial substrata.

Moreover, within shell hash the closed treatmentexhibited significantly greater settlement for P. barbiger( p=0.002, SNK). The abundance of C. setosus mega-lopae was low throughout and we did not detectsignificant differences among microhabitats (Fig. 5). T.dentatus megalopae showed significantly higher pref-erence for simulated algal turf only when predators wereexcluded ( p=0.03 between exclusion treatment inartificial grass, SNK). For this species, megalopaeabundance, for both control and closed treatments, didnot differ in the remaining microhabitats (Fig. 6). Thelatter is reflected in a significant interaction betweenmicrohabitat and exclusion treatments (Table 3). In thecase of P. perlatus, the abundance of megalopae insimulated algal turf with exclusion of predators wastwice as high compared to trays without exclusion( p=0.003, SNK). In turn, megalopae in algal turf insideopen trays were significantly more abundant than underany other condition (Fig. 7).

In general, the abundance of recruits exhibited asimilar pattern as the one described for megalopae of thefour species considered in the different experimentaltreatments. Recruits of P. barbiger, regardless of theexclusion condition were significantly more abundancein shell hash (Fig. 4), while other microhabitats did notdiffer from the control. The mean abundance of recruits

Fig. 4. Mean abundance of P. barbiger along three ontogenetic stages.Data correspond to those collected from experimental trays with artificialsubstrata with and without access of large predators. Recruits are definedby the first class mark of the size structure, corresponding to the firststages after settlement. For the sake of clarity, annual data were pooled.Different letters indicate significant differences with pb0.05.

Fig. 5. Mean abundance of C. setosus along three ontogenetic stages.Data correspond to those collected from experimental trays withartificial substrata with and without access of large predators. Recruitsare defined by the first class mark of the size structure, correspondingto the first stages after settlement. For the sake of clarity, annual datawere pooled. Different letters indicate significant differences withpb0.05.


within a given microhabitat was very similar betweenopen and closed trays (Fig. 4). For C. setosus simulatedalgal turf and shell hash recorded higher and simi-lar recruitment rates inside the closed exclusion trays( p=0.8, SNK). In contrast, large differences in re-cruitment abundance were observed between these twomicrohabitats when they were exposed to predators( pb0.001, SNK) (Fig. 5). Higher recruitment rates ofT. dentatus were recorded in simulated algal turf whenlarge predators were excluded ( pb0.01, differenceswere only significant within exclusion treatment in

artificial grass, SNK), otherwise treatments did notdiffer significantly (Fig. 6). Recruits of P. perlatusexhibited a significant triple interaction among trials,microhabitat type and exclusion condition (Fig. 7 andTable 3). The mean abundance of recruits in simulatedalgal turf was around five- to seven-fold higher than inthe other microhabitats, regardless of the presence orexclusion of predators. In addition, we detectedsignificant differences in open versus closed trays inthe shell-hash microhabitat, with recruits being moreabundant in closed trays (p=0.01, SNK).

Fig. 6. Mean abundance of T. dentatus along three ontogenetic stages.Data correspond to those collected from experimental trays withartificial substrata with and without access of large predators. Recruitsare defined by the first class mark of the size structure, correspondingto the first stages after settlement. For the sake of clarity, annual datawere pooled. Different letters indicate significant differences withpb0.05.

Fig. 7. Mean abundance of P. perlatus along three ontogenetic stages.Data correspond to those collected from experimental trays withartificial substrata with and without access of large predators. Recruitsare defined by the first class mark of the size structure, correspondingto the first stages after settlement. For the sake of clarity, annual datawere pooled. Different letters indicate significant differences withpb0.05.


In general, the distribution of juveniles among expe-rimental microhabitats was not a reflection of the dis-tribution of recruits. For P. barbiger and C. setosus,colonization rates showed a significant double interac-tion among the trial and microhabitat factors (Table 2).The post-hoc comparison analysis (SNK) indicated thatin all trials but one, colonization rates were significantlyhigher for a given microhabitat. In the case of P. barbiger,it was more abundant in cobbles while abundance of

juveniles in other microhabitats did not differ from thecontrol (Fig. 4). For C. setosus, densities were signifi-cantly higher on simulated algal turf and the abundance ofjuveniles on cobbles was higher than on shell hash andthis in turn was higher than in control (Fig. 5). Despitetheir generally low abundance, juvenile T. dentatus ex-hibited a significantly higher colonization rate into cobblescompared to other microhabitats, including control(Fig. 6). Although the abundance of juvenile P. perlatus

Table 3Three-factor ANOVA testing settlement

Source of variation Paraxhantus barbiger Cancer setosus Taliepus dentatus Pilumnoides perlatus

df Error F P df Error F P df Error F P df Error F P

MegalopaeTrial (T) 4 0.28 18.76 0.00 3 0.02 2.99 0.03 1 0.022 0.14 0.71 1 0.06 12.92 0.00Microhabitat (M) 3 1.97 4.05 0.01 3 0.03 2.32 0.08 3 0.005 7.37 0.00⁎⁎ 3 0.19 5.70 0.00Exclusion (E) 1 0.33 2.67 0.10 1 0.01 0.55 0.46 1 0.005 0.03 0.86 1 1.06 0.28 0.60T×M 12 0.28 6.93 0.00 9 0.02 2.12 0.03⁎⁎ 3 0.022 0.22 0.88 3 0.06 2.86 0.04T×E 4 0.28 1.15 0.33 3 0.02 0.51 0.68 1 0.022 0.23 0.64 1 0.06 16.37 0.00M×E 3 0.70 1.32 0.27 3 0.03 0.07 0.97 3 0.019 3.15 0.03⁎ 3 0.90 0.25 0.86T×M×E 12 0.28 2.47 0.01⁎ 9 0.02 1.84 0.07 3 0.022 0.88 0.46 3 0.06 13.85 0.00⁎⁎

Residual 151 119 64 62

RecruitsTrial (T) 4 0.17 19.81 0.00 3 0.11 15.08 0.00 1 0.42 64.25 0.00 1 0.15 95.92 0.00Microhabitat (M) 3 1.93 3.18 0.03 3 0.37 6.95 0.00 3 3.36 2.81 0.05 3 0.40 14.46 0.00Exclusion (E) 1 0.25 0.00 0.99 1 0.03 50.99 0.00 1 0.99 1.68 0.20 1 6.72 0.01 0.91T×M 12 0.17 11.58 0.00⁎⁎ 9 0.11 3.34 0.00 3 0.42 8.04 0.00⁎⁎ 3 0.15 2.62 0.06T×E 4 0.17 1.53 0.20 3 0.11 0.26 0.85 1 0.42 2.37 0.13 1 0.15 44.10 0.00M×E 3 0.12 1.84 0.14 3 0.31 1.40 0.24 3 0.46 3.73 0.02⁎ 3 1.98 0.17 0.91T×M×E 12 0.17 0.73 0.73 9 0.11 2.80 0.01⁎ 3 0.42 1.10 0.35 3 0.15 13.01 0.00⁎⁎

Residual 151 119 64 62

JuvenilesTrial (T) 6 0.12 3.13 0.01 6 0.12 7.35 0.00 6 0.03 2.40 0.03⁎ 6 0.01 4.20 0.00⁎⁎

Microhabitat (M) 3 0.20 4.54 0.00 3 0.46 8.67 0.00 3 0.02 2.29 0.01⁎ 3 0.01 1.59 0.20T×M 18 0.12 1.74 0.04⁎ 18 0.12 3.75 0.00⁎⁎ 18 0.03 1.35 0.17 18 0.01 1.63 0.07Residual 106 106 106 106

Recruitment and colonization rate differences as function of trial, exclusion condition and microhabitat.⁎ p-valuesb0.05; ⁎⁎ p-valuesb0.01.


was in general very low inside the experimental trays, thiswas equivalent to that of T. dentatus (Fig. 7).

4. Discussion

This study highlights the relative influences of threeof the most relevant processes affecting the abundanceand distribution patterns of four brachyuran speciessoon after settlement, namely (1) habitat selection bymegalopae, (2) microhabitat-specific mortality due topredation and (3) habitat selection by post-settlers. Allprocesses have been previously studied separately underlaboratory conditions but few studies, including thepresent research, have evaluated them as part of an in-tegral question using a field approach (Eggleston andArmstrong, 1995; Palma et al., 1999; Moksnes, 2002;Moksnes and Heck, 2006). Our research represents thefirst attempt to consider these factors simultaneously ona brachyuran species assemblage (4 species) allowingthe opportunity of direct comparison among species in-habiting the same environment. Additional general con-clusions can be reached since these species belong to 4different families with different trophic characteristicsand life history traits (i.e., mean size, size at maturity).

Our approach included field manipulative experimentsaimed at detecting direct relationships between observedpatterns and the likely responsible processes.

The patterns of abundance found in different naturalmicrohabitats showed that the four crab species undergoontogenetic shifts in their use of space; none of themused microhabitats in the same proportion throughouttheir early benthic existence. In all cases, the micro-habitats utilized by megalopae did not correspond tothose utilized by later stages. Therefore, active micro-habitat selection by post-larvae cannot, by itself, explainthe early demography in these species. In fact, the re-sults of the experimental trays, particularly from thosethat excluded predators, indicate that most megalopaeactively select specific microhabitats and this only hasan influence on the pattern of recruits in two out of thefour species considered (i.e., P. barbiger and P. perlatus).However, such influence disappears soon after settlement.Similar results were observed for Carcinus maenas andC. sapidus, species displaying active habitat selectionat the time of settlement; however, this selection wasresponsible only for the initial distribution on a structur-ally complex habitat (Moksnes, 2002;Moksnes andHeck,2006).


A frequently used argument in favor of post-larvalhabitat selection has been the escape from visual pre-dators by using complex habitats as refuge (Palma et al.,1999). This is in agreement with the existence of win-dows of vulnerability for benthic invertebrates aftermetamorphosis (Hunt and Scheilbling, 1997; Casariegoet al., 2004), evidenced by strong predation on newlysettled crabs by fishes demonstrated empirically in thefield (e.g. Gonzáles and Oyarzun, 2003). Thus, the ca-pability of selecting habitats with lower risk of predationis considered a strong selective factor (Gosselin andQian, 1997). However, in our study, higher predationpressure (i.e., differences in abundance between exclu-sion treatments) was evidenced in microhabitats forwhich recruits had the strongest preference, a patternthat was especially evident for all species but P. barbiger.This apparent contradictory finding can be explainedsince dense aggregations of predators typically occur incomplex habitats (Moksnes and Heck, 2006). However,in this study, we did not find a relationship between theabundance of large predators and that of crabs (covari-ation with predators was not significant). An alternativeexplanation could be a density-dependent change in thebehavior of predators (i.e., type III functional response),where predators switch from one prey species to anotherin habitats where certain resources are less abundant(Schmitt and Holbrook, 1984; Palma and Ojeda, 2002).

Field results showed strong ontogenetic changes inthe use of habitat and this was especially evident forP. barbiger and C. setosus. The former switched from amixed use of microhabitats to an almost exclusive use ofcobbles, while the latter switched from an intensive useof algal turf to a predominantly use of cobbles. On theother hand, while earlier stages of T. dentatus andP. perlatus occurred mainly on algal turf, larger sizeswere not detected in any of the microhabitats surveyed.In the case of T. dentatus, which is mainly a herbivore(Palma, personal communication), algal turf mayrepresent both a source of food as well as a shelter forearly stages, but algal turf may not represent anappropriate source of food and shelter for largerindividuals (i.e., juveniles). In fact, juveniles and adultsof this species have been documented in associationwith fronds and holdfasts of the large brown kelp Les-sonia trabeculata (Vásquez et al., 2001).

When field assessments and experimental result werecontrasted, direct effect of predation on the distributionof post-settlers does not seem to be the main factorexplaining the observed patterns of distribution exhib-ited by P. barbiger. This result should consider the factthat this species displays cryptic patterns of colorationin their early benthic stages (Palma et al., 2003), thus

conferring them a defense against visual predators (Palmaand Steneck, 2001). Only the abundance of megalopae onshell hash was significantly different between exclusiontreatments. However, these differences were unexpectedsince higher abundance was detected inside the opentreatment. This somewhat contradictory finding could beexplained considering the indirect effect of the exclusiontreatment which allowed higher survivorship ofC. setosusrecruits, a species known to prey upon newly settledP. barbiger (Palma, unpublished data).

When field and experimental data for P. barbiger arecompared, a clear migration pattern emerges. In the field,cobble substrate became increasingly important asmicro-habitat for larger individuals. Similarly, in our experi-ments juveniles actively colonized the open trays filledwith cobbles, a pattern that differed from the one dis-played by megalopae and recruits. Thus, the observeddistribution pattern of P. barbiger could result from ac-tive migration of juveniles.

Our results described a different situation forC. setosus.Inside the experimental trays, recruits of this speciesshowed preferences for simulated algal turf and shell hashwhen large predatorswere excluded, but they also occurredat high numbers in open-to-predators trays only for si-mulated algal turf treatment. This finding matches wellwith field observations of high aggregations of thisspecies' recruits in algal turf and not in shell hash. Despitethe small size of algal turf patches, compared to othermicrohabitats, we show that algal turf can serve as potentialnursery area forC. setosus. This type of substrata share, atleast, two of the characteristics that have been considereda nursery environment (Beck et al., 2001), i.e., reducedlosses due to predation and comparatively high abundanceof recruits.

Field observations also showed that large juveniles ofC. setosus (N12 mm CW) tended to aggregate mainlyinside cobble patches. This change in habitat use couldbe better explained by the occurrence of active migrationbetween microhabitats. Compared to the other species,high colonization rates of juvenile C. setosus were ob-served inside open trays, suggesting that this species iscapable of an active foraging behavior typical of activepredators (Wolff and Cerda, 1992). In fact, field observa-tions and laboratory experiments have demonstrated thatboth adult and juvenile C. setosus can exert an intensivepredation upon P. barbiger, T. dentatus as well as on co-specifics (Cerda and Wolff 1993; Palma, unpublisheddata).

The case of T. dentatus represents a mixed condition.On the one hand, their abundance on the experimentalcobble microhabitat as juveniles was not explained byan important effect due to predation on megalopae and


recruits, which might explain the increasing importanceof cobbles as microhabitat in our field observation. Onthe other hand, predation had a stronger effect in re-ducing abundance of megalopae and recruits in simu-lated algal turf compared to the other microhabitats;however, field patterns showed that algal turf repre-sented a preferred microhabitat (Fig. 2). Thus, the directeffect of predation does not seem to explain the distri-bution pattern of T. dentatus in the field. Active migra-tion again represents a reasonable explanation sincejuveniles actively moved into our open experimentalcobble plots, which agrees with the field observations.

For P. perlatus, habitat-specific post-settlementlosses due to predation could produce some of the ob-served distribution pattern; however, our results indicatethat predation is a poor explanatory factor for the dis-tribution of all ontogenetic stages. As no differenceswere detected in microhabitat colonization rates ofjuveniles among experimental trays, active migrationappears as a satisfactory explanation for the proportionaluse of all microhabitats surveyed in the field.

Increased active dispersal along ontogenetic stages isan expected behavior for mobile animals, particularlysince the value of food and refuge change with size (e.g.,Werner and Gilliam, 1984; Gosselin, 1997). Ontoge-netic changes in habitat use appear as the most importantfactor determining the ecological patterns in the earlybenthic life of the brachyuran decapod assemblagestudied. Our results indicate that this active microhabitatredistribution occurs in the early benthic stages ofP. barbiger, T. dentatus and P. perlatus and in olderrecruits of C. setosus. Although both habitat selectionby megalopae and the effect of predation was detected insome cases, the subsequent movement among habitatslargely outweighed their influence on field distribution.Thus, the dispersion of juvenile crabs along the bottomafter they settle seems to be poorly influenced by dif-ferences in the ecological and life history traits of thespecies considered in this study.

Active migration has traditionally not been included indemographic studies of marine populations (Palmer et al.,1996). However, recent studies recall that in populationsthat are recruitment-limited, active migration, directlyor indirectly influenced by predation (Moksnes, 2002;Moksnes and Heck, 2006) or competition (Moksnes,2004), appears as an important regulatory factor for benthicmarine invertebrate populations. This study represents acontribution in this direction.

The regulation of active migration in post-settlementstages of crabs has mainly been associated to density-dependent processes (Eggleston et al., 1998; Etheringtonet al., 2002; Moksnes, 2004). In this study we did not

control for density, however, and regardless that selectivesettlement was evaluated for each specie's settlementperiod, “trial" was an important source of variation. Thisfactor was generally in double or triple interaction withother factors in our analyses, demonstrating that settlersare exposed to an important range of densities. A closerlook at these interactions evidenced that they occurreddue to the normal fluctuations in the abundance ofcompetent larvae arriving to the benthos. For example,differences of abundance between microhabitats were notdetected in some trails due to natural low abundances. Forcoastal invertebrates, time-limited episodes of settlementare common and can be attributed to the occurrence ofoceanographic dynamics such as upwelling relaxationevents (Shanks et al., 2000; Narváez et al., 2006). Suchprocesses could also explain trial variations in recruit'sabundance, given that these patterns generally were theresult of megalopae distribution.

Finally, this study determined that in species withcomplex life cycles, active migration could represent astrong post-settlement factor responsible for modifyingpatterns of settlement at local spatial scales. The latteragrees with studies in other species where density-de-pendent migration is as important as mortality for ex-plaining losses of individuals that occur soon after theysettle (Eggleston et al., 1998; Etherington et al., 2002).The recognition that active migration among habitatscan represent an important factor in regulating popula-tions, demonstrated here by patterns of differential habi-tat use in this four-species assemblage of brachyurancrabs, highlights the importance of conservation of hab-itat patchiness even at small, local scales.

Acknowledgements

We thank Fernando Ogalde, Javier Infante, FreddyVeliz and Jose Miguel Rojas for their important supportduring field work. We thank the anonymous reviewersfor their excellent and detailed input and suggestions,which greatly improved the manuscript. This study wasfunded by FONDAP-FONDECYT grant 1501-0001 toFPO. [SS]

References

Beck, M., Heck, K., Able, K., Childers, D., Eggleston, D., Gillanders,Halpern, B., Hays, C., Hoshino, K., Minello, T., Orth, R.,Sheridan, P., Weinstein, M., 2001. The identification, conservationand management of estuarine and marine nurseries for fish andinvertebrates. BioScience 51, 633–641.

Casariego, A.M., Schwindt, E., Iribarne, O., 2004. Evidence of habitatstructure-generated bottleneck in the recruitment process of the SWAtlantic crab Cyrtograpsus angulatus. Mar. Biol. 145, 259–264.


Cerda, G., Wolf, M., 1993. Feeding ecology of Cancer polyodon in LaHerradura bay, northern Chile: 2. Food spectrum and prey consump-tion. Mar. Ecol. Prog. Ser. 100, 119–125.

Connell, J.H., 1985. The consequences of variation in initial settlementvs. post-settlement mortality in rocky intertidal communities. J. Exp.Mar. Biol. Ecol. 93, 11–45.

Eggleston, D.B., Armstrong, D.A., 1995. Pre- and post-settlement deter-minants of estuarine Dungeness crab recruitment. Ecol. Monogr. 65,193–216.

Eggleston, D.B., Etherington, L.L., Elies,W.E., 1998. Organism responseto habitat patchiness: species and habitat-dependence recruitment ofdecapod crustaceans. J. Exp. Mar. Biol. Ecol. 223, 111–132.

Etherington, L.L., Eggleston, D.B., Stackhausen, W.T., 2002. Parti-tioning loss rates of early juveniles blue crabs from seagrass habitatinto mortality and emigration. Bull. Mar. Sci. 72, 371–391.

Fagetti, G.E., Campodonico, I., 1971. Desarrollo larval en el laboratoriode Taliepus dentatus (Milne-Edwards) (Crustacea Brachyura:Majidae, Acanthocyclinae). Rev. Biol. Mar. (Valparaíso) 14, 1–14.

Fagetti, G.E., Campodonico, I., 1973. Larval development ofPilumnoidesperlatus (Brachyura: Xanthidae) under laboratory conditions. Mar.Biol. 18, 129–139.

Fariña, J.M., Palma, A.T., Ojeda, F.P., in press. A food web perspectiveof the subtidal communities off the temperate Chilean coast. In:McClanahan, T.R., Branco, G. (Eds.), Food webs and trophic dyna-mics of marine benthic ecosystems. Oxford University Press, Oxford.

Fernández, M., Castilla, J.C., 1997. The Chilean artisanal stone crab(Homalaspis plana) fishery: catch trends in open access zones andthe effect of management areas in Central Chile. J. Shellfish Res.16, 371–377.

Fernández, M., Castilla, J.C., 2000. Recruitment of Homalaspis planain intertidal habitats of central Chile and implications for thecurrent use of management and marine protected areas. Mar. Ecol.Prog. Ser. 208, 157–170.

Fernández, M., Iribarne, O., Armstrong, D., 1993. Habitat selection byyoung-of-the-year ungeness crab Cancer magister and predationrisk in intertidal habitats. Mar. Ecol. Prog. Ser. 92, 171–177.

Forward Jr., R.B., Reyns, N.B., Diaz, H., Cohen, J.H., Eggleston, D.B.,2005. Endogenous swimming rhythms underlying secondary dis-persal of early juvenile blue crabs, Callinectes sapidus. J. Exp. Mar.Biol. Ecol. 316, 91–100.

Gonzáles, P., Oyarzun, C., 2003. Diet of the Chilean sandperch,Pinguipes chilensis (Perciformes, Pinguipedidae) in southernChile. J. Appl. Ichthyol. 19, 371–375.

Gosselin, L.A., 1997. An ecological transition during juvenile life in amarine snail. Mar. Ecol. Prog. Ser. 157, 185–194.

Gosselin, L.A., Qian, P.E., 1997. Juvenile mortality in benthic marineinvertebrates. Mar. Ecol. Prog. Ser. 146, 265–282.

Jesse, S., Stotz, W., 2003. Spatio-temporal distribution patterns of crabassemblage in the shallow subtidal of the north Chilean pacificcoast. Crustaceana 75, 1161–1200.

Jones, G.P., 1997. Relationships between recruitment and postrecruitment processes in lagoonal populations of two coral reeffishes. J. Exp. Mar. Biol. Ecol. 213, 231–246.

Keough, M.J., Downes, B.J., 1982. Recruitment of marine invertebrates:the role of active larval choice and early mortality. Oecologia 54,348–352.

Hedvall, O., Moksnes, P., Phil, L., 1998. Active habitat selection bymegalopae and juveniles shore crab Carcinus maenas, a laboratorystudy in an annular flume. Hydrobiologia 375/376, 89–100.

Herbert, R.J., Hawkins, S.J., 2006. Effect of rock type on the recruitmentand early mortality of the barnacle Chthamalus montagui. J. Exp.Mar. Biol. Ecol. 334, 96–108.

Hunt, H.L., Scheilbling, R.E., 1997. Role of early post-settlementmortality in recruitment of benthic marine invertebrates. Mar. Ecol.Prog. Ser. 155, 269–301.

Lancellotti, D., Vásquez, J.A., 2000. Zoogeografía de macroinvertebra-dos bentónicos de la costa de Chile: contribución para la conser-vación marina. Rev. Chil. Hist. Nat. 73, 99–129.

Leon, R.I., Stotz, W.B., 2004. Diet and prey selection dynamics ofCancer polyodon in three different habitat types in Tongoy Bay,Chile. J. Mar. Biol. Assoc. U.K. 84, 751–756.

Manriquez, P.H., Cancino, J.M., 1991. Depredación deMembraniporaisabellaeana (Bryozoa) por Taliepus dentatus (Crustacea: Dec-apada). Rev. Biol. Mar. (Valparaíso) 26, 309–323.

Menge, B.A., Sutherland, J.P., 1987. Community regulation: variationin disturbance, competition, and predation in relation to environ-mental stress and recruitment. Am. Nat. 130, 730–757.

Moksnes, P.-O., 2002. The relative importance of habitat-specificsettlement, predation and juvenile dispersal for distribution andabundance of young juvenile shore crabs Carcinus maenas L.J. Exp. Mar. Biol. Ecol. 271, 41–73.

Moksnes, P.-O., 2004. Interference competition for space in nurseryhabitats: density-dependent effects on growth and dispersal in juvenileshore crabs Carcinus maenas. Mar. Ecol. Prog. Ser. 281, 181–191.

Moksnes, P-O., Heck Jr., J.L., 2006. Relative importance of habitatselection and predation for the distribution of blue crab megalopaeand young juveniles. Mar. Ecol. Prog. Ser. 308, 165–181.

Morales, C., Antezana, T., 1983. Diet selection of the Chilean stonecrab Homalaspis plana. Mar. Biol. 77, 79–83.

Narváez, D.A., Navarrete, S.A., Largier, J., Vargas, C.A., 2006. Onshoreadvection ofwarmwater, larval invertebrate settlement, and relaxationof upwelling off central Chile. Mar. Ecol. Prog. Ser. 309, 159–173.

Ojeda, F.P., Fariña, J.M., 1996. Temporal variations in the abundance,activity, and trophic patterns of the rockfish, Sebastes capensis, offthe central Chilean coast. Rev. Chil. Hist. Nat. 69, 205–211.

Palma, A.T., Ojeda, F.P., 2002. Abundance, distribution and feedingpatterns of a temperate reef fish in subtidal environments of theChilean coast: the importance of understory algal turf. Rev. Chil.Hist. Nat. 75, 189–200.

Palma, A.T., Steneck, R.S., 2001. Does variable coloration in juvenilemarine crabs reduce risk of visual predation?Ecology 82, 2961–2967.

Palma, A.T., Wahle, R.A., Steneck, R.S., 1998. Different earlypost-settlement strategies between American lobsters Homarusamericanus and rock crabs Cancer irroratus: in the Gulf ofMaine. Mar. Ecol. Prog. Ser. 162, 215–225.

Palma, A.T., Steneck, R.S., Wilson, C.J., 1999. Settlement-driven,multiscale demographic patterns of large benthic decapods in theGulf of Maine. J. Exp. Mar. Biol. Ecol. 241, 107–136.

Palma, A.T., Orrego, C., Arriagada, M., 2003. Crypsis in early benthicphases of brachyuran decapod crustaceans in central Chile. Rev.Chil. Hist. Nat. 76, 149–156.

Palma, A.T, Pardo, L.M., Veas, R.I., Cartes, C., Silva, M., Manriquez,K., Diaz, A., Muñoz, C., Ojeda, F.P., 2006. Coastal brachyurandecapods: settlement and recruitment under contrasting coastalgeometry conditions. Mar. Ecol. Prog. Ser. 316, 139–153.

Palmer, M.A., Allan, D., Butman, C.A., 1996. Dispersal as a regionalprocess affecting the local dynamics of marine and stream benthicinvertebrates. TREE 11, 322–326.

Pardieck, R.A., Orth, R.J., Diaz, R.J., Lipcius, R.N., 1999.Ontogenetic changes in habitat use by postlarvae and youngjuveniles of the blue crab. Mar. Ecol. Prog. Ser. 186, 227–238.

Quintana, R., Saelzer, H., 1986. The complete larval development ofthe edible crab, Cancer setosus Molina and observations on theprezoeal and first zoeal stages of C. coronatus Molina (Decapoda:


Brachyura, Cancridae). J. Fac. Sci. Hokkaido Univ. Ser. VI Zool.24, 267–303.

Retamal, M.A., 1999. Decápodos de Chile. Biodiversity Center ofETI, Multimedia Interactive Software. Centro ETI hile, Universi-dad de Concepción.

Richards, R.A., 1992. Habitat selection and predator avoidance –ontogenic shifts in habitat use by the Jonah crab Cancer borealis(Stimpson). J. Exp. Mar. Biol. Ecol. 156, 187–197.

Roughgarden, J., Gaines, S., Possingham, H., 1988. Recruitmentdynamics in complex life cycles. Science 241, 1460–1466.

Schmitt, J.R., Holbrook, S.J., 1984. Ontogeny of prey selection byblack surfperch Embiotoca jacksoni (Pisces: Embiotocidae): tworoles of fish morphology, foraging behavior and patch selection.Mar. Ecol. Prog. Ser. 18, 225–239.

Shanks, A.L., Largier, J., Brink, L., Brubaker, J., Hooff, R., 2000.Demonstration of the onshore transport of larval invertebrates byshoreward movement of an upwelling front. Limnol. Oceanogr. 45,230–236.

Sokal, R.R., Rohlf, F.J., 1995. Biometry: The Principles and Practice ofStatistics in Biological Research, 3rd ed. FreemanW.H, New York.

Stachowicz, J.J., Hay, M.E., 1996. Facultative mutualism between anherbivorous crab and a coralline alga: advantages of eating noxiousseaweeds. Oecologia 105, 377–387.

Stokes, M.E., Davis, C.S., Koch, G.G., 1995. Categorical DataAnalysis Using the SAS System. SAS Institute, Cary.

Strathmann, R.R., Barnscomb, E.S., Vedder, K., 1981. Fatal errors inset as a cost of dispersal and the influence of intertidal flora on setof barnacles. Oecologia 48, 13–18.

Tupper,M., Boutilier, R.G., 1996. Effects of habitat on settlement, growthand post-settlement mortality of Atlantic cod (Gadus morhua). Can.J. Fish. Aquat. Sci. 52, 2091–2096.

Underwood, A., 2004. Landing on one's foot: small-scale topographicfeatures of habitat and the dispersion of juvenile intertidal gastropods.Mar. Ecol. Prog. Ser. 268, 173–182.

Underwood, A., Keough, M.J., 2001. Supply-side ecology the nature andconsequences of variations in recruitment of intertidal organism. In:Berteness, M., Gaines, S., Hay, M. (Eds.), Marine CommunityEcology. Sinauer Associates Inc, Sunderland, pp. 183–200.

Vásquez, J.A., Véliz, D., Pardo, L.M., 2001. Diversidad bajo las grandesalgas pardas. In: Alveal, K., Antezana, T. (Eds.), Sustentabilidad de laBiodiversidad: Un Problema Actual, Base CientíFico-Técnicas,Teorización y Perspectivas. Publicaciones de la Universidad de Con-cepción, pp. 208–293.

Wahle, R.A., Steneck, R.S., 1991. Recruitment habitats and nurserygrounds of the American lobsterHomarus americanus: a demograph-ic bottleneck? Mar. Ecol. Prog. Ser. 69, 231–243.

Wahle, R.A., Steneck, R.S., 1992. Habitat restrictions in early benthiclife: experiments on habitat selection and in situ predation with theAmerican lobster. J. Exp. Mar. Biol. Ecol. 157, 91–114.

Werner, E.E., Gilliam, J.F., 1984. The ontogenetic niche and speciesinteractions in size-structured populations. Ann. Rev. Ecolog. Syst.15, 393–425.

Wolff, M., Cerda, G., 1992. Feeding ecology of the crab Cancerpolyodon in La Herradura Bay, northern Chile: II. Food spectrumand prey consumption. Mar. Ecol. Prog. Ser. 100, 119–125.

Processes regulating early post-settlement habitat use in a subtidal assemblage of brachyuran decapods

Documents