Demography and Feeding Behavior of the Kelp Crab Taliepus Marginatus in Subtidal Habitats Dominated by the Kelps Macrocystis Pyrifera Or Lessonia Trabeculata

Demography and feeding behavior of the kelp crab Taliepus marginatus in subtidal habitats dominated by the kelps Macrocystis pyrifera or Lessonia trabeculata

David Jofre Madariaga,1,2 Marco Ortiz,2 and Martin Thiel1,3,a

1 Facultad de Ciencias del Mar, Universidad Catolica del Norte, Coquimbo, Chile 2 Instituto Antofagasta de Recursos Naturales Renovables (IARnR), Instituto de Investigaciones Oceanologicas, Facultad de Recursos del Mar, Universidad de Antofagasta, Antofagasta, Chile 3 Centro de Estudios Avanzados en Zonas Aridas, Coquimbo, Chile

Abstract. We studied the population and feeding ecology of the kelp crab Taliepus marginatus in subtidal kelp forests dominated by either of two morphologically different kelp species (Macrocystis pyrifera or Lessonia trabeculata) in northern Chile. The sizes and abundances of T. marginatus differed between the two kelp habitats. Kelp crabs were more abundant in the M. pyrifera forest than in the L. trabeculata forest. Size-frequency distribu- tions showed that juvenile and immature crabs were more common in the M. pyrifera forest than in the L. trabeculata forest, where reproductive adults predominated. The smaller crabs in the M. pyrifera habitat also consumed a higher proportion of kelp tissues than the larger crabs in the L. trabeculata habitat, which had a higher proportion of animal food in their diet. In both kelp forests, individuals of T. marginatus showed a similar pattern of nocturnal feeding over a 24-h period, consuming more food at night than during the day. The more complex and dense forests of M. pyrifera appear to present better nursery habitats for juvenile kelp crabs than the more open and less dense forests dominated by L. trabeculata. These results suggest that the role of the two kelp habitats for T. marginatus varies during the life cycle of the kelp crabs, with M. pyrifera tending to have nursery function and L. trabeculata being more suitable as a reproductive habitat. Additional key words: habitat use, nursery habitat, reproductive habitat, diet

Habitats differ in their availability of food items, prey, and thus represent an important trophic link predation pressure, and physical disturbance (e.g., hab- (Hines 1982). As consumers, they obtain most of itat-specific environmental factors), resulting in specific their nutrition from algal material (Hines 1982; costs and benefits for individuals (Amaral et al. 2008). Kilar & Lou 1986; Daly & Konar 2010). The nutri- For example, habitats with higher structural complex- tional value and abundance of this algal material ity offer more shelter and resources, thereby support- are factors that determine its consumption (Wolcott ing high densities of individuals. On the other hand, &O′Connor 1992), and multiple structural and high densities may result in interference competition chemical adaptations have evolved that might for resources and space, leading to emigration and reduce the palatability of seaweeds (Duffy & Hay occupation of less favorable habitats by some individu- 1990). Despite this, some herbivores from temperate als, allowing for more feeding time, albeit possibly on kelp beds consume large amounts of algal biomass less preferred food items (Amaral et al. 2008). (Leighton 1966; Vasquez & Buschmann 1997). Kelp beds are highly productive benthic ecosys- Some algal consumers, primarily the smaller, less tems (Vasquez 1992; Steneck et al. 2002; Ortiz mobile ones, commonly specialize in one type of 2008), which provide shelter for a wide variety of food (e.g., the habitat-forming macroalgae) and live species (e.g., fishes, seastars, crabs, and snails) with directly on it, using it as both food and shelter different feeding behaviors (Smith et al. 1996). Maj- (Hines 1982; Woods 1993; Stachowicz & Hay 1999; id crabs living on kelps act as both consumers and Gutow et al. 2012). In contrast, larger individuals that roam farther also incorporate other food items aAuthor for correspondence. in their diet. This may lead to spatial segregation E-mail: [email protected] between different ontogenetic groups. 134 Jofre, Ortiz, & Thiel

In many habitats, crab consumers have special Methods feeding rhythms to avoid interspecific competition for food and predation risk (Jesse 2001; Kronfeld- Study area Schor & Dayan 2003). For example, members of many brachyuran species feed mainly during the This study was carried out in austral fall during night, with no or low foraging activity during the the months of April 2006 and April 2007 near Isla day (Aris et al. 1982; Jesse 2001; Novak 2004; Santa Marıa at the southern tip of the Mejillones Almeida et al. 2008). By limiting their activity per- Peninsula (SE Pacific coast, Antofagasta, Chile: iod to times of low predation pressure, these crabs 23°27′S–70°36′W). The study area is close to an reduce the amount of time available for foraging important upwelling center that supplies nutrients to (Aris et al. 1982). the coastal ecosystem (Escribano et al. 2004). These Taliepus marginatus (BELL 1835) (Superfamily nutrient-rich coastal areas are dominated by subtidal Majoidea) is a decapod crab widely distributed kelp forests of Macrocystis pyrifera and Lessonia along the Chilean coast, but little is known about trabeculata, which develop in different environmen- its life cycle. Recently settled juveniles of its conge- tal conditions (Ortiz 2008). Macrocystis pyrifera ner T. dentatus (MILNE EDWARDS 1834) are com- forms extensive and complex patches in coastal monly found in shallow subtidal habitats, often areas mostly protected from direct wave exposure, dominated by turf algae, and while the habitats of where they often grow on boulders extending over growing juveniles and subadults are not known, the depths of 2–10 m (Vega et al. 2005; Villegas et al. reproductive adults generally occur in subtidal kelp 2008). Blades of M. pyrifera grow along the entire forests (Pardo et al. 2007; Palma et al. 2011). Talie- stipe, forming a dense mesh of algal canopy from pus marginatus has been reported from deeper sub- the bottom to the sea surface (Villegas et al. 2008). tidal waters (Antezana et al. 1965) in subtidal kelp In contrast, L. trabeculata grows mostly on more systems dominated by either Macrocystis pyrifera exposed coasts on bedrock ranging in depth from 8 (LINNAEUS)C.AGARDH 1820 or by Lessonia trabecu- to 14 m (Vega et al. 2005; Villegas et al. 2008). The lata VILLOUTA &SANTELICES 1986 (Villegas et al. blades of older kelp stands usually originate from 2008). These two kelp species show important dif- the stipes at some distance above the bottom, but ferences; they develop in different environmental the individual kelp plants rarely exceed 1.5 m in conditions, have contrasting morphological struc- height and thus do not reach the sea surface. Given tures (Ortiz 2008; Villegas et al. 2008), and likely their different bathymetric distribution, at the study differ substantially in their tissue consistency and site the kelp forests composed of M. pyrifera are attractivity for grazers (for differences between M. closer to the shore, growing over a gentle slope pyrifera and Lessonia spp. see for example Pansch down to ~6 m depth, while stands of L. trabeculata et al. 2008). Based on these ecological differences, start at the seaward edge of the M. pyrifera patches, we expected that the natural diet, density, size dis- extending in depth from ~5 m down to ~12 m tribution, and foraging activity of T. marginatus (Fig. 1). might differ between these two kelp habitats. Specif- Both kelp forests have a high standing biomass, ically, we hypothesized that smaller crabs would be but system throughput (a measure of ecosystem more common in the shallow and dense kelp beds metabolism) in M. pyrifera kelp beds is higher than of M. pyrifera, while larger crabs would dominate in L. trabeculata (Ortiz 2008, 2010). A wide diversity in the deeper and more open beds of L. trabeculata. of different invertebrates is found in the kelp forests We also expected to find differences in the feeding around Isla Santa Marıa; among these, sea urchins ecology of crabs between the two kelp systems. and their seastar predators dominate in biomass This study thus aims to contribute to knowledge (Vasquez et al. 2006; Gaymer et al. 2010). Fish pre- of the demography and feeding ecology of T. mar- dators are also important in these kelp forests (Ortiz ginatus in two subtidal kelp habitats. The objectives 2008), which offer more refuge and food than sur- were to compare the (i) size structure, (ii) density, rounding barren grounds (e.g., Vasquez et al. 2006; (iii) natural diet, and (iv) daily feeding cycle of T. Villegas et al. 2008; Perez-Matus et al. 2012). marginatus inhabiting beds dominated by either M. In these kelp systems, Taliepus marginatus is one pyrifera or L. trabeculata. This work is part of an of the most abundant brachyuran crabs. Settlement extensive study examining the properties of the kelp occurs primarily during austral spring (personal forest ecosystems formed by M. pyrifera and L. tra- observations), and growing juveniles and subadults beculata (Ortiz 2008, 2010). were expected to be most abundant during late

Invertebrate Biology vol. 132, no. 2, June 2013 Demography and feeding behavior of a kelp crab 135

Fig. 1. Location of the study area on the Mejillones Peninsula close to Santa Marıa Island, SE Pacific coast (northern Chile), showing Macrocystis pyrifera (dark shading) and Lessonia trabeculata (light shading) kelp forests. Sampling plots in each kelp habitat are shown. summer and fall, as had also been reported for systems. The scuba divers collected crabs on the other kelp inhabitants (Gaymer et al. 2010) and for bottom and in the kelp canopy by carefully survey- kelp crabs in California (Hines 1982). Consequently, ing the entire fronds (e.g., searching on both sides sampling of crabs was conducted during early aus- of the blades). This was done to collect all crabs tral fall (April). Crabs collected during the fall of >10 mm carapace width (CW), including those that two subsequent years (2006 and 2007) were pooled were still camouflaged against the canopy back- herein. ground. Crabs were collected by hand and placed into thick-walled plastic bags with handles. These plastic bags were used because kelp crabs quickly Sampling procedure entangle in the mesh of regular dive bags. At the To estimate abundance and size structure of the end of each sampling, all crabs were transported to crabs in the two kelp habitats, we collected samples the shore for further processing. To preserve the of T. marginatus from three rectangular plots of stomach contents, each crab was injected with a 20 m2 (592 m) in each kelp forest (Fig. 1). The solution of 10% formalin in the oral and abdominal positions of these sampling plots were selected from regions. This method was used to rapidly stop diges- the areas with the highest kelp density, covering the tion and degradation processes. All samples were main depth range of each kelp. In the M. pyrifera frozen (20°C) for subsequent analysis. forest one of the three rectangular plots was placed close to the shore (~3 m deep) and the other two Sample analysis were oriented away from shore (~5–6 m deep), and in the L. trabeculata forest the plots were placed In the laboratory, all crabs were identified as T. between 8 and 12 m depth. All rectangular plots marginatus according to Retamal (1977). Sex was were linked and fixed by an anchor in the middle determined through visual observation of the abdo- (Fig. 1). Each plot was carefully sampled by two men. Females were distinguished by a wide abdo- scuba divers using semiautonomous (hookah) diving men, and all mature females were examined for

Invertebrate Biology vol. 132, no. 2, June 2013 136 Jofre, Ortiz, & Thiel brooded embryos. The total wet weight of each crab The crabs were collected by hand and placed in was measured with a semianalytical balance thick-walled plastic bags with a handle. The divers (precision0.1 g). Carapace width (CW) was mea- rapidly collected all visible crabs >40 mm CW, sured with a digital caliper (precision0.1 mm). In because crabs had to be collected quickly at the this study, we distinguish two size groups of crabs, defined sampling hours. During the nocturnal sam- juveniles and subadults (10.0–40.0 mm CW), and plings it was necessary to use a lantern, and the adult crabs (>40.0 mm CW). The smallest ovigerous restricted field of view limited the search efficiency female had a CW of 44.8 mm, suggesting that T. of divers. For this reason it was decided to focus on marginatus reaches sexual maturity at ~40 mm CW. larger crabs for both the day and the night samples The cardiac stomachs were extracted from each of the daily feeding cycle. These samples were pre- specimen and the wet weights of the full (before served and analyzed as described above. The feeding removing stomach contents) and empty stomachs periodicity of T. marginatus in each kelp forest was were measured after absorbing excess water with assessed by calculating the relative stomach content paper tissues. Stomach contents were diluted in dis- (SC%) over a 24-h cycle by equation (2): tilled water and 10% ethanol and analyzed visually gut content (g) with the aid of a binocular microscope. The con- SCð%Þ¼ 100 tents of each stomach were classified to the lowest body weight (g) possible taxonomic level. Finally, the importance of each food item in the diet of T. marginatus was This index describes the stomach fullness for each described in two ways: (i) frequency of occurrence, sampling hour (Jesse 2001; Baumann & Kwak and (ii) the wet weight of each food item. The fre- 2011). quency of occurrence was estimated based on the presence of a particular item in the stomach using Statistical analyses equation (1),

ni Prior to analyses, all data were tested for normal- F ¼ – – i N ity using Kolmogorov Smirnov and Shapiro Wilk’s tests. Homogeneity of variances was tested with F- tests or with Fligner–Killeen tests. If assumptions where ni is the number of occurrences (number of stomachs in which the prey i is present) and N is the were not met, data were transformed or a nonpara- total number of stomachs containing food. Then, metric test was applied (Zar 1996). To examine the each frequency was multiplied by 100 and expressed differences in the median size distribution between – as percentage of occurrence (%). A total of 361 kelp habitats, a nonparametric Mann Whitney crabs were examined in the two kelp habitats (258 U-test was performed. Density data were log-trans- in M. pyrifera and 103 in L. trabeculata) for fre- formed and a t-test was used to examine the quency of occurrence. For the wet weight propor- differences in crab densities between habitats. To tions, a total of 181 stomachs (97 from M. pyrifera determine whether the proportion of ovigerous and 84 from L. trabeculata) were randomly selected females differed between kelp habitats, a chi-squared v2 from the 361 examined stomachs (Table 1). For this test ( ) was used. Sizes of ovigerous females from analysis, all food items (by category) were weighed the two kelp habitats were compared with a on an analytical balance (precision0.0001 g), and Student’s t-test. the abundance of each item in each gut expressed as A two-way ANOVA was used to test for signiﬁ- a percentage of total stomach content weight. The cant differences in the daily feeding cycle, with the combination of the two methods was used to (i) two factors being time (hour) and site (habitat). determine whether food items were present in the When the ANOVAs revealed signiﬁcant differences, diet of crabs, and (ii) estimate the proportional wet a post hoc Tukey HDS was applied. The analyses weight (importance) of each food item in the were carried out using R (R Development Core stomachs. Team 2010). Nonmetric multidimensional scaling (nMDS) was performed to assess similarity/dissimilarity in diet Daily feeding cycle (species/taxa: using the proportion of wet weight per To determine the daily feeding cycle over 24 h, food item) between habitats (Macrocystis/Lessonia) we collected specimens of T. marginatus (n=20) at and size groups of the crabs (10.0–40.0 mm/ regular intervals of 6 h (16:00, 22:00, 04:00, 10:00). >40.0 mm CW) based on the Bray–Curtis similarity Each sampling was conducted by two scuba divers. index. Prior to the analysis, the data were square-

Invertebrate Biology vol. 132, no. 2, June 2013 Demography and feeding behavior of a kelp crab 137

Table 1. Volume proportions (%, meanSD) of each Results food category (wet weight in g) in the stomach contents of Taliepus marginatus from the two kelp habitats. Dashes Size distribution and abundance indicate absence of the respective food type in the analyzed stomachs. The size distribution of crabs differed significantly between the two kelp habitats (U=28185.5; df=1; Habitats p<0.001), even though the range of crab sizes was similar in the two habitats (Fig. 2). The size distri- Macrocystis Lessonia = = bution of females and males in Macrocystis pyrifera Stomachs analyzed (n 97) (n 84) % % varied from 14.0 to 63.9 mm CW (49.7 16.2 mm; meanSD) and 10.9–91.3 mm (48.39.2 mm), Food category respectively (Fig. 2). In contrast, in Lessonia trabe- Macrocystis 96.35.9 — culata, females ranged 20.9–72.2 mm CW Lessonia — 93.913.3 (53.48.4 mm) and males ranged 10.7–97.6 mm Rhodophyta 2.4 3.6 5.0 12.2 (70.618.4 mm) (Fig. 2). The total crab densities a Hydrozoa 0.1 0.4 0.1 were 3.11.6 ind. m2 in M. pyrifera, and 0.70.6 Detritus 0.10.7 0.72.0 2 — ind. m in L. trabeculata; these densities differed Crustacea 0.6 3.7 = Mollusca 0.10.8 0.11.3 significantly between the two kelp habitats (t 4.38; = < Bryozoa a0.1 a0.1 df 12; p 0.001). Gastropod egg capsules 0.11.0 a0.1 The proportion of females and males was 49.7% and 50.3%, respectively, in the M. pyrifera habitat, a 0.1=Weight < 0.1 g. and 28% of all females found in M. pyrifera were ovigerous. In the L. trabeculata habitat, the propor- root transformed. Three separate one-way analyses tion of females and males was 38.3% and 61.7%, of similarity (ANOSIM) were performed to test the respectively, and 66% of the females were ovigerous. following hypotheses: (A) there are no differences in The sizes of ovigerous females did not differ signifi- diet composition of T. marginatus from M. pyrifera cantly between the two kelp habitats (Macrocystis, and L. trabeculata habitats, and (B-i & B-ii) within n=52; Lessonia,n=27) (t=2.027; df=77; p=0.05). The each kelp habitat (i & ii) there are no differences in proportion of ovigerous females was significantly diet composition between immature and adult crabs. higher in L. trabeculata than in M. pyrifera 2 In addition, the SIMPER routine was used to estab- (v =21.263; df=1; p<0.001). lish which species/taxa contributed most to either the similarity or dissimilarity between habitats and Diet composition habitat per size (Clarke 1993). The multivariate analysis was carried out using PRIMER v.6 (Clarke In general, crabs from both habitats had similar & Gorley 2006). diets, but they differed in the proportional occurrence

Fig. 2. Size-frequency distribution of Taliepus marginatus in both kelp habitats. Open bars show the total frequency of males and females in both habitats. Black bars show the frequencies of ovigerous females. Data from both study years (April 2006 and April 2007) were pooled.

Invertebrate Biology vol. 132, no. 2, June 2013 138 Jofre, Ortiz, & Thiel

Fig. 3. Diet composition expressed as frequency of occurrence (as % of individuals) for Taliepus marginatus of two different size ranges (carapace width in mm) in each kelp habitat. n, number of stomachs examined per size range. of kelp tissues and other types of food (Fig. 3). of the two crab size categories (10.0–40.0 mm and Kelp remains (either M. pyrifera or L. trabeculata) >40.0 mm) had kelp tissues in their stomachs were found in the stomachs of all crabs. Tissues of (Table 2). The small crabs (10.0–40.0 mm) from L. red algae were also common (~70% of all samples) trabeculata had no animal tissues in their stomachs, in crab stomachs, except for the small crabs in L. while all other crabs, including the small crabs from trabeculata. The most common red algal genera M. pyrifera, also had ingested animal food. Diet occurring in stomach contents were Chondrus, Chon- composition significantly differed between crabs dria, Polysiphonia, Callophyllis, Pterosiphonia, Cryp- from both kelp habitats (ANOSIM, R Global=0.93; topleura, and Gastroclonium. Remains of a wide p<0.001) (Fig. 4). Within each habitat, the diet com- range of other taxa were also detected, particularly position of small crabs (10.0–40.0 mm) also differed Hydrozoa, Mollusca (small pieces of gastropod from that of the larger crabs (>40.0 mm); this differ- shells and bivalves, including occasionally complete ence was highly significant both in M. pyrifera individuals of Brachidontes granulatus (HANLEY (ANOSIM, R Global=0.68; p<0.001) and in L. trabe- 1843), Bryozoa, Polychaeta, Crustacea (parts of culata (ANOSIM, R Global=0.66; p<0.001) (Fig. 4). carapaces, pleopods, small shrimps, and amphi- The SIMPER routine showed that the two kelp spe- pods), and egg capsules of neogastropods (Mitrella cies contribute most to the dissimilarity in diets unifasciata (SOWERBY 1832), Nassarius sp., and between habitats and size categories (Table 3A,B). Crassilabrum crassilabrum (SOWERBY 1834)) (Fig. 3). Kelp tissues (M. pyrifera and L. trabeculata) were Daily feeding cycle highest in proportional volume (96.35.9 and 93.913.3) of the stomach contents for crabs from Taliepus marginatus showed a mostly nocturnal both kelp habitats, followed by red algae and other daily feeding cycle in both habitats (Fig. 5). There items (Table 1). In both kelp habitats, all individuals were no significant differences in feeding patterns

Table 2. Percentage of individuals of Taliepus marginatus per size class with algal tissues and other kinds of food (animal tissue and detritus) in the stomach contents, in both kelp habitats.

Habitats Macrocystis Lessonia Size range in mm (sample size) 10.0–40.0 (n=67) >40.0–92.0 (n=191) 10.0–40.0 (n=14) >40.0–92.0 (n=89) Percentage of crabs with algal tissue 100% 100% 100% 100% Percentage of crabs with other tissue 31% 46% 0% 58%

Invertebrate Biology vol. 132, no. 2, June 2013 Demography and feeding behavior of a kelp crab 139

Fig. 4. Nonmetric multidimensional scaling (nMDS) plot, resulting from classiﬁcation analysis (Bray–Curtis similari- ties) based on the wet weight (g) of item food (taxa/species) of Taliepus marginatus collected from Macrocystis pyrifera (circles) and Lessonia trabeculata (triangles). Data were square-root transformed for analysis. Open dots and triangles represent immature crabs, ﬁlled dots and triangles represent mature crabs. Number of crabs analyzed per habitat: M. pyrifera,n=97, and L. trabeculata,n=84. For SIMPER details see Table 3.

between the two kelp habitats (Fig. 5, Table 4). In decreased until 10:00. In L. trabeculata the feeding M. pyrifera, the peak stomach fullness was reached pattern was similar, but the peak stomach fullness at 22:00, remained high until 4:00, and then was reached at 4:00, after which it decreased sharply

Table 3. Results of SIMPER analysis of (dis)similarity in diet composition (A) within and between habitats, and (B) between immature and adult crabs (size ranges in mm, CW) from each kelp habitat (i & ii). Based on wet weight of food items (g wet weight per food item) from the stomach contents of Taliepus marginatus. Av. Biomass=mean biomass, Contribution %=contribution percentage of each taxon, Av. Diss.=average of dissimilarity, and Av. Sim.=average of similarity.

Av. Biomass (g ww) Contribution % Macrocystis Lessonia A. Habitats n=97 n=84 Between habitats (Av. Diss.=95.75) M. pyrifera 1.15 0 95.14 L. trabeculata 0 1.13 42.91 Rhodophyta 0.12 0.18 6.96 Within habitats Macrocystis (Av. Sim.=76.07) 1.15 93.84 Lessonia (Av. Sim.=68.77) 1.13 93.38 B. Crab sizes (range mm) 10.0–40.0 >40.0–92.0 ii-a. Macrocystis (Av. Diss.=43.40) n=14 n=83 M. pyrifera 0.54 1.25 74.15 Rhodophyta 0.06 0.13 11.92 Crustacea 0.05 0.01 5.93 ii-b. Lessonia (Av. Diss.=59.63) n=12 n=72 L. trabeculta 0.37 1.25 77.96 Rhodophyta 0 0.22 15.54

Invertebrate Biology vol. 132, no. 2, June 2013 140 Jofre, Ortiz, & Thiel

observed in adult crab stomachs may provide energy for reproductive purposes. Overall, these results suggest that the role of the two kelp habitats for T. marginatus varies during the life cycle of the kelp crabs, with M. pyrifera tending to have nursery function and L. trabeculata being more suitable as reproductive habitat.

Habitat use Many decapod crustaceans have changing habitat requirements during their life cycle. Our field data showed differences in the size distribution and den- Fig. 5. Relative stomach contents (in % body wet weight) sity of T. marginatus between the two kelp habitats, of Taliepus marginatus over a 24 h cycle in each kelp hab- which can be attributed to two possible causes. itat. Means and standard deviation (vertical bars) calcu- First, the two kelp species differ in habitat structure lated from 20 crab samples at each sampling hour. (Vega et al. 2005; Villegas et al. 2008), and second, Different letters indicate significant differences (p<0.001) the M. pyrifera kelp forests, due to their density, between sampling hours. For ANOVA details see complexity, and proximity to the coast, could play Table 4. an important role as nursery habitat for smaller crab individuals. In a field study on the most con- Table 4. Results from the statistical analysis of relative spicuous and abundant brachyuran species inhabit- stomach content volume (SC%) using two-factor analysis ing the shallow rocky subtidal zone in central Chile, of variance (ANOVA), with the factors Time (hour: Pardo et al. (2007) found that settlement and 16:00, 22:00, 4:00, and 10:00) and Site (habitat: Macrocys- recruitment of T. dentatus (the congener of T. mar- tis and Lessonia). ginatus) occurred mainly on shallow subtidal algal turfs. Furthermore, they could not find larger juve- Source df MS F value p-value niles in any of the habitats surveyed (e.g., algal turf, Hour 3 11.262 22.808 <0.001 cobbles, and shell hash). This might be due to an Habitat 1 1.886 3.820 0.052 ontogenetic habitat shift from turf algae to adjacent Hour: Habitat 3 1.448 2.934 <0.05 kelp forests. Small recruits find shelter and protec- Residuals 152 0.493 tion in shallow subtidal seaweeds (Pardo et al. 2007), which even may offer chemical defenses against fish predators (Palma et al. 2011). until the morning (10:00). In both kelp habitats, the Rockfishes are presumed the major predators of set- two night samplings differed significantly from the tlers and juveniles of most invertebrates in these day samplings (Fig. 5, Table 4). kelp habitats (Farina~ et al. 2008). Given the complex structure of M. pyrifera forests, juvenile crabs Discussion might escape predation in this habitat, enhancing their recruitment, as also reported for other inverte- The size distribution of the kelp crab Taliepus brate species inhabiting this kelp (Almanza et al. marginatus differed between kelp forests dominated 2012). by Macrocystis pyrifera and Lessonia trabeculata. The presence of the largest (and mostly repro- While the higher abundance, especially of smaller ductive) individuals of T. marginatus in the deeper crabs, in M. pyrifera forests suggests that this kelp (seaward) L. trabeculata kelp forests has also been could favor the survival and growth of juvenile reported by Palma et al. (2011), supporting our and subadult crabs, the higher proportion of large interpretation that this kelp has an important males and of ovigerous females in L. trabeculata reproductive function for kelp crabs. Similar obser- indicates that this kelp species has an important vations have also been reported for Pugettia pro- reproductive function for T. marginatus. In both ducta (RANDALL 1840) from California kelp forests kelp habitats, kelp tissues were the most abundant (Hines 1982; Hultgren & Stachowicz 2010). The food item of T. marginatus, but adult crabs (and size distribution of P. producta varies from the low small crabs in M. pyrifera) also ingested other food intertidal zone (Mastro 1981) to the deeper sections items. The larger proportion of animal food of the kelp forest (Hines 1982). Individuals of this

Invertebrate Biology vol. 132, no. 2, June 2013 Demography and feeding behavior of a kelp crab 141 species are apparently recruiting into intertidal algal diet may be 30% lower than those from an Phyllospadix zones, and subsequently migrate into animal diet (Wolcott & Wolcott 1984). Therefore, the subtidal kelp forests (Hines 1982). Several stud- a supplement of animal protein in the diet might ies on P. producta point out that adult crabs (52– benefit growth and possibly also the reproductive 72 mm CW) prefer kelps as reproductive habitat output of kelp crabs. (Hines 1982; Wicksten & Bostick 1983; Hultgren & Our results showed differences in the diet of T. Stachowicz 2010). Therefore, the predominance of marginatus with respect to crab size, similar to small recruits in nearshore habitats (turf algae, sea- results reported for many other crab species (Hines grass, cobble) and of reproductive individuals in 1982; Stevens et al. 1982; Choy 1986; Woods 1993; the seaward, deeper kelp habitats might be a gen- Woods & McLay 1996). In majid crabs, body size is eral pattern for kelp crabs (Hines 1982; Hultgren a limiting factor during resource utilization (Hines & Stachowicz 2010; Palma et al. 2011; this study). 1982). Different sizes at sexual maturity are also a The ultimate causes (e.g., food availability, shelter, reflection of species-specific abilities to use micro- lower competition, larval dispersal) for this appar- habitats (e.g., crevices) as refuges from predators ent habitat shift during ontogeny need to be inves- (Hines 1982). In our study, the analysis of diet com- tigated in future studies. position within each habitat suggested that immature crabs differ from adults in the use of the kelps. As members of the genus Taliepus live on the stipes Feeding ecology and blades of large kelps (Santelices 1989), kelp Individuals of T. marginatus from both habitats architecture may constrain their foraging behavior, consumed large amounts of kelp tissues (M. pyrif- especially of juvenile and subadult individuals. era or L. trabeculata), but they differed in the While there are extensive open spaces under the proportion of other food items. High prevalence of arborescent L. trabeculata, the closely spaced blades the habitat-forming algal species in stomachs has of M. pyrifera offer visual protection along the sti- also been observed for other majid species (Leigh- pes all the way down to the holdfasts (Villegas et al. ton 1966; Aracena 1971; Hines 1982; Woods 1993), 2008), allowing crabs to move up and down within and for other species associated with kelp habitats, the protective cover of the canopy. As the majority such as echinoids, gastropods, and fishes (Perez- of nonkelp food is found in the kelp holdfasts and Matus et al. 2012). Species that feed on habitat- on the surrounding seafloor (Santelices 1989; Thiel forming kelps obtain certain benefits, which include &Vasquez 2000; Ortiz 2008; Villegas et al. 2008), a large biomass and a predictable availability of it is likely that in M. pyrifera all crab sizes des- resources (Leighton 1966; Wolcott & O′Connor cend to the bottom to obtain these food items, 1992; Daly & Konar 2010; Hultgren & Stachowicz while the blade-less stipes of L. trabeculata restrict 2010). the movement of juvenile and subadult crabs. This Examination of the natural diet of T. marginatus interpretation is also supported by the fact that showed that in general it can be considered an her- small kelp crabs have a substantially smaller bivorous species, but that some individuals (espe- activity range than adult crabs (Hultgren & Stac- cially adults) also include animal material in their howicz 2010). diet. Other majid crabs, such as Loxorhynchus As kelp forest systems have high trophic com- crispatus STIMPSON 1857, Pugettia richii DANA 1851, plexity (e.g., multiple connections between a wide Notomithrax ursus (HERBST 1788), Taliepus denta- diversity of consumers and resources), development tus, and Eurynolambrus australis H. MILNE of adaptive foraging behaviors is important. Most EDWARDS &LUCAS 1841 also consumed some ani- organisms exhibit specific feeding rhythms mal material together with the algae dominating (Hines 1982; Jesse 2001; Kronfeld-Schor & Dayan their diets (Hines 1982; Manriquez & Cancino 2003; Hultgren & Stachowicz 2010). Despite the 1991; Woods 1993; Woods & McLay 1996; Cumil- differences in kelp forest structures, members of laf 2010). Animal food may play a role as a critical T. marginatus showed a similar daily feeding cycle nutrient supplement (Wolcott & O′Connor 1992) in both kelp habitats, with a mostly nocturnal for growth and reproduction. Algal tissues are feeding behavior. The tendency of T. marginatus to poor sources of essential amino acids, vitamins, feed more at night than during the day may also and sterols, compared with animal material. Nitro- help mitigate predation risks, very similar to gen is considered a limiting nutrient for herbivo- what has been suggested for other species of kelp rous crabs (Wolcott & O′Connor 1992). For crabs (Aris et al. 1982; Hultgren & Stachowicz instance, nitrogen assimilation efficiencies from an 2010).

Invertebrate Biology vol. 132, no. 2, June 2013 142 Jofre, Ortiz, & Thiel

Fig. 6. Hypothetical model of the life cycle of kelp crabs Taliepus spp., highlighting the importance of different species of macroalgae. Based on results from the present study and from Pardo et al. (2007) and Palma et al. (2011).

to several anonymous reviewers and to Lucas Eastman Conclusions and outlook for their constructive comments, which were of great help Kelp serves as both an important foundation spe- in improving the manuscript. During the preparation of cies and a significant energy source for much of the this manuscript, David Jofre Madariaga was partially nearshore food web (Dayton 1985; Graham 2004), supported by the CONICYT fellowship program (Becas para Estudios de Magister). This study was financed by both at the community and ecosystem level (Ortiz FONDECYT 1040293. 2008, 2010). The present results on habitat use and feeding behavior demonstrate that T. marginatus References uses the macroalgae as food and possibly as refuges during most part of its benthic life cycle (Fig. 6). Almanza V, Buschmann AH, Hernandez-Gonz alez MC, The results confirm that T. marginatus is an omni- & Henrıquez LA 2012. Can giant kelp (Macrocystis py- vore in both types of kelp habitat, maintaining a rifera) forests enhance invertebrate recruitment in strong dependence on kelp tissues throughout their southern Chile? Mar. Biol. Res. 8: 855–864. lives, albeit adding other food items as they grow. Almeida MJ, Flores A, & Queiroga H 2008. Effect of Both M. pyrifera and L. trabeculata are subject to crab size and habitat type on the locomotory activity > of juvenile shore crabs, Carcinus maenas. Estuar. Coast. intensive exploitation ( 70,000 tons in 2010) along – the Chilean coast (SERNAPESCA 2010). This has Shelf Sci. 80: 509 516. Amaral V, Cabral HN, & Paula J 2008. Implications of led to a notable reduction in the density and cover- habitat-specific growth and physiological condition on age of kelp beds. In addition, the effects from El juvenile crab population structure. Mar. Freshw. Res. ~ Nino Southern Oscillation (ENSO) also alter the 59: 726–734. extent of kelp beds (Vega et al. 2005; Gaymer et al. Antezana T, Fagetti E, & Lopez MT 1965. Observaciones 2010), even causing local kelp extinctions. Because bioecologicas en decapodos comunes de Valparaıso. these kelp forests are habitats for the reproductively Rev. Biol. Mar. 12: 1–60. active populations of T. marginatus, the continued Aracena O 1971. Algunos aspectos de la biologıa de la persistence of both kelps is likely to be important poblacion de Taliepus dentatus (Milne Edwards) 1834, for the population dynamics of this kelp consumer. en Caleta Leandro, Talcahuano. Tesis para optar al tıtulo de licenciado en Biologıa, Universidad de Con- cepcion, Chile. 145 pp. Acknowledgments. We thank Paula Ruz, Rosita Cha- Aris JP, Eisemann AD, & Moulton L 1982. The occur- vez, Manuel Rojo, Luis Rojas, Claudio Silva, Rodrigo rence of Pugettia richii (Crustacea: Decapoda) on Cys- Saavedra, Leonardo Campos, and Carlos Alvarez for toseira osmundacea follows a diel pattern. Bull. Mar. their help in the fieldwork. In particular, we are grateful Sci. 32: 243–249.

Invertebrate Biology vol. 132, no. 2, June 2013 Demography and feeding behavior of a kelp crab 143

Baumann JR & Kwak TJ 2011. Trophic relations of Kilar JA & Lou RM 1986. The subtleties of camouflage introduced flathead catfish in an Atlantic river. Trans. and dietary preference of the decorator crab, Micro- Am. Fish. Soc. 140: 1120–1134. phrys bicornutus Latreille (Decapoda: Brachyura). J. Choy SC 1986. Natural diet and feeding habits of the Exp. Mar. Biol. Ecol. 101: 143–160. crabs Liocarcinus puber and L. holsatus (Decapoda: Kronfeld-Schor N & Dayan T 2003. Partitioning of time Brachyura: Portunidae). Mar. Ecol. Prog. Ser. 31: 87– as an ecological resource. Annu. Rev. Ecol. Evol. Syst. 99. 34: 153–181. Clarke KR 1993. Non-parametric multivariate analyses of Leighton DL 1966. Studies of food preference in algivor- changes in community structure. Aust. J. Ecol. 18: ous invertebrates of southern California kelp beds. Pac. 117–43. Sci. 20: 104–113. Clarke KR & Gorley RN 2006. PRIMER v6: User Man- Manriquez PHR & Cancino JM 1991. Depredacion de ual/Tutorial. PRIMER-E, Plymouth. 187 pp. Membranipora isabelleana (Bryozoa) por Taliepus dent- Cumillaf JP 2010. Respuestas fisiologicas a la carnivorıa y atus (Crustacea: Decapoda). Rev. Biol. Mar. 26: herbivorıa durante el desarrollo bentonico del cangrejo 309–323. Taliepus dentatus (Milne Edwards, 1834) (Decapoda, Mastro E 1981. Algal preferences for decoration by the Majidae), una aproximacion bioenergetica. Tesis para Californian kelp crab, Pugettia producta (Randall) optar al tıtulo de ingeniero en Acuicultura, Universidad (Decapoda; Majidae). Crustaceana 41: 64–70. Austral de Chile, Puerto Montt, Chile. 61 pp. Novak M 2004. Diurnal activity in a group of Gulf of Daly B & Konar B 2010. Temporal trends of two spider Maine decapods. Crustaceana 77: 603–620. crabs (Brachyura: Majoidea) in nearshore kelp habitats Ortiz M 2008. Mass balanced and dynamical simulations in Alaska, USA. Crustaceana 83: 659–669. of trophic models of kelp ecosystems near the Mejill- Dayton PK 1985. Ecology of kelp communities. Annu. ones Peninsula of northern Chile (SE Pacific): compar- Rev. Ecol. Syst. 16: 215–45. ative network structure and assessment of harvest Duffy JE & Hay ME 1990. Seaweed adaptations to her- strategies. Ecol. Model. 216: 31–46. bivory. Bioscience 40: 368–376. ———— 2010. Dynamic and spatial model of kelp forest Escribano R, Rosales SA, & Blanco JL 2004. Under- of Macrocystis integrifolia and Lessonia trabeculata (SE standing upwelling circulation off Antofagasta (north- Pacific) for assessment harvest scenarios: short-term ern Chile): a three-dimensional numerical-modeling responses. Aquat. Conserv.: Mar. Freshw. Ecosyst. 20: approach. Cont. Shelf Res. 24: 37–53. 494–506. Farina~ JM, Palma AT, & Ojeda FP 2008. Subtidal kelp Palma AT, Soto-Gamboa M, & Ojeda FP 2011. Ontoge- communities off the temperate Chilean coast. In: Food netic habitat shift of an herbivorous crab: a chemically Webs and the Dynamics of Marine Reefs, Chapter 4. mediated defense mechanism? Rev. Biol. Mar. Ocea- McClanahan TR & Branch GM, eds., pp. 79–102. nogr. 46: 329–338. Oxford University Press, NewYork. Pansch C, Gomez I, Rothausler€ E, Veliz K, & Thiel M Gaymer CF, Palma AT, Vega JMA, Monaco CJ, & 2008. Species-specific defense strategies of vegetative Henrıquez LA 2010. Effects of La Nina~ on recruitment versus reproductive blades of the Pacific kelps Lessonia and abundance of juveniles and adults of benthic com- nigrescens and Macrocystis integrifolia. Mar. Biol. 155: munity-structuring species in northern Chile. Mar. 51–62. Freshw. Res. 61: 1185–1196. Pardo LM, Palma AT, Prieto C, Sepulveda P, Valdivia Graham MH 2004. Effects of local deforestation on the I, & Ojeda FP 2007. Processes regulating early post- diversity and structure of southern California giant settlement habitat use in a subtidal assemblage of kelp forest food webs. Ecosystems 7: 341–357. brachyuran decapods. J. Exp. Mar. Biol. Ecol. 344: Gutow L, Long JD, Cerda O, Hinojosa IA, Rothausler€ E, 10–22. Tala F, & Thiel M 2012. Herbivorous amphipods inhabit Perez-Matus A, Pledger S, Dıaz FJ, Ferry LA, & Vasquez protective microhabitats within thalli of giant kelp Mac- JA 2012. Plasticity in feeding selectivity and trophic rocystis pyrifera. Mar. Biol. 159: 141–149. structure of kelp forest associated fishes from northern Hines AH 1982. Coexistence in a kelp forest: size, popula- Chile. Rev. Chil. Hist. Nat. 85: 29–48. tion dynamics, and resource partitioning in a guild of R Development Core Team 2010. R: A Language and spider crabs (Brachyura, Majidae). Ecol. Monogr. 52: Environment for Statistical Computing. R Foundation 179–198. for Statistical Computing, Vienna, Austria. ISBN 3- Hultgren KM & Stachowicz JJ 2010. Size-related habitat 900051-07-0, URL http://www.R-project.org/. shifts facilitated by positive preference induction in a Retamal MA 1977. Los crustaceos decapodos chilenos de marine kelp crab. Behav. Ecol. 21: 329–336. importancia economica. Gayana Zool, 39: 1–49. Jesse S 2001. Comparative ecology of sympatric brachyu- Santelices B 1989. “Algas Marinas de Chile.” Distri- ran crab species in the shallow subtidal of the Pacific bucion, Ecologıa, Utilizacion y Diversidad. Ediciones coast of north Chile and their importance for the arti- Universidad Catolica de Chile, Santiago, Chile. 399 pp. sanal fishery Puerto Aldea. PhD Dissertation, Univer- SERNAPESCA 2010. Anuario Estadıstico de Pesca. Ser- sity of Bremen, Germany. 125 pp. vicio Nacional de Pesca, Santiago, Chile.

Invertebrate Biology vol. 132, no. 2, June 2013 144 Jofre, Ortiz, & Thiel

Smith SDA, Simpson RD, & Cairns SC 1996. The macro- northern Chile and the 1997–98 ENSO. J. Appl. Phy- faunal community of Ecklonia radiata holdfasts: col. 18: 505–519. description of the faunal assemblage and variation Vega JMA, Vasquez JA, & Buschmann AH 2005. Popu- associated with differences in holdfast volume. Aust. J. lation biology of the subtidal kelps Macrocystis integri- Ecol. 21: 81–95. folia and Lessonia trabeculata (Laminariales, Stachowicz JJ & Hay M 1999. Reducing predation through Phaeophyceae) in an upwelling ecosystem of northern chemically mediated camouﬂage: indirect effects of plant Chile: interannual variability and El Nino~ 1997–1998. defenses on herbivores. Ecology 80: 495–509. Rev. Chil. Hist. Nat. 78: 33–50. Steneck R, Graham M, Bourque B, Corbett D, Erlandson Villegas MJ, Laudien J, Sielfeld W, & Arntz WE 2008. J, Estes J, & Tegner M 2002. Kelp forest ecosystems: Macrocystis integrifolia and Lessonia trabeculata (Lami- biodiversity, stability, resilience and future. Environ. nariales; Phaeophyceae) kelp habitat structures and Conserv. 29: 436–459. associated macrobenthic community off northern Chile. Stevens BG, Armstrong DA, & Cusimano R 1982. Feed- Helgol. Mar. Res. 62: 33–43. ing habits of the Dungeness crab Cancer magister as Wicksten MK & Bostick CR 1983. Migration of kelp determined by the index of relative importance. Mar. crabs (Pugettia producta) at San Pedro, California. J. Biol. 72: 135–145. Crust. Biol. 3: 364–366. Thiel M & Vasquez JA 2000. Are kelp holdfasts island on Wolcott DL & O′Connor NJ 1992. Herbivory in crabs: the ocean ﬂoor?- indication for temporarily closed adaptations and ecological considerations. Am. Zool. aggregations of peracarid crustaceans. Hydrobiologia 32: 370–381. 440: 45–54. Wolcott DL & Wolcott TG 1984. Food quality and can- Vasquez JA 1992. Lessonia trabeculata, a subtidal bottom nibalism in the red land crab, Gecarcinus lateralis. kelp in northern Chile: a case study for a structural Physiol. Zool. 57: 318–324. and geographical comparison. In: Coastal Plant Com- Woods CMC 1993. Natural diet of the crab Notomithrax munities of Latin America. Seeliger U, ed. 392 pp. Aca- ursus (Brachyura, Majidae) at Oaro, South Island, New demic, New York. Zealand. N. Z. J. Mar. Freshw. Res. 27: 309–315. Vasquez JA & Buschmann AH 1997. Herbivore-kelp Woods CMC & McLay CL 1996. Diet and cryptic col- interactions in Chilean subtidal communities: a review. ouration of the crab Eurynolambrus australis (Brachy- Rev. Chil. Hist. Nat. 70: 41–52. ura: Majidae) at Kaikoura. Crustac. Res. 25: 34–43. Vasquez JA, Vega JMA, & Buschmann AH 2006. Long Zar JH 1996. Biostatistical Analysis, 3rd ed. Prentice term variability in the structure of kelp communities in Hall, Upper Saddle River, New Jersey. 662 pp.

Invertebrate Biology vol. 132, no. 2, June 2013