Marine Ecology. ISSN 0173-9565
ORIGINAL ARTICLE Diversity of faunal assemblages associated with ribbed mussel beds along the South American coast: relative roles of biogeography and bioengineering Roger D. Sepulveda 1,2, Patricio A. Camus3 & Carlos A. Moreno1
1 Instituto de Ciencias Ambientales y Evolutivas, Facultad de Ciencias, Universidad Austral de Chile, Valdivia, Chile 2 Research Center, South American Research Group on Coastal Ecosystems (SARCE), Caracas, Venezuela 3 Centro de Investigacion en Biodiversidad y Ambientes Sustentables, Departamento de Ecolog ıa, Universidad Catolica de la Sant ısima Concepcion, Concepcion, Chile
Keywords Abstract Associated fauna; biogeography; ecosystem engineering; historical effect; Southeastern Benthic organisms are among the most diverse and abundant in the marine Pacific; Southwestern Atlantic. realm, and some species are a key factor in studies related to bioengineering. However, their importance has not been well noted in biogeographic studies. Correspondence Macrofaunal assemblages associated with subtidal beds of the ribbed mussel Roger D. Sepulveda, Instituto de Ciencias (Aulacomya atra) along South America were studied to assess the relationship Ambientales y Evolutivas, Facultad de between their diversity patterns and the proposed biogeographic provinces in Ciencias, Universidad Austral de Chile, the Southeastern Pacific and Southwestern Atlantic Oceans. Samples from Casilla 567, Valdivia, Chile E-mail: [email protected] ribbed mussel beds were obtained from 10 sites distributed from the Peruvian coast (17°S) to the Argentinean coast (41°S). The sampling included eight beds Accepted: 25 May 2015 in the Pacific and two in the Atlantic and the collections were carried out using five 0.04 m2 quadrants per site. Faunal assemblages were assessed through clas- doi: 10.1111/maec.12301 sification analyses using binary and log-transformed abundance data. Variation in the size and density of mussels, and in the species richness, abundance and structure of their faunal assemblages were tested using a permutational multi- variate analysis of variance. Faunal assemblages showed a north–south latitudi- nal gradient along both the Pacific and Atlantic coasts. Binary and abundance data showed a difference in the resulting clustering arrangement of Pacific sites between 40°S and 44°S, indicating a pattern of continuity in the species distri- bution associated with biological substrates. At a regional scale, the distribution of species along the South American coast matched the general provincial pat- tern shown by prior studies, which show two biogeographic units on the Paci- fic coast separated by an intermediate (probably transitional) zone and a single province on the Atlantic coast extending up to Northern Argentina. Biological substrates such as ribbed mussel beds play an important ecological role by making a similar habitat type available on a large scale for a variety of inverte- brate species. Despite such habitat homogeneity, however, the associated fauna exhibit marked distribution breaks, suggesting strong constraints on dispersal. This therefore suggests that macrofaunal assemblages could possibly be used as biogeographic indicators.
determining their colonization success is the availability Introduction of physical and biological substrata. However, the impor- Benthic organisms are among the most diverse and abun- tance of available substrate for colonization has not been dant in the marine realm (Angel 1993), and a key factor given much attention in biogeographic studies. The heter-
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 943 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno ogeneity of physical substrata has long been known to upwelling can strongly affect the structure of mussel beds. have several effects on the provision of food, habitat and However, these factors have little or no impact on faunal refuge for benthic species (e.g. Underwood & Chapman composition, which instead, shows clear correspondence 1989; Chapman & Underwood 1994; Guichard & Bourget with regional and biogeographic patterns (Hammond & 1998; Camus et al. 1999). However, live substrata may Griffiths 2006; Cole & McQuaid 2010). play similar roles, particularly those species known as Mussels thus provide their associated fauna with a suite ‘ecosystem engineers’ (Jones et al. 1994) that develop of microenvironments, which constitute a relatively stable complex body architectures or three-dimensional aggrega- and homogenous habitat that is available at large scales tions. By generating suitable microhabitats that allow for (see Hammond & Griffiths 2006). Additionally, because feeding and reproduction of a wide variety of invertebrate of the stability and scale of this system, it is a suitable taxa, bioengineering may significantly affect the diversity model to assess the effects of large-scale environmental and distribution of species in the community and even gradients and/or biogeographic processes. Such analyses community membership (e.g. Bertness & Callaway 1994; are so far scarce and are mostly restricted to inter-tidal Jones et al. 1994; Seed 1996; Borthagaray & Carranza systems. This motivated us to conduct the first large-scale 2007; Kelaher et al. 2007a; Sepulveda et al. 2015). This is survey of subtidal populations of the ribbed mussel Au- of particular importance considering that the spatial or lacomya atra (Molina) along the coast of South America temporal scale of the biosubstrate exceeds that involved with the aim of characterizing their beds and their associ- in direct organismal interactions (Hastings et al. 2007). ated fauna in a biogeographic context. The complex beds assembled by marine mussels are an outstanding example of a widespread biogenic habitat The ribbed mussel A. atra hosting a high diversity of sessile and mobile animals. These animals can be epibiotic, epibenthic or endoben- Subtidal mussels have been an important food source for thic, and they form significantly distinct assemblages humans in South America since the Holocene (Carranza from that found on ambient substrata (Suchanek 1992; et al. 2009), but the beds of these organisms and their Thiel & Ullrich 2002; Buschbaum et al. 2009). On some associated fauna have not been studied in depth. The rocky intertidal shores, for instance, there can be high genus Aulacomya originated during the Paleogene, and redundancy of species hosted by different bioengineers South American fossils of A. atra date back to the Pleisto- (mussels versus algal turfs) in the same site (Kelaher et al. cene (~2 mya; Castilla & Guinez~ 2000); this suggests that 2007b), or by the same bioengineer mussel in different this mussel is a resilient component of the southern biota sites (Lee & Castilla 2012). On other shores, different as it has overcome Quaternary fluctuations (Camus 2001; mussel species can host somewhat different fauna, but the Thiel et al. 2007; Gordillo et al. 2008). The ribbed mussel abundance and biomass of this fauna is independent of has a disjointed distribution in the Southern Hemisphere, mussel identity or bed structure (e.g. Hammond & Grif- including Southern Africa (Castilla & Guinez~ 2000; Ham- fiths 2006). However, the same intertidal mussel can host mond & Griffiths 2006), the Falkland (Malvinas) Islands different fauna in beds assembled on rocky versus soft and Southern South America where it ranges from Central bottoms as these substrates are functionally different in Peru (Chimbote, ~9°S; Southeast Pacific) to Southern Bra- terms of the resources that they provide (Thiel & Ullrich zil (more likely to Uruguay: La Coronilla, ~34°S; South- 2002). In addition, while mussel beds on rocky shores west Atlantic) (Osorio et al. 1979; Castilla & Guinez~ 2000; usually host much higher species richness than surround- Lancellotti & V asquez 2000; Scarabino & Ortega 2004). ing substrata, comparisons among soft-bottom beds from The ribbed mussel occurs mainly in shallow habitats several regions (North Sea, Yellow Sea, Southern Chile (~0–25 m depth, with records of >100 m; Moreno & Moli- and Australia) showed that the faunal richness on soft- net 2013) on both hard substrata and coarse-grained sedi- bottom beds is often similar or even lower than outside ments. Furthermore, because of its large size (>155 mm the bed (Thiel & Ullrich 2002). However, at each of the length), this mussel has been intensely exploited through- four locations there was a markedly different species out most of its range. Such apparently high resilience composition (see Buschbaum et al. 2009 and references may be due to several life-history characteristics of this therein). These findings suggest that faunal species rich- organism, including the ability to maintain positive growth ness in mussel beds can be subject to multiple controls using ambient suspended organic matter (Stuart et al. that involve both environmental and local factors. In 1982), its extended breeding period and rapid gonad addition, these results highlight the importance of bioen- regeneration (Griffiths 1977), the dependence of reproduc- gineering itself over bioengineer identity in determining tion upon food availability rather than on temperature the composition of faunal assemblages. Indeed, key fac- (Jaramillo & Navarro 1995) and the high dispersal and fix- tors such as exploitation and mesoscale variation in ation abilities of its larvae (Uribe & Lopez 1980). This
944 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel latter characteristic is reflected, for instance, in the high physical features, which suggests that they are stronger gene flow and genetic similarity (95%) among its popula- determinants than previously thought. Along the Chilean tions in Southern Chile (Mena et al. 2001). coast, for instance, recent analyses (Hormazabal et al. By retaining sediments and shell fragments from live 2004) have detected a mesoscale oceanographic transition and dead mussels (Zaixso 2004; Thiel et al. 2007) and by between regions of low and high eddy kinetic energy attaining fecal production that may reach energy outputs located north and south of ~30°S, respectively. Further- equivalent to 90% of macroalgal production (Stuart et al. more, it has been shown that these are coincident with 1982), ribbed mussel beds represent a unique habitat that changes in mesoscale upwelling regimes and a break provides faunal assemblages with an abundant supply of between offshore versus coastal productivity (Thiel et al. organic matter and substrata. However, the bioceanic 2007). The biogeographic role of such transitions (e.g. as occurrence of ribbed mussels ranges from subtropical to a barrier to larval transport and dispersal) is increasingly subantarctic areas, and the increasing knowledge of the being acknowledged and may help to explain the marked biogeography of South America suggests there is little breaks in species richness and distribution as shown by likelihood that the effect of a homogeneous bioengineered widely different taxa, from littoral benthic groups with habitat on faunal composition may be overlaid on the high endemism levels at 30°S to the cephalopod fauna effect of large-scale factors, particularly when external formed entirely by non-endemic, wide-ranging, and forcings are involved. mostly eurithermic species (Thiel et al. 2007; Ib anez~ et al. 2009). The importance of such factors suggests that the diversity patterns of faunal assemblages inside ribbed Biogeography of South America: oceanographic patterns mussel beds should be similarly affected on a large scale. and transitions Recent studies about the biogeography and oceanography Biogeographic patterns of South American coasts have revealed the key roles played by topography, nearshore circulation and upwelling As recent advances shed new light on the biogeographic in nutrient enrichment, dispersal and recruitment patterns, processes and units along Southern South American which are processes that can now be better understood, coasts, the traditionally identified provinces remain con- particularly on mesoscales (Lagos et al. 2007; review by gruent for many different taxa (Fern andez et al. 2000; Thiel et al. 2007). On larger scales, in mid–low latitudes, Camus 2001; Mart ınez & del R ıo 2002; Thiel et al. 2007; the importance of physical-biological couplings and their Balech & Ehrlich 2008). However, one must acknowledge relationship with inter-annual variation and remote equa- discrepancies related to the idiosyncratic histories of par- torial forcing have been well established (Takesue et al. ticular taxa and species (e.g. Gonz alez & Moreno 2005; 2004; Thiel et al. 2007; Camus 2008). Additionally, in Haussermann & Forsterra 2005; C ardenas et al. 2009; mid–high latitudes, the importance of the Antarctic Cir- S anchez et al. 2011; Sepulveda & Ib anez~ 2012). For lit- cumpolar Current (ACC) and its interaction with physical toral benthic organisms in particular, the main biogeo- geography and basin-scale processes has been studied in graphic units (using their most frequent names in the depth (Ahumada et al. 2000; Camus 2001; Arhan et al. literature) recognized so far (but see Briggs & Bowen 2002; Thompson & Alder 2005; Sassi & Palma 2006; Thiel 2012 for fishes) are, from west to east (see Fig. 1): (i) the et al. 2007; Balech & Ehrlich 2008; Gat ıca et al. 2009). The warm-temperate Peruvian Province ranging from North- Southern Cone is dominated by cold flows derived from ern Peru (~5–6°S) to North-Central Chile (~30–33°S); the ACC. This movement of water reaches the Pacific coast (ii) a transitional, biogeographically heterogeneous area at ~41°S and branches into the equatorward Humboldt from ~30–33°S to Chilo e Island (~42°S; but see Lance- Current extending up to ~5°S. The poleward Cape Horn llotti & V asquez 1999; Haussermann & Forsterra 2009); Current then turns northeastwards to form the equator- (iii) the cold-temperate Magellanic Province, usually ward Falkland (Malvinas) Current in the Atlantic, which encompassing the South American cone from ~42°S collides with the warm poleward Brazilian current around (Southern Chile) to the Valdes Peninsula (~42.5°S, ~38–40°S. In addition, the Pacific–Atlantic connection Southern Argentina); (iv) the Argentinean Province from through the Strait of Magellan and the resulting equator- 42.5°S to Santa Catarina (~28°S, Southern Brazil), a ward Patagonian current (~52.5–38°S) in the Atlantic (Pi- rather transitional area characterized by the overlap of ola & Rivas 1997; Sassi & Palma 2006) most likely also subantarctic and subtropical components; and (v) the play important roles that remain understudied (Sielfeld & warm Brazilian Province, from 28°S to the Amazon riv- Vargas 1999; Gordillo et al. 2008). er’s mouth at the equator. Noteworthily, most provincial The major biogeographic features along Southern transitions appear to be associated with major oceanic South America appear to be closely linked to prominent and hydrologic features that are apparently related to
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 945 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno
sition and biogeographic zonation in Southern South America, considering the role of ribbed mussels as a bio- engineered habitat.
Material and Methods
Study area and sample collection During the austral summer of 2007 and 2008, we sam- pled 10 natural populations of the ribbed mussel A. atra from shallow subtidal sites along the Southeastern Pacific and Southwestern Atlantic coasts of Southern South America (Fig. 1), from west to east: (i) Peru: Ilo (17°400S, 71°250W); (ii) Chile: Iquique (21°190S, 70°050W), Antofagasta (23°450S, 70°280W), Coquimbo (29°570S, 71°220W), Valpara ıso (33°020S, 71°370W), Val- divia (39°460S, 73°180W), Melinka (43°540S, 73°450W), Punta Arenas (54°010S, 71°150W); (iii) Argentina: Puerto Madryn (42°460S, 64°560W), San Antonio Oeste (41°340S, 64°580W). Particularly in Chile, the selection of mussel beds was made to avoid the inclusion of stock-enhanced or re-introduced populations resulting from recent man- agement actions triggered by the intense exploitation of Fig. 1. Scheme of the main biogeographic units (delimited by +), coastal circulation patterns (filled arrows) and sample sites (filled A. atra in some areas (Jerez & Figueroa 2008). circles) of natural populations of ribbed mussel beds of Aulacomya All sampled mussel beds were located between 8 and atra along South American coasts. Provinces: PP, Peruvian Province; 25 m in depth on rocky substrates embedded in fine- and TA, Transitional Area; MP, Magellanic Province; AP, Argentinean medium-grain sandy matrices. Similar studies on mussel Province; BP, Brazilian Province. Currents: WWD, West Wind Drift; O- species smaller than A. atra (Tsuchiya & Nishihira 1985; HC, Oceanic Humboldt Current; C-HC, Coastal Humboldt Current; Thiel & Ullrich 2002; Buschbaum et al. 2009) have deter- CHC, Cape Horn Current; CCC, Coastal Counter Current; FC, mined that sampling units between 50 and 200 cm2 in Falkland (Malvinas) Current; BCC, Brazilian Counter Current. Localities: ILO, Ilo, Peru; IQU, Iquique; ANT, Antofagasta; COQ, area were appropriate for estimating faunal diversity. Coquimbo; VAP, Valpara ıso; VAL, Valdivia; MEL, Melinka; PAR, Punta However, the total length of adult ribbed mussels may Arenas, Chile; MAD, Puerto Madryn; SAO, San Antonio Oeste, often have a modal size close to 70 or 80 mm, and there- Argentina. fore we used a larger sampling unit (400 cm2) in order to minimize sampling biases. In each locality, five 20 9 20- boundary currents and oceanic gyres. This highlights the cm quadrats were randomly placed on a ribbed mussel latitudinal symmetry of the eastern and western coasts of bed, and mussel clumps were removed from each quad- the northern ends of the Magellanic Province. In addi- rat, fixed in a 4% formaldehyde solution and later pre- tion, recent studies have noted the impact of different served in a 70% ethanol solution. In the laboratory, each historical events, occurring from the late Tertiary to the quadrat sample was disaggregated over a white tray. We early Holocene, on traditional biogeographic discontinu- then measured the maximum length ( 1 mm) of the ances along the Chilean coast. They have shown that mussels and sorted all macrofaunal individuals >0.5 mm many common taxa are marked by phylogeographic and/ length, which were counted and identified to the species or distribution breaks despite the absence of major dis- level. Counts considered only individual organisms, persal barriers (see Thiel et al. 2007; Haye et al. 2014, excluding colonial groups and inter-connected individuals and references therein). such as Porifera, Hydrozoa and Bryozoa. At the time of In such a context, this study was aimed at assessing the sampling, however, the coasts of two Chilean sites (Co- geographic variation of ribbed mussel beds and the diver- quimbo and Valpara ıso) had recently undergone an sity patterns of their macrofaunal assemblages along the intense episodic exploitation by artisanal fishermen (see Southern South American coast between Southern Peru Discussion), causing a drastic reduction in the density and Southern Argentina, encompassing most of the and spatial coverage of ribbed mussels. As a side effect, Pacific–Atlantic range of the species. In particular, we the fauna remaining in the extant beds was extremely examined the relationship between faunal species compo- depauperate and provided highly deficient data, which
946 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel forced us to exclude these two sites from further analysis information for this fauna prevented us from knowing in order to avoid artifacts. whether these records corresponded to short-range (i.e. endemic) or undersampled species. Single records may introduce non-random biases across space (Valentine Data analyses et al. 2002) or lead to the overestimation of composi- The size (maximum length) and density (individu- tional differences among sites. Therefore, singletons were als 0.04 m 2) of ribbed mussels, as well as the species removed from the cluster analyses in order to avoid richness (number of species) and abundance (number of biased results. All probability values (Pperm) of significant individuals) of their associated fauna were compared model fits were derived from a pseudo-F distribution cal- among sites using a one-way permutational analysis of var- culated using 10,000 permutations of the original data iance (PERMANOVA) with an unrestricted permutation of set; when the simulated permutations were <1000, the the Euclidean distance matrices of the raw data (Anderson probability value was obtained by Monte-Carlo simula- et al. 2008). The association of latitude with mussel size, tions (PMC). All analyses were performed using PERMA- mussel density, and species richness and abundance was NOVA+ for the PRIMER statistical package (Clarke & evaluated using Spearman rank correlations (rS). Gorley 2006; Anderson et al. 2008). The geographic variation of faunal assemblages associ- ated with ribbed mussel beds was assessed through cluster Results analysis using the unweighted pair group method with arithmetic mean (UPGMA) algorithm (Quinn & Keough Size and density of ribbed mussels 2002). The analyses were performed both on a presence- absence matrix using the Jaccard similarity index and on The mean size (mm) of the mussels (Fig. 2A) was signifi- an abundance matrix using the Bray–Curtis similarity cantly different among sites (Table 1a), varying (mean SD) index. For the abundance matrix, the data were log-trans- from 36.81 17.49 in Ilo to 117.07 21.33 in Punta formed due to the high dispersion of values. Significant Arenas, with an overall mean of 66.79 6.76 ( SE). The clusters were identified by conducting a similarity profile density (individuals 0.04 m 2) of ribbed mussels (Fig. 2B) test (SIMPROF) using 10,000 permutations. In addition, was highly variable and also differed significantly among a one-way PERMANOVA was used to test for differences sites (Table 1b), ranging (mean SD) from 1.20 0.45 in in the community structure among biogeographic regions Coquimbo to 77.33 46.36 in Ilo, with an overall for the presence-absence and log-transformed abundance mean local density (n = 10 populations) of 22.57 7.29 data matrices (Anderson et al. 2008). Finally, our field ( SE). Latitude was not significantly associated with mussel data included several single-site records, but the lack of size (rS = 0.08, n = 10, P = 0.830) or with density
AB
Fig. 2. Population characteristics of the ribbed mussel beds of Aulacomya atra and CD community variables for their associated fauna. (A): Size (maximum length; mean 1 SD) and (B): density (individuals 0.04 m 2; mean 1 SD) of ribbed mussels beds. (C): Species richness (number of species; mean 1 SD) and (D): abundance (number of individuals; mean 1 SD) of associated fauna of ribbed mussels in sampling sites along the South American coast. For abbreviations, see Fig. 1. Gray dotted lines separate sites between oceans.
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 947 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno
Table 1. Permutational analysis of variance outputs for ribbed mussels (mean size and density) and community (species richness and abundance) variables among sites; and for the community structure [presence-absence (P/A), and log-transformed abundance data] among biogeographic regions.
Variables source of variation SS df MS pseudo-F Pperm
(a) mussel size site 21,317 9 2369 15.227 <0.001 residual 6222 40 156 (b) mussel density site 18,529 9 2059 13.812 <0.001 residual 5963 40 149 (c) species richness site 5129 7 733 24.838 <0.001 residual 944 32 30 (d) abundance site 511,510 7 73,073 22.618 <0.001 residual 103,380 32 3231 (e) community structure (P/A) region 12,489 2 6245 2.849 0.026a residual 10,958 5 2192 (f) community structure (log) region 11,715 2 5858 3.189 0.022a residual 9184 5 1837
SS = Sum of squares; df = degree of freedom; MS = Mean squares. a Probability values were obtained through Monte-Carlo simulations.
(rS = 0.26, n = 10, P = 0.47) of the mussels, although size abundances of Aciculata and Amphipoda were also and density were negatively correlated (rS = 0.66, n = 10, found in Valdivia. Furthermore, the orders Arcoida, P = 0.038). Solemyoidea and Veneroidea (Mollusca) were highly rep- Along the geographic range of the ribbed mussel, pop- resented in the Ilo site in comparison to the other sites; ulation abundance was highest in the northernmost bed moreover, the order Dentallida was only found in the in the Pacific (Ilo), lowest around the range center in the northernmost sample site. By contrast, these abundant Pacific (Coquimbo and Valpara ıso, Central Chile), and faunal components found in the northern sites had a close to average in the Atlantic (San Antonio Oeste and very low representation in the southern site of Punta Puerto Madryn), although population abundance Arenas. However, other orders such as Neomeniamor- remained below average in most sites. The geographic pha, Myoida (Mollusca), Valvatida (Echinodermata) and distribution of abundance was clearly not normal and Terebratulida (Brachiopoda) were only present in the there was no apparent trend (Fig. 2B). Punta Arenas site, but at low abundance. On the Atlan- tic coast, low to medium abundances of Amphipoda, Aciculata and Archaeogastropoda were found at Puerto Species richness and abundance of the associated fauna Madryn; nevertheless, the abundances of the most com- Excluding the fauna associated with the mussel beds from mon groups in the Atlantic were less than 50% of those Coquimbo and Valparaiso, we found a total of 7795 indi- reached in northernmost sites on the Pacific coast viduals belonging to 197 species, 44 orders and 11 phyla. (Table 2). Table 2 summarizes the taxonomic composition of faunal From the total of 197 species, 61 (31%) were recorded assemblages at the phyla and order level, showing the only from a single site. Mussel beds along the Chilean species richness of orders and their abundance at each coast, particularly in Antofagasta, Iquique and Valdivia, mussel bed. had higher species richness than those in Peru and Along the South American coast, the orders most rep- Argentina (Fig. 2C). Faunal abundance (total number of resented in terms of number of species were Decapoda individuals) was also higher in Antofagasta, Iquique and (Arthropoda) and Aciculata (Annelida) with 27 and 21 Valdivia, while lower values were found in San Antonio species, respectively. Secondly, between 10 and 14 species Oeste, Argentina, and Punta Arenas, Chile (Fig. 2). There represented the orders Archaeogastropoda, Neoloricata were significant differences among sites in both species (Mollusca), Canalipalpata (Annelida), Amphipoda and richness and abundance (Table 1c and d), but neither of Isopoda (Arthropoda). The most abundant orders the two variables exhibited a significant linear latitudinal among all of the sites were Aciculata (21.33%), Amphi- trend ( 0.14 ≤ rS ≤ 0.62, 0.10 ≤ P ≤ 0.75, n = 8). In poda (13.87%), Neotaeniglossa (10.98%) and Ophiurida addition, the richness and abundance of the associated
(10.12%). Particularly in the northernmost sites, there fauna were positively correlated (rS = 0.76, P = 0.028, were high abundances of Aciculata, Neotaenioglossa and n = 8), although neither of them showed a significant Ophiurida (i.e. Ilo, Iquique and Antofagasta); high association with the size or density of ribbed mussels
948 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel
Table 2. Taxonomic composition (phyla, orders) of faunal assemblages associated with ribbed mussel beds in South America, indicating the total number of species (S) per order, and the local and total (last column) abundance (number of individuals) per order and site. For abbreviations, see Fig. 1.
Intermediate Peruvian Province areaa Magellanic Province phylum order S ILO IQU ANT VAL MEL PAR MAD SAO total
Cnidaria Actinaria 4 16 123 160 1 0 0 14 36 350 Nemertea Heteronemertea 2 0 0 3 16 1 11 16 0 47 Nematoda Enoplida 1 0 0 0 0 5 1 0 8 14 Mollusca Arcoida 2 112 24 12 0 0 0 0 0 148 Mytiloida 4 0 0 19 52 6 0 2 0 79 Solemyoida 1 13 1 1 0 0 0 0 0 15 Veneroida 8 38 6 14 21 1 4 0 0 84 Pholadomyoida 1 0 0 0 10 0 0 0 0 10 Myoida 1 0 0 0 0 0 1 0 0 1 Octopoda 1 0 0 0 0 0 0 0 1 1 Archaeogastropoda 12 8 13 163 8 98 5 120 12 427 Neotaenioglossa 8 272 106 230 78 154 1 7 8 856 Neogastropoda 9 67 58 146 26 76 8 0 1 382 Patellogastropoda 3 0 0 4 1 9 2 7 0 23 Nudibranchia 3 0 0 0 0 11 1 1 0 13 Neoloricata 11 5 52 4 8 33 11 2 3 118 Dentallida 1 8 0 0 0 0 0 0 0 8 Neomeniamorpha 2 0 0 0 0 0 5 1 1 7 Sipunculida Golfingiaformes 1 0 0 0 4 1 0 0 0 5 Echiura Echiuroinea 1 0 0 0 0 0 0 0 1 1 Annelida Aciculata 21 95 483 256 449 117 92 148 23 1663 Canalipalpata 12 42 82 58 228 57 9 45 3 524 Scolecida 3 0 8 91 1 1 1 4 1 107 Arthropoda Pycnogonomorpha 2 0 0 0 67 0 0 5 1 73 Nymphomorpha 1 0 0 0 0 1 1 0 0 2 Insecta indet. 1 0 0 0 0 0 1 0 0 1 Brachypoda 1 0 1 0 1 0 0 0 0 2 Sessilia 1 14 13 61 125 1 0 0 0 214 Ostracoda indet. 2 1 1 4 1 1 2 2 0 12 Decapoda 27 15 37 56 21 9 2 68 15 223 Amphipoda 14 27 217 123 482 7 11 184 30 1081 Isopoda 10 5 2 2 28 2 1 31 6 77 Tanaidacea 2 0 0 1 4 22 0 114 2 143 Echinodermata Valvatida 1 0 0 0 0 0 1 0 0 1 Spinulosida 2 1 4 5 1 0 8 0 1 20 Forcipulatida 4 0 1 1 0 0 13 5 2 22 Ophiurida 3 245 258 187 25 17 33 14 10 789 Arbacioida 2 1 5 37 0 40 0 24 11 118 Echinoida 2 0 0 0 7 9 10 21 1 48 Dendrochirotida 2 0 4 0 3 0 1 0 0 8 Brachiopoda Acrotretida 1 0 8 9 4 0 0 0 0 21 Terebratulida 1 0 0 0 0 0 1 0 0 1 Chordata Enterogona 3 0 0 0 14 0 1 3 4 22 Pleurogona 3 0 0 0 30 0 1 0 3 34 a See also Lancellotti & Vasquez (1999), Camus (2001) and Haussermann & Forsterra (2009) for variations in the biogeographic boundaries.
( 0.41 ≤ r ≤ 0.50, 0.20 ≤ P ≤ 0.82, n = 8). Notewor- S Biogeographic pattern of faunal assemblages thily Noteworthy, the three richest and most abundant faunal assemblages occurred in beds of low or intermedi- The cluster analyses (excluding Coquimbo and Valpa- ate mussel density. raiso) revealed a differential classification of the faunal
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 949 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno assemblages along the geographic distribution of the Chile; (ii) Melinka and Punta Arenas (30.7%), in Austral ribbed mussel (Fig. 3). The results based on presence- Chile; and (iii) San Antonio Oeste and Puerto Madryn absence data showed three clusters differentiating among (50.6%), in Argentina. In this case, Valdivia was not clas- Northern Pacific and Southern Pacific sites, and these sified with any of the other sites although it appeared to latter differed from the Atlantic site (Fig. 3A): (i) Ilo, be more similar to northern sites. The analysis showed Iquique and Antofagasta, in Peru to Northern Chile (sim- significant differences in the log-abundance matrices of ilarity: 45.0%); (ii) Valdivia and Melinka (31.4%), con- the faunal community structure among biogeographic necting Southern and Austral Chile; and (iii) San regions (Table 1f). Antonio Oeste and Puerto Madryn (45.3%), in Southern The presence-absence and abundance results both had Argentina. Punta Arenas, in Austral Chile, was not classi- a similar basic structure. However, they differed in the fied with any of the other sites although it was more sim- clustering of Punta Arenas and Valdivia with other Pacific ilar to Pacific than Atlantic sites. The analysis showed sites. These differences in clustering may underlie some significant differences in the presence-absence matrices of real biogeographic distinctions, but they could also be an the faunal community structure among biogeographic artifact due to the exclusion of Coquimbo (30°S) and regions (Table 1e). The results based on abundance data Valpara ıso (33°S) that are located in central Chile. These showed a similar pattern, identifying three clusters with two sites usually indicate the southern end of the Peru- minor differences in some sites (Fig. 3B): (i) Ilo, Iquique vian province and the beginning of a transition zone and Antofagasta (similarity: 54.2%), in Peru–Northern reaching the northern end of the Magellanic province at 41–42°S, near the Valdivia site (40°S). Given the absence of Central Chilean sites, Valdivia thus showed a higher ° A compositional affinity to Melinka (44 S) than with the northernmost sites. Seemingly, the high variability of abundance data has forced Valdivia to appear more simi- lar to northern than to austral sites although it is not dis- tinct enough to form a significant cluster. Valdivia was also the site with the highest faunal richness, which could have an additional effect owing to the proportional aver- aging of distances (weighted by the number of species in clusters) performed by UPGMA. Independent of these related effects, in the case of Punta Arenas (with the low- est faunal richness in Chile), the fact that this site formed an isolated cluster in the presence-absence results could be due to either the absence of sample sites south of Me- linka, or to the existence of a secondary biogeographic B unit (see Discussion). Nonetheless, even considering the above-described limitations, the two analyses were robust in revealing a large dissimilarity (≥90%) between the Pacific and Atlantic assemblages and a clear faunal differ- entiation along the Southeastern Pacific, which separates at least two groups of distinct assemblages in Peru– Northern Chile and South–Austral Chile.
Discussion
Diversity of faunal assemblages: biogeography versus bioengineering Faunal assemblages associated with A. atra exhibited high Fig. 3. Classification analyses (unweighted pair group method with taxonomic diversity, and their species richness and abun- arithmetic mean algorithm) of faunal assemblages associated with ribbed mussel (Aulacomya atra) beds. (A): Presence-absence data, using dance varied independently of the size and density of Jaccard similarity index. (B): Log-abundance data, using the Bray–Curtis mussels although the latter differed significantly among similarity index. Dashed lines indicate non-significant clusters resulting localities. These findings suggest that, except in extreme from a similarity profile test (SIMPROF). For abbreviations, see Fig. 1. cases (e.g. the beds were defaunated by harvesting in
950 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel
Coquimbo and Valpara ıso), faunal assemblages may be of the Magellanic province (see discussions by Sielfeld & highly resilient to the impacts of recurrent disturbances Vargas 1999; Lancellotti & V asquez 2000; Camus 2001; such as harvesting and El Nino~ Southern Oscillation. Fur- Haussermann & Forsterra 2009). Shallow subtidal habitats thermore, the positive correlation between faunal richness in the Strait are dominated by low-salinity surface waters, and abundance indicates that few species attain domi- which flow from the central part of the Strait to both the nance in these assemblages, and disturbances may be Pacific and Atlantic (Sievers & Silva 2006). However, the responsible for maintaining or enhancing their diversity. strong influence of winds and tides generates a net trans- The faunal richness of South American ribbed mussel port of low-salinity water towards the Atlantic continental beds (local range: 41–76, total: 197) is comparable to that shelf at 52.5°S (Sassi & Palma 2006). This, coupled with recorded in their South African counterparts (Griffiths advection and diffusion, promotes a northward flow (the et al. 1992), and ranks high in comparison with other so-called Patagonian current), which moves along the coast mussel species (see Thiel & Ullrich 2002; Hammond & until 47°S, although it also continues offshore as far as Griffiths 2006; Buschbaum et al. 2009). However, the 38°S (Piola & Rivas 1997). The Strait may therefore play a numerous single-site records in this first survey suggest twofold role by acting as (i) an effective barrier to the lati- that species richness was underestimated (Lancellotti & tudinal spread of species and (ii) an inter-oceanic connec- V asquez 2000), and therefore faunal diversity could be tion favoring the eastward dispersal of species from the characteristically high in ribbed mussel beds. Pacific. If the Strait does indeed act as a barrier, this might Along the Southeastern Pacific, faunal assemblages partly explain the peculiarity of Punta Arenas. Further- showed relatively high turnover rates. The fauna found more, the Strait acting as a connection between oceans sug- also showed marked differences between the northern gests that faunal assemblages along the Strait, and to some area (Peru + Northern Chile) and Southern Chile. These extent north of its Atlantic entrance, might not show major differences reflect the long-established Peruvian and Mag- compositional differences. In this context, the beds of San ellanic provinces (see Fern andez et al. 2000; Camus 2001; Antonio Oeste and Puerto Madryn, far north of the Strait, Thiel et al. 2007), although the gap in our analyses pre- are located at two gulfs surrounding the Valdes Peninsula vented us from examining the transition between the two (~42.5°S), which is an area of biogeographic turnover that areas. Thus, in agreement with prior studies (e.g. Ham- usually defines the northern limit of the Magellanic prov- mond & Griffiths 2006; Cole & McQuaid 2010), our ince for non-pelagic taxa (Balech & Ehrlich 2008). Thus, results show that the mussel-associated fauna from differ- the distinctiveness of these faunal assemblages in our ent regions is highly sensitive to biogeographic signals analyses could be due to the loss of Magellanic components and follows the characteristic provincial pattern docu- and the addition of species from the Northern Argentinean mented for each region. Furthermore, it is noted that province. This explanation would allow them to have a these differences occur despite the fact that the associated mixed character, which is reflected in our finding that their invertebrates make use of similar bioengineered habitat similarity did not exceed 50% despite that they are just 12 throughout their geographic ranges. These results also latitude minutes apart. imply that faunal biogeographic patterns may be decou- While A. atra is considered to be a common compo- pled from large-scale environmental gradients. In addi- nent of the Patagonian region in Chile and Argentina, lit- tion, faunal distributions are not dependent upon the tle is known about its occurrence south of the Strait of structure of the mussel beds as the relationship of faunal Magellan in the Tierra del Fuego archipelago. This leaves richness with faunal and mussel abundance was respec- an important question unanswered regarding the biogeo- tively positive and null in South America but negative in graphic homogeneity of the faunal assemblages of this South Africa (Hammond & Griffiths 2006). The above species throughout the Magellanic province. The repeated suggests a superimposition of biogeographic factors on effects of Quaternary glaciations in Austral South Amer- ecological patterns, and at least for the Southeastern Paci- ica disconnected the Pacific and Atlantic Oceans. This led fic we speculate that such factors could be related to to the eventual removal of significant components of the external physical forcing that could be inducing marked marine fauna although common cold-temperate bivalves distribution breaks on the Chilean coast at 30°S and 33°S such as Venus antiqua and Ensis macha, which are ecolog- (see Thiel et al. 2007). ically comparable to A. atra, exhibit similar present and The presence-absence data also showed the unique char- past distribution ranges (Gordillo et al. 2008). The stabil- acteristics of the faunal assemblage of Punta Arenas (54°S). ity in bivalve ranges could be explained in three ways, by This site is located in the Strait of Magellan where prior potential recolonization from populations that survived works have detected the endpoint of the latitudinal distri- in marine refuges during Pleistocene glaciations, by mid- bution of diverse taxa. These range limits have been inter- Holocene immigrations from the Pacific via the Beagle preted as due to internal biogeographic differentiation Channel after the last deglaciation (Gordillo et al. 2008)
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 951 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno or through immigration via the Magellan Strait. In all regional patterns of local abundance and site occupancy scenarios, the ribbed mussel would also have been present (Castilla & Camus 1992; Thiel et al. 2007; Camus 2008). during the post-glacial reassembly of mollusk communi- Second, at least in some areas in Northern Chile, the ties in the Beagle Channel (Gordillo et al. 2008), and absence or reduction of mussel beds is likely a conse- since then the ecological and climatic conditions in the quence of limited larval supply (Thiel et al. 2007), and it area have remained quite stable (Gordillo 1999). In addi- is noted that both larval supply and recruitment can be tion, the Atlantic and the Magellanic province mollusk highly affected by El Nino~ or La Nina~ events (Gamarra & faunas have not shown significant changes since at least Cornejo 2002; Thatje et al. 2008). The third and by far the late Pleistocene. Furthermore, the ribbed mussel has most important factor is exploitation, which may in turn been a characteristic species of the marine Quaternary in interact with El Nino~ events with unpredictable conse- Patagonia (Aguirre 2003). However, the reassembly of the quences on large-scale patterns (e.g. Castilla & Camus associated fauna may not have followed a similar pattern, 1992). In Peru, Chile and Argentina, ribbed mussels have and it is possible that these assemblages are not the same been harvested by coastal gatherers since at least the mid- north and south of the Strait of Magellan. Holocene. They continue to be heavily exploited through- out the whole distribution range (e.g. Ciocco et al. 1998; Bonavia et al. 2001; Laudien et al. 2007; Torres et al. Density and size of ribbed mussels 2007; Jerez & Figueroa 2008; Castro et al. 2011), and The population density of the ribbed mussel A. atra was there are evident signs of overexploitation in Peru and found to be highly heterogeneous throughout its range. Chile (Carranza et al. 2009). In exceptional cases, size- Although this study was not designed to test hypotheses selective harvesting might even contribute indirectly to on geographic abundance patterns, the extent (17–54°S) causing a biased abundant-centered distribution of indi- of our survey around the latitudinal midpoint (~33°S) of viduals by increasing the density of small individuals at A. atra’s range allows us to conclude that the observed the range center (e.g.inLottia gigantea; Fenberg & Rivad- pattern does not fit the traditional abundant-center eneira 2011). However, ribbed mussels are not always model (Brown 1984), which predicts decreasing abun- harvested in accordance with the minimum legal sizes dance toward range edges. While deviations from this (Carranza et al. 2009). In addition, the short-term deple- model are not unusual in coastal marine invertebrates tion of our sampled populations in Coquimbo and Val- (e.g. Sagarin & Gaines 2002; Mieszkowska & Lundquist para ıso, although not usual, is an example of the 2011), we acknowledge that our survey may include potential role of harvesting in preventing the formation potential biases that should be corrected by sampling a (or explaining the absence) of a center of abundance. greater number of more evenly spaced sites. However, it Exploitation also seems to be an important factor affect- seems unlikely that new sampling would unveil a previ- ing mussel fauna especially given the size variation of the ously undetected center of abundance. Numerous official mussels. The smallest and largest mussel sizes occurred at surveys of benthic resources have consistently shown the the lowest (Ilo) and highest (Punta Arenas) latitudes, low overall abundance of the ribbed mussel on the Chil- respectively. As such, there is (Fig. 2A) an apparent (non- ean coast (e.g. Galleguillos et al. 2002; Stotz et al. 2007). significant) latitudinal trend, which, albeit suggestive, has Conversely, we recorded a comparatively higher abun- no relationship with the rules of ecogeography or tempera- dance of A. atra at the sites at the extreme ends of our ture–size relationships (Karl & Fischer 2008). The average survey. However, it should be noted that we did not sam- size in Ilo (<40 mm) was much smaller than in other pop- ple the species’ range edges, and our data cannot be used ulations in the same area (around 70 mm; IMARPE 2007). to support or reject alternative patterns such as ramped- Furthermore, this bed had the highest faunal density in north or abundant-edge (see Sagarin & Gaines 2002). our survey, which suggests that it was subject to intense The historical data for A. atra are not sufficient to exploitation. By contrast, the large size of mussels in Punta infer past geographic trends in abundance. We suggest Arenas (~120 mm) may reflect a low exploitation level that any early observed pattern had to have been super- due to the relative isolation of this bed, which is located seded by the interaction among at least three factors of away from areas that are recurrently exploited by gatherers natural and anthropogenic origin. First, El Nino~ events and where ribbed mussels have been shown to occur in are well known to cause mass mortality of ribbed mussels low frequencies (R ıos et al. 2003). Evidence of harvesting in Peru and Northern Chile, which leads to either drastic was also apparent in most of the Pacific beds. This is most decreases in abundance or local extinctions (Gamarra & visible in the depleted areas of Coquimbo and Valparaiso, Cornejo 2002; Laudien et al. 2007; Thiel et al. 2007; which have right-skewed size distributions in comparison Thatje et al. 2008). Such El Nino~ perturbations can with the large variance of some of the Chilean beds, or indeed produce long-lasting or even permanent effects on others that showed no major differences in respect to the
952 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel overall size average. However, the example of Punta Are- Environmental Analysis. Pergamon Press, Oxford, United nas (large size versus low density) also highlights the role Kingdom: 699–717. of compensation effects. We suggest that the observed geo- Anderson M.J., Gorley R.N., Clarke K.R. (2008) graphic correlation between mussel size and density might PERMANOVA+ for PRIMER: Guide to Software and involve an interaction between size-selective harvesting Statistical Methods. PRIMER-E, Plymouth: 214 pp. and self-thinning processes (Hughes & Griffiths 1988; Angel M.V. (1993) Biodiversity of the pelagic ocean. Biological 7 – Guinez~ & Castilla 1999; Carranza et al. 2009). Unfortu- Conservation, , 760 772. nately, the poor knowledge of the ecology of A. atra in Arhan M., Naveira-Garabato A.C., Heywood K.J., Stevens D.P. South America prevents further insights regarding the (2002) The Antarctic Circumpolar Current between the Falkland Islands and South Georgia. Journal of Physical influence of other factors on mussel beds in this region. Oceanography, 32, 1914–1931. While many other things are known to affect mussel beds Balech E., Ehrlich M. (2008) Esquema biogeogr afico del mar (e.g. physical disturbance, predation, inter-specific compe- Argentino. Revista de Investigaci on y Desarrollo Pesquero, 19, tition; Noda 1999; Hamilton 2000; Thiel & Ullrich 2002; 45–75. Navarrete et al. 2010), with the currently available data it Bertness M.D., Callaway R. (1994) Positive interactions in is only possible to discuss the unpredictable changes in size communities. Trends in Ecology and Evolution, 9, 191–193. structure caused by mass mortalities and recruitment fail- Bonavia D., Johnson-Kelly L.W., Reitz E.J., Wings E.S. (2001) ~ ures during strong El Nino events in Peru and Northern El precer amico medio de Huarmey: historia de un sitio Chile (Gamarra & Cornejo 2002; Carranza et al. 2009). (PV35-106). Bulletin de l’Institut Francais d’ etudes Andines, Even with this lack of data, nonetheless, we conclude 30, 265–333. that ribbed mussel beds show no sign of any prior histori- Borthagaray A.I., Carranza A. (2007) Mussels as ecosystem cal patterns that might affect size or density, but rather, engineers: their contribution to species richness in a rocky these seem to be largely controlled by contemporary fac- littoral community. Acta Oecologica, 31, 243–250. tors operating on ecological time scales. By contrast, by Briggs J.C., Bowen B.W. (2012) A realignment of marine providing a suitable habitat for macrofaunal organisms on biogeographic provinces with particular reference to fish a large scale, A. atra contributes to the enhancement of distributions. Journal of Biogeography, 39,12–30. local diversity and allows for the assembly of distinctive Brown J.H. (1984) On the relationships between abundance and fauna. However, bioengineering of A. atra seems to play distribution of species. American Naturalist, 124, 255–279. no role in the biogeographic variation of its faunal assem- Buschbaum C., Dittmann S., Hong J.-S., Hwang I.-S., Strasser blages, which apparently respond to the same factors M., Thiel M., Valdivia N., Yoon S.-P., Reise K. (2009) affecting other organisms inhabiting ambient substrata. Mytilid mussels: global habitat engineers in coastal sediments. Helgoland Marine Research, 63,47–58. Camus P.A. (2001) Biogeograf ıa marina de Chile Continental. Acknowledgements Revista Chilena de Historia Natural, 74, 587–617. Camus P.A. (2008) Understanding biological impacts of ENSO This work was partially funded by Fondo Nacional de on the eastern Pacific: an evolving scenario. International Desarrollo Cient ıfico y Tecnologico, FONDECYT Journal of Environment and Health, 2,5–19. #3100145 and Direccion de Investigacion y Desarrollo, Camus P.A., Andrade Y.N., Broitman B. (1999) Effects of Universidad Austral de Chile, DID #01-2006 grants to substratum topography on species diversity and abundance R.D. Sepulveda. We are grateful to Federico Marquez in Chilean rocky intertidal communities. Revista Chilena de (Centro Nacional Patagonico, CENPAT, Puerto Madryn, Historia Natural, 72, 377–388. Argentina), Enrique Morsan and Sandro Acosta (Instituto C ardenas L., Castilla J.C., Virad F. (2009) A phylogeographical de Biologıa Marina y Pesquerıa Almirante Storni, San analysis across three biogeographical provinces of the south- Antonio Oeste, Argentina) and Raul Castillo (Instituto eastern Pacific: the case of the marine gastropod Concholepas del Mar del Peru, IMARPE, Ilo, Peru) for their assistance concholepas. Journal of Biogeography, 36, 969–981. in sample collection. We also thank Melissa Pav ez for her Carranza A., Defeo O., Beck M., Castilla J.C. (2009) Linking help with the figures. fisheries management and conservation in bioengineering species: the case of South American mussels (Mytilidae). Reviews in Fish Biology and Fisheries, 19, 349–366. References Castilla J.C., Camus P.A. (1992) The Humboldt-El Nino~ Aguirre M.L. (2003) Late Pleistocene and Holocene scenario: coastal benthic resources and anthropogenic palaeoenvironments in Golfo San Jorge, Patagonia: influences, with particular reference to the 1982/83 ENSO. molluscan evidence. Marine Geology, 194,3–30. South African Journal of Marine Science, 12, 703–712. Ahumada R.B., Pinto L., Camus P. (2000) The Chilean coast. Castilla J.C., Guinez~ R. (2000) Disjoint geographical In: Sheppard C.R.C. (Ed), Seas at the Millennium: An distribution of intertidal and nearshore benthic invertebrates
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 953 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno
in the Southern Hemisphere. Revista Chilena de Historia Gordillo S., Rabassa J., Coronato A. (2008) Paleoecology and Natural, 73, 585–603. paleobiogeographic patterns of mid-Holocene mollusks Castro A., Zubimendi M.A., Ambrustolo P. (2011) from the Beagle Channel (southern Tierra del Fuego, Archaeological littoral sites on the northern coast of Santa Argentina). Revista Geol ogica de Chile, 35, 321–333. Cruz: valuable evidence of sea level changes on the Griffiths R.J. (1977) Reproductive cycles in littoral populations continental Patagonian coasts (Argentina). Quaternary of Choromytilus meridionalis (Kr.) and Aulacomya ater International, 245, 111–121. (Molina) with a quantitative assessment of gamete Chapman M.G., Underwood A.J. (1994) Dispersal of intertidal production in the former. Journal of the Experimental snail, Littorina pyramidalis, in response to the topographic Marine Biology and Ecology, 30,53–71. complexity of the substratum. Journal of Experimental Griffiths C.L., Hockey P.A.R., van Erkom Schurink C., Le Marine Biology and Ecology, 179, 145–169. Roux P.J. (1992) Marine invasive aliens on South African Ciocco N.F., Lasta M.N., Bremec C.S. (1998) Pesquer ıas de shores: implications for community structure and trophic bivalvos: mejillon, vieiras (tehuelche y patagonica) y otras functioning. In A.I.L. Payne, K.H. Brink, K.H. Mann & R. especies. In: Boschi E.E. (Ed), El mar argentino y sus recursos Hilborn (Eds), Benguela Trophic Functioning. South African pesqueros, Tomo 2. Los moluscos de inter es pesquero. Cultivo y Journal of Marine Science, 12, 713–722. estrategias reproductivas de bivalvos y equinoideos. INIDEP, Guichard F., Bourget E. (1998) Topographic heterogeneity, Mar del Plata, Argentina: 143–166. hydrodynamics, and benthic community structure: a scale- Clarke K.R., Gorley R.N. (2006) PRIMER v6: User Manual/ dependent cascade. Marine Ecology Progress Series, 171,59–70. Tutorial. PRIMER-E Ltd, Plymouth Marine Laboratory, Guinez~ R., Castilla J.C. (1999) A tridimensional self-thinning Plymouth, UK: 190 pp. model for multilayered intertidal mussels. American Cole V.J., McQuaid C.D. (2010) Bioengineers and their Naturalist, 154, 341–357. associated fauna respond differently to the effects of Hamilton D.J. (2000) Direct and indirect effects of predation biogeography and upwelling. Ecology, 91, 3549–3562. by common eiders and abiotic disturbance in an intertidal Fenberg P.B., Rivadeneira M.M. (2011) Range limits and community. Ecological Monograph, 70,21–43. geographic patterns of abundance of the rocky intertidal owl Hammond W., Griffiths C. (2006) Biogeographical patterns in limpet, Lottia gigantea. Journal of Biogeography, 38, 2286– the fauna associated with southern African mussel beds. 2298. African Zoology, 41, 123–130. Fern andez M., Jaramillo E., Marquet P.A., Moreno C.A., Hastings A., Byers J.E., Crooks J.A., Cuddington K., Jones Navarrete S.A., Ojeda F.P., Valdovinos C.R., V asquez J.A. C.G., Lambrinos J.G., Talley T.S., Wilson W.G. (2007) (2000) Diversity, dynamics and biogeography of Chilean Ecosystem engineering in space and time. Ecology Letters, benthic nearshore ecosystems: an overview and guidelines 10, 153–164. for conservation. Revista Chilena de Historia Natural, 73, Haussermann V., Forsterra G. (2005) Distribution patterns 797–830. of Chilean shallow-water sea anemones (Cnidaria: Galleguillos R., C espedes I., Leon R., Pinto A., Catrilao M., Anthozoa: Actiniaria, Corralimorpharia), with a discussion Chacon V., Herold E., Veloso C., Herrera C., Smith E. of the taxonomic and zoogeographic relationships (2002) Prospecci on de Recursos Bent onicos en las Provincias between the actinofauna of the South East Pacific, the de Concepci on-Nuble~ . Informe final, Fondo Nacional de South West Atlantic and Antarctic. Scientia Marina, 69 Desarrollo Regional, VIII Region, Chile: 129 pp. (S2), 91–102. Gamarra A., Cornejo O. (2002) Study of the mussel Haussermann V., Forsterra G. (2009) Marine Benthic Fauna of Aulacomya ater Molina, 1782 (Bivalvia: Mytilidae), near Chilean Patagonia. Nature in Focus, Puerto Montt, Chile: Santa Rosa Island, Independence Bay, Peru, during the El 1000 pp. Nino~ Phenomenon 1997-98. Investigaciones Marinas, 30(S Haye P.A., Segovia N.I., Munoz-Herrera~ N.C., G alvez F.C., Symp), 140–140. Mart ınez A., Meynard A., Pardo-Gandarillas M.C., Poulin Gat ıca C., Quinones~ R.A., Figueroa D., Wiff R., Navarro E., E., Faugeron S. (2014) Phylogeographic structure in benthic Donoso M. (2009) Asociacion entre la Corriente de Deriva marine invertebrates of the Southeast Pacific coast of Chile de los Vientos del Oeste y la abundancia relativa del pez with differing dispersal potential. PLoS ONE, 9, e88613. espada (Xiphias gladius) frente a la costa de Chile. Latin Hormazabal S., Shaffer G., Leth O. (2004) Coastal transition American Journal of Aquatic Research, 37,97–105. zone off Chile. Journal of Geophysical Research, 109 Gonz alez M.T., Moreno C.A. (2005) The distribution of the (C01021), 1–13. ectoparasite fauna of Sebastes capensis from the southern Hughes R.N., Griffiths C.L. (1988) Self-thinning in barnacles hemisphere does not correspond with zoogeographical and mussels: the geometry of packing. American Naturalist, provinces of free-living marine animals. Journal of 132, 484–491. Biogeography, 32, 1539–1547. Ib anez~ C.M., Camus P.A., Rocha F.J. (2009) Diversity and Gordillo S. (1999) Holocene molluscan assemblages in the distribution of cephalopod species off the coast of Chile. Magellan region. Scientia Marina, 63(S1), 15–22. Marine Biology Research, 5, 374–384.
954 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH Sepulveda, Camus & Moreno Biogeography and bioengineering of the ribbed mussel
IMARPE. (2007) Seguimiento de pesquer ıas y evaluaci on de Moreno C.A., Molinet C. (2013) El paradigma de la recursos pesqueros. Instituto del Mar del Peru, Callao: 184 pp. distribucion batim etrica de Loxechinus albus (Molina) en Jaramillo R., Navarro J. (1995) Reproductive cycle of the Chile. Revista Chilena de Historia Natural, 86, 225–227. Chilean ribbed mussel Aulacomya ater (Molina, 1782). Navarrete S.A., Gelcich S., Castilla J.C. (2010) Long-term Journal of Shellfish Research, 14, 165–171. monitoring of coastal ecosystems at Las Cruces, Chile: Jerez G., Figueroa M. (2008) Desaf ıos y perspectivas de la defining baselines to build ecological literacy in a world of repoblacion de moluscos bivalvos en Chile. In: Lovatelli A., change. Revista Chilena de Historia Natural, 83, 143–157. Far ıas A., Uriarte I. (eds.) Estado actual del cultivo y manejo Noda T. (1999) Within- and between-patch variability of de moluscos bivalvos y su proyecci on futura: factores que predation intensity on the mussel Mytilus trossulus Gould afectan su sustentabilidad en Am erica Latina. FAO Actas de on a rocky intertidal shore in Oregon, USA. Ecological Pesca y Acuicultura N°12, Roma: 223–235. Research, 14, 193–203. Jones C.G., Lawton J.H., Shachak M. (1994) Organisms as Osorio C., Atria J., Mann S. (1979) Moluscos marinos de ecosystem engineers. Oikos, 69, 373–386. importancia economica en Chile. Biolog ıa Pesquera, 11,3–47. Karl I., Fischer K. (2008) Why get big in the cold? Towards a Piola A.R., Rivas A.L. (1997) Corrientes en la plataforma solution to a life-history puzzle. Oecologia, 155, continental. El Mar Argentino y sus Recursos Pesqueros, 1, 215–225. 119–132. Kelaher B.P., Castilla J.C., Prado L., York P., Schwindt E., Quinn G.P., Keough M.J. (2002) Experimental Design and Bortolus A. (2007a) Spatial variation in molluscan Data Analysis for Biologists. Cambridge University Press, assemblages from coralline turfs of Argentinean Patagonia. Cambridge, UK: 537 pp. Journal of Molluscan Studies, 73, 139–146. R ıos C., Mutschke E., Morrison E. (2003) Biodiversidad Kelaher B.P., Castilla J.C., Prado L. (2007b) Is there bentonica sublitoral en el estrecho de Magallanes. Revista de redundancy in bioengineering for molluscan assemblages on Biolog ıa Marina y Oceanograf ıa, 38,1–12. the rocky shores of central Chile? Revista Chilena de Historia Sagarin R.D., Gaines S.D. (2002) Geographical abundance Natural, 80, 173–186. distributions of coastal invertebrates: using one-dimensional Lagos N.A., Tapia F.J., Navarrete S.A., Castilla J.C. (2007) ranges to test biogeographic hypotheses. Journal of Spatial synchrony in the recruitment of intertidal Biogeography, 29, 985–997. invertebrates along the coast of central Chile. Marine S anchez R., Sepulveda R.D., Brante A., C ardenas L. (2011) Ecology Progress Series, 350,29–39. Spatial pattern of genetic and morphological diversity in the Lancellotti D.A., V asquez J.A. (1999) Biogeographical patterns direct developer Acanthina monodon (Gastropoda: of benthic macroinvertebrates in the Southeastern Pacific Mollusca). Marine Ecology Progress Series, 434, 121–131. littoral. Journal of Biogeography, 26, 1001–1006. Sassi M.G., Palma E.D. (2006) Modelo hidrodin amico del Lancellotti D.A., V asquez J.A. (2000) Zoogeograf ıa de Estrecho de Magallanes. Mec anica Computacional, 25, 1461– macroinvertebrados bentonicos de la costa de Chile: 1477. contribucion para la conservacion marina. Revista Chilena Scarabino F., Ortega L. (2004) Registros uruguayos de de Historia Natural, 73,99–129. Aulacomya atra atra (Bivalvia: Mytilidae): rol de Laudien J., Rojo M.E., Oliva M.E., Arntz W.E., Thatje S. condiciones oceanogr aficas anomalas y de dispersion por (2007) Sublittoral soft bottom communities and diversity of feofitas flotantes. Comunicaciones de la Sociedad Mejillones Bay in northern Chile (Humboldt Current Malacol ogica del Uruguay, 8, 299–304. upwelling system). Helgoland Marine Research, 61, 103–116. Seed R. (1996) Patterns of biodiversity in the Lee M.R., Castilla J.C. (2012) Do changes in microhabitat macroinvertebrate fauna associated with mussel patches on availability within a Marine Reserve reduces the species rocky shores. Journal of the Marine Biological Association of richness of small mobile macrofauna and meiofauna? the United Kingdom, 76, 203–210. Journal of the Marine Biological Association of the United Sepulveda R.D., Ib anez~ C.M. (2012) Clinal variation in the Kingdom, 92, 1283–1288. shell morphology of intertidal snail Acanthina monodon in Mart ınez S., del R ıo C. (2002) Las provincias malacologicas the southeastern Pacific Ocean. Marine Biology Research, 8, miocenas y recientes del Atl antico sudoccidental. Anales de 363–372. Biolog ıa, 24, 121–130. Sepulveda R.D., Rozbaczylo N., Ib anez~ C.M., Flores M., Mena C., Gonz alez C., Clasing E., Gallardo M. (2001) Cancino J.M. (2015) Ascidian-associated polychaetes: Variabilidad gen etica en Aulacomya ater (Molina, 1792) en ecological implications of aggregation size and tube-building el sur de Chile. Ciencia y Tecnolog ıa Marina, 24,71–79. chaetopterids on assemblage structure in the Southeastern Mieszkowska N., Lundquist C.J. (2011) Biogeographical Pacific Ocean. Marine Biodiversity, 45, 733–741. patterns in limpet abundance and assemblage composition Sielfeld W., Vargas M. (1999) Review of marine fish in New Zealand. Journal of Experimental Biology and zoogeography of Chilean Patagonia (42°–57°S). Scientia Ecology, 400, 155–166. Marina, 63(S1), 451–463.
Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH 955 Biogeography and bioengineering of the ribbed mussel Sepulveda, Camus & Moreno
Sievers H.A., Silva N. (2006) Masas de agua y circulacion en M., Dumont C.P., Escribano R., Fern andez M., Gajardo J.A., los canales y fiordos australes. In: Silva N., Palma S. (Eds), Gaymer C.F., Gomez I., Gonz alez A.E., Gonz alez H.E., Haye Avances en el conocimiento oceanogr afico de las aguas P.A., Illanes J.E., Iriarte J.L., Lancellotti D.A., Luna-Jorquera interiores chilenas, Puerto Montt a Cabo de Hornos. Comit e G., Luxoro C., Manr ıquez P.H., Mar ın V., Munoz~ P., Oceanogr afico Nacional , Pontificia Universidad Catolica de Navarrete S.A., P erez E., Poulin E., Sellanes J., Sepulveda Valpara ıso, Valpara ıso, Chile: 53–58. H.H., Stotz W., Tala F., Thomas A., Vargas C.A., V asquez Stotz W., Olivares C., Valdebenito M., Lancellotti D., Aburto J.A., Vega A. (2007) The Humboldt Current System of J., Cerda C., Morales J., Guajardo P., Jim enez V., Mondaca Northern-Central Chile: oceanographic processes, ecological C. (2007) Criterios de explotaci on de recursos bent onicos interactions and socioeconomic feedback. Oceanography and secundarios en areas de manejo. Informe Final Proyecto FIP Marine Biology an Annual Review, 45, 195–345. 2005–42, Fondo de Investigacion Pesquera, Chile: 424 pp. Thompson G.A., Alder V.A. (2005) Patterns in tintinnid Stuart V., Newell R.C., Lucas M.I. (1982) Conversion of Kelp species composition and abundance in relation to debris and faecal material from the mussel Aulacomya ater hydrological conditions of the southwestern Atlantic during by marine microorganisms. Marine Ecology Progress Series, 7, austral spring. Aquatic Microbial Ecology, 40,85–101. 47–57. Torres J., Silva C., Lucero M. (2007) El rol de la pesca en la Suchanek T.H. (1992) Extreme biodiversity in the marine intensificacion de las ocupaciones costeras durante el environment: mussel bed communities of Mytilus Holoceno Medio-Tard ıo (Bah ıa de Concepcion, Region del californianus. Northwest Environmental Journal, 8, 150–152. B ıo-B ıo, Chile). Magallania, 35,71–93. Takesue R.K., Van Geen A., Carriquiry J.D., Ort ız E., Tsuchiya M., Nishihira M. (1985) Islands of Mytilus edulis as a God ınez-Orta L., Granados I., Sald ıvar M., Ortlieb L., habitat for small intertidal animals: effect of island size on Escribano R., Guzm an N., Castilla J.C., Varas M., community structure. Marine Ecology Progress Series, 25,71– Salamanca M., Figueroa C. (2004) Influence of coastal 81. upwelling and El Nino-Southern~ Oscillation on nearshore Underwood A.J., Chapman M.G. (1989) Experimental analyses water along Baja California and Chile: shore-based of the influences of topography of the substratum on monitoring during 1997-2000. Journal of Geophysical movements and density of an intertidal snail, Littorina Research-Oceans, 109,1–14. unifasciata. Journal of Experimental Marine Biology and Thatje S., Heilmayer O., Laudien J. (2008) Climate variability Ecology, 134, 175–196. and El Nino~ Southern Oscillation: implications for natural Uribe J., Lopez D. (1980) Fijacion primaria y variaciones coastal resources and management. Helgoland Marine morfologicas, durante la metamorfosis de algunos bivalvos Research, 62(S1), 5–14. chilenos. Brazilian Journal of Oceanography, 29, 367–369. Thiel M., Ullrich N. (2002) Hard rock versus soft bottom: the Valentine J.W., Roy K., Jablonsky D. (2002) Carnivore/non- fauna associated with intertidal mussel beds on hard carnivore ratios in northeastern Pacific marine gastropods. bottoms along the coast of Chile, and considerations on the Marine Ecology Progress Series, 228, 153–163. functional role of mussel beds. Helgoland Marine Research, Zaixso H.E. (2004) Bancos de cholga Aulacomya atra atra 56,21–30. (Molina) (Bivalvia: Mytilidae) del golfo San Jos e (Chubut, Thiel M., Macaya E.C., Acuna~ E., Arntz W.E., Bastias H., Argentina): Diversidad y relaciones con facies afines. Revista Brokordt K., Camus P.A., Castilla J.C., Castro L.R., Cort es de Biolog ıa Marina y Oceanograf ıa, 39,61–78.
956 Marine Ecology 37 (2016) 943–956 ª 2016 Blackwell Verlag GmbH