FACULTAD DE CIENCIAS DEL MAR UNIVERSIDAD CATÓLICA DEL NORTE DOCTORADO EN BIOLOGÍA Y ECOLOGÍA APLICADA

“Factors influencing the potential of rafting dispersal - analysis of patterns and causal

mechanisms along the geographic range of intertidal species on the coast of Chile”

Boris Alejandro López Arriagada

Profesor Guía: Dr. Martin Thiel

COQUIMBO, 2018 FACULTAD DE CIENCIAS DEL MAR UNIVERSIDAD CATÓLICA DEL NORTE DOCTORADO EN BIOLOGÍA Y ECOLOGÍA APLICADA

“Factors influencing the potential of rafting dispersal - analysis of patterns and causal

mechanisms along the geographic range of intertidal species on the coast of Chile”

Por: Boris Alejandro López Arriagada

Departamento de Biología Marina

Fecha: 17‐10‐2018

Aprobado Comisión de Calificación

______Dr. Juan Macchiavello Armengol Decano Facultad de Ciencias del Mar

______Dr. Martin Thiel Profesor Guía

______Dra. Fadia Tala Dr. Marcelo Rivadeneira Comité Tutorial Comité Tutorial

______Dr. Moisés Aguilera Dr. Erasmo Macaya Comité Tutorial Profesor Externo

Tesis entregada como un requisito para obtener el título de Doctor en Biología y Ecología Aplicada en la Facultad de Ciencias del Mar. Universidad Católica del Norte. Sede Coquimbo.

2018

ii

FACULTAD DE CIENCIAS DEL MAR UNIVERSIDAD CATÓLICA DEL NORTE DOCTORADO EN BIOLOGÍA Y ECOLOGÍA APLICADA

Departamento de Biología Marina

“Factors influencing the potential of rafting dispersal - analysis of patterns and causal

mechanisms along the geographic range of intertidal species on the coast of Chile”

Actividad de Titulación presentada para optar al Título de Doctor en Biología y Ecología Aplicada

Boris Alejandro López Arriagada

Coquimbo, Octubre de 2018

iii

FACULTAD DE CIENCIAS DEL MAR UNIVERSIDAD CATÓLICA DEL NORTE DOCTORADO EN BIOLOGÍA Y ECOLOGÍA APLICADA

DECLARACIÓN DEL AUTOR

Se permiten citas breves sin permiso especial de la Institución o autor, siempre y cuando se otorgue el crédito correspondiente. En cualquier otra circunstancia, se deberá solicitar permiso de la Institución o el autor.

Boris Alejandro López Arriagada Firma

2018

iv

RESUMEN

Varios estudios, usando evidencia molecular y ecológica, han confirmado que el rafting contribuye a la conectividad poblacional y la expansión de rangos geográficos de muchas especies. Sin embargo, también hay indicios de que la efectividad del transporte de especies vía rafting es limitada al momento de establecerse en nuevos hábitats. Esto sugiere que el transporte puede no ser el factor limitante para la dispersión de especies, pero que otros factores (ecológicos o fisiológicos) podrían restringir la colonización potencial de estas especies. Asimismo, la dispersión efectiva de una especie podría variar dentro de su rango geográfico. El objetivo de este estudio fue evaluar los patrones de dispersión del alga flotante Durvillaea antarctica (Chamisso) Hariot 1892 y sus epibiontes asociados a lo largo de la costa continental de Chile (28°S-42°S), así como examinar cuales mecanismos afectan la dispersión y conectividad de sus poblaciones dentro del rango geográfico del clado continental de D. antarctica. Se propuso que los factores ecológicos son más importantes en la dispersión efectiva de individuos de D. antarctica en poblaciones centrales de su rango geografico, mientras que en poblaciones periféricas características reproductivas y funcionales de las especies serán más importantes; desviaciones a partir de esta predicción de dispersión efectiva dentro del rango geográfico podría deberse a condiciones oceanográficas locales. La variabilidad espacio-temporal de ejemplares de D. antarctica varados en playas (Capítulo 1), mostró que los varamientos y características de los individuos variaron fuertemente de acuerdo a los distritos biogeográficos a lo largo de la costa de Chile durante tres años consecutivos. El estudio bimensual de varamientos en tres playas (32°S, 36°S y 39°S) durante tres años consecutivos (Capítulo 2) confirmó el patrón espacial de que altas biomasas varadas fueron observadas en playas del norte y sur en comparación a playas de la zona central. Además, se observó un fuerte patrón estacional que estuvo relacionado con la intensidad de tormentas locales. Una prospección a gran escala de ejemplares dispersados por rafting (28°S-42°S) y una a menor escala de ejemplares bentónicos (31°S-32°S) (Capítulo 3) evidenció que la proporción de machos y hembras que viajan juntos en discos coalescentes de D. antarctica fue baja (entre 5% - 17%). Asimismo, el análisis de capacidades funcionales indicó que no hubo diferencias en las concentraciones de pigmentos y florotaninos según el sexo de los individuos. La v evaluación espacio-temporal de la diversidad de epibiontes en discos de ejemplares varados de D. antarctica (Capítulo 4) evidenció que especies sésiles fueron más frecuentes. La riqueza taxonómica de epibiontes fue más alta en los distritos biogeográficos del sur (33°S- 42°S) que en aquellos del norte (28°S-33°S), incrementándose en verano en comparación a invierno. Especies formadoras de hábitat (e.g. mitílidos, algas no flotantes) favorecieron las co-ocurrencias con otros epibiontes dentro de un disco de D. antarctica, estando asociadas a expansiones de rangos geográficos de especies de epibiontes, particularmente en la zona sur (33°S-42°S). Los patrones filogeográficos de dos especies de algas epifitas del género Gelidium, frecuentes en discos de ejemplares varados de D. antarctica fueron contrastantes (Capítulo 5). Gelidium lingulatum tuvo alta diversidad genética, pero su estructura genética no siguió un claro patrón geográfico, lo que puede ser explicado por dispersión vía rafting de individuos flotantes de D. antarctica. Por otro lado, G. rex tuvo menor diversidad genética con una débil estructura genética, y un quiebre filogeográfico que coincide con la discontinuidad filogeográfica descrita para la región entre los 29°S-33°S, lo que sugiere eventos de dispersión a corta distancia. Los resultados indicaron que D. antarctica posee una importante potencial de dispersión mediante ejemplares flotantes y de transporte de especies asociadas, pero existirían fuertes fluctuaciones espacio-temporales en la disponibilidad de ejemplares flotantes que retornan a la costa que afectarían su eventual inmigración exitosa en otras áreas. El ciclo anual de crecimiento de las poblaciones bentónicas de D. antarctica, la disponibilidad de sustrato primario, la explotación de sus bancos naturales, y características oceanográficas y topográficas a escala regional (i.e. surgencias, plumas de ríos), y de mayor escala (i.e. el fenómeno del Niño) podrían explicar los patrones observados en este estudio. En general, los resultados sugieren que las características reproductivas y funcionales de los individuos flotantes no serían limitantes en la capacidad de dispersión efectiva y consecuente baja conectividad de sus poblaciones bentónicas, en comparación con factores oceanográficos locales y demográficos de las poblaciones bentónicas residentes. Por lo tanto, este estudio muestra que la efectividad de la dispersión por rafting mediante algas flotantes en zonas templadas varía fuertemente a lo largo del rango geográfico de las especies, así como en escalas temporales. Corrientes locales y regímenes de vientos generan zonas de retención para suministros flotantes que limitan el arribo hacia otras áreas. En general, las resultados de esta tesis sugieren que en

vi los límites de baja latitud del rango de una especie, el stress fisiológico suprime el potencial de dispersión de algas flotantes, mientras que en los límites de alta latitud del rango, las fluctuaciones de la disponibilidad de individuos flotantes y las interacciones ecológicas con comunidades bentónicas establecidas afectan la conectividad poblacional.

Palabras Claves: algas flotantes; biogeografía marina; cochayuyos; dispersión por rafting; varamientos.

vii

ABSTRACT

Several studies, using both molecular and ecological evidence, confirm that rafting contributes to population connectivity and range expansion of many species. However, there is also indication that the effectiveness of rafting transport of species to establish in new habitats is limited. This suggests that transport might not be the limiting factor for the dispersal of species but that other factors (ecological or physiological) may restrict the colonization potential of these species. Also, this effective dispersal of a species could vary within its geographic range. The aim of this study was to assess the dispersal patterns of the floating bull kelp Durvillaea antarctica (Chamisso) Hariot 1892 and its associated epibionts along the continental coast of Chile (28°S-42°S) as well as to examine which mechanisms affect dispersal and connectivity of their populations within the geographic range of the continental clade of D. antarctica. It was proposed that ecological factors are more important in the effective dispersal of individuals of D. antarctica in central populations of its geographic range, while in peripheral populations reproductive and functional characteristics of the species are more important; deviations from the predictions of the effective dispersal within the geographic range could be due to local oceanographic conditions. The spatio-temporal variability of stranded bull kelps on beaches (Chapter 1), showed that strandings and morphometric characteristics of rafts varied strongly according to biogeographic districts along the coast of Chile during three consecutive years. The bimonthly study of strandings on three beaches (32°S, 36°S and 39°S) during three consecutive years (Chapter 2) confirmed the spatial pattern that higher stranded biomasses were observed in northern and southern beaches than central beaches. Furthermore, there was a strong seasonal pattern of kelp strandings that was also related to the intensity of local storms. A large-scale survey of rafted holdfasts (28°S-42°S) and a small-scale survey of benthic holdfasts (31°S-32°S) (Chapter 3) showed that the proportion of males and females traveling together in coalescent holdfasts of D. antarctica was low (between 5%- 17%). Likewise, the functional capacity analysis indicated that there were no differences in pigment and florotanine concentrations according to the sex of the individuals. The spatio- temporal evaluation of the diversity of epibionts on holdfasts of stranded bull kelps (Chapter 4) showed that sessile species were the most frequent associated epibionts. Also, taxonomic richness of epibionts was higher in southern biogeographic districts (33°S-42°S)

viii than northern ones (28°S-33°S), increasing in summer compared to winter. Habitat-forming species (e.g. mytilid mussels, non-buoyant seaweeds) favoured co-occurrences with other epibionts within a holdfast of D. antarctica, being associated with range extensions of epibionts, particularly in the southern zone (33°S-42°S). Phylogeographic patterns of two epiphytic seaweeds from the genus Gelidium frequent on holdfasts of stranded D. antarctica were contrasting (Chapter 5). Gelidium lingulatum had high genetic diversity, but its genetic structure did not follow a geographic pattern, which could be explained mainly by rafting dispersal via floating bull kelps. On other hand, G. rex had less genetic diversity with a shallow genetic structure, and a phylogeographic break coinciding with the phylogeographic discontinuity described for the region between 29°S–33°S, which suggests short-distance dispersal events. The results indicate that D. antarctica has an important dispersal potential through floating specimens and for transport of associated species, but there are strong spatio-temporal fluctuations in the availability of floating specimens that return to the coast that would affect their eventual successful immigration in other areas. Annual cycles of growth of the benthic populations of bull kelps, the availability of potential primary substratum, the exploitation of their natural beds, and oceanographic and topographic features at regional (i.e. local upwelling, river plumes), and larger scales (i.e. El Niño) might explain the patterns observed in this study. Overall, the results suggest that the reproductive and functional characteristics of floating individuals are not limiting in the effective dispersal capacity and consequent low connectivity of their benthic populations along the Chilean coast compared to local oceanographic features, and demographic factors of resident benthic populations. Hence, this study shows that the effectiveness of rafting dispersal via floating seaweeds in temperate regions vary strongly along the geographic range of the species, as well as on time scales. Local currents and wind regimes generate retention zones for floating supplies that limit the arrival to other areas. In general, the findings of this thesis suggest that at low-latitude range limits of a species the physiological stress suppresses the dispersal potential of floating seaweeds, while at high-latitude range limits, the fluctuations in raft availability and ecological interactions with established benthic communities affect population connectivity.

Key Words: bull kelps; floating seaweeds; marine biogeography; rafting dispersal; strandings.

ix

Acknowledgments

Quiero agradecer a mi profesor guía, Dr. Martin Thiel por su dedicación en la supervisión de esta tesis doctoral. Gracias por la exigencia, y por sacarme de mi área de confort académica, ya que en mi opinión, eso permite los aprendizajes significativos.

También quiero agradecer a la Dra. Fadia Tala de la Universidad Católica del Norte (UCN) por su gran apoyo en varios aspectos de la ejecución de este estudio. Gracias por tu sincera amistad y espero que continuemos trabajando juntos en el futuro. Me gustaría también agradecer a mis compañeros del laboratorio BEDIM, especialmente a Óscar Pino, José Pantoja, Vieia Villalobos, Ulyces Urtubia, Tim Kiessling y Nicolás Riquelme quienes colaboraron en actividades de terreno y laboratorio. Asimismo, a mis colegas Felipe Sáez y David Yáñez del laboratorio de Botánica Marina de la UCN. A los estudiantes de intercambio Dominic Lizée-Prynne, Hannes Reinwald, Callum Blake y especialmente a Lennart Schreiber por su cordial interacción y colaboración en esta tesis.

No puedo olvidar de agradecer a mis compañeros del Programa de Doctorado en Biología y Ecología Aplicada: Isabel, Liss, Ricardo, Cristián, Naiti, Avia y James, por compartir los primeros años de estudios, y luego a Johanne Dobringer, Solange Vargas y María José Hevia por su amistad e incondicional apoyo. Les deseo lo mejor en sus vidas personales y profesionales. Agradezco también a mis colegas Daniel Piñones y Jessica Vargas del Programa PAR-Explora Coquimbo por introducirme en el bello mundo de la Ciencia Escolar.

Quiero agradecir también a la Dra. Florence "Flo" Tellier de la Universidad Católica de la Santísima Concepción (UCSC) y al personal del Laboratorio de Ecología Molecular, Juan Carlos Retamal y Karla Pérez por colaborar en el desarrollo de este estudio. A mi amigo Dr. Ángel Urzúa (UCSC) por facilitar mis estadías en Concepción. Al Dr. Erasmo Macaya (UdeC), Dr. Marcelo Rivadeneira (CEAZA), Dr. Nelson Valdivia (UACH), y Dra. Ceridwen Fraser (ANU) por colaborar en aspectos puntuales de esta tesis.

x

Al Departamento de Acuicultura y Recursos Agroalimentarios de la Universidad de Los Lagos, Osorno por depositar su confianza en mi desarrollo profesional. A la Beca de Doctorado CONICYT (Folio 21140010), y al proyecto FONDECYT 1131082 por financiar este estudio.

Finalmente, y lo más importante, quiero agradecer a mi familia, a mi preciosa hija Celeste López, y a Mónica Gallardo por todo su amor y apoyo durante estos años de estudio.

xi

Table of Contents

Resumen v Abstract viii Acknowledgments x Table of Contents xii List of Tables xv List of Figures xvii List of Appendices xxiii

I. General Introduction 1 Dispersal patterns of floating seaweeds 1 Possible causes of low connectivity among benthic populations of seaweeds 2 The abundant-centre hypothesis 4 General Hypothesis 5 Model species 5 General and specific objectives 8

Chapter 1: The variable routes of rafting: stranding dynamics of floating 10 bull-kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Pacific 1.1. Abstract 11 1.2. Introduction 12 1.3. Materials and Methods 13 1.4. Results 15 1.5. Discussion 18 1.6. Acknowledgements 23 1.7. References 23

xii

Chapter 2: Spatio-temporal variability of strandings of the southern bull 26 kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches on the coast of Chile - linking with local storms 2.1. Abstract 27 2.2. Introduction 27 2.3. Materials and Methods 30 2.4. Results 36 2.5. Discussion 43 2.6. Acknowledgements 49 2.7. References 49

Chapter 3: No sex-related dispersal limitation in a dioecious, oceanic long- 62 distance traveller: the bull lelp Durvillaea antarctica 3.1. Abstract 63 3.2. Introduction 63 3.3. Materials and Methods 65 3.4. Results 67 3.5. Discussion 69 3.6. Acknowledgements 72 3.7. References 72

Chapter 4: Epibiont communities on stranded kelp rafts of Durvillaea 75 antarctica (Fucales, Phaeophyceae) – do positive interactions facilitate range extensions? 4.1. Abstract 76 4.2. Introduction 77 4.3. Materials and Methods 78 4.4. Results 80 4.5. Discussion 82 4.6. Acknowledgements 86 4.7. References 86

xiii

Chapter 5: Phylogeography of two intertidal seaweeds, Gelidium lingulatum 89 and G. rex (Rhodophyta: Gelidiales), along the South East Pacific – patterns explained by rafting dispersal? 5.1. Abstract 90 5.2. Introduction 91 5.3. Materials and Methods 93 5.4. Results 97 5.5. Discussion 101 5.6. Acknowledgements 104 5.7. References 104

II. General Discussion 109 Dispersal patterns of floating bull kelps along Chilean coast - implications 109 for its population connectivity Durvillaea antarctica as dispersal agent for associated species 114 Conclusions 116

III. General References 118

Appendices 127 Appendix 1: Supplementary Material 1: The variable routes of rafting: 128 stranding dynamics of floating bull-kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Pacific. Appendix 2: Supplementary Material 2: Spatio-temporal variability of 135 strandings of the southern bull kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches on the coast of Chile - linking with local storms. Appendix 3: Supplementary Material 3: Epibiont communities on stranded 145 kelp rafts of Durvillaea antarctica (Fucales, Phaeophyceae) – do positive interactions facilitate range extensions?. Appendix 4: Supplementary Material 4: Phylogeography of two intertidal 184 seaweeds, Gelidium lingulatum and G. rex (Rhodophyta: Gelidiales), along the South East Pacific – patterns explained by rafting dispersal?.

xiv

List of Tables

Table 1.1. Results of three-way PERMANOVA for response variables per 16 family of stranded Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S). Pair-wise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Pair-wise tests show districts or seasons that differed (P < 0.05). Nonsignificant pair- wise tests are not shown

Table 1.2. Results of three-way ANOVAs testing the effect of district, year 17 and season of each response variable of stranded Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S). Pair-wise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Pair-wise tests show districts or seasons that differed (P < 0.05). Nonsignificant pair-wise tests are not shown.

Table 2.1. Summary of GLM for stranded biomass and number of rafts of 38 Durvillaea antarctica on beaches from the continental coast of Chile (32°S- 39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Significant values (P < 0.05) are shown in bold.

Table 2.2. Summary of GLM for length and wet weight of stranded 40 individuals of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Significant values (P < 0.05) are shown in bold.

Table 3.1. Durvillaea antarctica: summary of correlations and ANOVAs of 68 simple linear regression between pigments and colour values R (red), G (green), and B (blue) in the blades of stipes samples. Stipes from both sites (Puerto Oscuro and Totoralillo Sur) were pooled for these analyses.

Table 3.2. Durvillaea antarctica: summary of two-way ANOVAs for colour 69 (R, G and B) according to site (Puerto Oscuro and Totoralillo Sur) and sexual stage (vegetative, reproductive male, and reproductive female) of algal samples.

Table 3.3. Durvillaea antarctica: summary of two-way ANOVAs for 70 variables (pigments and phlorotannins) according to site (Puerto Oscuro and Totoralillo Sur) and sexual stage (vegetative, reproductive male, and reproductive female) of algal samples.

xv

Table 5.1. Sampling sites of benthic populations of Gelidium lingulatum and 96 G. rex. The code used to identify each sampling site, coordinates (latitude and longitude), indices of genetic diversity found for the mitochondrial marker COI, results of neutrality tests and their significance are indicated. N: number of individuals sequenced; h: number of haplotypes; hpriv: number of private haplotypes; S: number of polymorphic sites; H: standardized haplotype diversity and SD (after rarefaction); %π: percentage of nucleotide diversity and SD; D: Tajima’s test; Fs: Fu’s test. Mean and overall values correspond to the average values of the locations and the total for each species, respectively. Significant values (P < 0.05) are shown in bold. N.A. = not available or uncalculated (locations with N < 14).

xvi

List of Figures

Fig. I.1. Conceptual overview of biotic and abiotic factors (intrinsic and 3 extrinsic to source populations) that might explain the low connectivity between benthic seaweed populations.

Fig. 1.1. Geographical distribution of sampling sites and biogeographic 13 districts described for the coast of Chile distinguished in this study (Coquimbo-Choros District: 28°S-30°S, Septentrional District: 30°S-33°S, Mediterranean District: 33°S-37°S, Meridional District: 37°S-42°S). The geographic distribution of Durvillaea antarctica within the study area is also indicated. RS/SS = ratio rocky shoreline (km) versus sandy shoreline (km). The number of beaches sampled within each biogeographic district is shown.

Fig. 1.2. Box plot of response variables of stranded individuals of Durvillaea 16 antarctica on beaches from the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). (A) Stranded biomass, and (B) Percentage of plants with Lepas. Different letters above the box-plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, outliers.

Fig. 1.3. Box plot of response variables of stranded individuals of Durvillaea 19 antarctica on beaches from the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). (A) Mean length of plants, (B) Mean wet weight of plants, (C) Mean number of stipes of plants, and (D) Percentage of plants in reproductive stage. Different letters above the box-plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, outliers.

Fig. 1.4. Percentages of males and females of stranded individuals of 20 Durvillaea antarctica on beaches from the continental coast of Chile (28°S- 42°S), according to marine biogeographic districts during winters in three consecutive years. (A) winter 2013, (B) summer 2013/2014, (C) winter 2014, (D) summer 2014/2015, (E) winter 2015, (F) summer 2015/2016. CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. The numbers on top of each column correspond to frequencies of females and males, respectively.

xvii

Fig. 1.5. Frequency of stranded individuals of Durvillaea antarctica on 21 beaches from the continental coast of Chile (28°S-42°S), according to three categories of floating times (short, intermediate, long). (A) Number of individuals sampled in non-reproductive and reproductive stage, (B) Number of individuals sampled ≤ 50 cm and > 50 cm frond length.

Fig. 2.1. Geographical distribution of surveyed beaches and biogeographic 32 districts described for the coast of Chile distinguished in this study (Coquimbo-Choros District: 28°S-30°S, Septentrional District: 30°S-33°S, Mediterranean District: 33°S-37°S, Meridional District: 37°S-42°S). The geographic distribution of Durvillaea antarctica within the study area is also indicated.

Fig. 2.2. Stranded biomass and number of rafts of Durvillaea antarctica on 37 three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Details of post-hoc analyses are shown in Table S2.1.

Fig. 2.3. Relationships between stranded biomass of Durvillaea antarctica 39 (kg per km of shoreline) and Douglas sea scale (mean and maximum values of weekly lag according to the date of each survey) on three beaches from the continental coast of Chile. A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated.

Fig. 2.4. Average (mean ± SD) length of stranded individuals Durvillaea 41 antarctica on three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A: Pichicuy (32°S), B: Itata Norte (36°S) and C: Curiñanco (39°S). Only stranded specimens > 50 cm long were considered (n = 1,904). Details of post-hoc analyses are shown in Table S2.2.

Fig. 2.5. Average (mean ± SD) wet weight of stranded individuals 42 Durvillaea antarctica on three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A: Pichicuy (32°S), B: Itata Norte (36°S) and C: Curiñanco (39°S). Only stranded specimens > 50 cm long were considered (n = 1,904). Details of post-hoc analyses are shown in Table S2.2.

Fig. 2.6. Percentage of stranded individuals of Durvillaea antarctica with 44 Lepas spp. attached, and sizes (mm of capitular length, mean ± SD) of specimens of Lepas spp. on three beaches from the continental coast of Chile. A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S), according to bimonthly surveys from 2014 to 2017.

xviii

Fig. 3.1. Study area and spatial distribution of the stranded and benthic D. 65 antarctica in central Chile. PTOS: Puerto Oscuro; TOT: Totoralillo Sur.

Fig. 3.2. Durvillaea antarctica: frequency of holdfasts with different 68 numbers of stipes, and with only vegetative stipes, only male stipes, only female stipes, and both sexes within one holdfast. Samples collected at (A) 33 beaches on the coast of Chile between 28°S and 42°S (n = 1044 stranded holdfasts during the winters 2013, 2014 and 2015), and (B) two sites near the northern distribution limit of D. antarctica (n = 56 holdfasts from Puerto Oscuro and n = 62 holdfasts from Totoralillo Sur).

Fig. 3.3. Durvillaea antarctica: average (mean ± SD) colour values of algal 68 samples of different sexual stages and from two localities, Puerto Oscuro and Totoralillo Sur. Different letters above the columns indicate differences between sexual stages significant at p = 0.05. Numbers of stipes from each site and sexual stage are listed at the bottom of each column.

Fig. 3.4. Durvillaea antarctica: average (mean ± SD) concentration (mg/g 69 wet wt) of pigments in blade samples of different sexual stage and locality, Puerto Oscuro and Totoralillo Sur. Different letters above the columns indicate differences between sexual stages significant at p = 0.05. Numbers of stipes from each site and sexual stage are listed at the bottom of each column.

Fig. 3.5 Durvillaea antarctica: average (mean ± SD) concentration (% dry 70 wt) of phlorotannins in blade samples of different sexual stage and locality, Puerto Oscuro and Totoralillo Sur. Different letters above the columns indicate differences between sexual stages significant at p = 0.05. Numbers of stipes from each site and sexual stage are listed at the bottom of each column.

Fig. 4.1. Geographical distribution of sampling sites and biogeographic 79 districts described for the coast of Chile described in this study (Coquimbo- Choros District: 28°S-30°S, Septentrional District: 30°S-33°S, Mediterranean District: 33°S-37°S, Meridional District: 37°S-42°S). The geographic distribution of Durvillaea antarctica within the study area is also indicated. The number of beaches sampled within each biogeographic district is shown. The size of gray circles represents the sampled individuals of D. antarctica in each biogeographic district.

xix

Fig. 4.2. Average (mean ± SD) taxonomic richness of epibionts attached on 80 stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to marine biogeographic districts (CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District). (A) Taxonomic richness at plant-level and accumulated per beach in the different districts (Chao 2 index). (B) Average taxonomic richness of sessile and mobile epibionts. Letters (a-b-c-d) above the columns indicate significant differences between biogeographic districts (P< 0.05). Letters in italics correspond to the results of accumulated taxonomic richness (in A) or mobile species (in B). Number of plants (at plant-level) and beaches (total accumulated) from each biogeographic district are listed at the bottom of each column.

Fig. 4.3. Average (mean ± SD) taxonomic richness of epibionts attached on 81 stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to floating time (short, < 2 days; intermediate, 2 - 10 days and long, > 10 days). (A) Taxonomic richness at plant-level and accumulated per floating time (Chao 2 index). (B) Average taxonomic richness of sessile and mobile epibionts. Letters (a-b-c) above the columns indicate significant differences between floating time categories (P< 0.05). Letters in italics correspond to the results of accumulated taxonomic richness (in A) or mobile species (in B). Number of plants (at plant-level) and beaches (total accumulated) from each floating time category are listed at the bottom of each column.

Fig. 4.4. Species co-occurrence matrix of frequent epibionts attached on 82 stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to positive, negative and random species co-occurrences. The numbers of positive/negative co- occurrences for each species are shown.

Fig. 4.5. Comparison of literature and rafting ranges of epibiont species 83 attached on stranded bull kelp Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S). (A) Species with rafting ranges within their literature ranges (RO). (B) Species with rafting ranges that surpass the southern edges of their literature ranges, southward extension (SE). (C) Species with rafting ranges that surpass the northern edges of their literature ranges, northward extension (NE). Sessile and mobile species are indicated with triangles and circles, respectively. Study area (light gray area) and northern edge of the distribution range of D. antarctica (dark gray line) are also indicated.

xx

Fig. 4.6. Box plot of percentages of stranded individuals of Durvillaea 83 antarctica with epibionts outside their literature ranges according to absence and presence of Lepas spp. for SE (southward extension) and NE species (northward extension) on beaches along the continental coast of Chile (28°S- 42°S). Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range.

Fig. 4.7. Box plot of percentages of stranded individuals of Durvillaea 84 antarctica with epibionts outside their literature ranges (% per beach) according to absence and presence of epibiont species with positive co- occurrences, (A) Gelidium lingulatum and (B) Semimytilus algosus, and epibiont species with negative co-occurrences, (C) Limnoria chilensis and (D) Scurria scurra for SE (southward extension) and NE species (northward extension) on beaches along the continental coast of Chile (28°S-42°S). Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range.

Fig. 5.1. Geographic location of the coast of Chile, showing the main 92 biogeographic zones (provinces and districts) and breaks (30°S-33°S and 41°S-42°S) (modified from Camus 2001). The sampled distributions of Gelidium lingulatum and G. rex in the Chilean coast are also indicated (shaded bars), as well as the geographic distribution according to the literature (lines). Also, local strandings (stranded biomass and length) of floating bull kelps Durvillaea antarctica in different biogeographic districts are shown (correspondence between circle sizes and stranded biomasses, and kelps sizes and lengths are indicated) (extracted from López et al. 2017).

Fig. 5.2. Geographic distribution of haplotypes and haplotype networks of 98 Gelidium lingulatum for mitochondrial COI and chloroplastic rbcL markers. Sampling locations where no individuals of the species were found from the northern part of the study area are also indicated. Photographs of a specimen and intertidal patches of species are shown in the lower right. The within- location diversity and the geographical extent of each haplotype are shown. On the map each circle represents a location and the proportion of pie chart indicates the frequency of individuals for each haplotype. The pie chart color-code corresponds to the one used in haplotype networks of each marker. In the networks, each circle represents a haplotype and its size is proportional to the frequency in which the haplotype was encountered (correspondence between circle sizes and numbers of individuals is indicated). Perpendicular bars between each haplotype pair correspond to the number of mutational steps among them. Abbreviations for location codes are as in Table 1.

xxi

Fig. 5.3. Geographic distribution of haplotypes and haplotype networks of 99 Gelidium rex for mitochondrial COI and chloroplastic rbcL markers. Sampling locations where no individuals of the species were found from the northern and southern sites of the study area are also indicated. Photographs of a specimen and intertidal patches of species are shown in the lower right. See legend of Fig. 6.2 for details.

Fig. 5.4. Scatter plot of genetic differentiation and geographic distance of 100 pairwise locations for COI marker. (A) Gelidium lingulatum, and (B) G. rex. Pairwise genetic distances, represented as D, are plotted against pairwise geographic distances (km). Each point corresponds to a pairwise comparison of locations. The results of the statistical analyses and the regression line for significant relationship (G. rex) are also shown. Locations with N < 14 were excluded from analyses.

Fig. 5.5. Mismatch distribution for COI datasets for Gelidium lingulatum 101 (A), G. rex north (B), G. rex north-f (C) and G. rex south (D), according to spatial expansion models. The observed distributions of the number of pairwise differences (bars) are contrasted to their expected distributions (solid lines) under a model of spatial expansion.

Fig. 6.1. Map of the continental coast of Chile, showing the main results of 113 this study. The main biogeographic zones (provinces and districts) and breaks (30°S and 41°S–42°S) (modified from Camus 2001) are indicated.

Fig. 6.2. Stranded biomass, extension of potentially suitable habitat (PSH) 114 and benthic biomass of Durvillaea antarctica along the continental coast of Chile (30°S – 43°S). Mean stranded biomass values for 28 beaches were obtained from Chapter 1 (Extracted from Schreiber 2018).

xxii

List of Appendices

Appendix 1: Fig. S1.1. Box plot of stranded biomass of Durvillaea 128 antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer and years (2013-2015). Different letters above the box plot indicate differences between biogeographic districts and years (P < 0.05). CCD: Coquimbo- Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

Appendix 1: Fig. S1.2. Box plot of maximum length of stranded individuals 129 of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo- Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

Appendix 1: Fig. S1.3. Box plot of maximum wet weight of stranded 130 individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo- Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

Appendix 1: Fig. S1.4. Box plot of maximum number of stipes per plant of 131 stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

xxiii

Appendix 1: Fig. S1.5. Box plot of maximum size of Lepas spp. attached in 132 stranded individuals of Durvillaea antarctica on beaches to the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

Appendix 1: Table S1.1. Beaches sampled in the study, according to 133 biogeographic districts (Coquimbo-Choros District and Septentrional District) of continental coast of Chile (28°S-33°S). Coordinates and distances surveyed are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth.

Appendix 1: Table S1.2. Beaches sampled in the study, according to 134 biogeographic districts (Mediterranean District and Meridional District) of continental coast of Chile (33°S-42°S). Coordinates and distances surveyed are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth.

Appendix 2: Table S2.1. Summary of GLM (only source of variations and 136 P-values) for stranded biomass and number of rafts of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Pairwise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Nonsignificant pairwise tests are not shown. Pi: Pichicuy; It: Itata Norte; Cu: Curiñanco; Y1: year 1 (2014/15), Y2: year 2 (2015/16); Y3: year 3 (2016/17); Ju: July; Se: September; No: November; Ja: January; Ma: March; May: May.

Appendix 2: Table S2.2. Summary of GLM (only source of variations and 138 P-values) for length and wet weight of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Pairwise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Nonsignificant pairwise tests are not shown. Pi: Pichicuy; It: Itata Norte; Cu: Curiñanco; Y1: year 1 (2014/15), Y2: year 2 (2015/16); Y3: year 3 (2016/17); Ju: July; Se: September; No: November; Ja: January; Ma: March; May: May.

xxiv

Appendix 2: Fig. S2.1. Summary of main change tendencies on studied 140 responses of bull kelp Durvillaea antarctica strandings. Letters (a-b-c-d) in the right column indicate significant differences between factor categories (P < 0.05). Letters on the right side correspond to the results of pairwise comparisons for each response variable. Horizontal solid gray lines correspond to reference for significant groups.

Appendix 2: Fig. S2.2. Relationships between stranded biomass of 141 Durvillaea antarctica (kg per km of shoreline) and Douglas sea scale (mean and maximum values of biweekly lag according to the date of each survey) on three beaches from the continental coast of Chile (28°S-42°S). A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated.

Appendix 2: Fig. S2.3. Relationships between stranded biomass of 142 Durvillaea antarctica (kg per km of shoreline) and Douglas sea scale (mean and maximum values of monthly lag according to the date of the each survey) on three beaches from the continental coast of Chile (28°S-42°S). A- B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated. Appendix 2: Fig. S2.4. Percentages of stranded biomass of bull kelps of 143 Durvillaea antarctica, according to complete plants (i.e. individuals with intact fronds that are coalesced within a single holdfast) and fragments (i.e. parts of a frond without holdfasts or only holdfasts) on three beaches from the continental coast of Chile (28°S-42°S) during bimonthly surveys on three years (2014/15-2016/17). A-B-C: Pichicuy (32°S), D-E-F: Itata Norte (36°S) and G-H-I: Curiñanco (39°S). Appendix 2: Fig. S2.5. Bimonthly accumulated landings of artisanal 144 fisheries of bull kelps (Durvillaea antarctica) on three political- administrative regions of Chile which include the surveyed beaches, between July 2014 to June 2017. A: Vth Region (for Pichicuy, 32°S), B: VIIIth Region (for Itata Norte, 36°S) and C: XIVth Region (for Curiñanco, 39°S). Source: Fisheries Statistical Yearbook 2014-2016. National Fisheries and Aquaculture Service (Anuario Estadístico de Pesca. Servicio Nacional de Pesca y Acuicultura). Gobierno de Chile. (http://www.sernapesca.cl/informes/estadisticas).

xxv

Appendix 3: Table S3.1.1. Beaches sampled in the study, according to 146 biogeographic districts (Coquimbo-Choros District, CCD and Septentrional District, SED) along the continental coast of Chile (28°S-33°S). Coordinates, distances surveyed and number of sampled individuals of Durvillaea antarctica are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth. Number of sampled specimens of D. antarctica are shown in parentheses.

Appendix 3: Table S3.1.2. Beaches sampled in the study, according to 147 biogeographic districts (Mediterranean District, MED and Meridional District, MD) along the continental coast of Chile (33°S-42°S). Coordinates, distances surveyed and number of sampled individuals of Durvillaea antarctica are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth. Number of sampled specimens of D. antarctica are shown in parentheses.

Appendix S3.2: Expanded Materials and Methods 148

Appendix 3: Table S3.3.1: Epibiont species attached on stranded individuals 152 of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S) selected for analysis of comparison of geographical ranges. Type of epibiont (seaweed or animal), motility (sessile or mobile), geographic range limits (northern and southern edges), percentage of individuals of D. antarctica with Lepas spp. attached (total records of each species are also indicated in parentheses), autonomous dispersal ability (high or low, according to the criterion based on the presence of planktonic larvae or propagules in the water column for more than 7 days, a period previously described in Haye et al. 2014 as important for the genetic differentiation of marine invertebrate populations across the phylogeographic break at 30°S on the coast of Chile), reproduction type (DD: direct development; PL: pelagic larval; PP: propagules such as spores, gametes, zygotes, papillae or sporophylls; VR: vegetative reproduction, such as fragmentation or prostrate thalli; pelagic larval duration are also mentioned) and references for each species are indicated.

Appendix S3.4: Expanded Results 171

Appendix 3: Table S3.4.1. Seaweed species found on holdfasts of stranded 172 bull kelp Durvillaea antarctica on sandy beaches from the continental coast of Chile (28°S - 42°S) during winters and summers 2014/15-2015/16. The percentage is based on a total of 5,219 stranded individuals of D. antarctica. Only species with more than one record were included in statistical analyses. Species that were incorporated in (*) co-occurrences and () geographical range analysis. The unidentified species corresponded to individuals that due to their small size, absence of reproductive structures and/or tissue deterioration, it was not possible to reach a more specific taxonomic level.

xxvi

Appendix 3: Table S3.4.2. Invertebrate species found on holdfasts of 175 stranded bull kelp Durvillaea antarctica on sandy beaches from the continental coast of Chile (28°S - 42°S) during winters and summers 2014/15-2015/16. The percentage is based on a total of 5,219 stranded individuals of D. antarctica. Only species with more than one record were included in statistical analyses. (*) Species that were incorporated in co- occurrence analysis. (#) Species that were incorporated in geographical ranges analysis. (+) Mobile taxa. The unidentified species corresponded to individuals that due to their small size, absence of reproductive structures and/or tissue deterioration, it was not possible to reach a more specific taxonomic level.

Appendix 3: Table S3.4.3. Summary of three-way ANCOVA for taxonomic 177 richness of epibionts attached (total, sessile and mobile) on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to biogeographic district (CCD, SED, MED and MD), year (2014 and 2015), season (winter and summer) and holdfast wet weight as covariate. Significant values (P < 0.05) are shown in bold.

Appendix 3: Table S3.4.4. Summary of one-way ANCOVA for taxonomic 179 richness of epibionts attached (total, sessile and mobile) on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to floating time (short, intermediate and long) and holdfast wet weight as covariate. Significant values (P < 0.05) are shown in bold.

Appendix 3: Fig. S3.4.1. Average (mean ± SD) taxonomic richness of 180 epibionts attached on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to years (2014 and 2015) and seasons (winter and summer). (A) Taxonomic richness at plant-level and accumulated per season and year (Chao 2 index). (B) Average taxonomic richness of sessile and mobile epibionts. Letters (a-b-c-d) above the columns indicate significant differences between season and year (P < 0.05). Letters in italics correspond to the results of accumulated taxonomic richness (in A) or mobile species (in B). Number of plants (at plant-level) and beaches (total accumulated) from each year and season are listed at the bottom of each column.

Appendix 3: Fig. S3.4.2. Relationship between taxonomic richness of 181 associated epibionts and holdfast wet weight (kg) on stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S- 42°S). Summary of ANOVA of simple linear regression between both variables is indicated.

xxvii

Appendix 3: Table S3.4.5. Summary of GLMs with a negative binomial 182 distribution of the frequencies of raft-associated species of stranded D. antarctica found outside their known benthic ranges according to the presence/absence of two species with high positive co-occurrences (Gelidium lingulatum and Semimytilus algosus) and two species with high negative co- occurrences (Limnoria chilensis and Scurria scurra) for SE and NE species from the continental coast of Chile (28°S-42°S). Significant values (P< 0.05) are shown in bold.

Appendix 3: Fig. S3.4.3. Number of stranded rafts of Durvillaea antarctica 183 per km surveyed shoreline (white bars) and their epibiont species attached on beaches along the latitudinal gradient (bins of 1°) of the Chilean coast (28°S- 42°S), according to literature (light gray bars) and rafting ranges (dark gray bars). Epibiont species are shown in three categories, RO: rafting occurs within their literature ranges, SE: rafting surpasses the southern edges of their literature ranges, NE: rafting surpasses the northern edges of their literature ranges.

Appendix 4: Table S4.1. Sampling sites from the northern part of the study 185 area, where no individuals of Gelidium lingulatum and G. rex (*) were found. The code used to identify each sampling site and coordinates (latitude and longitude) are indicated.

Appendix 4: Table S4.2. Variable sites and sequence frequencies for rbcL 186 of Gelidium lingulatum and G. rex.

Appendix 4: Table S4.3. Spatial analysis of molecular variance (SAMOVA) 187 for 2 to 10 groups of locations of Gelidium lingulatum, and for 2 to 5 groups of locations of G. rex. The codes represent the sampling sites. Locations with N ˂ 14 were excluded from analysis. G1, G2 and Gn corresponding to the groups of locations formed after analysis, according to the number of groups required. In bold, the best fit of groups is shown.

Appendix 4: Table S4.4. Population pairwise φST values for mitochondrial 190 marker COI in Gelidium lingulatum. The codes represent 15 sampling sites (locations with N < 14 were excluded from analysis). Below: φST / Above: P- values. Significant φST values (P-values after Bonferroni correction) are shown in bold.

Appendix 4: Table S4.5. Population pairwise φST values for mitochondrial 191 marker COI in Gelidium rex. The codes represent 10 sampling sites (locations with N < 14 were excluded from analysis). Below φST /Above: P- values. Significant φST values (P-values after Bonferroni correction) are shown in bold.

xxviii

I. General Introduction

Dispersal patterns of floating seaweeds Rafting is one of the most effective dispersal mechanisms in marine ecosystems, facilitating the connectivity and gene flow among marine populations over large scales (> 100 - 1,000 km, Thiel and Haye 2006). Various types of floating substrata have been reported, with buoyant seaweeds being one the most common ones (Thiel and Gutow 2005a). Macroalgae have very short-lived planktonic dispersal stages (spores, gametes) (Santelices 1990a), and most of them are negatively buoyant and sink to the seafloor when detached from the primary substratum, but some species possess high buoyancy, mainly brown algae (e. g. species from the genus Macrocystis, Sargassum, Ascophyllum, Carpophyllum and Fucus), allowing individuals to travel long distances at the sea surface (Thiel and Gutow 2005a). Several studies have reported large densities of floating seaweeds in coastal areas, including occasional events of transoceanic rafting (e.g. Ingólfsson 1998, Macaya et al. 2005, Fraser et al. 2010, Coyer et al. 2011a, 2011b, Garden et al. 2011, Rothäusler et al. 2012, Wichmann et al. 2012, Saunders 2014, Gutow et al. 2015, Fraser et al. 2018). Also, there are records for several raft-associated species (i.e. non-buoyant seaweeds, invertebrates), suggesting that rafting dispersal via floating seaweeds would allow population connectivity over long distances (e.g. Maggs et al. 2008, Nikula et al. 2010, Fraser et al. 2011, Nikula et al. 2013, Cumming et al. 2014, Macaya et al. 2016, Waters et al. 2018a). Different biotic and abiotic factors affect the buoyancy and transport capabilities of floating seaweeds. High temperatures and strong solar radiation tend to reduce the floating time, particularly at low latitudes (Rothäusler et al. 2012), although acclimation and persistence at the sea surface may vary seasonally among species (Graiff et al. 2013, Tala et al. 2016, 2017). Also, associated grazers (i.e. isopods and amphipods) as well as abundant epibionts may suppress the survival of floating seaweeds (Rothäusler et al. 2011a, 2011b, Graiff et. al. 2016). Moreover, local oceanographic factors, such as currents and winds, strongly influence the dispersal trajectories of floating seaweeds (Garden et al. 2014,

1

Rothäusler et al. 2015, Hawes et al. 2017, Craw and Waters 2018, Waters et al. 2018b, Waters and Craw 2018). Although rafting has been one of the most studied dispersal mechanisms in recent years (see reviews by Macaya et al. 2016, and Thiel and Fraser 2016), empirical support for rafting contributing to ongoing population connectivity is scarce. For example, in the case of the seaweed Carpophyllum maschalocarpum, there is indication for recent dispersal, contributing to higher genetic diversity in some local populations (Buchanan and Zuccarello 2012). For some kelp-associated benthic invertebrates (isopods from the genus Limnoria) algal rafting also appears to contribute to population connectivity after initial establishment (Haye et al. 2012). However, there is also strong indication that rafted propagules do not immigrate into established benthic populations. For example, in southern New Zealand, Collins et al. (2011), Fraser et al. (2011) and Bussolini and Waters (2015) found genetic patterns in beach-cast rafts of Durvillaea antarctica that do not coincide with local populations, suggesting a very low effective transport of individuals. Similar results have been reported along the Chilean coast, where the geographic distribution of genetic diversity of D. antarctica suggests that there is very little migration between relatively close populations (i.e. < 200 km, Fraser et al. 2010). Also, molecular studies support the effectiveness of rafting dispersal by floating seaweeds in the recolonization of previously disturbed areas (e.g. after earthquakes or covered by ice during the last glacial period, Fraser et al. 2009a, 2009b, 2009c, Neiva et al. 2012, Moon et al. 2017), particularly at high latitudes (i.e. > 40° S). A recent study showed rafting dispersal of D. antarctica in Antarctica, evidencing by molecular analysis that specimens had traveled > 20,000 km from mid-latitude source populations (Fraser et al. 2018). At temperate latitudes, at higher temperatures and higher solar radiation, successful immigration by this mechanism could be limited, even though several studies reported high abundances of floating seaweeds in coastal areas (Hobday 2000a, 2000b, 2000c, Smith 2002, Hinojosa et al. 2011).

Possible causes of low connectivity among benthic populations of seaweeds Several explanations have been proposed for understanding the low connectivity among populations, specifically in buoyant species that can perform long-distance dispersal by rafting: (i) reproductive traits: floating individual lose their potential reproductive capacity

2

(e.g. generate less propagules) when afloat in certain areas and seasons, or be differential between the sexes, (ii) physiological capacity: organisms or dispersed propagules might have depressed capacities for survival when environmental conditions are less favourable, for example, at the sea surface, (iii) oceanographic dynamics: local conditions affect transport and arrival of propagules to the coast, and (iv) ecological interactions: density- dependent effects reduce the immigration of propagules into established populations (Fig. A.1). Thus, many of the factors that could explain the low connectivity of species would be related to the intrinsic characteristics of propagules at the sea surface. However, considering the extensive geographic ranges of certain species, local adaptive attributes and other larger-scale phenomena (e.g. ENSO) could generate variability in production and migration of new individuals to other areas. Therefore, it could be expected that the dispersal capacity of a species could vary strongly throughout its geographical range.

Fig. I.1. Conceptual overview of biotic and abiotic factors (intrinsic and extrinsic to source populations) that might explain the low connectivity between benthic seaweed populations.

3

The abundant-centre hypothesis One way to explain this variability in the dispersal capacity of a species within its geographical range is the Abundant-Centre Hypothesis (ACH) (Brown 1984). The ACH predicts that within the range of a species the highest abundances are found in the middle of the species geographic range with optimum environmental conditions. As distance from the optimal point increases in any direction, the species experiences less favourable conditions, and its population declines up to a point where the environment becomes too extreme for the species’ survival, and therefore, physiological constraints limit populations at the edges of their geographic range (e.g. Sagarin and Gaines 2002a, 2002b, Fenberg and Rivadeneira 2011, Pironon et al. 2017). This hyphotesis has been used to contrast changes in life-history traits, ecological interactions and processes (e.g. reproduction, population dynamics, competition, extinction risk, resistant to climate change, among others) between central and edge populations of the distributional range of species from marine and terrestrial environments (see review Sagarin and Gaines 2002b, Sagarin et al. 2006, Pironon et al 2017). However, there are few studies that directly or indirectly assess the dispersal ability of species within its geographic range, particularly in coastal environments. Peripheral populations are often geographically isolated, smaller and more fragmented than central locations (Araújo et al. 2011, 2015, Cahill and Levinton 2016, Iacchei et al. 2016), therefore, it is expected that abiotic and biotic factors differentially affect the dispersal potential the floating (and/or rafting) species throughout their geographical ranges. In this manner, if the predictions of the ACH are fulfilled and rafting dispersal does not have an effective contribution to the connectivity of populations, it is expected that the low effective dispersal observed in both central and edge areas are due to different constraints. In central populations, high abundances of viable propagules are expected, but because of the greater abundance of the resident populations, strong density-dependent effects (i.e. density blocking; see Waters et al. 2013, Fraser et al. 2017) should reduce the successful immigration from distant populations. On the other hand, towards peripheral areas where the conditions are suboptimal and populations are smaller, the abundance and quality of propagules is suppressed which would mean less immigration, mainly explained by biological factors (i.e. functional or reproductive traits) of the species (Sagarin and

4

Gaines 2006). However, this expected pattern could also vary due to local oceanographic factors (i.e. coastal currents and winds) that affect the dispersal, particularly in certain species with an extensive geographical distribution (Garden et al. 2014, Rothäusler et al. 2015). Coastal species are a good model to test these specific hypotheses framed within the context of the ACH, because many have wide distributional ranges, but typically narrow in width (< 0.1 km), reducing it to one-dimensional ranges with center and edge areas more easily defined (Sagarin et al. 2006). The characteristics of the continental coast of Chile, considering its geographical extent, a linear topography, a strong latitudinal gradient of environmental variables (e.g. temperature and solar radiation, Thiel et al. 2007), different biogeographical zones that indicate local and historical oceanographic conditions that may affect dispersal (Camus 2001, Hormazábal et al. 2004, Haye et al. 2014), and a latitudinal gradient in the abundance of floating macroalgae along the coast (Hinojosa et al. 2010, 2011, Wichmann et al. 2012) that can carry other marine organisms (Thiel 2003, Wichmann et al. 2012, Macaya et al. 2016), makes this an ideal study area to evaluate the effectiveness of rafting dispersal.

General hypothesis Ecological factors (i.e. intraspecific competition) are more important in the effective dispersal of individuals of D. antarctica in central populations of its geographic range, while in populations from the edges of the geographic range reproductive and functional characteristics of the specimens will be more important. Deviations from this prediction of the effective dispersal within the geographic range can be explained by local oceanographic conditions.

Model species To test the hypothesis of this study, the southern bull kelp Durvillaea antarctica and its associated macroalgae were used as model organisms. Durvillaea antarctica (Chamisso) Hariot 1892 is a brown seaweed that has a wide distribution in the southern hemisphere, with a predominantly subantarctic distribution (Fraser et al. 2009a, 2010, 2011, Batista et al. 2018). It is a dioecious species that exhibits holdfast coalescence, joining multiple

5 conspecifics (González et al. 2015). In Chile, benthic populations occur from 30ºS, the northern distribution limit, to the Strait of Magellan (55°S) in the south (Ramírez and Santelices 1991, Hoffmann and Santelices 1997). Genetic studies indicate that there are two distinct clades of D. antarctica, a clade from southern-central Chile that is distributed along the continental coast of Chile (30°S-44°S), and the subantarctic clade that occurs throughout the entire subantarctic region, including southern Chile (49°S-56°S) (Fraser et al. 2009a, 2010). The continental clade features distinct genetic differences between adjacent populations (i.e. < 200 km), indicating that oceanographic, ecological or biological factors may suppress rafting dispersal and limit effective connectivity between populations (Fraser et al. 2010). Durvillaea antarctica inhabits rocky areas, exposed or semi-exposed to waves, between the lower intertidal and subtidal zone, up to 10-15 m depth (Hoffmann and Santelices 1997). Many different associate species inhabit the coalesced holdfasts of benthic individuals (Santelices et al. 1980). Durvillaea antarctica shows a strong seasonal pattern of abundance mainly in summer, when massive recruitment can occur (Castilla and Bustamante 1989, Bustamante and Castilla 1990, Westermeier et al. 1994, Castilla et al. 2007). The abundance of D. antarctica increases with wave-exposure, sharing its habitat with Lessonia nigrescens (currently L. spicata) (Santelices et al. 1980, Westermeier et al. 1994). The phenology of gametogenesis, maturity stages, embryogenesis and ontogenetic development of zygotes of D. antarctica have been described for central, and southern Patagonia of Chile (Collantes et al. 1997, 2002, Mansilla et al. 2017). Collantes et al. (2002) indicated that along the continental coast of Chile the reproductive period is mainly between autumn and early spring, followed by a vegetative period in summer. Durvillaea antarctica has positive buoyancy, being a common floating kelp along the Chilean coast (Hinojosa et al. 2010, 2011, Wichmann et al. 2012) with the potential to travel long distances (i.e. > 1,000’s of km, see Fraser et al. 2011). The morphological, physiological and reproductive stage of floating individuals varies seasonally (Graiff et al. 2013, Tala et al. 2013, 2016, 2017). Furthermore, there are abundant strandings of kelp rafts on rocky and sandy beaches of southern-central Chile (Rodríguez 2003, Duarte et al. 2008, 2009, Quintanilla-Ahumada et al. 2018), but there is no systematic information on the spatio-temporal dynamics of their strandings across the extensive distribution of the

6 continental clade of D. antarctica, which is an important basis to evaluate restrictions in population connectivity. Another important factor that should be considered is that D. antarctica frequently harbors several species of non-buoyant seaweeds (mainly red algae) and other invertebrates that can be found attached to the holdfast (Macaya et al. 2016). The main associated seaweed species, according to their relative abundances, are Lessonia spicata, Corallina officinallis var chilensis and several species from the genus Gelidium. Two species from this latter genus (G. lingulatum and G. rex), which have a similar geographic range as D. antarctica, were chosen to determine whether rafting dispersal via D. antarctica affects the genetic patterns of these epiphytic species. Gelidium lingulatum (Kützing 1868) and G. rex (Santelices & Abbott 1985) are two endemic red seaweeds from the Southeast Pacific (Ramírez and Santelices 1991). They form mono-specific patches at wave-exposed sites (Ortega et al. 2001), mainly in the intertidal zone, where they grow attached to rocks and calcareous shells, but also to holdfasts of large kelps (Santelices 1990b, Macaya et al. 2016, Otaíza et al. 2018). Particularly, G. lingulatum is present at 1-2 m above MLLW (mean lower low water) and G. rex is most often found at lower heights, about 0-0.5 m above MLLW (Santelices 1986). The geographical distribution of G. lingulatum extends from Antofagasta (23°S) to Tierra del Fuego (56°S) (Hoffmann and Santelices 1997). However, the current distribution of the species is not clearly established, because previous identifications were based solely on morphological characters, regardless of high potential phenotypic plasticity and the morphological similarity to other species of the genus (Santelices 1990b). On the other hand, G. rex has morphological features facilitating the identification in the field and it is distributed more narrowly between Coquimbo (30°S) and Concepción (36°S) (Hoffmann and Santelices 1997), although it has been reported that its northern distribution could extend to 16°S (Santelices and Abbott 1985). Both species present a reproductive cycle with an alternation of haploid and diploid phases (Hernández 1997) and have limited ability to spread by spores (Bobadilla and Santelices 2005). Therefore, the dispersal as epiphytes on floating seaweeds could affect the genetic diversity along the latitudinal gradient.

7

General and specific objectives The information presented above indicates that the benthic populations of D. antarctica from the continental coast of Chile would have low connectivity and gene flow, even though high abundances of floating specimens have been reported in coastal areas. Therefore, the general objective of this thesis was to examine the dispersal patterns of the floating algae D. antarctica along the continental coast of Chile, as well as to evaluate its role as a disperser of other associated species. For this, the following specific objectives have been proposed:

a) Analyze the spatio-temporal dynamics of D. antarctica strandings on beaches of continental Chile (28°S-42°S). b) Determine the variability of floating bull kelp D. antarctica supplies to beaches of the coast of Chile over shorter (bimonthly) time scales, relating to the local storm intensity. c) Identify possible reproductive and functional constraints that may exist in coalesced individuals from benthic and floating populations of D. antarctica along the coast of Chile. d) Evaluate the diversity, co-occurrences and range extensions of epibionts on holdfasts of stranded specimens of southern bull kelp D. antarctica along the continental coast of Chile (28°S-42°S). e) Determine the phylogeographic patterns of two frequent epiphytic algae found attached on holdfast of bull kelps D. antarctica along the coast of Chile.

This study has been divided into five chapters. In the first two chapters, the temporal (semiannually and bimonthly) and spatial (several locations) variability of supplies of bull kelps on beaches and the characteristics of stranded individuals along the continental coast of Chile were evaluated. In the next chapter, reproductive and physiological characteristics of stranded and benthic individuals of D. antarctica were analyzed. The fourth chapter evaluated how the observed stranding dynamics of bull kelps may have consequences on the diversity of raft-associated species, as well as the biological interactions among epibionts can affect their transport and dispersal to other areas. In the last chapter the

8 phylogeographic patterns of benthic populations of two epiphytic seaweeds from the genus Gelidium, commonly found on holdfasts of stranded bull kelps across the continental coast of Chile, was determined. Finally, the main results and conclusions obtained in the chapters are evaluated in a General Discussion.

9

Chapter 1

The variable routes of rafting: stranding dynamics of floating bull-kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Pacific

(manuscript published in J. Phycol. 53: 70-84, 2017)

Boris A. López1,2, Erasmo C. Macaya3,4,5, Fadia Tala1,6, Florence Tellier7,8 & Martin Thiel1,4,9*

1Facultad de Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 2Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos, Avenida Fuchslocher 1305, Osorno, Chile. 3Laboratorio de Estudios Algales (ALGALAB), Departamento de Oceanografía, Universidad de Concepción, Casilla 160-C, Concepción, Chile. 4Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile. 5Centro FONDAP de Investigaciones en Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL). 6Centro de Investigación y Desarrollo Tecnológico en Algas (CIDTA), Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 7Departamento de Ecología, Facultad de Ciencias, Universidad Católica de la Santísima Concepción, Casilla 297, Concepción, Chile. 8Centro de Investigación en Biodiversidad y Ambientes Sustentables (CIBAS), Universidad Católica de la Santísima Concepción, Casilla 297, Concepción, Chile. 9Centro de Estudios Avanzados en Zonas Áridas, CEAZA, Coquimbo, Chile.

*Corresponding author: [email protected].

10

J. Phycol. 53, 70–84 (2017) © 2016 Phycological Society of America DOI: 10.1111/jpy.12479

THE VARIABLE ROUTES OF RAFTING: STRANDING DYNAMICS OF FLOATING BULL KELP DURVILLAEA ANTARCTICA (FUCALES, PHAEOPHYCEAE) ON BEACHES IN THE SE PACIFIC1

Boris A. Lopez Facultad de Ciencias del Mar, Universidad Catolica del Norte, Larrondo 1281, Coquimbo, Chile Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos, Avenida Fuchslocher 1305, Osorno, Chile Erasmo C. Macaya Laboratorio de Estudios Algales (ALGALAB), Departamento de Oceanografıa, Universidad de Concepcion, Casilla 160-C, Concepcion, Chile Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile Centro FONDAP de Investigaciones en Dinamica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Valdivia, Chile Fadia Tala Facultad de Ciencias del Mar, Universidad Catolica del Norte, Larrondo 1281, Coquimbo, Chile Centro de Investigacion y Desarrollo Tecnologico en Algas (CIDTA), Universidad Catolica del Norte, Larrondo 1281, Coquimbo, Chile Florence Tellier Departamento de Ecologıa, Facultad de Ciencias, Universidad Catolica de la Santısima Concepcion, Casilla 297, Concepcion, Chile Centro de Investigacion en Biodiversidad y Ambientes Sustentables (CIBAS), Universidad Catolica de la Santısima Concepcion, Casilla 297, Concepcion, Chile and Martin Thiel2 Facultad de Ciencias del Mar, Universidad Catolica del Norte, Larrondo 1281, Coquimbo, Chile Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile Centro de Estudios Avanzados en Zonas Aridas, CEAZA, Coquimbo, Chile

Dispersal on floating seaweeds depends on which would facilitate successful rafting dispersal, availability, viability, and trajectories of the rafts. In gamete release, and reproduction upon arrival. the southern hemisphere, the bull kelp Durvillaea High biomasses of stranded kelps occurred in the antarctica is one of the most common floating northern-central (30°S–33°S) and southernmost seaweeds, but phylogeographic studies had shown districts (37°S–42°S), and lower biomasses in the low connectivity between populations from northernmost (28°S–30°S) and southern-central continental Chile, which could be due to limitations districts (33°S–37°S). The highest percentages and in local supply and dispersal of floating kelps. To sizes of epibionts (Lepas spp.), indicative of test this hypothesis, the spatiotemporal dynamics of prolonged floating periods, were found on stranded kelp strandings were examined in four kelps in the northernmost and southernmost biogeographic districts along the Chilean coast (28°– districts. Based on these results, we conclude that 42°S). We determined the biomass and demography rafting dispersal can vary regionally, being more of stranded individuals on 33 beaches for three common in the northernmost and southernmost subsequent years (2013, 2014, 2015) to examine districts, depending on intrinsic (seaweed biology) whether rafting is restricted to certain districts and and extrinsic factors (shore morphology and seasons (winter or summer). Stranded kelps were oceanography) that affect local supply of kelps and found on all beaches. Most kelps had only one stipe regional hydrodynamics. (one individual), although we also frequently found Key index words: floating kelps; marine biogeography; coalesced holdfasts with mature males and females, rafting; stalked barnacles; strandings Abbreviations: ANOVA, analysis of variance; CCD, 1Received 12 April 2016. Accepted 5 September 2016. Coquimbo-Choros District; CSC, Coastal System of 2Author for correspondence: e-mail [email protected]. Editorial Responsibility: A. Buschmann (Associate Editor) Coquimbo; MD, Meridional District; MED,

70

11 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 71

Mediterranean District; PERMANOVA, permuta- F. vesiculosus based on ocean circulation and surface tional multivariate analysis of variance; PERMDISP, winds, Roth€ausler et al. (2015) revealed strong spa- permutational analysis of multivariate dispersions; tial and seasonal variability of trajectories in the RS, rocky shores; SED, Septentrional District; SS, Northern Baltic Sea, indicating that some coastal sandy shores areas were functioning as sources and others as sinks. Genetic studies on floating seaweeds indicate that gene flow between populations varies greatly In marine environments, floating seaweeds are among locations, suggesting that transport and con- one of the most common natural substrata that nectivity via seaweed rafts may be highly effective in serve as transport vehicle for other organisms (Hob- some areas and limited in others (Fraser et al. 2010, day 2000, Thiel and Gutow 2005a,b, Vandendriess- Guillemin et al. 2016). che et al. 2007). Furthermore, they support trophic While all these methods contribute to our knowl- subsidies and provide spatial refugia for both pela- edge of seaweed rafting, they provide no informa- gic and benthic communities (Duarte et al. 2009, tion on the quantity and status (reproductive stage, Duggins et al. 2016). While the role of floating sea- size, sex) of the seaweeds traveling between adjacent weeds as rafting vehicle for a wide range of organ- or distant beaches. This information, however, is isms has been widely described (Helmuth et al. fundamental to determine rafting dynamics, 1994, Ingolfsson 1995, Thiel 2003a, Nikula et al. because the likelihood of successful dispersal is 2010, Cumming et al. 2014), the supply, pelagic per- greater if the individuals reaching a local shore are sistence, and oceanic trajectories have received less (i) reproductive, (ii) of both sexes (in dioecious sea- attention. weeds such as some Fucales) or the sporophyte Rafting dispersal of floating seaweeds, which phase that can produce male and female gameto- includes traveling, arrival, and establishment in new phytes (Laminariales), (iii) of large size with high habitats, depends on intrinsic and extrinsic factors reproductive capacity (for sufficient propagules to (reviewed in Macaya et al. 2016). Over large scales be released and settled), and (iv) the new habitat is (>1,000 km), latitudinal gradients of temperature suitable for colonization. Furthermore, in order to and solar radiation affect buoyancy and survival of understand mesoscale rafting dynamics, it is funda- seaweed rafts, with higher pelagic persistence at mental to estimate floating times of rafts as this high latitudes (Macaya et al. 2005, Roth€ausler et al. information allows inferring potential floating dis- 2012, Tala et al. 2016). On smaller, regional scales, tances and source regions. The presence and sizes the connectivity among benthic populations can be of lepadid barnacles allow inferring floating times influenced by rafting supplies from adjacent popula- and distances of rafts from different sources (Thiel tions (Muhlin et al. 2008, Garden et al. 2011) and and Gutow 2005b). local oceanographic factors (Hinojosa et al. 2010, Seaweed strandings are frequent on many coasts, Garden et al. 2014, Roth€ausler et al. 2015), but the and are particularly important on boulder and knowledge of mesoscale (~1,000 km) dynamics of sandy beaches (McLachlan and Brown 2006), where floating seaweeds is limited. For example, Collins the seaweeds represent trophic subsidies to these et al. (2010) reported that in Canterbury Bight environments with low primary productivity, driving (southern New Zealand), there is a strong influence a number of ecological processes, such as degrada- of the north-flowing Southland Current on the dis- tion, consumption, habitat supply, and biogeochem- persal and local strandings of bull kelp Durvillaea ical processing (Kirkman and Kendrick 1997, antarctica (Chamisso) Hariot 1892. Also, Roth€ausler Duarte et al. 2008, 2009, Lastra et al. 2014). These et al. (2015) showed that currents and surface winds strandings on beaches can also be an indicator for in the Northern Baltic Sea affect the seasonal trajec- the arrival of floating kelps in a particular area, tories of floating Fucus vesiculosus (Linnaeus 1753), where some areas appear to receive higher supplies dispersing tens to hundreds of kilometers away from than others (Garden et al. 2011). The amounts of their sources. stranded seaweeds may also differ between seasons, Different approaches have been used to infer the due to more storm- and herbivore-induced seaweed sources and sinks of floating seaweeds, including detachments during fall and winter (Marsden genetic analyses (e.g., Muhlin et al. 2008, Collins 1991). Seasonal differences in raft survival and et al. 2010, Neiva et al. 2014, Bussolini and Waters reproductive activity, especially between summer 2015), hydrographic modeling (Roth€ausler et al. and winter (Graiff et al. 2013, Tala et al. 2013, 2015), radio telemetry (Harrold and Lisin 1989), 2016), might also affect dispersal dynamics and pop- and geomorphological indicators (Garden et al. ulation connectivity. 2011, Garden and Smith 2015). Using the type of Herein we used the bull kelp D. antarctica as rock attached to holdfasts of beach-cast D. antarctica model organism to determine rafting dynamics on as an indicator of source regions in southern New the regional scale. D. antarctica has a wide distribu- Zealand, Garden et al. (2011) found that some indi- tion in the southern hemisphere, with a predomi- viduals had traveled over 200 km from the original nantly subantarctic distribution (Fraser et al. 2009, source regions. Modeling dispersal of floating 2010, 2011). It is a dioecious species that exhibits

12 72 BORIS A. LOPEZ ET AL. holdfast coalescence joining multiple conspecifics linear topography with a north–south orientation, a latitudi- (Gonzalez et al. 2015). In Chile, benthic popula- nal temperature gradient in surface waters, as well as the tions occur from 30°S, the northern distribution absence of distinct geographic barriers to dispersal of marine organisms (Thiel et al. 2007). The Humboldt Current System limit (Hoffmann and Santelices 1997, Tala et al. ° (cold and nutrient-rich waters) influences the coastal area, 2013) to the Strait of Magellan (55 S) in the south. with extensive upwelling zones affecting coastal communities Genetic studies indicate that there are two distinct (Hormazabal et al. 2004, Thiel et al. 2007, Lachkar and Gru- clades of D. antarctica, a clade from southern-central ber 2012, Aravena et al. 2014). Chile that is distributed along the continental coast Based on the contrasting topological and oceanographic of Chile (30°S–44°S), which in the following will be characteristics, three major biogeographic regions have been called “continental clade”, and the subantarctic described for the marine biota: a northern area which com- prises a warm-temperate biota (Peruvian Province, 18°S– clade that occurs throughout the entire subantarctic ° ° – ° 30 S), a southern area with the austral biota (Magellan Pro- region, including southern Chile (49 S 56 S; Fraser vince, 42°S–56°S), and a nontransitional, intermediate area et al. 2009, 2010). The continental clade features (30°S–42°S) including mixed components of biota (Camus distinct genetic differences between adjacent popu- 2001). In this intermediate area, Camus (2001) distinguished lations, indicating that oceanographic, ecological, or several subregions (districts) based on local oceanographic biological factors may suppress rafting dispersal and features, such as for example differences in freshwater inputs from rivers. Also, the total extent and average length of the limit effective connectivity between populations sandy shores (SS) tends to be higher in southern-central (Fraser et al. 2010, Waters et al. 2013). Chile (Thiel et al. 2007) with a decreasing proportion of Durvillaea antarctica has positive buoyancy, being a rocky shores (RS) in the southern part of the study area common floating kelp along the Chilean coast (33°S–42°S; Fig. 1). (Hinojosa et al. 2010, Wichmann et al. 2012) with the potential to travel over distances of thousands of kilometers (Fraser et al. 2011). The morphologi- cal, physiological, and reproductive stage of floating individuals varies seasonally (Graiff et al. 2013, Tala et al. 2013, 2016), although there is no evidence for functional differences (i.e., pigment and phlorotan- nin concentrations) between sexes (Lizee-Prynne et al. 2016). Furthermore, there are abundant strandings of kelp rafts on rocky and sandy beaches of southern-central Chile (~39°S, see also Duarte et al. 2008, 2009). In New Zealand, the stranding dynamics of D. antarctica varied throughout the year with a slight tendency of higher biomass of stranded kelps in summer (Marsden 1991). However, in Chile, there is no systematic information on the spa- tiotemporal dynamics of bull kelp strandings across the extensive distribution of the continental clade of D. antarctica. This coastal zone has biogeographic regions and subregions with boundaries based on topographical and oceanographic features (Camus 2001, Hormazabal et al. 2004). Consequently, it is likely that the characteristics and floating times of stranded individuals vary between these regions, affecting the dispersal dynamics of bull kelp rafts. The current study thus aims to (i) determine the spatiotemporal pattern in the amounts of bull kelp D. antarctica strandings along the continental coast of Chile (28°S–42°S), (ii) characterize the status (size, sex, reproductive stage) and floating times (using epibionts Lepas spp.) of stranded bull kelps, and (iii) evaluate whether local stranding patterns coincide with the marine biogeographic districts described for the continental coast of Chile. FIG. 1. Geographic distribution of sampling sites and biogeo- graphic districts described for the coast of Chile distinguished in this study (Coquimbo-Choros District: 28°S–30°S, Septentrional MATERIAL AND METHODS District: 30°S–33°S, Mediterranean District: 33°S–37°S, Meridional District: 37°S–42°S). The geographic distribution of Durvillaea Oceanographic and biogeographic characteristics of the study antarctica within the study area is also indicated. RS/SS = ratio area. The South-East Pacific, particularly the southern-cen- rocky shoreline (km) versus sandy shoreline (km). The number tral coast of Chile (from ~28°Sto42°S), is characterized by a of beaches sampled within each biogeographic district is shown.

13 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 73

This study was conducted on 33 sandy and boulder bea- Reproductive stage: for each plant, tissue samples were ches (28°S–42°S) across the benthic and pelagic geographic taken for analysis of the reproductive stage. If a plant con- range of the continental clade of D. antarctica (Fig. 1). The tained more than one stipe, we took one tissue sample distance between beaches varied from 30 to 100 km and the per stipe up to a maximum of five stipes per plant (for each case, only the five longest stipes were considered). extension of the stretches that were surveyed on each beach â ranged from 0.28 to 11.08 km, depending on beach length Each sample was placed in a Ziploc bag, covered with salt and/or amounts of stranded kelps (Table S1 in the Support- to dehydrate and preserve the reproductive tissues. ing Information). All beaches were delimited by at least one Depending on the number of complete plants per beach, rocky shore and many beaches also contained small, rocky we took tissue samples from up to 10 plants per beach outcrops. While the kelp individuals stranded on sandy bea- and survey to determine the reproductive status and sex ches have very limited possibilities of colonization, they can of stranded kelps. be considered representative of stranding dynamics on adja- Floating time: for each plant we determined whether it cent rocky shores, the typical habitat of the bull kelp had been colonized by Lepas spp. or not. Most plants were D. antarctica. colonized by L. australis Darwin, 1851, which dominates in ° – ° Beaches were distributed across the four biogeographic dis- southern Chile (38 S 42 S; Hinojosa et al. 2006), and tricts defined by Camus (2001): the southern edge of the only relatively few samples in the northernmost district Peruvian Province, 28°S–30°S, hereafter termed Coquimbo- (CCD) contained L. anatifera Linnaeus, 1758. As these two Choros District (CCD); Septentrional District (SED), 30°S– species have similar sizes and growth rates (Thiel and 33°S; Mediterranean District (MED), 33°S–37°S; Meridional Gutow 2005b), we herein consider them as Lepas spp. and District (MD), 37°S–42°S (Fig. 1). Another criterion consid- refer to them in the following simply as Lepas. If the ered for this categorization corresponds to the genetic struc- plants contained only cyprids (recently settled larvae) of ture observed for benthic populations of D. antarctica (Fraser Lepas this was recorded in situ but no samples were taken. et al. 2010), which had suggested limited population connec- If a plant contained already metamorphosed individuals of – tivity and restriction of some haplotypes to particular biogeo- Lepas, we took samples of the 10 20 largest individuals in graphic districts, especially between ~33°S and ~37°S. order to measure their sizes, which are indicative of float- Sampling of Durvillaea antarctica. Recently stranded indi- ing time (see below). viduals of D. antarctica were collected on 33 beaches (28°S– Analysis of reproductive stage. Durvillaea antarctica is a dioe- 42°S) (Fig. 1) during winter and summer in three consecu- cious species and sex and reproductive stage can only be tive years (2013, 2014, 2015). Surveys were made on foot fol- determined by histological observations (Collantes et al. lowing the coastline, collecting all parts and entire 2002). In the laboratory, thin transverse sections were cut of individuals of recently stranded D. antarctica, along the most each tissue sample and examined microscopically to deter- recent flotsam lines (from the last 2–3 high tides). Care was mine both the sex and maturity. For this, 30 conceptacles per taken to only collect recently stranded kelp parts or plants; sample were analyzed, considering the protocols described by during the summer kelps rapidly dried out and only kelps Collantes et al. (2002). Given the purpose of the study and to with greenish or dark-brown color that maintained some flex- simplify subsequent analyses, herein we only distinguished ibility (indicative of freshness) were collected. Kelp samples two maturity stages of tissues, namely “vegetative” and “repro- were categorized into plants and fragments. Herein we use ductive” (see also Lizee-Prynne et al. 2016): the expression “plant” to refer to one or more individuals with intact fronds that are coalesced within a single holdfast, Vegetative: absence of cellular differentiation or initial dif- while fragments corresponded to parts of a frond without ferentiation of immature conceptacle between the subcor- holdfasts or holdfasts with badly damaged (i.e., non cortical tex and the medulla. In some cases, there may be newly layers) fronds. formed conceptacles but, as the gametes are undeveloped and the sex is unidentifiable, these samples were consid- On each beach, the start and endpoints of a survey were â ered as vegetative. Also, any senescent individual (with georeferenced with a portable GPS Garmin eTrex 209. For almost empty conceptacles) was equally categorized as veg- each sampling, the total distance that was surveyed on each etative. beach was determined with the online tool Google Earth, Reproductive: presence of mature conceptacles that are which allowed taking into account the beach curvature for well developed (male or female). These individuals have distance estimates. The biomass of stranded D. antarctica per identifiable sexes where gametes may already be in the beach (kg wet per km of shoreline) at a given sampling date process of being released. was calculated based on the total weight of plants and frag- ments found on a beach and the distance surveyed at this An individual was considered reproductive when at least beach. 50% of the examined conceptacles were found to be repro- Measurements and samples of complete plants. A total of 7,252 ductively mature. Then, the percentage of reproductive indi- complete plants were measured during the study. For each viduals was calculated for each beach. In D. antarctica, each complete plant of D. antarctica, the following variables were stipe within a coalesced holdfast represents a single individual measured: (Lizee-Prynne et al. 2016), and therefore the percentage was Total length: the length in centimeters was measured as calculated based on the total number of stipes analyzed for the rectilinear distance from the holdfast to the distal end each beach. Also, the number of plants that had stipes at of the longest frond. least 50 cm long was quantified, because at this size the indi- Biomass: the wet weights of the frond, stipes, and holdfast viduals start to become sexually mature (Collantes et al. of a plant were measured separately, using a portable elec- 2002). tronic hanging digital scale of 1 g accuracy. The total bio- Estimates of floating time. Detached seaweeds (and any mass of a plant was calculated by adding the frond, stipes, other objects) that are starting to float in surface waters are and holdfast weights. immediately colonized by stalked barnacles from the genus Number of stipes: the total number of stipes of each plant Lepas. The availability of competent larvae of these stalked was counted. It is noteworthy that in the case of D. antarc- barnacles may have some seasonality (i.e., higher in spring), tica, each stipe of a holdfast represents a single individual but generally they can be found throughout the year (Ander- (Gonzalez et al. 2015). son 1994). Therefore, the sizes of Lepas can be used as a

14 74 BORIS A. LOPEZ ET AL. proxy of floating time because they only adhere to buoyant To examine whether floating time affects floating plants, substrata (Helmuth et al. 1994), and size is a good estimator we compared the proportions of plants in the three temporal of the time an item has been afloat (Macaya et al. 2005, Thiel categories (short, intermediate, and long floating times) that and Gutow 2005b, Fraser et al. 2011). were reproductive and those that were nonreproductive using To evaluate the minimum floating time of stranded bull a29 3 contingency table. Significant differences were ana- kelp D. antarctica, juvenile or adult Lepas were sampled from lyzed with a chi-square test of independence (Zar 2010). A colonized plants. At the end of each survey day, we took pho- similar test was conducted to examine whether floating time tographs of the 10–20 largest specimens of Lepas collected affects plant sizes, i.e., we compared the proportions of plants from a plant. The Lepas individuals were carefully laid out ≤50 cm in the three temporal categories with the correspond- next to a scale and photographed. Subsequently, the capitu- ing proportion of plants >50 cm, i.e., sizes at which plants lar length (rectilinear distance between the distal angle of start to become sexually mature. the carina plate and the beginning of the peduncle) of each All statistical analyses were run with the statistical packages specimen was measured, using Image Pro Plus v6 (Media Primer v6 (Clarke and Warwick 2001) and GraphPad Prism Cybernetics Inc., Rockville, USA). version 6.00 for Windows (GraphPad Software Inc 2012). According to the presence and size of stalked barnacles attached, all the plants of D. antarctica were categorized in three groups: (i) short floating time (<2d)– plants without RESULTS any Lepas; (ii) intermediate floating time (2–10 d) – plants with cyprid recruits or small, juvenile Lepas (<5 mm capitular Strandings and morphometric characteristics of length); and (iii) long floating time (>10 d) – plants with plant. Strandings of D. antarctica in the study area large, adult Lepas (≥5 mm capitular length). According to occurred during all seasonal surveys and on most Thiel and Gutow (2005b), the growth rates of L. anatifera beaches. On the northernmost beaches in the CCD, and L. australis range from 0.22 to 0.46 mm d 1. Therefore, no strandings were observed during some surveys. > plants of D. antarctica with Lepas 5 mm are equivalent to The total stranded biomass varied from 0 to more than 10 d of floating times. 1,700 kg per km of shoreline, with an overall aver- Statistical analyses. For each beach, we calculated the fol- lowing dependent variables: average wet biomass per km of age of 88 kg per km shoreline and a high variability shoreline, percentage of stranded plants with Lepas, maxi- between beaches. The sizes of stranded plants ran- mum and mean length of plants, maximum and mean weight ged from 30 to 960 cm with an overall average of plants, maximum and mean number of stipes per plant length of 145 cm. The mean weight of holdfasts was and percentage of reproductive plants. In order to examine 0.24 kg, while the largest holdfast reached a maxi- whether biogeographic district, sampling year, or season had mum weight of 5.74 kg. Also, the weight of the an effect on the dynamics of stranded D. antarctica, we con- fronds fluctuated between 0.005 and 32 kg, with an ducted permutational multivariate analysis of variance (PER- MANOVA; Anderson 2001). The response variables were overall average of 0.96 kg. About half of all col- divided into two ‘families’ of related dependent variables lected plants had a holdfast with only one stipe, (Chandler 1995), those associated with (i) stranding dynam- with an average and maximum of 2 and 16 stipes ics, which included two variables: stranded biomass per km of per holdfast, respectively. shoreline and proportion of stranded plants with Lepas, and The analysis of the full model for each family of (ii) plant biology, with seven variables that are related to mor- response variables showed that in the case of the phometric and phenological characteristics of stranded stranding ‘family’, there were significant differences D. antarctica on each beach (mean and maximum length, = mean and maximum weight, mean and maximum number of between biogeographic districts (F3,174 25.05; stipes, proportion of reproductive individuals). P < 0.001). The CCD and MED were different from = < Three-way PERMANOVA tests were conducted, considering the SED and MD (t174 6.19; P 0.05), although the factors district (fixed factor: four levels), year (fixed fac- the latter two did not differ between them tor: three levels), and season (fixed factor: two levels). + (Table 1). There were neither significant differ- Dependent variables such as weight were log 1 transformed, ences between years and seasons nor interactions of while for those expressed in percent arcsine transformation was used. Data were normalized before Euclidean distances first and second order (Table 1). were calculated. Permutations (9,999) were applied to residu- With respect to the morphometric and phenologi- als under the full model for PERMANOVA. Post hoc pair-wise cal variables of individuals of D. antarctica (plant comparisons were then used to explore significant factor ‘family’), significant differences were found between = < effects using 9,999 permutations. We tested for differences in biogeographic districts (F3,174 13.56; P 0.001). multivariate dispersion between factors using the PERMDISP The SED and MD areas were different from each routine (Anderson et al. 2008). other as well as from the CCD and MED If the full model revealed significant effects of either fac- = < tor, response variables were analyzed individually with (t174 5.09; P 0.05). However, there were no dif- ANOVA using the same model as above. With a single ferences between CCD and MED (Table 1). Also, response variable on an Euclidean distance matrix, the result- significant differences between seasons were evident. ing F ratio is the same as in the traditional ANOVA (Ander- There were no significant differences between years, son et al. 2008). Post hoc pair-wise comparisons were done or interactions of factors (Table 1). with significant factor effects. The individual analysis of the variables showed To determine whether the proportion of male and female that for almost all response variables evaluated individuals differs among the four biogeographic districts and two seasons, a 2 9 2 9 4 three-level contingency table was there were significant differences between biogeo- constructed. Significant differences were analyzed by a chi- graphic districts (Table 2). Stranded biomass was square test of independence (Zar 2010). low in the CCD, high in the SED, intermediate in

15 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 75

TABLE 1. Results of three-way PERMANOVA for response variables per family of stranded Durvillaea antarctica on beaches from the continental coast of Chile (28°S–42°S).

‘Family’ Source of variation df MS Pseudo F P-value Significant pair-wise comparisons Stranding District (D) 3 36.196 25.05 <0.001 CCD 6¼ SED; CCD 6¼ MED; CCD 6¼ MD; SED 6¼ MED; MED 6¼ MD Year (Y) 2 0.971 0.67 0.599 Season (S) 1 2.199 1.52 0.225 D 9 Y 6 2.386 1.65 0.080 D 9 S 3 1.195 0.82 0.554 Y 9 S 2 2.969 2.05 0.086 D 9 Y 9 S 6 1.262 0.87 0.565 Residuals 174 1.445 Plant District (D) 3 74.387 13.56 <0.001 CCD 6¼ SED; CCD 6¼ MD; SED 6¼ MED; SED 6¼ MD; MED 6¼ MD Year (Y) 2 7.373 1.34 0.220 Season (S) 1 71.107 12.96 <0.001 W 6¼ S D 9 Y 6 7.187 1.31 0.173 D 9 S 3 4.533 0.83 0.578 Y 9 S 2 8.712 1.59 0.141 D 9 Y 9 S 6 6.456 1.18 0.259 Residuals 174 5.486 Pair-wise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Pair-wise tests show districts or seasons that differed (P < 0.05). Nonsignificant pair-wise tests are not shown. W, winter, S, summer. = the MED, and high in the MD districts (t174 4.09; SED, low in the central MED, and also intermediate P < 0.05; Fig. 2A). There was a significant interac- in the southernmost district, MD (Fig. 2B). There tion between districts and years, showing that the was a significant interaction for this variable among stranded biomasses in year 2 (2014) were similar the factors year and season, showing differences between CCD and MED, whereas this was not seen between winter and summer of year 2, while in in the other years (2013, 2015; Fig. S1 in the Sup- years 1 and 3 there were no differences between porting Information). On the other hand, the pro- seasons (Table 2). portion of plants with Lepas was highest in the In general, the length and weight of the stranded northernmost district (CCD), intermediate in the plants tended to be higher in the SED, while in the

FIG. 2. Box plot of response variables of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S–42°S), according to marine biogeographic districts during winter and summer (2013–2015). (A) Stranded biomass, and (B) percent- age of plants with Lepas. Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD, Coquimbo-Choros District; SED, Septentrional District; MED, Mediterranean District; MD, Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.59 of interquartile range; circles, outliers.

16 76 BORIS A. LOPEZ ET AL.

TABLE 2. Results of three-way ANOVAs testing the effect of district, year, and season on each response variable of stranded Durvillaea antarctica on beaches from the continental coast of Chile (28°S–42°S).

Variable Source of variation df MS Pseudo F P-value Significant pair-wise comparisons Biomass District (D) 3 65.068 39.32 <0.001 CCD < SED; CCD < MED; CCD < MD; SED > MED; MED < MD Year (Y) 2 2.017 1.22 0.300 Season (S) 1 5.814 3.51 0.060 D 9 Y 6 5.033 3.04 <0.01 CCD < MED: Y1, Y3 D 9 S 3 0.075 0.04 0.986 Y 9 S 2 0.840 0.51 0.602 D 9 Y 9 S 6 1.580 0.95 0.455 Residuals 174 1.655 Lepas (%) District (D) 3 8,370.7 14.59 <0.001 CCD < SED; CCD < MED; CCD < MD; MED < MD Year (Y) 2 155.84 0.27 0.767 Season (S) 1 36.03 0.06 0.808 D 9 Y 6 365.91 0.63 0.702 D 9 S 3 803.38 1.40 0.238 Y 9 S 2 1,829.7 3.19 <0.05 S < W: Y2 D 9 Y 9 S 6 466.69 0.81 0.559 Residuals 174 573.76 Length (mean) District (D) 3 68,880 18.61 <0.001 CCD < SED; MED < SED; MD < SED Year (Y) 2 3,614.7 0.98 0.368 Season (S) 1 877.67 0.24 0.620 D 9 Y 6 3,459.9 0.93 0.467 D 9 S 3 4,774.3 1.29 0.277 Y 9 S 2 8,844 2.39 0.092 D 9 Y 9 S 6 4,090.6 1.10 0.363 Residuals 174 3,702.2 Length (max) District (D) 3 0.00007 22.55 <0.001 CCD < SED; CCD < MD; MED < SED; MED < MD Year (Y) 2 9,391.1 0.29 0.746 Season (S) 1 21.946 0.0004 0.980 D 9 Y 6 35,054 1.098 0.365 D 9 S 3 13,719 0.43 0.735 Y 9 S 2 33,742 1.06 0.349 D 9 Y 9 S 6 20,249 0.63 0.698 Residuals 174 31,913 Weight (mean) District (D) 3 2.900 13.03 <0.001 MED < CCD; MED < SED; MD < SED Year (Y) 2 0.235 1.06 0.348 Season (S) 1 0.014 0.06 0.801 D 9 Y 6 0.271 1.22 0.306 D 9 S 3 0.092 0.41 0.749 Y 9 S 2 0.097 0.43 0.641 D 9 Y 9 S 6 0.349 1.56 0.166 Residuals 174 0.222 Weight (max) District (D) 3 15.921 19.82 <0.001 CCD < SED; PP < MD; MED < MD; MED < SED Year (Y) 2 0.320 0.39 0.677 Season (S) 1 0.336 0.42 0.513 D 9 Y 6 1.036 1.29 0.255 D 9 S 3 0.077 0.09 0.961 Y 9 S 2 0.004 0.005 0.995 D 9 Y 9 S 6 1.259 1.56 0.165 Residuals 174 0.803 Stipes (mean) District (D) 3 2.422 13.03 <0.001 MED < CCD; MED < SED; MED < MD; SED < CCD; MD < CCD Year (Y) 2 1.477 1.06 0.076 Season (S) 1 3.850 0.06 <0.01 W < S D 9 Y 6 0.556 1.22 0.429 D 9 S 3 0.069 0.41 0.945 Y 9 S 2 0.0002 0.43 0.999 D 9 Y 9 S 6 0.392 1.56 0.651 Residuals 174 0.558 Stipes (max) District (D) 3 111.46 10.62 <0.001 MED < SED; MED < MD Year (Y) 2 5.437 0.52 0.591 Season (S) 1 68.677 6.54 <0.05 W < S D 9 Y 6 21.524 2.05 0.062 D 9 S 3 15.900 1.51 0.220 Y 9 S 2 2.240 0.21 0.806 D 9 Y 9 S 6 8.854 0.84 0.533 Residuals 174 10.499

(continued)

17 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 77

TABLE 2. (continued)

Variable Source of variation df MS Pseudo F P-value Significant pair-wise comparisons Repro District (D) 3 636.04 1.74 0.197 Year (Y) 2 796.88 2.18 0.093 Season (S) 1 39,747 108.73 <0.001 S < W D 9 Y 6 635.1 1.74 0.116 D 9 S 3 948.97 2.59 0.055 Y 9 S 2 427.68 1.17 0.089 D 9 Y 9 S 6 80.418 0.22 0.562 Residuals 174 365.54 Biomass: total stranded biomass; Lepas (%): proportion of plants with Lepas; Length (mean): mean length of plants; Length (max): maximum length of plants; Weight (mean): mean wet weight of plants; Weight (max): maximum wet weight of plants; Stipes (mean): mean number of stipes of plants; Stipes (max): maximum number of stipes of plants; Repro: proportion of repro- ductive individuals. W: winter, S: summer; Y1: year 1; Y2: year 2; Y3: year 3. Pair-wise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Pair-wise tests show districts or seasons that differed (P < 0.05). Nonsignificant pair-wise tests are not shown.

MED plants were smaller and of low weight (Figs. 3, to the total were higher in plants with intermediate A and B, S2 and S3 in the Supporting Information). (92%) and long floating times (91%) than among Length and weight of stranded plants was not the plants with short floating times (85%; v2 = < affected by other factors (Table 2). The number of 2 37.36; P 0.001; Fig. 5B). stipes also varied between biogeographic districts, with more stipes per plant in the CCD, while the lowest number was observed in the central MED DISCUSSION (Figs. 3C and S4 in the Supporting Information). The current study revealed that there are coastal Likewise, stranded plants in summer had a higher areas receiving higher supplies of floating seaweeds number of stipes than in winter. There were no sig- than others, and throughout this study, the differ- nificant effects of year or interaction between fac- ences between districts remained consistent, regard- tors (Table 2). less of sampling year and season. Also, the Reproductive stage. The proportion of reproductive morphometric and phenological characteristics of individuals did not differ between biogeographic dis- stranded kelps differed between districts, indicating tricts, showing only significant differences between that the dispersal capabilities and reproductive seasons (Table 2). The percentage of reproductive potential of the individuals vary on the regional scale. individuals was higher in winter than in summer Spatial and temporal variations of strandings. There (Fig. 3D). There were differences in the percentages was a high variability in the biomass of stranded kelps of male and female individuals of D. antarctica, along the Chilean coast, being higher on beaches according to biogeographic districts and seasons between 30°S–33°S and 38°S–42°S, with high variabil- v2 = < ( 10 33.085; P 0.001). Particularly, during the ity between beaches within each district. No annual or winters, no difference was observed in the percent- seasonal effects were detected in the strandings, which ages of male and female specimens between biogeo- is consistent with observations of the temporal variabil- v2 = = graphic districts ( 7 12.010; P 0.213; Fig. 4, A, C ity of floating kelps in the study area (Hinojosa et al. and E). However, the percentage of reproductive 2011) and strandings of D. antarctica in New Zealand females tended to be lower in summer compared to (Marsden 1991). Stranded biomasses observed in this winter, and particularly, no reproductive female indi- study are within the range of those reported in previ- viduals were observed in the CCD and MED in sum- ous studies (Duarte et al. 2008, 2009) during spring mer 2015/2016 (Fig. 4F). For summers 2013/2014 and summer on sandy beaches of southern-central and 2014/2015, there was no difference in the per- Chile (i.e., Valdivia, ~39°S). The arrival of seaweeds is centage of both sexes between the biogeographic dis- an extremely dynamic process, both in space and time v2 = = tricts ( 7 7.548; P 0.374; Fig. 4, B and D). (Rodrıguez 2003, Orr et al. 2005), with frequent Floating time. The number of reproductive indi- events of resuspension and new deposition during dif- viduals differed between the three categories of ferent tidal cycles (Colombini et al. 2000, Orr et al. v2 = < floating times considered ( 2 101.90; P 0.001). 2005). Monthly samplings in subsequent studies The proportions of reproductive individuals with would be useful to detect pulses of arrival of seaweeds, intermediate (11.6%) and long floating times for example, after storm events (Hobday 2000). (32.5%) were higher than that observed in the We observed low-stranded biomasses on beaches reproductive individuals with short floating time in the CCD (28°S–30°S). The northern distribution (6.7%; Fig. 5A). Also, there were significant differ- limit of D. antarctica at the continental coast of ences in the frequencies of large individuals Chile reaches to ~30°S (Hoffmann and Santelices between the different floating times, showing that 1997, Tala et al. 2013). Given the absence of ben- the proportions of individuals >50 cm with respect thic populations in the CCD, low biomass was

18 78 BORIS A. LOPEZ ET AL.

FIG. 3. Box plot of response variables of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S–42°S), according to marine biogeographic districts during winter and summer (2013–2015). (A) Mean length of plants, (B) mean wet weight of plants, (C) mean number of stipes of plants, and (D) percentage of plants in reproductive stage. Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD, Coquimbo-Choros District; SED, Septentrional District; MED, Mediterranean District; MD, Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.59 of interquartile range; circles, outliers. expected, but the regular occurrence of stranded open ocean (Marın and Delgado 2003) may keep kelp plants on all sampled beaches within this dis- rafts from returning to the shore, thereby further trict confirms that they are frequently transported contributing to the low biomasses of stranded kelps into this area. Previous studies confirmed the pres- in the CCD. ence of floating bull kelps in the CCD (Tala et al. In the CCD, the percentage of rafts with Lepas was 2013), but many plants had been collected in off- higher compared to the other districts. Also, in this shore waters, distant from the coast. Offshore cur- district the largest sizes of Lepas were observed (see rents that transport coastally upwelled waters to the Fig. S5 in the Supporting Information). Hence, this

19 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 79

FIG. 4. Percentages of males and females of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S–42°S), according to marine biogeographic districts during three consecutive years. (A) Winter 2013, (B) summer 2013/2014, (C) winter 2014, (D) summer 2014/2015, (E) winter 2015, (F) summer 2015/2016. CCD, Coquimbo-Choros District; SED, Septentrional District; MED, Mediter- ranean District; MD, Meridional District. The numbers on top of each column correspond to frequencies of females and males, respectively. indicates that rafts of D. antarctica can float over rela- inferred from genetic and attachment substratum tively long distances (~200–300 km) from their near- analyses on the coast of southern New Zealand (Col- est benthic populations (in this case, located in the lins et al. 2010, Garden et al. 2011, Bussolini and SED district). Similar dispersal distances had been Waters 2015). Studies of floating D. antarctica and

20 80 BORIS A. LOPEZ ET AL.

Biomass of stranded kelps was also low in the MED (33°S–37°S) even though Hinojosa et al. (2011) had reported relatively high densities of floating kelps in offshore waters of this district. Sim- ilar as in the CCD, strong upwelling currents in the MED may cause the dense accumulation of kelp rafts in offshore retention zones preventing return to the shore (Sobarzo et al. 2007, Hinojosa et al. 2011). The low proportion of stranded kelps with Lepas in the MED further suggests that few kelps from the large floating pool return to the coast. Pelagic retention of floating seaweeds due to the geomorphology of the coastline and local oceano- graphic processes has also been described from other areas (Komatsu et al. 2007, 2008, 2014). Finally, the low proportion of rocky shores in the MED (see RS/SS index in Fig. 1) indicates that there is comparatively little suitable habitat for ben- thic populations of D. antarctica, further limiting local supplies of floating kelps. This suggests that the low biomass of stranded kelps in the MED is due to local oceanographic dynamics and limited supply from benthic source populations. However, other factors such as harvest of this bull kelp from natural populations, which is highest in the MED (i.e., ~80% of total landings; Subsecretarıa de Pesca y Acuicultura 2015), could also affect the amount of source material, although we observed few detached (by knife) plants during surveys on beaches within the MED. On the other hand, in the SED and MD, high biomasses of stranded bull kelps were observed. In the MD, the supply of floating kelps is high and strong westerly winds may periodically push floating kelps shoreward (Hinojosa et al. 2010), thereby con- tributing to the observed high stranding biomasses. Furthermore, moderate UV-radiation and lower sea surface temperature in this area may facilitate sur- FIG. 5. Frequency of stranded individuals of Durvillaea antarc- tica on beaches from the continental coast of Chile (28°S–42°S), vival of kelp rafts, as had previously been shown for according to three categories of floating times (short, intermedi- M. pyrifera (Roth€ausler et al. 2009), preventing disin- ate, long). (A) Number of individuals sampled in non-reproduc- tegration and sinking at sea, thereby allowing more tive and reproductive stage, (B) number of individuals sampled kelp rafts to return to the shore. For the SED, sev- ≤ > 50 cm and 50 cm frond length. eral oceanographic discontinuities have been described, caused by different wind stress, eddy Macrocystis pyrifera (Linnaeus) Agardh 1820 in the kinetic energetic, and local upwelling (Hormazabal Coastal System of Coquimbo (CSC) in Chile (i.e., et al. 2004, Yuras et al. 2005, Thiel et al. 2007, ~30°S) (Roth€ausler et al. 2011, Tala et al. 2013) had Tapia et al. 2014), which can affect transport of shown that physiological performance tends to kelp rafts. In this area, the presence of stranded decline dramatically in kelp rafts with prolonged kelps with Lepas probably indicates offshore accumu- floating times. This is accentuated in the summer, lations of floating individuals, which might be trans- due to the increase of temperature and solar radia- ported to the shore during periods of upwelling tion (Graiff et al. 2013). Furthermore, the annual relaxation. These considerations suggest that a pro- growth season of benthic kelps modulates the avail- portion of kelp rafts of D. antarctica in these dis- ability of floating plants, particularly in the spring tricts, and in particular in the MD, may be and summer months (Kingsford 1992, Hirata et al. transported away from source sites by currents and 2001, Hinojosa et al. 2010). However, in our study surface winds and later deposited on distant shores. no seasonal differences were found in the biomass of This has also been described on the southern coast stranded kelps in the CCD, suggesting low temporal of New Zealand, where short-distance movements of variability in supply from benthic sources. kelp rafts in coastal waters are strongly influenced

21 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 81 by wind and wind-induced surface current (Hawes in the CCD and MED, and their proportion was low 2008). in the SED. Unlike previous summers, the summer Morphological and reproductive features of stranded 2015/2016 was affected by the El Nino~ phe- individuals. The morphological and phenological nomenon, with higher sea surface temperatures in characteristics of D. antarctica varied between bio- northern Chile (between 1.8°C and 2.8°C higher geographic districts, which has important implica- than average), mainly near 30°S (Song et al. 2015). tions for rafting dispersal. In general, larger This suggests that female individuals of D. antarctica individuals tend to produce more gametes (Col- are more susceptible to high temperatures than lantes et al. 1997), and also larger sizes of floating male individuals, probably due to higher functional seaweeds might enhance their floating persistence costs of producing and protecting female gametes. and consequently their dispersal potential (Thiel While in species from the genus Fucus, biomass and Gutow 2005b, Graiff et al. 2013). investment in gamete production is less than 0.5% The number of stipes and individual biomass of of the total biomass (Vernet and Harper 1980), it stranded plants were lowest in the MED compared to cannot be ruled out that the increase in tempera- the other districts. A plant with more stipes enhances ture might trigger the release of female (but not of the likelihood that both sexes are present in the same male) gametes during the floating period, which coalesced holdfast. Lizee-Prynne et al. (2016) showed would explain why no female individuals were found for stranded and benthic holdfasts of D. antarctica on those beaches. As these effects could influence that between 5% and 17% of all plants contain both the dispersal potential of floating kelps, this sexes in reproductive stage, although only 50% deserves future investigation. In general, though, tended to have multiple stipes. Consequently, the and especially during the winter, rafted kelps main- probability of successful rafting dispersal is lower in tained a high reproductive potential after extensive areas with small plants, such as the MED. rafting trips, confirming that biological factors do The main reproductive season of D. antarctica is not limit the dispersal potential of D. antarctica. between autumn and spring (Collantes et al. 2002), Regional dispersal patterns and population connectiv- which was confirmed herein. Some of the individu- ity. Throughout our study most stranded individu- als that had been floating for relatively long times als (> 80%) had very short floating times, indicating (>10 d) were found to be reproductive, potentially that most kelp rafts had local sources. However, the contributing to the connectivity within and between presence of rafts with Lepas in all biogeographic dis- districts. This suggests that prolonged floating does tricts indicates that there can be rafting exchange not affect their reproductive potential, because they between these areas. Factors other than local remain viable after long rafting trips. However, Tala oceanographic conditions and reproductive features et al. (2013), at 30°S, showed for floating specimens of floating specimens have also been suggested as of D. antarctica that those with large Lepas had high an explanation for the low connectivity among pop- frequencies of conceptacles in senescent stage ulations. In the case of sessile species like seaweeds, (empty conceptacles) and increased tissue damage, density-dependent factors such as competition for which would compromise the reproductive potential space may be influencing the successful incorpora- of floating kelps, particularly in summer months. tion of immigrants to benthic populations, indepen- Similar results also have been observed during the dent of the dispersal capabilities of the species, and disintegration of sporophylls for floating M. pyrifera the number of propagules that arrive in an area (Macaya et al. 2005, Hernandez-Carmona et al. (i.e., density-blocking, see also Waters et al. 2013). 2006, Roth€ausler et al. 2011). This suggests that the Although, these effects were not measured in this release of gametes could also be occurring before study, these could be even more critical in the case the specimens arrive at the shore, and given the of dioecious species such as D. antarctica, particu- restricted movement of zygotes and the absence of larly in the colonization of new habitats, where the primary substrata, this would result in low dispersal presence of both sexes enhances the possibility of success. Viability of gametes from floating individu- successful dispersal. However, while dispersal poten- als might also be different from those in benthic tial might be lower in dioecious species, gametes populations, although release of viable zoospores from rafters could also be fertilized by gametes of has been reported for floating rafts of Hormosira the opposite sex from the benthic population, thus banksii (Turner) Decaisne 1842 (McKenzie and Bell- incorporating these immigrant genes in the resident grove 2008). Previous studies in D. antarctica had population. reported that benthic samples had a higher repro- The relatively low proportion of rafts with large ductive potential than stranded samples, although Lepas (indicative of prolonged floating, i.e., more there were no differences between males and than 10 d) suggests that few individuals migrate females (Lizee-Prynne et al. 2016). between populations from different districts (i.e., In this study, the percentages of female and male considering continuous drift in the same direction individuals did not vary between biogeographic dis- at conservative current velocities of 10 cm s 1, tricts or season, except in summer 2015/2016, ~0.36 km h 1, see Thiel 2003b), limiting the possi- where reproductive females were completely absent bility of successful dispersal (Waters et al. 2013,

22 82 BORIS A. LOPEZ ET AL.

Neiva et al. 2014). This could explain the low our study, this is likely to occur within the SED and genetic connectivity evidenced in benthic popula- MD districts compared with other areas. tions of D. antarctica along the continental coast of Chile (Fraser et al. 2010). This study was financed by CONICYT/FONDECYT 1131082 As discussed above, local oceanographic factors to M.T., F. Tellier and F. Tala and PhD-fellowship Beca CON- ICYT-PCHA/DoctoradoNacional/2014-21140010 to BL. We appear to affect dispersal of kelp rafts of D. antarc- wish to express our gratitude to Oscar Pino, Vieia Villalobos, tica on the regional scale. While other studies had Jose Pantoja, Alvaro Gallardo, Ulyces Urtubia, Felipe Saez, shown seasonal variability in the effect of local cir- Tim Kiessling, and Callum Blake for their excellent field and culation and surface winds on the abundance and laboratory assistance. We also thank them for companionship dispersal trajectories of floating seaweeds (Thiel on lengthy field trips and especially our friends in southern et al. 2011, Roth€ausler et al. 2015), herein no Chile for hosting us after exhausting days on the beach. Com- strong seasonal signals were detected. In contrast, ments from Lars Gutow and an anonymous referee helped to improve the original manuscript. The authors declare that persistent oceanographic conditions (e.g., upwel- they have no conflict of interests. ling-related offshore transport) appear to affect the geographic dispersal patterns of floating bull kelps, with only minor temporal variability. Nevertheless, Anderson, D. T. 1994. Barnacles. Structure, Function, Development and Evolution. Chapman and Hall, London, UK, 357 pp. successful dispersal, if it occurs, seems to be lim- Anderson, M. J. 2001. A new method for non-parametric multi- ited to the winter season, when most D. antarctica variate analysis of variance. Austral Ecol. 26:32–46. are reproductive. Anderson, M. J., Gorley, R. N. & Clarke, K. R. 2008. PERMANOVA+for PRIMER: Guide to Software and Statistical Methods. PRIMER-E Ltd., Plymouth, UK, 214 pp. CONCLUSIONS AND OUTLOOK Aravena, G., Broitman, B. & Stenseth, N. C. 2014. Twelve years of change in coastal upwelling along the central-northern coast The results of this study, using the strandings of of Chile: spatially heterogeneous responses to climatic vari- D. antarctica on sandy and boulder beaches as proxy ability. PLoS ONE 9:e90276. for kelp arrival on adjacent rocky shores, allow us to Bussolini, L. T. & Waters, J. M. 2015. Genetic analyses of rafted macroalgae reveal regional oceanographic connectivity pat- understand how the dynamics of floating kelps terns. J. Biogeogr. 42:1319–26. could explain the observed patterns of low genetic Camus, P. A. 2001. Biogeografıa marina de Chile continental. connectivity between benthic populations of this Rev. Chil. Hist. Nat. 74:587–617. species along the continental coast of Chile (Fraser Chandler, C. R. 1995. Practical considerations in the use of simul- taneous inference for multiple tests. Anim. Behav. 49:524–7. et al. 2010). Our results confirm that biological fac- Clarke, K. R. & Warwick, R. M. 2001. Change in Marine Communi- tors (i.e., joint dispersal of the two sexes and sea- ties: An Approach to Statistical Analysis and Interpretation. PRI- sonal phenological variability) are not as important MER-E, Plymouth, UK, 172 pp. to explain the low connectivity of coastal popula- Collantes, G., Merino, A. & Lagos, V. 2002. Fenologıa de la game- tions (Lizee-Prynne et al. 2016) and that oceano- togenesis, madurez de conceptaculos, fertilidad y embriogene- € sis en Durvillaea antarctica (Chamisso) Hariot (Phaeophyta, graphic factors (Rothausler et al. 2015) are most Durvillaeales). Rev. Biol. Mar. Oceanogr. 37:83–112. relevant to explain these patterns. However, ecologi- Collantes, G., Riveros, R. & Acevedo, M. 1997. Fenologıa repro- cal processes (i.e., high-density blocking, see Waters ductiva de Durvillaea antarctica (Phaeophyta, Durvillaeales) del intermareal de caleta Montemar, Chile central. Rev. Biol. et al. 2013, Neiva et al. 2014) could also be impor- – tant and they should be analyzed in future studies. Mar. Oceanogr. 32:111 6. Collins, C. J., Fraser, C. I., Ashcroft, A. & Waters, J. M. 2010. Additional genetic studies of stranded individuals of Asymmetric dispersal of southern bull-kelp (Durvillaea antarc- this species would allow inference of dispersal dis- tica) adults in coastal New Zealand: testing an oceanographic tances and trajectories in coastal areas (Collins et al. hypothesis. Mol. Ecol. 19:4572–80. 2010, Bussolini and Waters 2015). Colombini, I., Aloia, A., Fallaci, M., Pezzoli, G. & Chelazzi, L. 2000. Temporal and spatial use of stranded wrack by Examination of stranded specimens of D. antarctica the macrofauna of a tropical sandy beach. Mar. Biol. had shown that more than 40 species of nonbuoyant 136:531–41. seaweeds, especially Rhodophyta, are transported on Cumming, R. A., Nikula, R., Spencer, H. G. & Waters, J. M. 2014. fronds or holdfast (Macaya et al. 2016). In particular, Transoceanic genetic similarities of kelp-associated sea slug the recurrence of nonbuoyant species, Lessonia spi- populations: long-distance dispersal via rafting? J. Biogeogr. 41:2357–70. cata (Suhr) Santelices and Mazzaella laminarioides Duarte, C., Jaramillo, E. & Contreras, H. 2008. Stranded algal (Bory) Fredericq in areas further north of their main wracks on a sandy beach of south central Chile: feeding and geographic range (29°S) had suggested that dispersal habitat preferences of juveniles and adults of Orchestoidea occurs via rafting (see also Tellier et al. 2009, Monte- tuberculata (Nicolet), (Amphipoda, Talitridae). Rev. Chil. Hist. Nat. 81:69–81. cinos et al. 2012). Since there is not much informa- Duarte, C., Jaramillo, E., Contreras, H., Acuna,~ K. & Navarro, J. tion on these secondary rafters (epibiont organisms M. 2009. Importance of macroalgae subsidy on the abun- which are transported by floating seaweeds), future dance and population biology of the amphipod Orchestoidea studies should focus on assessing the effects of dis- tuberculata (Nicolet) in sandy beaches of south central Chile. Rev. Biol. Mar. Oceanogr. 44:691–702. persal by rafting on these organisms at regional Duggins, D. O., Gomez-Buckley, M. C., Buckley, R. M., Lowe, A. scales. If rafting dispersal is effective, it would be T., Galloway, A. W. E. & Dethier, M. N. 2016. Islands in the expected that species and genetic diversity is greater stream: kelp detritus as faunal magnets. Mar. Biol. 163:16– in areas with more rafting exchange. In the case of 25.

23 DYNAMICS OF FLOATING DURVILLAEA ANTARCTICA 83

Fraser, C. I., Nikula, R., Spencer, H. G. & Waters, J. M. 2009. Kelp Hormazabal, S., Shaffer, G. & Leth, O. 2004. Coastal transition genes reveal effects of subantarctic sea ice during the Last zone off Chile. J. Geophys. Res. Oceans 109:C01021. Glacial Maximum. Proc. Natl. Acad. Sci. USA 106:3249–53. Ingolfsson, A. 1995. Floating clumps of seaweed around Iceland - Fraser, C. I., Nikula, R. & Waters, J. M. 2011. Oceanic rafting by a natural microcosms and a means of dispersal for shore coastal community. Proc. R. Soc. Biol. Sci. 278:649–55. fauna. Mar. Biol. 122:13–21. Fraser, C. I., Thiel, M., Spencer, H. G. & Waters, J. M. 2010. Con- Kingsford, M. J. 1992. Drift algae and small fish in coastal temporary habitat discontinuity and historic glacial ice drive waters of northeastern New Zealand. Mar. Ecol. Prog. Ser. genetic divergence in Chilean kelp. BMC Evol. Biol. 10:203. 80:41–55. Garden, C. J., Craw, D., Waters, J. M. & Smith, A. 2011. Rafting Kirkman, H. & Kendrick, G. A. 1997. Ecological significance rocks reveal marine biological dispersal: a case study using and commercial harvesting of drifting and beach-cast clasts from beach-cast macroalgal holdfasts. Estuar. Coast. macro-algae and seagrasses in Australia: a review. J. Appl. Shelf Sci. 95:388–94. Phycol. 9:311–26. Garden, C. J., Currie, K., Fraser, C. I. & Waters, J. M. 2014. Raft- Komatsu, T., Matsunaga, D., Mikami, A., Sagawa, T., Boisnier, E., ing dispersal constrained by an oceanographic boundary. Tatsukawa, K., Aoki, M. et al. 2008. Abundance of drifting Mar. Ecol. Prog. Ser. 501:297–302. seaweeds in eastern East China Sea. J. Appl. Phycol. 20:801–9. Garden, C. J. & Smith, A. M. 2015. Voyages of seaweeds: the Komatsu, T., Mizuno, S., Natheer, A., Kantachumpoo, A., Tanaka, role of macroalgae in sediment transport. Sediment. Geol. K., Morimoto, A., Hsiao, S. T., Roth€ausler, E. A., Shishidou, 318:1–9. H., Aoki, M. & Ajisaka, T. 2014. Unusual distribution of Gonzalez, A. V., Beltran, J., Flores, V. & Santelices, B. 2015. Mor- floating seaweeds in the East China Sea in the early spring phological convergence in the inter-holdfast coalescence pro- of 2012. J. Appl. Phycol. 26:1169–79. cess among kelp and kelp-like seaweeds (Lessonia, Macrocystis, Komatsu, T., Tatsukawa, K., Filippi, J. B., Sagawa, T., Mat- Durvillaea). Phycologia 54:283–91. sunaga, D., Mikami, A., Ishida, K. et al. 2007. Distribution Graiff, A., Karsten, U., Meyer, S., Pfender, D., Tala, F. & Thiel, M. of drifting seaweeds in eastern East China Sea. J. Mar. Syst. 2013. Seasonal variation in floating persistence of detached 67:245–52. Durvillaea antarctica (Chamisso) Hariot thalli. Bot. Mar. 56:3– Lachkar, Z. & Gruber, N. 2012. A comparative study of biological 14. production in eastern boundary upwelling systems using an GraphPad Software Inc. 2012. GraphPad Prism. Version 6 for Win- artificial neural network. Biogeosciences 9:293–308. dows. La Jolla, California. Lastra, M., Rodil, I. F., Sanchez-Mata, A., Garcıa-Gallego, M. & Guillemin, M. L., Valero, M., Tellier, F., Macaya, E. C., Destombe, Mora, J. 2014. Fate and processing of macroalgal wrack subsi- C. & Faugeron, S. 2016. Phylogeography of seaweeds in the dies in beaches of Deception Island, Antarctic Peninsula. South East Pacific: complex evolutionary processes along a J. Sea Res. 88:1–10. latitudinal gradient. In Hu, Z. M. & Fraser, C. I. [Eds.] Sea- Lizee-Prynne, D., Lopez, B., Tala, F. & Thiel, M. 2016. No sex- weed Phylogeography. Springer, Dordrecht, the Netherlands, related dispersal limitation in a dioecious, oceanic long-dis- pp. 251–78. tance traveller: the bull kelp Durvillaea antarctica. Bot. Mar. Harrold, C. & Lisin, S. 1989. Radio-tracking rafts of giant-kelp - 59:39–50. local production and regional transport. J. Exp. Mar. Biol. Macaya, E. C., Boltana, S., Hinojosa, I. A., Macchiavello, J. E., Ecol. 130:237–51. Valdivia, N. A., Vasquez, N. R., Buschmann, A. H., Vasquez, J. Hawes, N. A. 2008. Nearshore dispersal and reproductive viability A., Vega, J. M. A. & Thiel, M. 2005. Presence of sporophylls in of intertidal fucoid algae: how effective is drift in local to floating kelp rafts of Macrocystis spp. (Phaeophyceae) along regional dispersal? MSc dissertation, University of Canter- the Chilean Pacific coast. J. Phycol. 41:913–22. bury, Christchurch, New Zealand, 99 pp. Macaya, E. C., Lopez, B., Tala, F., Tellier, F. & Thiel, M. 2016. Helmuth, B., Veit, R. R. & Holberton, R. 1994. Long-distance dis- Float and raft: role of buoyant seaweeds in the phylogeogra- persal of a sub-antarctic brooding bivalve (Gaimardia trape- phy and genetic structure of non-buoyant associated flora. In sina) by kelp-rafting. Mar. Biol. 120:421–6. Hu, Z. M. & Fraser, C. I. [Eds.] Seaweed Phylogeography. Hernandez-Carmona, G., Hughes, B. & Graham, M. H. 2006. Springer, Dordrecht, the Netherlands, pp. 97–130. Reproductive longevity of drifting kelp Macrocystis pyrifera Marın, V. H. & Delgado, L. E. 2003. S-chlorophyll squirts at 30°S (Phaeophyceae) in Monterey Bay, USA. J. Phycol. 42:1199– off the Chilean coast (eastern South Pacific): feature-tracking 207. analysis. J. Geophys. Res. Oceans 108:1–6. Hinojosa, I., Boltana, S., Lancellotti, D., Macaya, E., Ugalde, P., Marsden, I. D. 1991. Kelp-sandhopper interactions on a sand Valdivia, N., Vasquez, N., Newman, W. A. & Thiel, M. 2006. beach in New Zealand. I. Drift composition and distribution. Geographic distribution and description of four pelagic bar- J. Exp. Mar. Biol. Ecol. 152:61–74. nacles along the south east Pacific coast of Chile - a zoogeo- McKenzie, P. F. & Bellgrove, A. 2008. Dispersal of Hormosira bank- graphical approximation. Rev. Chil. Hist. Nat. 79:13–27. sii (Phaeophyceae) via detached fragments: reproductive via- Hinojosa, I. A., Pizarro, M., Ramos, M. & Thiel, M. 2010. Spatial bility and longevity. J. Phycol. 44:1108–15. and temporal distribution of floating kelp in the channels McLachlan, A. & Brown, A. C. 2006. The Ecology of Sandy Shores. and fjords of southern Chile. Estuar. Coast. Shelf Sci. 87:367– Academic Press, Burlington, Massachusetts, 373 pp. 77. Montecinos, A., Broitman, B. R., Faugeron, S., Haye, P. A., Tel- Hinojosa, I. A., Rivadeneira, M. M. & Thiel, M. 2011. Temporal lier, F. & Guillemin, M. L. 2012. Species replacement along and spatial distribution of floating objects in coastal waters a linear coastal habitat: phylogeography and speciation in of central-southern Chile and Patagonian fjords. Cont. Shelf the red alga Mazzaella laminarioides along the south east Paci- Res. 31:172–86. fic. BMC Evol. Biol. 12:97. Hirata, T., Tanaka, J., Iwami, T., Ohmi, T., Dazai, A., Aoki, M., Muhlin, J. F., Engel, C. R., Stessel, R., Weatherbee, R. A. & Braw- Ueda, H., Tsuchiya, Y., Sato, T. & Yokohama, Y. 2001. Eco- ley, S. H. 2008. The influence of coastal topography, circula- logical studies on the community of drifting seaweeds in the tion patterns, and rafting in structuring populations of an south-eastern coastal waters of Izu Peninsula, central Japan. intertidal alga. Mol. Ecol. 17:1198–210. I. Seasonal changes of plants in species composition, appear- Neiva, J., Assis, J., Fernandes, F., Pearson, G. A. & Serr~ao, E. A. ance, number of species and size. Phycol. Res. 49:215–29. 2014. Species distribution models and mitochondrial DNA Hobday, A. J. 2000. Abundance and dispersal of drifting kelp phylogeography suggest an extensive biogeographical shift in Macrocystis pyrifera rafts in the Southern California Bight. the high-intertidal seaweed Pelvetia canaliculata. J. Biogeogr. Mar. Ecol. Prog. Ser. 195:101–16. 41:1137–48. Hoffmann, A. & Santelices, B. 1997. Flora marina de Chile Central. Nikula, R., Fraser, C. I., Spencer, H. G. & Waters, J. M. 2010. Cir- Ediciones Universidad Catolica de Chile, Santiago de Chile, cumpolar dispersal by rafting in two subantarctic kelp-dwell- Chile, 434 pp. ing crustaceans. Mar. Ecol. Prog. Ser. 405:221–30.

24 84 BORIS A. LOPEZ ET AL.

Orr, M., Zimmer, M., Jelinski, D. E. & Mews, M. 2005. Vernet, P. & Harper, J. L. 1980. The costs of sex in seaweeds. Biol. Wrack deposition on different beach types: spatial and J. Linn. Soc. 13:129–38. temporal variation in the pattern of subsidy. Ecology 86:1496– Waters, J. M., Fraser, C. I. & Hewitt, G. M. 2013. Founder takes 507. all: density-dependent processes structure biodiversity. Trends Rodrıguez, S. R. 2003. Consumption of drift kelp by intertidal Ecol. Evol. 28:78–85. populations of the sea urchin Tetrapygus niger on the central Wichmann, C. S., Hinojosa, I. A. & Thiel, M. 2012. Floating kelps Chilean coast: possible consequences at different ecological in Patagonian Fjords: an important vehicle for rafting inver- levels. Mar. Ecol. Prog. Ser. 251:141–51. tebrates and its relevance for biogeography. Mar. Biol. Roth€ausler, E., Corell, H. & Jormalainen, V. 2015. Abundance 159:2035–49. and dispersal trajectories of floating Fucus vesiculosus in the Yuras, G., Ulloa, O. & Hormazabal, S. 2005. On the annual cycle Northern Baltic Sea. Limnol. Oceanogr. 60:2173–84. of coastal and open ocean satellite chlorophyll off Chile (18 Roth€ausler, E., Gomez, I., Hinojosa, I. A., Karsten, U., Miranda, degrees-40 degrees S). Geophys. Res. Lett. 32:L23604. L., Tala, F. & Thiel, M. 2011. Kelp rafts in the Humboldt Zar, J. H. 2010. Biostatistical Analysis. Prentice-Hall Inc, Upper Current: interplay of abiotic and biotic factors limit their Saddle River, New Jersey, 960 pp. floating persistence and dispersal potential. Limnol. Oceanogr. 56:1751–63. Roth€ausler, E., Gomez, I., Hinojosa, I. A., Karsten, U., Tala, F. & Thiel, M. 2009. Effect of temperature and grazing on growth Supporting Information and reproduction of floating Macrocystis spp. (Phaeophyceae) along a latitudinal gradient. J. Phycol. 45:547–59. Additional Supporting Information may be Roth€ausler, E., Gutow, L. & Thiel, M. 2012. Floating seaweed and their communities. In Wiencke, C. & Bischof, K. [Eds.] Sea- found in the online version of this article at the weed Biology. Springer, Berlin, Germany, pp. 109–34. publisher’s web site: Sobarzo, M., Bravo, L., Donoso, D., Garces-Vargas, J. & Schneider, W. 2007. Coastal upwelling and seasonal cycles that influence Figure S1. Box plot of stranded biomass of the water column over the continental shelf off central Durvillaea antarctica on beaches along the conti- Chile. Prog. Oceanogr. 75:363–82. nental coast of Chile (28°S–42°S), according to Song, Z. Y., Shu, Q., Bao, Y., Yin, X. Q. & Qiao, F. L. 2015. The ~ marine biogeographic districts during winter and prediction on the 2015/16 El Nino event from the perspec- summer (2013–2015). tive of FIO-ESM. Acta Oceanol. Sin. 34:67–71. Subsecretarıa de Pesca y Acuicultura. 2015. Veda biologica para Figure S2. Box plot of maximum length of el recurso cochayuyo Durvillaea antarctica en la VI Region del Libertador Bernardo O‘Higgins. Informe Tecnico (RPESQ) stranded individuals of Durvillaea antarctica on No. 151/2015. Unidad de Recursos bentonicos. Gobierno de beaches along the continental coast of Chile Chile. 35 pp. (28°S–42°S), according to marine biogeographic Tala, F., Gomez, I., Luna-Jorquera, G. & Thiel, M. 2013. Morpho- districts during winter and summer (2013–2015). logical, physiological and reproductive conditions of rafting bull kelp (Durvillaea antarctica) in northern-central Chile (30 Figure S3. Box plot of maximum wet weight of degrees S). Mar. Biol. 160:1339–51. stranded individuals of Durvillaea antarctica on Tala, F., Velasquez, M., Mansilla, A., Macaya, E. C. & Thiel, M. beaches along the continental coast of Chile 2016. Latitudinal and seasonal effects on short-term acclima- ° – ° tion of floating kelp species from the South-East Pacific. J. (28 S 42 S), according to marine biogeographic Exp. Mar. Biol. Ecol. 483:31–41. districts during winter and summer (2013–2015). Tapia, F. J., Largier, J. L., Castillo, M., Wieters, E. A. & Navarrete, S. A. 2014. Latitudinal discontinuity in thermal conditions along Figure S4. Box plot of maximum number of the nearshore of central-northern Chile. PLoS ONE 9:e110841. stipes per plant of stranded individuals of Durvil- Tellier, F., Meynard, A. P., Correa, J. A., Faugeron, S. & Valero, M. laea antarctica on beaches along the continental 2009. Phylogeographic analyses of the 30 degrees S south-east coast of Chile (28°S–42°S), according to marine Pacific biogeographic transition zone establish the occurrence of a sharp genetic discontinuity in the kelp Lessonia nigrescens: biogeographic districts during winter and summer vicariance or parapatry? Mol. Phylogenet. Evol. 53:679–93. (2013–2015). Thiel, M. 2003a. Rafting of benthic macrofauna: important fac- tors determining the temporal succession of the assemblage Figure S5. Box plot of maximum size of Lepas on detached macroalgae. Hydrobiologia 503:49–57. spp. attached in stranded individuals of Durvillaea Thiel, M. 2003b. Extended parental care in crustaceans - an antarctica on beaches to the continental coast of update. Rev. Chil. Hist. Nat. 76:205–18. Chile (28°S–42°S), according to marine biogeo- Thiel, M. & Gutow, L. 2005a. The ecology of rafting in the mar- graphic districts during winter and summer ine environment. I. The floating substrata. Oceanogr. Mar. – Biol. Ann. Rev. 42:181–263. (2013 2015). Thiel, M. & Gutow, L. 2005b. The ecology of rafting in the mar- ine environment. II. The rafting organisms and community. Table S1. Beaches sampled in the study, Oceanogr. Mar. Biol. Ann. Rev. 43:279–418. according to biogeographic districts (Coquimbo- Thiel, M., Hinojosa, I. A., Joschko, T. & Gutow, L. 2011. Spatio- Choros District and Septentrional District) of con- temporal distribution of floating objects in the German tinental coast of Chile (28°S–33°S). Bight (North Sea). J. Sea Res. 65:368–79. Thiel, M., Macaya, E. C., Acuna,~ E., Arntz, W. E., Bastias, H., Table S2. Beaches sampled in the study, Brokordt, K., Camus, P. A. et al. 2007. The Humboldt Cur- according to biogeographic districts (Mediter- rent System of northern and central Chile. Oceanogr. Mar. – ranean District and Meridional District) of conti- Biol. Ann. Rev. 45:195 344. ° – ° Vandendriessche, S., Vincx, M. & Degraer, S. 2007. Floating sea- nental coast of Chile (33 S 42 S). weed and the influences of temperature, grazing and clump size on raft longevity - A microcosm study. J. Exp. Mar. Biol. Ecol. 343:64–73.

25

Chapter 2

Spatio-temporal variability of strandings of the southern bull kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches on the coast of Chile - linking with local storms.

(manuscript submitted to J. Appl. Phycol.)

Boris A. López1,2*, Erasmo C. Macaya3,4,5, Ricardo Jeldres3, Nelson Valdivia5,6, César C. Bonta6, Fadia Tala1,7 and Martin Thiel1,4,8

Running title: Spatio-temporal variability of strandings of bull kelp Durvillaea antarctica

1Departamento de Biología Marina, Facultad de Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 2Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos, Avenida Fuchslocher 1305, Osorno, Chile. 3Departamento de Oceanografía, Universidad de Concepción, Casilla 160-C, Concepción, Chile. 4Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile. 5Centro FONDAP de Investigaciones en Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Valdivia, Chile. 6Instituto de Ciencias Marinas y Limnológicas, Facultad de Ciencias, Universidad Austral de Chile, Valdivia, Chile. 7Centro de Investigación y Desarrollo Tecnológico en Algas (CIDTA), Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 8Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Coquimbo, Chile.

*Corresponding author: [email protected]

26

Abstract

Availability of floating seaweeds may depend on the seasonal cycles of benthic populations, but our ability to predict temporal patterns of stranded biomasses is still limited. Season-dependent, local storms favor detachment of seaweeds from the substratum, which can be reflected in the stranded biomasses on adjacent beaches after these events. Hence, we hypothesized that seaweed strandings are positively correlated with storm intensity. Using as a model the southern bull kelp Durvillaea antarctica (Chamisso) Hariot, a species frequently found in seaweed strandings in Chile, bimonthly surveys were carried out on three beaches: Pichicuy (32°S), Itata Norte (36°S), and Curiñanco (39°S) for three years (2014 to 2017). Stranded biomass, total length, and wet weight of specimens were quantified and related to local storms (using the Douglas sea scale). Stranded biomasses decreased in the spring months of each year, being higher in Pichicuy and Curiñanco than Itata Norte. Regression models showed a better fit with recent storms in Curiñanco compared to other beaches. An interannual decrease of beach-cast raft size was observed, showing smaller specimens in Itata Norte than Pichicuy and Curiñanco. Reduced habitat availability and the exploitation of natural beds in the central zone (34°-37°S) might explain the decrease of biomasses and sizes of stranded bull kelps. Also, oceanographic features at intermediate (i.e. local winds and currents) and large scales (i.e. El Niño) can help to explain the temporal variability, particularly in spring and summer. Our results suggest that harvesting of stranded bull kelps might be most favorable in summer and autumn.

Keywords: bull kelps; Douglas sea scale; floating seaweeds; sandy beaches; strandings.

2.1. Introduction

Detached buoyant seaweeds are one of the most common floating natural substrata in marine environments, facilitating the dispersal and population connectivity of many associated species (Thiel and Gutow 2005a, 2005b; Thiel and Fraser 2016). The spatio- temporal variability in the abundance of floating seaweeds, and the effect of abiotic and 27

biotic factors explaining these fluctuations, has been focus of several studies in recent years (see review by Rothäusler et al. 2012). Particularly, inter- and intra-annual studies have suggested a clear seasonal tendency in the availability of floating seaweeds, produced mainly by the annual growth season of benthic populations (e.g. Hobday 2000; Hirata et al. 2001; Hinojosa et al. 2010). For example, Kingsford (1992) showed that the abundance of floating Sargassum sinclairii on the northeastern coast of New Zealand was higher from November to January (i.e. austral spring-early summer) than during other months, suggesting that this pattern is related to the benthic demographic changes of this species.

After the detachment from the primary substratum and subsequent pelagic state, a fraction of floating seaweed populations returns to the coast through wind and local currents (Garden et al. 2014; Hawes et al. 2017), resulting in local strandings with high spatio-temporal variability (Marsden 1991; McLachlan and Brown 2006). These accumulations of stranded seaweeds constitute trophic subsidies for the macrofauna (e.g. amphipods and isopods) and shorebirds that inhabit boulder and sandy beaches (Dugan et al. 2003; Orr et al. 2005; MacMillan and Quijón 2012; Lastra et al. 2014); stranded seaweeds can also be used by fishermen and shore gatherers in the case of economically important seaweeds (Kirkman and Kendrick 1997; Piriz et al. 2003).

Studies on the stranding dynamics of floating seaweeds in coastal areas have focused on factors that operate mainly during the pelagic stage and that affect processes such as dispersal trajectories (e.g. winds and local currents, Garden et al. 2014; Rothäusler et al. 2015) and survival of floating individuals at the sea surface (e.g. temperature and solar radiation, Graiff et al. 2013; Tala et al. 2013; 2016; 2017; grazing, Rothäusler et al. 2009; 2018; epibiosis, Graiff et al. 2016). However, for a better understanding it is necessary to document the temporal variability of pulses of beach-cast rafts and examine the factors that cause seasonal changes in the availability of floating individuals.

Most buoyant seaweeds grow in benthic habitats (Hurd et al. 2014). Entire individuals or parts thereof can be detached from the substratum and enter the floating population due to different mechanisms, such as stipe breakage, detachment of the holdfast, and lifting of the attachment substratum (Thiel and Gutow 2005a; Garden et al. 2014). Stipe breakage can occur throughout the life time of seaweeds, although it is more common in

28

senescent individuals and during the spring months (Arenas et al. 1995; Hinojosa et al. 2010). In addition, grazing activities contribute to weakening and later breaking of the stipe (Pfister and Betcher 2018), mainly in spring and summer months (Santelices et al. 1980; Duggins et al. 2001). Likewise, grazers can weaken the holdfast and facilitate its detachment from the substratum, especially those organisms that excavate burrows inside the holdfast (Thiel 2003). The occurrence of local storms is another factor that has been associated with the availability of floating seaweeds, but that has received less attention (Duggins et al. 2001). These events can produce mechanical failure of the holdfast and or attachment substratum, which tends to increase depending on storm intensity. For example, ZoBell (1971) observed higher detachment rates of giant kelp Macrocystis pyrifera during winter storm season in California. This suggests a strong seasonality of storm effects on the benthic and floating parts of seaweed populations.

Storm-driven events can also influence long-distance dispersal of seaweed rafts (Gillespie et al. 2012), which could generate higher accumulations of stranded seaweeds in some coastal areas than in others, albeit depending on storm direction and intensity there might be high spatio-temporal variability. For example, Waters et al. (2018) indicated that severe storms disrupt oceanographic barriers (e.g. fronts) by transporting floating bull kelp Durvillaea antarctica over longer distances than under less intense storms. Similar results of storm-driven dispersal of floating bull kelps have been recently described by Craw and Waters (2018), and Waters and Craw (2018). Therefore, the relationship between local storm events may contribute importantly to the stranding dynamics of floating seaweeds.

Herein we used as model the bull kelp Durvillaea antarctica (Chamisso) Hariot 1892, a species with positive buoyancy and a wide geographical distribution in the southern hemisphere, particularly in subantarctic regions (Fraser et al. 2009; 2010; Batista et al. 2018). In Chile, it is distributed from 30°S to 56°S (Hoffmann and Santelices 1997), but stranded specimens can also be found on beaches beyond its northern distribution limit (López et al. 2017). Durvillaea antarctica is frequently found floating in coastal areas (Hinojosa et al. 2010; 2011; Wichmann et al. 2012), and in seaweed strandings on beaches along the continental coast of Chile (Duarte et al. 2008; 2009), with strong variations in stranded biomasses at a regional scale, being higher at 30°S-33°S (i.e. northern zone) and 37°S-42°S (i.e. southern zone) compared to 33°S-37°S (i.e. central zone) (López et al.

29

2017). Short estimated floating times (i.e. < 2 days) of stranded individuals (based on small sizes and low frequency of stalked barnacles attached) indicated that most beach-cast rafts are subsidized by adjacent benthic populations, especially in the central zone of Chile (López et al. 2017). This is congruent with previous phylogeographic studies that indicated a low genetic connectivity of bull kelp populations in that area (Fraser et al. 2010).

In Chile, benthic populations of D. antarctica strongly vary between years with a marked seasonal pattern, and higher abundances during austral summer (Castilla and Bustamante 1989; Westermeier et al. 1994; Castilla et al. 2007). Also, D. antarctica is a commercially exploited seaweed, being extracted from its natural beds mainly in central Chile (34°S-36°S) during the spring and summer months (Durán et al. 1987; Gelcich et al. 2006). The temporal variability in benthic supply and of rafting-generating events (i.e. storms) should thus mirror the temporal dynamics of stranded kelp biomass on local beaches. Based on these considerations, we hypothesized that (i) stranded kelp biomass increases with local storm intensity, (ii) stranded kelp biomass increases in winter months, and (iii) stranded kelp biomasses are higher on northern and southern beaches than on beaches from the central zone. In order to test these hypotheses, herein we (i) characterized the morphometric features (lengths, wet weights) and floating times (using epibionts Lepas spp.) of beach-cast bull kelps, and (ii) examined the relationship between the stranding dynamics of D. antarctica with local storms.

2.2. Materials and Methods

2.2.1 Study area

The southern-central coast of Chile (from ca. 32°S-39°S) is characterized by a linear topography and a latitudinal temperature gradient in surface waters (Thiel et al. 2007). The Humboldt Current System (cold and nutrient-rich waters) influences the coastal area, with extensive upwelling zones affecting coastal communities (Hormazábal et al. 2004; Aravena et al. 2014). This area belongs to the Intermediate Area (30°S-42°S), a biogeographic region of marine biota that has mixed components between the northern Peruvian Province (18°S to 30°S) and the southern Magellan Province (42°S to 56°S; Camus 2001). In this

30

Intermediate Area, Camus (2001) distinguished several sub-regions (districts) based mainly on local oceanographic features (e.g. differences in freshwater inputs from rivers).

The present study was conducted on three sandy beaches across the benthic geographical range of D. antarctica on the continental coast of Chile: Pichicuy (32°21’S; 71°27’W), Itata Norte (36°18’S; 72°49’W), and Curiñanco (39°44’S; 73°23’W; Fig. 2.1). The distance between beaches varied between 300-500 km and the extension of the stretches that were surveyed on each beach ranged from 2.1 to 2.7 km, depending on beach length and/or amounts of stranded kelps. On each beach, the start and end points of each survey were georeferenced with a portable GPS Garmin eTrex® 20x (each beach was delimited by at least one rocky shore and it also contained small rocky outcrops).

Beaches were distributed across the three biogeographic districts defined by Camus (2001): Septentrional District (SED), 30°S to 33°S; Mediterranean District (MED), 33°S to 37°S; Meridional District (MD), 37°S to 42°S (Fig. 2.1). In this area, López et al. (2017) showed strong variations in bull kelp strandings, being the mean strandings higher on beaches from SED and MD, intermediate in MED, and lowest in the northern district (i.e. 28°S-30°S, Coquimbo-Los Choros District, CCD). That study also indicated that there were only minor seasonal differences in kelp stranding dynamics across these districts and the northern-most district (CCD) was exposed to El Niño effects.

3.2.2. Sampling of Durvillaea antarctica

Recently stranded individuals of D. antarctica were collected on each beach during bimonthly surveys over three consecutive years, starting in July 2014 and ending in May 2017. Surveys were made on foot following the coastline, collecting all parts and entire individuals of D. antarctica found along the most recent flotsam lines (from the last 2 - 3 high tides). Care was taken to only collect recently stranded kelp parts or plants; during the summer months, kelps rapidly dried out and only kelps with greenish or dark-brown color that maintained some flexibility (indicative of freshness) were collected. Kelp samples were categorized into plants and fragments. Herein we use the expression “plant” to refer to one or more individuals with intact fronds that are coalesced within a single holdfast, and

31

“fragments” to refer parts of a frond without holdfasts or holdfasts with badly damaged (i.e. non-cortical layers) fronds.

Fig. 2.1. Geographical distribution of surveyed beaches and biogeographic districts described for the coast of Chile distinguished in this study (Coquimbo-Choros District: 28°S-30°S, Septentrional District: 30°S-33°S, Mediterranean District: 33°S-37°S, Meridional District: 37°S-42°S). The geographic distribution of Durvillaea antarctica within the study area is also indicated.

32

For each sampling, the total distance that was surveyed on each beach was determined with the online tool Google Earth, which allowed taking into account the beach curvature for distance estimates. Stranded biomass of D. antarctica per beach (kg wet weight per km of shoreline) at a given sampling date was calculated based on the total weight of plants and fragments found on a beach and the distance surveyed at this beach. Also, the number of rafts (i.e. entire plants) per km of shoreline and the percentages of stranded biomass, according to complete plants and fragments, were calculated.

2.2.3. Measurements and samples of plants

A total of 2,105 complete specimens were measured during the study. For each complete plant of D. antarctica, the following variables were measured:

(i) Total length: the length in centimeters was measured as the rectilinear distance from the holdfast to the distal end of the longest frond.

(ii) Biomass: the wet weights of the frond, stipes and holdfast of a plant were measured separately, using a portable digital hanging scale of 1g accuracy. The total biomass of a plant was calculated by pooling the frond, stipes, and holdfast weights.

(iii) Floating time: for each plant we determined whether it had been colonized by Lepas spp. or not (mainly L. australis and L. pectinata; Hinojosa et al. 2006). Detached seaweeds that are starting to float in surface waters are immediately colonized by stalked barnacles from the genus Lepas. Cyprid larvae of these stalked barnacles may have some seasonality (i.e. higher in spring), but generally they can be found throughout the year (Anderson 1994). Therefore, the sizes of Lepas can be used as a proxy of floating time because they only adhere to buoyant substrata (Helmuth et al. 1994), and size is a good estimator of the time an item has been afloat (Thiel and Gutow 2005b; Fraser et al. 2011). If the plants contained only cyprids (recently settled larvae) of Lepas, this was recorded in situ, but no samples were taken. If a plant contained already metamorphosed individuals of Lepas, we took samples and photographs of the 10-20 largest individuals in order to measure their sizes (i.e. capitular length, the rectilinear distance between the distal angle of the carina plate and the beginning of the peduncle), which are indicative of floating time. Each Lepas individual was measured using the scaled photographs with the software Image Pro Plus v6 33

(Media Cybernetics Inc., Maryland, USA). For each survey, the percentage of individuals of D. antarctica with Lepas spp. was calculated, as well as the size of attached specimens. According to Thiel and Gutow (2005b), the growth rates of L. anatifera and L. australis range from 0.22 to 0.46 mm ꞏ day-1 at 14ºC. Therefore, individuals of D. antarctica with Lepas > 5 mm are equivalent to more than 10 days of floating times.

2.2.4. Estimate of local storm intensity

One of the main physical factors that cause extensive detachment of floating seaweeds is the local sea condition (e.g. seasonal storms; Thiel and Gutow 2005a). Herein we examined whether local weather conditions affect the amounts of stranded kelps on each beach. To do this, we used data recorded by the Meteorological Service of the Chilean Navy (http://meteoarmada.directemar.cl/site/estadopuertos/estadopuertos.html) of daily weather conditions of the main port closest to the surveyed beaches: (i) Los Vilos (31°54’S; 71°30’W) for Pichicuy, (ii) Talcahuano (36°43’S; 73°06’W) for Itata Norte, and (iii) Corral (39°53’S; 73°25’W) for Curiñanco. In each case, the local sea state was summarized using the Douglas sea scale. This descriptive scale uses a discrete range of variation considering the height of the waves and the swell (i.e. wind-driven waves), expressed in 10 degrees, from 0 (i.e. calm, no swell) to 9 (i.e. phenomenal, very high swell; Owens 1982).

Since the local accumulation of stranded rafts may be consequence of sea conditions days or weeks earlier, the Douglas sea scale values for each main port before the day of sampling date were recorded, using daily values for three different lag periods: (i) the last week, (ii) the last 15 days, and (iii) the last month before each beach survey. For each lag period (i.e. sampling date and surveyed beach), the maximum and mean values of Douglas sea scale were calculated.

2.2.5. Statistical analyses

For each surveyed beach and sampling date, we calculated the following dependent variables: total stranded wet biomass per km of shoreline, number of stranded rafts per km shoreline, percentage of stranded biomass of complete plants and fragments, percentage of

34

complete plants with Lepas, size (i.e. capitular length) of Lepas spp., mean length and mean wet weight of stranded complete plants of D. antarctica. In the case of percentage of stranded individuals with Lepas and size of attached specimens of Lepas spp., only descriptive analyses (mean, standard deviations, and ranges) were done due to the low frequency of D. antarctica with Lepas throughout the study. Also, the percentages of stranded biomass (distinguishing complete plants and fragments) were only analyzed descriptively.

To investigate the relationship between the stranded biomass of D. antarctica and local storms, simple linear regressions were carried out for each beach separately, using maximum and mean lag values (monthly, biweekly, and weekly) of Douglas sea scale for each sampling date. In each case, the mean and maximum values were calculated after pooling the monthly, biweekly, and weekly values and a positive lag was used: x = Douglas (time), y = stranded biomass (time + 1).

In order to examine whether the above-mentioned response variables (stranded biomass, number of stranded rafts, mean length, and mean wet weight of complete stranded plants) varied according to beach, sampling year, and sampling month, we conducted generalized linear models (GLM) with a Gaussian distribution of errors (Zuur et al. 2009). A Gaussian fit was chosen, because preliminary checks were made to normal, log-normal and Gamma distributions, as well as residual analyses of the models showed a better fit of the distribution of errors using a normal distribution. Three-way GLM tests were run separately to compare total stranded biomass and number of rafts (i.e. plants), considering beach (fixed factor: three levels), year (fixed factor: three levels), and sampling month (fixed factor: six levels) as factors, and the maximum weekly value of Douglas sea scale as covariate; stranded biomass and number of rafts were logarithm x+1 transformed for the analyses. Maximum weekly values of Douglas sea scale were chosen because it presented the highest values of association with the stranded biomass (see Results). We used a variant of three factor analysis without replication that does not have interaction of the three factors, but there are pairwise interactions (Underwood 1997).

Full three-way GLMs were done in the cases of total length and wet weight of stranded D. antarctica, applying the same procedure as described above. For effects of

35

these statistical analyses, only stranded specimens >50 cm long were considered, because above this size, bull kelp starts to become sexually mature (Collantes et al. 2002), and potentially contribute to population connectivity. The goodness of fit of these fixed factor models was estimated through Pseudo-R2 for GLM according to Nakagawa and Schielzeth (2013). Lastly, if there were significant effects (P < 0.05), the differences between categories of factors were examined a posteriori with Tukey HSD tests (Zar 2010). All statistical analyses were done using “fitdistrplus“, “lme4”, and “multcomp” packages in R 3.4 (R Development Core Team 2018).

2.3. Results

2.3.1. Spatio-temporal variability of stranded biomasses and rafts on beaches

There was high inter and intra-annual variability in stranded biomass of bull kelps on each beach. Significant differences in the stranded biomasses were found between surveyed beaches, years and sampling months, as well as interactions among factors (Pseudo-R2 = 0.646; P < 0.001; Table 2.1). In general, higher stranded biomasses were observed on the beaches of Pichicuy and Curiñanco compared to Itata Norte (P < 0.05) (Fig. 2.2). Also, stranded biomass tended to decrease in the third year of study (i.e. 2016/17), particularly in Pichicuy. There was a clear seasonal pattern, with stranded biomass decreasing significantly during austral spring (i.e. September-December) and increasing in summer and autumn (January-May; P < 0.05). This seasonal pattern was much more marked on the beaches of Pichicuy and Curiñanco than in Itata Norte, particularly in the years 2015-2016 (Figs. 2.2, and S2.1, Table S2.1).

The number of rafts per km shoreline also varied significantly among surveyed beaches, years and sampling months, as well as interactions among factors were observed (Pseudo-R2 = 0.575; P < 0.001; Table 2.1). Higher numbers of beach-cast rafts of D. antarctica were found on the beaches of Pichicuy and Curiñanco compared to Itata Norte (P < 0.05), although the temporal pattern varied depending on each beach. In Pichicuy, the number of stranded rafts decreased in the third year (i.e. 2016/17), with low values during austral spring (i.e. September-November). In Itata Norte, the number of rafts was higher during the second year, with no variations among sampling months, but in Curiñanco the

36

number of stranded rafts was significantly lower during austral spring of each year (September-December; P < 0.05; Figs. 2.2, and S2.1, Table S2.1).

Fig. 2.2. Stranded biomass and number of rafts of Durvillaea antarctica on three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A- B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Details of post-hoc analyses are shown in Table S2.1.

2.3.2. Relationship between stranded biomasses and local storms

The positive relationship between stranded biomass of D. antarctica and local storm intensity varied according to the lag used and the surveyed beach. Stronger associations

37

between both variables were observed using the weekly lag, intermediate with the biweekly lag, and lower with the monthly lag (Figs. 2.3, S2.2 and S2.3). At the northern-most site (Pichicuy), a significant relationship was found with the maximum values of storm intensities, while there was no significant association with the mean values of intensity (Figs. 2.3A and B). Regression models for both variables were not significant at the central site (Itata Norte) (Figs. 2.3C and D), while at the southern site (Curiñanco) a significant relationship and better fit between the maximum and mean values of Douglas sea scale and stranded biomass was observed compared to the other beaches (Figs. 2.3E and F).

Table 2.1. Summary of GLM for stranded biomass and number of rafts of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Significant values (P < 0.05) are shown in bold.

Stranded Biomass Number of rafts Source of df F P-value df F P-value variation Beach, (B) 2;54 9.16 < 0.001 2;54 9.09 < 0.001 Year, (Y) 2;54 5.29 < 0.050 2;54 3.75 < 0.050 Month, (M) 5;54 6.61 < 0.001 5;54 5.79 < 0.001 Douglas, (D) 1;54 4.78 < 0.050 1;54 4.65 < 0.050 B x Y 4;54 4.97 < 0.001 4;54 4.90 < 0.010 B x M 10;54 5.68 < 0.010 10;54 2.53 < 0.010 Y x M 10;54 4.18 < 0.010 10;54 3.89 < 0.010 B x D 2;54 3.76 < 0.050 2;54 4.12 < 0.050 Y x D 2;54 2.09 0.133 2;54 1.61 0.209 M x D 10;54 1.15 0.337 10;54 1.31 0.228 B x Y x D 4;54 2.44 0.065 4;54 2.11 0.092 B x M x D 10;54 1.37 0.196 10;54 0.99 0.478 Y x M x D 10;54 1.12 0.203 10;54 0.84 0.512

2.3.3. Morphometric characteristics of stranded individuals

During all surveys complete plants were found, although fragments of bull kelps dominated on all three beaches (see Fig. S2.4). The differences in the average length of stranded specimens of D. antarctica between beaches significantly depended on both, month and year of sampling, evidenced by statistically significant interactions among these factors

38

(Pseudo-R2 = 0.483; P < 0.001; Table 2.2). In Pichicuy, the average length of kelps decreased during the third sampling year (i.e. 2016/17), without significant variations between sampling months (Fig. 2.4A). In Itata Norte, there was a high inter and intra- annual variability in the lengths of stranded bull kelps, showing the highest values in July 2017, while there were no strong variations in the remaining sampling months (Fig. 2.4B). On the other hand, in Curiñanco the average length of stranded bull kelps tended to increase during the sampling years, particularly in austral spring (i.e. September) of the last year (Figs. 2.4C, and S2.1, Table S2.2).

Fig. 2.3. Relationships between stranded biomass of Durvillaea antarctica (kg per km of shoreline) and Douglas sea scale (mean and maximum values of weekly lag according to the date of each survey) on three beaches from the continental coast of Chile. A-B:

39

Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated.

Table 2.2. Summary of GLM for length and wet weight of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Significant values (P < 0.05) are shown in bold.

Length Wet Weight Source of df F P-value df F P-value variation Beach, (B) 2;1850 2.23 0.107 2;1850 4.89 < 0.010 Year, (Y) 2; 1850 2.61 0.073 2; 1850 5.29 < 0.010 Month, (M) 5; 1850 1.03 0.420 5; 1850 1.21 0.256 Douglas, (D) 1; 1850 1.55 0.213 1; 1850 1.78 0.182 B x Y 4; 1850 5.37 < 0.001 4; 1850 6.97 < 0.001 B x M 10;1850 4.88 < 0.001 10;1850 5.68 < 0.001 Y x M 10;1850 5.11 < 0.001 10;1850 6.19 < 0.001 B x Y x M 20;1850 2.14 0.096 20;1850 1.13 0.563 B x D 2;1850 1.68 0.186 2;1850 1.76 0.172 Y x D 2;1850 1.14 0.320 2;1850 0.71 0.491 M x D 10;1850 0.90 0.563 10;1850 1.35 0.163 B x Y x D 4;1850 1.38 0.238 4;1850 0.98 0.417 B x M x D 10;1850 0.86 0.609 10;1850 1.17 0.288 Y x M x D 10;1850 0.98 0.581 10;1850 1.68 0.195 B x Y x M x D 20;1850 1.18 0.318 20;1850 0.73 0.374

Average wet weights of stranded bull kelps varied significantly between the surveyed beaches, depending on the month and year of sampling, according to the significant interactions among these factors (Pseudo-R2 = 0.492; P < 0.001; Table 2.2). Plants were larger on the beaches of Pichicuy and Curiñanco compared to Itata Norte (P < 0.05) (Fig. 2.5). In Pichicuy, wet weights decreased in the third year of the study (i.e. 2016/17), with lower values during austral autumn and winter months (i.e. May-July) (Fig. 2.5A). In Itata Norte, wet weights were higher during austral winter and spring (July- September) with strong variability among sampling years (Fig. 2.5B), while in Curiñanco, wet weights increased between sampling years, with higher values in austral spring and autumn (i.e. September and March; Figs. 2.5C, and S2.1, Table S2.2).

40

Fig. 2.4. Average (mean ± SD) length of stranded individuals Durvillaea antarctica on three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A: Pichicuy (32°S), B: Itata Norte (36°S) and C: Curiñanco (39°S). Only stranded specimens > 50 cm long were considered (n = 1,904). Details of post-hoc analyses are shown in Table S2.2.

41

Fig. 2.5. Average (mean ± SD) wet weight of stranded individuals Durvillaea antarctica on three beaches from the continental coast of Chile, according to bimonthly surveys from 2014 to 2017. A: Pichicuy (32°S), B: Itata Norte (36°S) and C: Curiñanco (39°S). Only stranded specimens > 50 cm long were considered (n = 1,904). Details of post-hoc analyses are shown in Table S2.2.

42

2.3.4. Floating time of stranded individuals

The presence of beach-cast rafts of D. antarctica with specimens of Lepas spp. attached varied among beaches. The highest percentages of individuals with Lepas (i.e. 50% - 75%) were observed in Pichicuy, particularly in austral winter (July-September 2016/17) (Fig. 2.6A). On this beach, the mean sizes of Lepas spp. fluctuated between 1 and 7.5 mm, although specimens with more than 10 mm were occasionally observed (Fig. 2.6B). On the contrary, in Itata Norte there was a very low presence of stranded bull kelps with Lepas (i.e. < 10%), mostly of small sizes (i.e. < 4 mm), throughout the study (Fig. 2.6C-D). In Curiñanco the proportion of stranded specimens with Lepas spp. strongly varied among sampling months, with higher values in March and September (i.e. 30% - 40%) than in the other periods (< 20%) (Fig. 2.6E). The mean sizes of Lepas on this beach varied between 1 and 5.5 mm, not exceeding 10 mm (Fig. 2.6F).

2.4. Discussion

There was a high temporal variability (inter- and intra-annual) in the supplies of stranded southern bull kelps on the sampled beaches, as well as in the morphometric characteristics of the specimens. The relationship between stranded biomass of D. antarctica with local storms was stronger on the southern beach compared to the other two beaches. Also, beach- cast rafts in Pichicuy (northern zone) and Curiñanco (southern zone) showed evidence of prolonged floating times, suggesting that floating kelps may be supplied from other areas contributing to the relatively high stranding biomasses on these beaches. In contrast, bull kelp strandings in Itata Norte (central zone) depend mostly on adjacent benthic populations, probably because the Concepción upwelling system (Sobarzo et al. 2007; Hinojosa et al. 2011) and the Itata river plume block the return of floating kelps to the coast (Vargas et al. 2006). These results are congruent with previous studies about bull kelp strandings (López et al. 2017), and phylogeographic patterns of this species (i.e. low genetic connectivity among populations; Fraser et al. 2010) described for the study area.

43

Fig. 2.6. Percentage of stranded individuals of Durvillaea antarctica with Lepas spp. attached, and sizes (mm of capitular length, mean ± SD) of specimens of Lepas spp. on three beaches from the continental coast of Chile. A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S), according to bimonthly surveys from 2014 to 2017.

2.4.1. Temporal variability of bull kelp strandings - influence of local storms

Stranded biomasses of D. antarctica showed a clear seasonal pattern (i.e. high values in austral summer and autumn months, and low in spring months). This trend is in agreement with the temporal variability observed in benthic abundances of D. antarctica along the Chilean coast (Santelices et al. 1980; Castilla and Bustamante 1989; Westermeier et al.

44

1994), with high densities/biomasses during summer and autumn months. Also, multi-year surveys of floating kelps along the continental coast of Chile showed that rafts of D. antarctica are less abundant during spring (Hinojosa et al. 2011), which coincides with the low strandings in this season.

Recent local storms (i.e. weekly) were generally better predictors of the amounts of stranded bull kelps, although our results suggest that the local storm intensity does not increase in the winter months, being frequent also in summer and autumn months. Breakage of bedrock from the low intertidal zone (where D. antarctica lives) due to storms and subsequent detachment of bull kelps have been described on the coasts of Macquarie Island and New Zealand (Smith and Bayliss-Smith 1998; Garden et al. 2011). Local storm effects on detachment of seaweeds can depend on how the latter are attached to the substratum and the rock strength (Garden and Smith 2011, 2015). For example, strong storms are necessary to generate a bedrock fracture on primary substratum of high hardness, while individuals with coalescent holdfasts that overgrow other organisms (e.g. acorn barnacles) have less adhesion, and therefore storms of less intensity can dislodge them from the rocks (Garden and Smith 2015). In the case of D. antarctica, it has been observed that most of beach-cast rafts have a variety of epibionts attached at the base of the holdfast (e.g. barnacles, mussels, seaweeds; López et al. 2018), which might have reduced the attachment strength of the holdfast to the rock substratum. This suggests that intermediate-intensity storms could be contributing to the availability of floating bull kelps (at least complete plants) due to failure of the underlying epibionts.

Also, the high presence of fragments in strandings throughout the study could be caused by other factors besides storms. For example, several studies indicated as dispersal strategy in Sargassum spp. that benthic individuals maintain the attachment strength through their reproductive phase and decline thereafter, favoring the breakage of stipes and thus the significant increment of floating rafts in summer months (Norton 1976; 1977; Deysher and Norton 1982; Ohno 1984; Hirata et al. 2003; Xu et al. 2016). Although this mechanism has not been described in D. antarctica, there may be some susceptibility of senescent individuals after the maturity season (i.e. austral winter, Collantes et al. 1997; 2002), generating weakening of the stipe and subsequent breakage during summer. López et al. (2017), evidenced high presence of vegetative specimens (included senescent) in D.

45

antarctica strandings during summer on beaches in this study area. Also, grazers tend to increase their activity in the summer season, opening cavities in the holdfast (e.g. excavating snails and burrowing isopods, Santelices et al. 1980; Miranda and Thiel 2008) or weakening the stipe (e.g. snails, sea urchins, Santelices et al. 1980), which can also contribute to the high amounts of stranded rafts during that period. For example, López et al. (2018), found the excavating snail Scurria scurra to be common on holdfasts (and on stipes, pers. observ.) of stranded bull kelps on beaches of this same study area.

Our results indicate that the seasonality of D. antarctica strandings is determined by the storm intensity, as well as the growth cycle of its benthic populations, and other biotic factors such as grazing that favor the detachment of individuals (or parts of these) from the substratum. Therefore, our results support the hypothesis that the stranded biomasses increase with the storm intensity, nevertheless, there is no support to the hypothesis that beach-cast rafts increase in the winter months.

Stranded biomasses and raft sizes of D. antarctica decreased in the third year of study (i.e. 2016/2017), mainly on the northern beach (i.e. Pichicuy). This is consistent with the change in environmental conditions (e.g. increase of sea surface temperature) in this area of the Humboldt Current System during 2015-2016 (including early 2017) caused by El Niño (EN) (Song et al. 2015; Adams and Flores 2016). During these periods, extinctions of kelp beds can be observed over large spatial extensions (Castilla and Camus 1992; Camus et al. 1994; Vega et al. 2005). For example, Ladah et al. (1999) evidenced a massive disappearance of Macrocystis pyrifera in Baja California during the EN event of 1997- 1998. Also, López et al. (2017) observed that during summer 2015/2016 there were no stranded female specimens of D. antarctica on beaches of the northern coast of Chile (28°S-30°S). Likewise, small sizes of stranded rafts during 2016/2017 seem to be a supporting indicator that the previous EN event might have affected benthic kelp populations, because they were likely dominated by small individuals.

Variations in the productivity of benthic seaweed populations can also generate interannual changes in the availability of floating kelps. Castilla et al. (2007), using a long- term study on the central zone of the Chilean coast, observed strong changes of benthic abundances of D. antarctica over twenty years. Other studies on New Zealand and Chilean

46

coasts indicated similar variations of benthic abundances of bull kelps among years (Santelices et al. 1980; Westermeier et al. 1994; Taylor and Schiel 2005), suggesting that high local variability of seasonal recruitment and/or differential growth and mortality can affect the availability of floating rafts, and subsequent stranding dynamics among years. Based on this, the high interannual variability observed in the stranded biomasses of D. antarctica suggests a high dynamism in the annual growth cycles of their benthic populations that subsidize the availability of floating individuals in adjacent areas. Future studies that examine the relationship between benthic and pelagic abundances of D. antarctica over several years are necessary to understand the subsidy dynamics of this bull kelp on local beaches.

2.4.2. Spatial variability of strandings of beach-cast rafts

Stranded biomasses, sizes, and floating times of individuals of D. antarctica varied strongly across the studied beaches. Beaches from the northern and southern zones had consistently higher beach-cast raft abundances than the beach from the central zone. These spatial patterns of stranded biomass are congruent with those previously reported (López et al. 2017) as well as with observed values in other studies (Rodríguez 2003; Duarte et al. 2008; 2009).

Pichicuy beach (~32°S) is close to the northern limit of the D. antarctica distribution on the continental coast of Chile (~30°S) (Fig. 3.1). In areas near the edges of the geographic ranges of seaweed species, it is expected that the benthic populations are smaller and more fragmented than central populations (Sagarin and Gaines 2002; Araújo et al. 2011; 2015). Previous studies confirmed the presence of floating bull kelps in areas north (i.e. Coquimbo, 30°S) and south (i.e. Valparaíso, 33°S) of Pichicuy (Hinojosa et al 2011; Tala et al. 2013), but the physiological performance of floating rafts in this zone tends to decrease due to abiotic factors suppressing their floating times (i.e. temperature and solar radiation, Rothäusler et al. 2012; Graiff et al. 2013; Tala et al. 2016; 2017). Also, oceanographic discontinuities, in terms of wind stress, eddy kinetic energy, and local upwelling, have been described in this area (Hormazábal et al. 2004; Yuras et al. 2005; Tapia et al. 2014). This in turn can affect the transport of bull kelp rafts; for instance, local

47

upwelling can generate offshore accumulations of floating specimens (evidenced by individuals with Lepas spp. in this study), which might return to the shore during periods of upwelling relaxation. Our results suggest that the local oceanographic features of this area might favor transport and return of floating bull kelps towards the coast.

High stranded biomasses were also observed in Curiñanco (~39°S), where strong westerly winds may periodically push floating kelps shoreward (Atkinson et al 2002; Hinojosa et al. 2010). Also, moderate UV-radiation and lower sea surface temperature may facilitate long-term persistence of kelp rafts in this region (Rothäusler et al. 2009), increasing the possibility that they return to the coast. On both beaches (Pichicuy and Curiñanco), the high presence of stranded individuals with Lepas spp. and their large sizes suggest that an important fraction of the stranded biomasses are subsidized from more distant benthic populations (possibly from the south, due to the general northward flow of the Humboldt Current), which could explain the high abundances of stranded rafts.

On the other hand, the low stranded biomasses at Itata Norte (~36°S) contrasts with the high abundances of floating bull kelps described in coastal areas in this region (Hinojosa et al. 2011). Strong upwelling currents keep kelp rafts in offshore waters and these might block the return to the coast (Sobarzo et al. 2007; Hinojosa et al. 2011). The nearby river plumes (e.g. from the Itata river) could generate hydrogeographic barriers (Saldías et al. 2012; Rech et al. 2014; Saldías et al. 2016) that suppress seasonal transport and arrival of kelp rafts from other areas. The small proportion of rafts with small stalked barnacles attached suggests that most stranded specimens come from nearby benthic populations and have not passed much time at sea. Moreover, this area presents a lower habitat availability for D. antarctica (i.e. rocky shores) due to the longer extension of sandy beaches compared to other areas (Thiel et al. 2007). Finally, commercial exploitation of natural kelp beds is concentrated between 34°S-36°S (i.e. over 80% of total landings, SERNAPESCA 2017; see Fig. S2.5). Extraction reduces the biomasses and sizes of kelps in the benthic populations (Castilla and Bustamante 1989; Bustamante and Castilla 1990; Castilla et al. 2007), which would suppress the availability of rafts, and could explain the low stranding biomass and small sizes of beach-cast rafts in this area.

48

3.4.3. Conclusions and Outlook

Seasonal variability of bull kelp strandings coincides with the seasonal fluctuations reported for the abundances of benthic populations of D. antarctica in our study area, and this seasonality was consistent among years. Local storm intensity also affected seasonal variability of strandings of D. antarctica on beaches, although this can vary along the latitudinal gradient due to oceanographic factors acting at larger, regional scales.

Increasing extraction of D. antarctica in recent years (SERNAPESCA 2017) and the implementation of management measures (i.e. ban periods, Subsecretaría de Pesca y Acuicultura 2015; 2017) suggest that stranded biomass might play an important role in covering the economic demand for this kelp. Our results indicate that collection of stranded bull kelps in open access coastal areas by shore gatherers will be most efficient during austral summer and autumn. However, it should be considered that there are also other incipient economic activities associated with beach-cast seaweeds, such as the collection of sandhoppers for the aquarium trade (Tapia-Lewin et al. 2017). Furthermore, seasonal variation in kelp supply can affect secondary consumers on sandy beaches (Quintanilla- Ahumada et al. 2018) and rocky shores (Rodriguez 2003). Harvesting stranded kelps would likely affect these kelp consumers and therefore economic and ecological effects of kelp collection should be carefully evaluated.

Acknowledgments

This study was financed by CONICYT/FONDECYT 1131082 to M.T. and F.T. and PhD- fellowship Beca CONICYT-PCHA/DoctoradoNacional/2014-21140010 to BL. We wish to express our gratitude to Óscar Pino, Vieia Villalobos and Ulyces Urtubia from Universidad Católica del Norte for their field assistance, as well as all the participants of the Universidad de Concepción (Algalab) and Universidad Austral de Chile (Lafkenchelab) who collaborated during all field surveys in central and southern beaches, respectively. E.C.M. and N.V. were supported by FONDAP research grant 15150003 (IDEAL). Also, Marcelo Rivadeneira is recognized for his suggestions about statistical analyses.

49

2. 5. References

Adams GD, Flores D (2016) Influencia de El Niño Oscilación del Sur en la disponibilidad y abundancia de recursos hidrobiológicos de la pesca artesanal en Ica, Perú. Rev Biol Mar Oceanogr 51:265-272. https://doi:10.4067/s0718-19572016000200005

Anderson DT (1994) Barnacles: Structure, Function, Development and Evolution. Chapman and Hall, London, UK

Araújo R, Serrão EA, Sousa-Pinto I, Aberg P (2011) Phenotypic differentiation at southern limit borders: The case study of two fucoid macroalgal species with different life- history traits. J Phycol 47:451-462. https://doi:10.1111/j.1529-8817.2011.00986.x

Araújo R, Serrão EA, Sousa-Pinto I, Arenas F, Monteiro CA, Toth G, Pavia H, Aberg P (2015) Trade-offs between life-history traits at range-edge and central locations. J Phycol 51:808-818. https://doi:10.1111/jpy.12321

Aravena G, Broitman B, Stenseth NC (2014) Twelve years of change in coastal upwelling along the central-northern coast of Chile: spatially heterogeneous responses to climatic variability. PLoS One 9:e90276. https://doi:10.1371/journal.pone.0090276

Arenas F, Fernández C, Rico J, Fernández E, Haya D (1995) Growth and reproductive strategies of Sargassum muticum (Yendo) Fensholt and Cystoseira nodicaulis (Whit.) Roberts. Sci Mar 59:1-8

Atkinson LP, Valle‐Levinson A, Figueroa D, Pol‐Holz D, Gallardo VA, Schneider W, Blanco JL, Schmidt M (2002) Oceanographic observations in Chilean coastal waters between Valdivia and Concepción. J Geophys Res Oceans 107:181-193. https://doi.org/10.1029/2001JC000991

Batista MB, Batista A, Franzan P, Simionatto P, Lima TC, Velez-Rubio GM, Scarabino F, Camacho O, Schmitz C, Martinez A, Ortega L, Fabiano G, Rothman MD, Liu G, Ojeda J, Mansilla A, Barreto LM, Assis J, Serrão EA, Santos R, Horta PA (2018) Kelps’ long-distance dispersal: Role of ecological/oceanographic processes and implications to marine forest conservation. Diversity 10:11. https://doi:10.3390/d10010011

50

Camus PA (1990) Procesos regionales y fitogeografía en el Pacífico Suroriental: el efecto de "El Niño - Oscilación del Sur". Rev Chil Hist Nat 63:11-17

Camus PA (2001) Biogeografía marina de Chile continental. Rev Chil Hist Nat 74:587-617. http://dx.doi.org/10.4067/S0716-078X2001000300008

Castilla JC, Bustamante RH (1989) Human exclusion from rocky intertidal of Las Cruces, central Chile - effects on Durvillaea antarctica (Phaeophyta, Durvilleales). Mar Ecol Prog Ser 50:203-214. https://doi:10.3354/Meps050203

Castilla J, Camus P (1992) The Humboldt-El Niño scenario: coastal benthic resources and anthropogenic influences, with particular reference to the 1982/83 ENSO. S Afr J Mar Sci 12:703-712. https://doi.org/10.2989/02577619209504735

Castilla JC, Campo MA, Bustamante RH (2007) Recovery of Durvillaea antarctica (Durvilleales) inside and outside Las Cruces Marine Reserve, Chile. Ecol Appl 17:1511-1522. https://doi:10.1890/06-1285.1

Collantes G, Riveros R, Acevedo M (1997) Fenología reproductiva de Durvillaea antarctica (Phaeophyta, Durvillaeales) del intermareal de caleta Montemar, Chile central. Rev Biol Mar Oceanogr 32:111-116

Collantes G, Merino A, Lagos V (2002) Fenología de la gametogénesis, madurez de conceptáculos, fertilidad y embriogénesis en Durvillaea antarctica (Chamisso) Hariot (Phaeophyta, Durvillaeales). Rev Biol Mar Oceanogr 37:83-112. https://doi:10.4067/s0718-19572002000100009.

Craw D, Waters J (2018) Long distance kelp-rafting of rocks around southern New Zealand. N Z J Geol Geophys: 1-16. https://doi.org/10.1080/00288306.2018.1492424

Deysher L, Norton TA (1981) Dispersal and colonization in Sargassum muticum (Yendo) Fensholt. J Exp Mar Biol Ecol 56:179-195. https://doi.org/10.1016/0022- 0981(81)90188-X

Duarte C, Jaramillo E, Contreras H (2008) Stranded algal wracks on a sandy beach of south central Chile: feeding and habitat preferences of juveniles and adults of

51

Orchestoidea tuberculata (Nicolet), (Amphipoda, Talitridae). Rev Chil Hist Nat 81:69-81

Duarte C, Jaramillo E, Contreras H, Acuña K, Navarro JM (2009) Importance of macroalgae subsidy on the abundance and population biology of the amphipod Orchestoidea tuberculata (Nicolet) in sandy beaches of south central Chile. Rev Biol Mar Oceanogr 44:691-702

Dugan JE, Hubbard DM, McCrary MD, Pierson MO (2003) The response of macrofauna communities and shorebirds to macrophyte wrack subsidies on exposed sandy beaches of southern California. Estuar Coast Shelf Sci 58:25-40. https://doi:10.1016/s0272-7714(03)00045-3

Duggins D, Eckman JE, Siddon CE, Klinger T (2001) Interactive roles of mesograzers and current flow in survival of kelps. Mar Ecol Prog Ser 223:143-155. https://doi:10.3354/Meps223143

Durán LR, Castilla JC, Oliva D (1987) Intensity of human predation on rocky shores at Las Cruces in Central Chile. Environ Conserv 14:143-149. https://doi:10.1017/S0376892900011504

Fraser CI, Nikula R, Spencer HG, Waters JM (2009) Kelp genes reveal effects of subantarctic sea ice during the Last Glacial Maximum. Proc Natl Acad Sci USA 106:3249-3253. https://doi:10.1073/pnas.0810635106

Fraser CI, Nikula R, Waters JM (2011) Oceanic rafting by a coastal community. Proc R Soc B-Biol Sci 278:649-655. https://doi:10.1098/rspb.2010.1117

Fraser CI, Thiel M, Spencer HG, Waters JM (2010) Contemporary habitat discontinuity and historic glacial ice drive genetic divergence in Chilean kelp. BMC Evol Biol 10:203. https://doi:10.1186/1471-2148-10-203

Garden CJ, Smith AM (2011) The role of kelp in sediment transport: Observations from southeast New Zealand. Mar Geol 281:35-42. https://doi:10.1016/j.margeo.2011.01.006

52

Garden CJ, Craw D, Waters JM, Smith A (2011) Rafting rocks reveal marine biological dispersal: A case study using clasts from beach-cast macroalgal holdfasts. Estuar Coast Shelf Sci 95:388-394. https://doi:10.1016/j.ecss.2011.10.008

Garden CJ, Currie K, Fraser CI, Waters JM (2014) Rafting dispersal constrained by an oceanographic boundary. Mar Ecol Prog Ser 501:297-302. https://doi:10.3354/meps10675

Garden CJ, Smith AM (2015) Voyages of seaweeds: The role of macroalgae in sediment transport. Sediment Geol 318:1-9. https://doi:10.1016/j.sedgeo.2014.11.007

Gelcich S, Edwards-Jones G, Kaiser MJ, Castilla JC (2006) Co-management policy can reduce resilience in traditionally managed marine ecosystems. Ecosystems 9:951- 966. https://doi:10.1007/s10021-005-0007-8

Gillespie RG, Baldwin BG, Waters JM, Fraser CI, Nikula R, Roderick GK (2012) Long- distance dispersal: a framework for hypothesis testing. Trends Ecol Evol 27:47-56. https://doi:10.1016/j.tree.2011.08.009

Graiff A, Karsten U, Meyer S, Pfender D, Tala F, Thiel M (2013) Seasonal variation in floating persistence of detached Durvillaea antarctica (Chamisso) Hariot thalli. Bot Mar 56:3-14. https://doi:10.1515/bot-2012-0193

Graiff A, Pantoja JF, Tala F, Thiel M (2016) Epibiont load causes sinking of viable kelp rafts: seasonal variation in floating persistence of giant kelp Macrocystis pyrifera. Mar Biol 163:190-204. https://doi.org/10.1007/s00227-016-2962-3

Hawes NA, Taylor DI, Schiel DR (2017) Transport of drifting fucoid algae: nearshore transport and potential for long distance dispersal. J Exp Mar Biol Ecol 490:34-41. https://doi.org/10.1016/j.jembe.2017.02.001

Helmuth B, Veit RR, Holberton R (1994) Long distance dispersal of a subantarctic brooding bivalve (Gaimardia trapesina) by kelp rafting. Mar Biol 120:421-426. https://doi:10.1007/bf00680216

Hinojosa I, Boltana S, Lancellotti D, Macaya E, Ugalde P, Valdivia N, Vásquez N, Newman WA, Thiel M (2006) Geographic distribution and description of four

53

pelagic barnacles along the south east Pacific coast of Chile - a zoogeographical approximation. Rev Chil Hist Nat 79:13-27

Hinojosa IA, Pizarro M, Ramos M, Thiel M (2010) Spatial and temporal distribution of floating kelp in the channels and fjords of southern Chile. Estuar Coast Shelf Sci 87:367-377. https://doi:10.1016/j.ecss.2009.12.010

Hinojosa IA, Rivadeneira MM, Thiel M (2011) Temporal and spatial distribution of floating objects in coastal waters of central-southern Chile and Patagonian fjords. Cont Shelf Res 31:172-186. https://doi:10.1016/j.csr.2010.04.013

Hirata T, Tanaka J, Iwami T, Ohmi T, Dazai A, Aoki M, Ueda H, Tsuchiya Y, Sato T, Yokohama Y (2001) Ecological studies on the community of drifting seaweeds in the south-eastern coastal waters of Izu Peninsula, central Japan. I. Seasonal changes of plants in species composition, appearance, number of species and size. Phycol Res 49:215-229. https://doi.org/10.1046/j.1440-1835.2001.00242.x

Hirata T, Tanaka J, Iwami T, Ohmi T, Dazai A, Aoki M, Ueda H, Tsuchiya Y, Sato T, Yokohama Y (2003) Ecological studies on the community of drifting seaweeds in the south‐eastern coastal waters of Izu Peninsula, central Japan. II: Seasonal changes in plants showing maximum stipe length in drifting seaweed communities. Phycol Res 51:186-191. https://doi.org/10.1046/j.1440-1835.2003.00309.x

Hobday AJ (2000) Abundance and dispersal of drifting kelp Macrocystis pyrifera rafts in the Southern California Bight. Mar Ecol Prog Ser 195:101-116. https://doi:10.3354/meps195101

Hoffmann A, Santelices B (1997) Flora marina de Chile Central. Ediciones Universidad Católica de Chile, Santiago de Chile, Chile

Hormazábal S, Shaffer G, Leth O (2004) Coastal transition zone off Chile. J Geophys Res Oceans 109:C01021. https://doi:10.1029/2003jc001956.

Hurd CL, Harrison PJ, Bischof K, Lobban CS (2014) Seaweed ecology and physiology. Cambridge University Press, Cambridge, UK.

Kingsford MJ (1992) Drift algae and small fish in coastal waters of northeastern New Zealand. Mar Ecol Prog Ser 80:41-55. https://doi:10.3354/meps080041

54

Kirkman H, Kendrick GA (1997) Ecological significance and commercial harvesting of drifting and beach-cast macro-algae and seagrasses in Australia: a review. J Appl Phycol 9:311-326. https://doi:10.1023/A:1007965506873

Ladah LB, Zertuche‐González JA, Hernández‐Carmona G (1999) Giant kelp (Macrocystis pyrifera, Phaeophyceae) recruitment near its southern limit in Baja California after mass disappearance during ENSO 1997–1998. J Phycol 35:1106-1112. https://doi.org/10.1046/j.1529-8817.1999.3561106.x

Lastra M, Rodil IF, Sánchez-Mata A, García-Gallego M, Mora J (2014) Fate and processing of macroalgal wrack subsidies in beaches of Deception Island, Antarctic Peninsula. J Sea Res 88:1-10. https://doi:10.1016/j.seares.2013.12.011

López BA, Macaya EC, Tala F, Tellier F, Thiel M (2017) The variable routes of rafting: stranding dynamics of floating bull kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Pacific. J Phycol 53:70-84. https://doi:10.1111/jpy.12479

López BA, Macaya EC, Rivadeneira MM, Tala F, Tellier F, Thiel M (2018) Epibiont communities on stranded kelp rafts of Durvillaea antarctica (Fucales, Phaeophyceae) – do positive interactions facilitate range extensions? J Biogeogr 45: (in press). https://doi:10.1111/jbi.13375

MacMillan MR, Quijón PA (2012) Wrack patches and their influence on upper-shore macrofaunal abundance in an Atlantic Canada sandy beach system. J Sea Res 72:28- 37. https://doi:10.1016/j.seares.2012.05.004

Marsden ID (1991) Kelp sandhopper interactions on a sand beach in New Zealand .1. Drift composition and distribution. J Exp Mar Biol Ecol 152:61-74. https://doi:10.1016/0022-0981(91)90135-j

McLachlan A, Brown AC (2006) The Ecology of Sandy Shores. Academic Press, Burlington, MA, USA

Miranda L, Thiel M (2008) Active and passive migration in boring isopods Limnoria spp. (Crustacea, Peracarida) from kelp holdfasts. J Sea Res 60:176-183. https://doi:10.1016/j.seares.2008.06.002

55

Nakagawa S, Schielzeth H (2013) A general and simple method for obtaining R2 from generalized linear mixed‐effects models. Methods in Ecology and Evolution 4 (2):133-142. doi:https://doi.org/10.1111/j.2041-210x.2012.00261.x

Norton T (1976) Why is Sargassum muticum so invasive? Br Phycol J 11:197-198

Norton TA (1977) The growth and development of Sargassum muticum (Yendo) Fensholt. J Exp Mar Biol Ecol 26:41-53. https://doi.org/10.1016/0022-0981(77)90079-X

Ohno M (1984) Observation on the floating seaweeds of near-shore waters of southern Japan. Hydrobiologia 116:408-412. https://doi.org/10.1007/BF00027711

Orr M, Zimmer M, Jelinski DE, Mews M (2005) Wrack deposition on different beach types: Spatial and temporal variation in the pattern of subsidy. Ecology 86:1496- 1507. https://doi:10.1890/04-1486

Owens EH (1982) Sea conditions. In: Schwartz ML (ed) Beaches and Coastal Geology. Encyclopedia of Earth Sciences Series. Springer US, Boston, USA, pp 722-804. https://doi:10.1007/0-387-30843-1_397

Pfister CA, Betcher SP (2018) Climate drivers and animal host use determine kelp performance over decadal scales in the kelp Pleurophycus gardneri (Laminariales, Phaeophyceae). J Phycol 54:1-11. https://doi.org/10.1111/jpy.12601

Piriz ML, Eyras MC, Rostagno CM (2003) Changes in biomass and botanical composition of beach-cast seaweeds in a disturbed coastal area from Argentine Patagonia. J Appl Phycol 15:67-74. https://doi:10.1023/A:1022959005072

Quintanilla-Ahumada D, Quijón PA, Navarro JM, Pulgar J, Duarte C (2018) Living on a trophic subsidy: Algal quality drives an upper-shore herbivore’s consumption, preference and absorption but not growth rates. PloS One 13:e0196121. doi:https://doi.org/10.1371/journal.pone.0196121

Rech S, Macaya-Caquilpan V, Pantoja JF, Rivadeneira MM, Jofre Madariaga D, Thiel M (2014) Rivers as a source of marine litter - A study from the SE Pacific. Mar Pollut Bull 82:66-75. https://doi:10.1016/j.marpolbul.2014.03.019

56

Rodríguez SR (2003) Consumption of drift kelp by intertidal populations of the sea urchin Tetrapygus niger on the central Chilean coast: possible consequences at different ecological levels. Mar Ecol Prog Ser 251:141-151. https://doi:10.3354/meps251141

Rothäusler E, Gómez I, Hinojosa IA, Karsten U, Tala F, Thiel M (2009) Effect of temperature and grazing on growth and reproduction of floating Macrocystis spp. (Phaeophyceae) along a latitudinal gradient. J Phycol 45:547-559. https://doi:10.1111/j.1529-8817.2009.00676.x

Rothäusler E, Gutow L, Thiel M (2012) Floating seaweed and their communities. In: Wiencke C, Bischof K (eds) Seaweed Biology. Ecological Studies, Ecological Studies 219. Springer, Berlin, Germany, pp 109-134. https://doi.org/10.1007/978-3- 642-28451-9_17

Rothäusler E, Corell H, Jormalainen V (2015) Abundance and dispersal trajectories of floating Fucus vesiculosus in the Northern Baltic Sea. Limnol Oceanogr 60:2173- 2184. https://doi:10.1002/lno.10195

Rothäusler E, Reinwald H, López BA, Tala F, Thiel M (2018) High acclimation potential in floating Macrocystis pyrifera to abiotic conditions even under grazing pressure–a field study. J Phycol 54:368-379. https://doi.org/10.1111/jpy.12643

Sagarin RD, Gaines SD (2002) The 'abundant centre' distribution: to what extent is it a biogeographical rule? Ecol Lett 5:137-147. https://doi:10.1046/j.1461- 0248.2002.00297.x

Saldías GS, Sobarzo M, Largier J, Moffat C, Letelier R (2012) Seasonal variability of turbid river plumes off central Chile based on high-resolution MODIS imagery. Remote Sens Environ 123:220-233. https://doi:10.1016/j.rse.2012.03.010

Saldías GS, Largier JL, Mendes R, Pérez-Santos I, Vargas CA, Sobarzo M (2016) Satellite- measured interannual variability of turbid river plumes off central-southern Chile: Spatial patterns and the influence of climate variability. Prog Oceanogr 146:212- 222. https://doi:10.1016/j.pocean.2016.07.007

57

Santelices B, Castilla JC, Cancino J, Schmiede P (1980) Comparative ecology of Lessonia nigrescens and Durvillaea antarctica (Phaeophyta) in central Chile. Mar Biol 59:119-132. https://doi:10.1007/bf00405461

SERNAPESCA (2017) Anuario Estadístico de Pesca. Servicio Nacional de Pesca y Acuicultura, Ministerio de Economía, Fomento y Turismo, Gobierno de Chile, Chile.

Smith J, Bayliss-Smith T (1998) Kelp-plucking: coastal erosion facilitated by bull-kelp Durvillaea antarctica at subantarctic Macquarie Island. Antarct Sci 10:431-438. https://doi.org/10.1017/S0954102098000522

Sobarzo M, Bravo L, Donoso D, Garcés-Vargas J, Schneider W (2007) Coastal upwelling and seasonal cycles that influence the water column over the continental shelf off central Chile. Prog Oceanogr 75:363-382. https://doi.org/10.1016/j.pocean.2007.08.022

Song Z, Shu Q, Bao Y, Yin X, Qiao F (2015) The prediction on the 2015/16 El Niño event from the perspective of FIO-ESM. Acta Oceanol Sin 34:67-71. https://doi:10.1007/s13131-015-0787-4.

Subsecretaría de Pesca y Acuicultura (2015) Veda biológica para el recurso cochayuyo Durvillaea antarctica en la VI Región del Libertador Bernardo O‘Higgins. Informe Técnico (RPESQ), No. 151/2015. Unidad de Recursos Bentónicos. Gobierno de Chile, Chile

Subsecretaría de Pesca y Acuicultura (2017) Veda extractiva del recurso cochayuyo (Durvillaea antarctica), en la VII Región, 2017 Informe Técnico, No 8/2017. Comité Científico Técnico Bentónicos. Gobierno de Chile, Chile.

Tala F, Gómez I, Luna-Jorquera G, Thiel M (2013) Morphological, physiological and reproductive conditions of rafting bull kelp (Durvillaea antarctica) in northern- central Chile (30°S). Mar Biol 160:1339-1351. https://doi:10.1007/s00227-013- 2186-8

58

Tala F, Velásquez M, Mansilla A, Macaya EC, Thiel M (2016) Latitudinal and seasonal effects on short-term acclimation of floating kelp species from the South-East Pacific. J Exp Mar Biol Ecol 483:31-41. https://doi:10.1016/j.jembe.2016.06.003

Tala F, Penna-Díaz MA, Luna-Jorquera G, Rothäusler E, Thiel M (2017) Daily and seasonal changes of photobiological responses in floating bull kelp Durvillaea antarctica (Chamisso) Hariot (Fucales: Phaeophyceae). Phycologia 56:271-283. https://doi.org/10.2216/16-93.1

Tapia FJ, Largier JL, Castillo M, Wieters EA, Navarrete SA (2014) Latitudinal discontinuity in thermal conditions along the nearshore of central-northern Chile. PLoS One 9:e110841. https://doi:10.1371/journal.pone.0110841.

Tapia-Lewin S, Vergara K, De La Barra C, Godoy N, Castilla JC, Gelcich S (2017) Distal impacts of aquarium trade: Exploring the emerging sandhopper (Orchestoidea tuberculata) artisanal shore gathering fishery in Chile. Ambio 46:706-716. https://doi:10.1007/s13280-017-0906-x

Taylor DI, Schiel DR (2005) Self-replacement and community modification by the southern bull kelp Durvillaea antarctica. Mar Ecol Prog Ser 288:87-102. https://doi:10.3354/meps288087

Team R DC (2018) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: http://www.r-project.org/.

Thiel M (2003) Reproductive biology of Limnoria chilensis: another boring peracarid species with extended parental care. J Nat Hist 37:1713-1726. https://doi.org/10.1080/00222930210125416

Thiel M, Gutow L (2005a) The ecology of rafting in the marine environment. I. The floating substrata. Oceanogr Mar Biol Ann Rev 42:181-263. https://doi:10.1201/9780203507810

Thiel M, Gutow L (2005b) The ecology of rafting in the marine environment. II. The rafting organisms and community. Oceanogr Mar Biol Ann Rev 43:279-418. https://doi:10.1201/9781420037449

59

Thiel M, Macaya EC, Acuña E, Arntz WE, Bastias H, Brokordt K, Camus PA, Castilla JC, Castro LR, Cortés M, Dumont CP, Escribano R, Fernández M, Gajardo JA, Gaymer CF, Gómez I, González AE, González HE, Haye PA, Illanes JE, Iriarte JL, Lancellotti DA, Luna-Jorquera G, Luxoroi C, Manríquez PH, Marín V, Muñoz P, Navarrete SA, Pérez E, Poulin E, Sellanes J, Sepúlveda HH, Stotz W, Tala F, Thomas A, Vargas CA, Vásquez JA, Vega JMA (2007) The Humboldt Current System of northern and central Chile. Oceanogr Mar Biol Ann Rev 45:195-344. https://doi:10.1201/9781420050943

Thiel M, Fraser CI (2016) The role of floating plants in dispersal of biota across habitats and ecosystems. In: Olafsson E (ed) Marine Macrophytes as Foundation Species. CRC Press, Taylor & Francis Group, Florida, USA, pp 76-99

Underwood AJ (1997) Experiments in ecology: their logical design and interpretation using analysis of variance. Cambridge University Press, Cambrige, UK.

Vargas CA, Narváez DA, Piñones A, Navarrete SA, Lagos NA (2006) River plume dynamic influences transport of barnacle larvae in the inner shelf off central Chile. J Mar Biol Assoc U.K. 86:1057-1065. https://doi:10.1017/s0025315406014032

Vega JMA, Vásquez JA, Buschmann AH (2005) Population biology of the subtidal kelps Macrocystis integrifolia and Lessonia trabeculata (Laminariales, Phaeophyceae) in an upwelling ecosystem of northern Chile: interannual variability and El Niño 1997–1998. Rev Chil Hist Nat 78:33-50

Waters JM, Craw D (2018) Cyclone-driven marine rafting: storms drive rapid dispersal of buoyant kelp rafts. Mar Ecol Prog Ser 602:77-85. https://doi.org/10.3354/meps12695

Waters JM, King TM, Fraser CI, Craw D (2018) Crossing the front: contrasting storm- forced dispersal dynamics revealed by biological, geological and genetic analysis of beach-cast kelp. J R Soc Interface 15:20180046. https://doi:10.1098/rsif.2018.0046

Westermeier R, Muller DG, Gómez I, Rivera P, Wenzel H (1994) Population biology of Durvillaea antarctica and Lessonia nigrescens (Phaeophyta) on the rocky shores of southern Chile. Mar Ecol Prog Ser 110:187-194. https://doi:10.3354/meps110187

60

Wichmann CS, Hinojosa IA, Thiel M (2012) Floating kelps in Patagonian Fjords: an important vehicle for rafting invertebrates and its relevance for biogeography. Mar Biol 159:2035-2049. https://doi:10.1007/s00227-012-1990-x

Xu M, Sakamoto S, Komatsu T (2016) Attachment strength of the subtidal seaweed Sargassum horneri (Turner) C. Agardh varies among development stages and depths. J Appl Phycol 28:3679-3687. https://doi.org/10.1007/s10811-016-0869-5.

Yuras G, Ulloa O, Hormazábal S (2005) On the annual cycle of coastal and open ocean satellite chlorophyll off Chile (18°-40°S). Geophys Res Lett 32:L23604. https://doi:10.1029/2005gl023946

Zar JH (2010) Biostatistical analysis. 5th Edition edn. Prentice-Hall, Inc, Upper Saddle Rive, New Jersey, USA

ZoBell CE (1971) Drift seaweeds on San Diego county beaches. Nova Hedwigia 32 (Suppl):269-314

Zuur A, Ieno E, Walker N, Saveliev A, Smith G (2009) Mixed effects models and extensions in ecology with R. Statistics for Biology and Health. Springer, Science and Business Media, New York, USA

61

Chapter 3

No Sex-Related Dispersal Limitation in a Dioecious, Oceanic Long-Distance Traveller: the Bull Kelp Durvillaea antarctica

(manuscript published in Bot. Mar. 59: 39-50, 2016)

Dominic Lizée-Prynne1,2, Boris Lopez1,3, Fadia Tala1,4, Martin Thiel1,5,6*

1Facultad Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 2Département de Biologie, Université Laval, Québec, Québec, Canada. 3Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos Avenida Fuchslocher 1305, Osorno, Chile. 4Centro de Investigación y Desarrollo Tecnológico en Algas de la Universidad Católica del Norte (CIDTA-UCN), Larrondo 1281, Coquimbo, Chile. 5Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile. 6Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Coquimbo, Chile.

*Corresponding author: [email protected]

62

Botanica Marina 2016; 59(1): 39–50

Dominic Lizée-Prynne, Boris López, Fadia Tala and Martin Thiel* No sex-related dispersal limitation in a dioecious, oceanic long-distance traveller: the bull kelp Durvillaea antarctica

DOI 10.1515/bot-2015-0072 Chile. Instead, we suggest that ecological processes, such Received 1 September, 2015; accepted 15 December, 2015; online as density-blocking and physical factors (i.e. currents and first 22 January, 2016 winds), limit the potential for successful rafting dispersal.

Abstract: Dispersal of dioecious floating seaweeds could Keywords: chemical composition; dioecious seaweeds; be limited due to biological constraints. This study exam- dispersal potential; phlorotannins; pigments; rafting. ined for benthic and floating populations (stranded indi- viduals) of the rafting kelp Durvillaea antarctica whether male and female individuals cohabit within one holdfast. As a previous study had indicated colour differences Introduction between sexes, we also examined whether these colour For any successful dispersal to occur it is crucial that viable differences are consistent and possibly related to pigment propagules reach new environments. In sexually reproduc- and phlorotannin concentrations. Our large-scale sur- ing species this implies the arrival of members of both sexes vey of rafted holdfasts and a small-scale survey of ben- that can successfully reproduce and generate subsequent thic holdfasts at two sites found that reproductive males generations. Depending on the species, the probability of and females do travel together in coalesced holdfasts, successful colonisation of new habitats can be consider- although this proportion is relatively low (5–17%). There ably lower in sexually reproducing species than in asexu- were no sex-specific differences in pigment and phloro- ally reproducing species, where any single individual can tannin concentrations, but there were significant differ- produce multiple propagules. This is especially true for ences between the two benthic populations. There was species in which the sexes are separate due to their dioe- no relationship between the colouration of thalli and the cious sexual system. Among these dioecious organisms concentration of pigments but there was a slight colour are many brown seaweeds (Luthringer et al. 2014), some difference between vegetative and reproductive sexual of which have been reported as long-distance dispers- stages. Based on these results we conclude that biological ers based on genetic evidence (Fraser et al. 2010a, Coyer conditions are not the cause for the lack of genetic con- et al. 2011). Long-distance dispersal has been suggested nectivity between D. antarctica populations from central for some positively buoyant dioecious seaweeds that are able to maintain viable propagules over long periods *Corresponding author: Martin Thiel, Facultad Ciencias del of time, as is the case for Fucus vesiculosus ­Linnaeus Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile; Millennium Nucleus Ecology and Sustainable Management 1753 (­Vandendriessche et al. 2007, Coyer et al. 2011) and of Oceanic Island (ESMOI), Larrondo 1281, Coquimbo, Chile; ­Hormosira banksii (Turner) Decaisne 1842 (McKenzie and and Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Bellgrove 2008). Larrondo 1281,Coquimbo, Chile, e-mail: [email protected] Long-distance dispersal has allowed the spreading Dominic Lizée-Prynne: Facultad Ciencias del Mar, Universidad of many dioecious species, terrestrial or aquatic, which Católica del Norte, Larrondo 1281, Coquimbo, Chile; and employ diverse strategies to ensure transport and estab- Département de Biologie, Université Laval, Québec, 1045, Avenue de la Medecine, Québec, QC G1V 0A6, Canada lishment of viable propagules. For instance, certain sea- Boris López: Facultad Ciencias del Mar, Universidad Católica del weeds may be dispersed via the consumption of their Norte, Larrondo 1281, Coquimbo, Chile; and Departamento de spores by various herbivores and their release via the Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos excrements of the animal or by having their spores stick Avenida Fuchslocher 1305, Osorno, Chile to body appendages (Buschmann and Bravo 1990). In Fadia Tala: Facultad Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile; and Centro de Investigación dioecious animals that have internal fertilisation, sperm y Desarrollo Tecnológico en Algas de la Universidad Católica del storage may enhance the probability of simultaneous Norte (CIDTA-UCN), Larrondo 1281, Coquimbo, Chile travel of both sexes (Shine et al. 2001, Friesen et al. 2014).

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 63 40 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller

Other dioecious species feature sexual parasitism, where and biological traits that limit the dispersal potential of the significantly smaller males attach themselves to the dioecious seaweeds, even in species where both sexes can females for various reasons, such as nutrition, and release be united in the same holdfast. sperm whenever the dispersing female is ready to spawn To study the potential of long-distance dispersal in a (Vieira et al. 2013). dioecious species, the brown seaweed Durvillaea antarc- In dioecious seaweeds, each individual will either be tica (Fucales) was selected as a model species. Molecular male or female (in the haploid or diploid phase). Among studies have confirmed that long-distance dispersal of this the brown seaweeds, dioecy during the haploid phase is large kelp has occurred in the subantarctic region over most prevalent (Luthringer et al. 2014) and dioecy during distances of thousands of kilometres (Fraser et al. 2009, the diploid phase is exclusive to species of the order 2010b). However, for a distinct clade from the continental Fucales, which may or may not be buoyant. As certain of coast of Chile, there are extensive regions (200–300 km these dioecious species travel thousands of kilometres coastline) with unique haplotypes, and there appears to before reaching new potential colonisation sites, the prob- be very limited exchange between these regions (Fraser ability of simultaneous arrival of both sexes is limited. et al. 2010b). Thus, there are factors, ecological or biologi- However, some large brown seaweeds exhibit holdfast cal, that suppress the dispersal potential of this particular coalescence joining multiple conspecifics (González et al. clade of D. antarctica along the Chilean continental coast 2015). If this coalescence unites individuals of both sexes (Fraser et al. 2010b). in the same holdfast and both travelling sexes are sexu- In Chile extensive benthic populations of D. antarctica ally mature at the time of arrival, the probability of long- occur between 30°S, its most northerly location (Cheshire distance dispersal increases considerably. et al. 1995), and 44°S, the southernmost confirmed loca- Various environmental factors and biological traits tion of the continental clade (Fraser et al. 2010b). Conse- can also limit the colonisation potential of long-distance quently, it could be that dispersal is suppressed due to the travellers. For instance, a small number of propagules limited chances of successful immigration into dense pop- arriving from a distant population may have limited ulations. However, it is also possible that holdfasts rarely chances to establish themselves within a dense local pop- contain individuals of both sexes. Currently, no informa- ulation as a result of density-blocking (Waters et al. 2013, tion is available about the sex and maturity of individuals Neiva et al. 2014), when most available space is already that travel together with a coalesced holdfast. Furthermore, occupied and propagules of new arrivals will be strongly it is possible that one sex is more sensitive to environmen- outnumbered by local propagules. On the other hand, tal conditions than the other. In D. antarctica, males were the biological traits of a species, e.g. thermal tolerance reported to be of a yellowish colour whereas females have limits or tolerance to solar radiation at the sea surface, a darker, black-brownish colour (Collantes et al. 1997). This may reduce its potential for long-distance dispersal under could indicate important chemical differences between the certain conditions. For seaweeds it has been shown that sexes that could impact their dispersal potential. Male and warmer water temperatures or higher solar irradiance female individuals usually have differential energy alloca- limit their reproductive potential (Macaya et al. 2005, tion, where females invest a lot more energy into reproduc- Rothäusler et al. 2009). Furthermore, in dioecious sea- tion and protection of the gametes (Delph 1999, Cornelissen weeds, one sex may be more susceptible to environmen- and Stiling 2005, Vergés et al. 2008). This variation has a tal factors than the other (Delph 1999), and thus lose its direct impact on the chemical composition of each indi- reproductive potential during long-distance voyages. For vidual. For instance, one could expect that females might example, male individuals of the red seaweed Asparagop- produce more phlorotannins, as this phenolic compound sis armata Harvey 1855 are more susceptible to herbivory has repeatedly been proven to have a protective role, such than females because they do not invest as much energy as having high antioxidant activity in seaweeds (Li et al. in anti-herbivore defences (Vergés et al. 2008). Also, in 2009, Wang et al. 2009, Onofrejová et al. 2010). Addition- dioecious plants, males and females can present sub- ally, the concentration of pigments, which are essential for stantial physiological differences (­Cornelissen and Stiling photosynthesis and photoprotection, can vary between 2005). This phenomenon of sexual differentiation also sexes (Guillemin et al. 2014). occurs in other seaweeds, such as the red seaweed Graci- The current study thus aims to (i) determine whether laria chilensis Bird, McLachlan and Oliveira 1986 (Guil- D. antarctica rafts comprise coalesced holdfasts with both lemin et al. 2014). However, this variability remains to be male and female individuals, and (ii) evaluate if male properly quantified for brown seaweeds (Thornber 2006). or female individuals show different concentrations of In summary, there may be a variety of ecological factors various chemical components, reflecting poor resistance

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 64 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller 41 to floating conditions and interfering with long-distance dispersal. We hypothesize that a significant number of holdfasts contain individuals of both sexes that also mature in synchrony, but that male and female individu- als differ significantly in chemical composition of pig- ments and phlorotannins, resulting in sexual differences in dispersal potential.

Materials and methods

Study sites and sampling of benthic specimens

We studied the central Chilean clade of Durvillaea ­antarctica (Chamisso) Hariot 1892, which grows in dense benthic populations in the low intertidal zone of rocky shores between 30°S and 44°S (Cheshire et al. 1995, Fraser et al. 2010b). In order to obtain information on coalesced holdfasts, in particular the maturity stage and sex dis- tribution of travelling individuals, recently stranded rafts with holdfasts were collected on 33 beaches during winters 2013, 2014 and 2015 as the fertile phase of D. antarctica is primarily found between austral autumn and spring (­Santelices et al. 1980, Collantes et al. 2002) (Figure 1). A total of 1044 kelp rafts with holdfasts were evaluated for both their sex and their maturity stage (see below). For each intact organism with complete holdfast, the number of stipes per holdfast was determined. If a Figure 1: Study area and spatial distribution of the stranded holdfast had more than five stipes ( = individuals), only and benthic D. antarctica in central Chile. PTOS, Puerto Oscuro; the five largest stipes were sampled. In the following we TOT, Totoralillo Sur. will use the term stipe to refer to an individual (consist- ing of its stipe and corresponding blade) within a hold- fast. For a detailed examination of sex and biochemical of the study we only used individual holdfasts that had composition, kelps were sampled from benthic habitats at least 250 g total weight (including holdfast, stipes and at two locations in northern-central Chile, Puerto Oscuro blades) and individual stipes that were at least 50 cm long, (31°25′22″S; 71°35′48″W) and Totoralillo Sur (32°05′07″S; because at this size the individuals start to become sexu- 71°31′41″W). These samples were collected during June ally mature (Santelices et al. 1980, Collantes et al. 2002). 2015 as the individuals would mostly be in their fertile Three samples were taken from each individual stipe: phases (Figure 1). (1) a large blade sample in order to determine the sex and We collected 56 benthic D. antarctica holdfasts in maturity stage of the individual, (2) a small blade sample Puerto Oscuro and 62 holdfasts in Totoralillo Sur, in order to determine the colour, and (3) a small blade sample for to sample approximately 100 stipes, or individuals, per pigment and phlorotannin analyses. Of the approximately beach. Holdfasts were detached from rock surfaces in 100 stipes that were collected per beach, eight repro- the lower intertidal zone using a hatchet, taking care to ductive samples of each sex as well as eight vegetative minimize mechanical damage of the individual. The entire samples were selected for determination of phlorotannin organism was then weighed with the holdfast before concentrations. The large blade samples were individu- cutting each individual at the base of its stipe. Each indi- ally placed in large Ziploc® bags and covered with salt vidual was then weighed on its own and measured from to dehydrate and preserve the reproductive tissues. The the base of the stipe to the tip of the blade. For this part small pigment and phlorotannin samples were kept in

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 65 42 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller individual small bags, transported in darkness in a cooler, Colour determination and frozen immediately upon return to the laboratory (within 12 h after field sampling) in order to maintain the In order to examine whether colour is related to pigment pigments and phlorotannins intact. concentrations, we determined the colour of each indi- vidual; a small piece of the blade (approximately 5 × 5 cm) was photographed and the picture was processed with Sex and maturity stage determination the image analysis software Adobe Photoshop Lightroom 5.3. In this manner, the colour analysis remained objec- The sex of Durvillaea antarctica can only be determined tive and quantifiable, as opposed to a previous study with full accuracy when gametes are identified by means where colour variation was only described qualitatively of histological observation. Male conceptacles contain (Collantes et al. 1997), leading to possibly unreliable many small antheridia containing 64 sperm cells each, results based on the use of various systems and observ- and female conceptacles contain multiple larger oogonia, ers (Kendal et al. 2013). The blade pieces were photo- with four large, dark oospheres within each oogonium. graphed within 24 h after sampling in order to exclude Using the large blade samples, thin transverse cuts were the risk of colour variation caused by the drying of the examined microscopically to determine both the sex and samples. These pictures were taken in the laboratory and maturity stages of the conceptacles. As each stipe of a under constant light conditions to allow a more accu- holdfast represents a single individual (González et al. rate comparison between images. Four study lamps with 2015), the maturity of 30 conceptacles per stipe were 4 W white and cold lightbulbs were used to illuminate determined from different cuts following Collantes et al. the algae without creating shadows. The method used (2002) in order to describe the maturity stage of each to quantitatively express the colouration of the images stipe, which was done by identifying the most abundant was red, green, and blue (RGB) values, which can define conceptacle maturity stage. The current study aimed to any given colour (Zhang et al. 2014). Although this is a determine whether the stipe was vegetative or reproduc- rather common way of representing colour, it is impor- tive, and thus we did not distinguish between the differ- tant to note that RGB values are rarely standardised ent maturity stages described by Collantes et al. (2002). and vary according to the instrument used to obtain the As the conceptacles described by Collantes et al. (2002) as image (Kendal et al. 2013). All images were taken using a “immature” contain fully developed gametes and are very Canon EOS Rebel T3i Camera at a distance of 50 cm from close (within hours or days) to becoming “fully mature”, the sample and the RGB values of these samples were we considered these as reproductive. Also, since senes- obtained using Adobe Photoshop Lightroom 5.3. The cent individuals as described by Collantes et al. (2002) three colour values, R, G, and B, were evaluated individ- release very few gametes and the conceptacles are closed ually. A high colour value means the colour is extremely and moving to the medullar tissue (reverting to vegetative light. Therefore, the RGB colour value 0,0,0 is associated state), all senescent individuals were considered to be veg- with black. A perfectly white object has an RBG colour etative. In this manner, the study remained conservative value of 255,255,255. As colour seemed to vary somewhat as it is sometimes possible to find mature conceptacles within a single sample, the average colour of nine differ- within a generally senescent individual. Consequently, ent pixels within the sample was calculated. the two maturity stages “vegetative” and “reproductive” were distinguished as follows: –– Vegetative: Absence of cellular differentiation or ini- Pigment concentrations tial differentiation of immature conceptacle between the subcortex and the medulla. In some cases there The studied pigments were Chl a, Chl c and total carot- may be newly formed conceptacles but, as the gam- enoids, which includes mainly β-carotene and fucox- etes are undeveloped and the sex is unidentifiable, anthin. The method used to determine the pigment they were considered to be vegetative. Also, any senes- concentrations was as described in Tala et al. (2013). The cent individual was equally categorised as vegetative. samples were frozen at -80°C before being added to N,N- –– Reproductive: Presence of conceptacles that are well dimethylformamide (DMF) for 24 h at 4°C in darkness. As developed. These individuals have identifiable sexes described in Rothäusler et al. (2011), the concentrations of where gametes may already be in the process of being Chl a, Chl c and total carotenoids were calculated using released. The neck of the conceptacle, as well as the the dichromatic equations for Chl a (Inskeep and Bloom ostiole, may or may not be visible. 1985), Chl c and carotenoids (Henley and Dunton 1995)

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 66 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller 43 and expressed as mg per g wet weight. The absorbance sexes) using a χ2-test (Zar 2010). This analysis was done of all pigment supernatants was measured in a UV-visible using Microsoft Excel 2013. spectrophotometre (Rayleigh, model UV-1601, China) and In order to determine the relationship between pig- at various wavelengths (480, 510, 664.5, 647 and 750 nm) ments and the coloration of the kelp (R, G, and B, sepa- in order to calculate each pigment concentration. Total rately) for benthic samples from Puerto Oscuro and pigment concentration, when used, was calculated by Totoralillo Sur, the Pearson correlation coefficient was adding all individual pigments together and maintaining calculated, as well as ANOVA of simple linear regression units as mg per g wet weight. to verify the significance of the slope. To compare con- centrations of all the studied pigments and phlorotan- nins with the sexual stage (fixed factor: male, female, or Phlorotannin concentrations vegetative) and according to their localities (fixed factor: Puerto Oscuro, Totoralillo Sur), a two-way ANOVA was Phlorotannin levels were determined for a total of 24 applied after previous analyses of normality (Kolmogorov- samples for each location (eight vegetative samples, eight Smirnov’s test) and homoscedasticity (Levene’s test). All reproductive females, and eight reproductive males). pigment values were transformed using log10(x+1) and Phlorotannin levels, as soluble phenols, were obtained phlorotannin values were modified using the arcsine using the method described by Koivikko (2008) with some transformation. Lastly, if there were significant effects, moderate modifications. The standard used was purified the differences between treatments were examined a pos- phloroglucinol (Merck, Darmstadt, Germany). The frozen teriori with Tukey HSD tests (Zar 2010). For all analyses samples were lyophilised for 96 h before being pulverised with pigments, three stipes were considered to be outliers with an Oster stick mixer. In order to extract the phenols, and were therefore excluded from the tests. All statisti- 10 mg of each sample was kept shaking for 24 h in 3 ml cal tests were done with R 3.1.3, using the lme4 package of 70% acetone (Merck, Darmstadt, Germany) at 4°C. (R Development Core Team 2015). The samples were then centrifuged at 3300 g for 15 min at 4°C and 250 μl of the supernatant was tested. To this sample we added 1250 μl of deionised water, 500 μl of 1 N Folin-Ciocalteu reagent (Sigma-Aldrich, Steinheim, Results

Germany), and 1000 μl of 20% sodium carbonate (NaCO3; Sigma-Aldrich, Steinheim, Germany). After addition of Stipe distribution on holdfasts the reagents samples were incubated for 45 min at room temperature in complete darkness, and then centrifuged Based on the samplings of rafting (recently stranded) at 2000 g for 3 min at room temperature before reading the holdfasts from the winters 2013, 2014 and 2015, 52% of the absorbance at 730 nm using a spectrophotometre (Ray- 1044 holdfasts had only one stipe, the majority of which leigh UV-1601 UV/VIS, Beijing, China). The results were were in the vegetative state (Figure 2A). Of all stranded expressed as percent dry weight. Two successive extrac- holdfasts, only 5.7% had at least one stipe of each sex tions were done 24 h apart and the two values were added that was reproductive (Figure 2A). Results were similar to determine the final concentration. for the benthic holdfasts from winter 2015 as most hold- fasts only had one stipe. However, in this case stipes were mostly in the reproductive stage, whether male or female, Statistical tests and 17.7% of all holdfasts had reproductive stipes of both sexes (Figure 2B). The distribution of holdfasts with both The number of stipes per holdfast was determined for all sexes were significantly different from a binomial distri- stranded and benthic samples. For the coalesced holdfasts, bution both for stranded (χ2 = 23.35; df = 13; p < 0.05) and for it was noted whether they hosted only vegetative (V), only benthic holdfasts (χ2 = 14.55; df = 7; p < 0.05). reproductive males (M), only reproductive females (F), or both reproductive male and female stipes together (MF). In order to determine the probability that such a floating Pigments and colour holdfast includes both sexes with a specific number of stipes, the obtained distribution of sexes in stranded and The total pigments were negatively correlated with both benthic samples was compared with a binomial distribu- the R (r = -0.205; p = 0.003) and G (r = -0.159; p = 0.022) colour tion (presence of both sexes and absence of one or both values (Table 1). There was no relationship between Chl

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 67 44 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller

A 35 Stranded holdfasts a, Chl c, carotenoids and total pigment concentrations Vegetative (V) and the blue colour. Statistically significant colour differ- 30 Only female (F) Only male (M) ences were observed between the sexual stages, where the 25 Both MF vegetative stipes had higher red, green, and blue colour

20 values than female stipes (Figure 3). However, the male stipes did not differ from either vegetative or female stipes 15 (p < 0.05), except for the blue colour value where values

10 of males were lower than those of vegetative individu-

Frequency of holdfast (%) als (Figure 3). Furthermore, no location effect was found 5 for the red and green colour values, but the blue colour 0 values were higher in Puerto Oscuro than in Totoralillo B 35 Benthic holdfasts Sur (Figure 3; Table 2). Also, there was no significant inter- action between localities and sexual stages (Table 2). 30

25 Puerto Oscuro Totoralillo Sur 20 A 60 Red value

50 15 40 10 Frequency of holdfast (%) 30 5

Colour value 20 0 12345 6>7 10 Number of stipes 0 Figure 2: Durvillaea antarctica: frequency of holdfasts with differ- B 60 Green value ent numbers of stipes, and with only vegetative stipes, only male stipes, only female stipes, and both sexes within one holdfast. 50 Samples collected at (A) 33 beaches on the coast of Chile between 40 28°S and 42°S (n = 1044 stranded holdfasts during the winters 2013, 2014 and 2015), and (B) two sites near the northern distribution 30 limit of D. antarctica (n = 56 holdfasts from Puerto Oscuro and n = 62 holdfasts from Totoralillo Sur). Colour value 20

10 Table 1: Durvillaea antarctica: summary of correlations and ANOVAs of 0 simple linear regression between pigments and colour values R (red), G C 60 (green), and B (blue) in the blades of the different stipes samples. Blue value 50 Pigments Colour r df F p-Value 40 Total pigment R -0.205 1;203 8.93 0.003 G -0.159 1;203 5.28 0.022 30 B 0.013 1;203 0.03 0.844 Colour value 20 Chlorophyll a R -0.212 1;203 9.58 0.002 G -0.164 1;203 5.66 0.018 10 B 0.008 1;203 0.01 0.908 Chlorophyll c R -0.162 1;203 5.49 0.020 0 Male Female Vegetative G -0.120 1;203 2.97 0.086 B 0.045 1;203 0.42 0.516 Sexual stage Carotenoids R -0.205 1;203 8.94 0.003 Durvillaea antarctica: average (mean±SD) colour values G -0.163 1;203 5.54 0.019 Figure 3: B 0.006 1;203 0.01 0.926 of algal samples of different sexual stages and from two locali- ties, Puerto Oscuro and Totoralillo Sur. Different letters above the Significant values (p < 0.05) in bold face. columns indicate differences between sexual stages significant Stipes from both sites (Puerto Oscuro and Totoralillo Sur) were at p = 0.05. Numbers of stipes from each site and sexual stage are pooled for these analyses. listed at the bottom of each column.

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 68 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller 45

Table 2: Durvillaea antarctica: summary of two-way ANOVAs for concentrations between males and females, but pigment colour (R, G and B) according to site (Puerto Oscuro and Totoralillo concentrations were consistently higher in Puerto Oscuro Sur) and sexual stage (vegetative, reproductive male, and reproduc- than in Totoralillo Sur. However, there was no significant tive female) of algal samples. interaction between site and stipe stage (Table 3). Variable Effect df F p-Value

Red value, R Site (S) 1;202 0.41 0.522 Stipe stage (SS) 2;202 3.49 < 0.05 Phlorotannins, location, and sexual stage S × SS 2;202 0.95 0.390 Green value, G Site (S) 1;202 1.33 0.251 Phlorotannin concentrations were significantly higher Stipe stage (SS) 2;202 3.68 < 0.05 in Totoralillo Sur than in Puerto Oscuro, although there S × SS 2;202 0.82 0.440 was no difference for females (Figure 5). Furthermore, < Blue value, B Site (S) 1;202 7.58 0.01 phlorotannin concentrations of female individuals were Stipe stage (SS) 2;202 5.92 < 0.01 S × SS 2;202 0.12 0.888 significantly higher than those of vegetative individuals in Puerto Oscuro (Figure 5). There was no significant interac- Significant values (p < 0.05) in bold face. tion between the site and stipe stage (Table 3).

Pigments, location, and sexual stage Discussion For Chl a, Chl c and carotenoids, there was a significant dif- ference in pigment concentration between vegetative and The current study has confirmed that reproductive indi- reproductive individuals, whether male or female. Vegeta- viduals of Durvillaea antarctica do in fact travel together tive individuals had significantly higher concentrations of via coalesced holdfasts. Additionally, there is no signifi- total and specific pigments than reproductive individuals, cant biochemical difference between reproductive males although carotenoids for female stipes have similar levels and females when it comes to Chl a, Chl c, total carote- as vegetative stipes (Figure 4). There was no difference in noids, and phlorotannins. These results indicate that the

Puerto Oscuro Totoralillo Sur

1.2 A Total pigments 0.6 B Chlorophyll a

1.0 0.5

0.8 0.4

0.6 0.3

0.4 0.2

0.2 0.1

0.0 0.0

0.16 CDChlorophyll c 0.30 Carotenoids 0.14 0.25 0.12 0.20 0.10

Pigment concentration (mg/g wet wt) 0.08 0.15 0.06 0.10 0.04 0.05 0.02 0.00 0.00 Male Female Vegetative Male Female Vegetative Sexual stage Sexual stage

Figure 4: Durvillaea antarctica: average (mean±SD) concentration (mg/g wet wt) of pigments in blade samples of different sexual stage and locality, Puerto Oscuro and Totoralillo Sur. Different letters above the columns indicate differences between sexual stages significant at p = 0.05. Numbers of stipes from each site and sexual stage are listed at the bottom of each column.

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 69 46 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller

Table 3: Durvillaea antarctica: summary of two-way ANOVAs for holdfast (Smith 2002), potentially containing both sexes. variables (pigments and phlorotannins) according to site (Puerto Also along the Chilean coast, large numbers of D. antarc- Oscuro and Totoralillo Sur) and sexual stage (vegetative, reproduc- tica are floating at any given moment in time (Hinojosa tive male, and reproductive female) of algal samples. et al. 2011), many of which have intact holdfasts (Tala Variable Effect df F p-Value et al. 2013, this study). Of the 1000 recently stranded indi- viduals of D. antarctica with holdfasts, almost half had a Total pigment Site (S) 1;199 11.67 < 0.001 coalesced holdfast with at least two stipes. Thus, the prob- Stipe stage (SS) 2;199 8.97 < 0.001 S × SS 2;199 1.24 0.289 ability of long-distance dispersal is greatly enhanced as Chlorophyll a Site (S) 1;199 10.35 < 0.001 the presence of multiple individuals in a single holdfast Stipe stage (SS) 2;199 9.10 < 0.001 substantially improves the possibility that individuals of S × SS 2;199 1.31 0.272 both sexes are travelling together over great distances, < Chlorophyll c Site (S) 1;199 18.48 0.001 although the overall proportion of holdfasts with mature Stipe stage (SS) 2;199 10.33 < 0.001 S × SS 2;199 0.59 0.555 individuals of both sexes is relatively small (5–17%). Carotenoids Site (S) 1;199 11.69 < 0.001 It is likely that holdfast coalescence in D. antarctica Stipe stage (SS) 2;199 7.89 < 0.001 is responsible for the impressive long-distance dispersal S × SS 2;199 1.27 0.283 of this species throughout the subantarctic region (Fraser Phlorotannins Site (S) 1;42 7.04 < 0.05 et al. 2010b). In fact, other seaweeds that are capable of Stipe stage (SS) 2;42 3.53 < 0.05 long-distance dispersal, such as Macrocystis pyrifera (Lin- S × SS 2;42 1.64 0.205 naeus) Agardh 1820 (González et al. 2015) and Fucus vesic- Significant values (p < 0.05) in bold face. ulosus (Malm and Kautsky 2004), have also been proven to have coalesced holdfasts, contributing to the idea that holdfast coalescence may play a crucial role in the disper- 1.6 Puerto Oscuro Totoralillo Sur sal of certain species, particularly if they are dioecious. 1.4 The small proportion of D. antarctica holdfasts trav- 1.2 elling with both sexes may be further constrained by 1.0 another important condition, the timing of maturity of the 0.8 individuals of both sexes. As D. antarctica tends to mature 0.6 over a rather extensive time period of 3 months between 0.4 May and July (Santelices et al. 1980), it could also be pos- 0.2 sible that not all individuals within a holdfast are fully

Phlorotannin concentration (% dry wt) 0.0 mature at the same time. This relatively long reproductive Male Female Vegetative Sexual stage season and a generally simultaneous release of gametes within each individual limits the probability that male Figure 5: Durvillaea antarctica: average (mean±SD) concentration and female gametes are released at the same time, there- (% dry wt) of phlorotannins in blade samples of different sexual fore diminishing fertilisation probabilities. Furthermore, stage and locality, Puerto Oscuro and Totoralillo Sur. Different it is only during this fertile phase that D. antarctica rafts letters above the columns indicate differences between sexual stages significant at p = 0.05. Numbers of stipes from each site and containing both sexes can contribute to the genetic con- sexual stage are listed at the bottom of each column. nectivity between populations. As the benthic individuals had a higher proportion of reproductive individuals than the stranded ones, which were mainly vegetative, gamete lack of genetic connectivity of D. antarctica populations release might occur before floating individuals return to along the continental coast of Chile is not a result of males the shore thereby further limiting their dispersal potential. and females failing to disperse together.

Biochemical composition Dioecious species, but united in holdfasts The results from Collantes et al. (1997) indicated that, It has been extrapolated that within the entire Southern in Durvillaea antarctica, there might exist a correlation Ocean, over 70 million Durvillaea antarctica kelp rafts between the colour of an individual and its sex. They are afloat at any one time (Smith 2002). Furthermore, reported that lighter, yellower individuals were mostly around 20 million of these rafts are estimated to have a males whereas darker, blacker individuals were generally

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 70 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller 47 females. Such results suggest sexual dimorphism for It is interesting to note that the specimens from Puerto D. antarctica, meaning there are phenotypic differences Oscuro, which had higher pigment concentrations than between the males and females, which would possibly the samples from Totoralillo Sur, were collected from a site be caused by the fact that reproductive females carry mil- that receives very little direct sunlight under a high cliff. lions of large, dark eggs in the conceptacles whereas male Possibly, these bull kelps had adjusted pigment concen- antheridia are small and light. However, the results of the trations to low light conditions as also reported for other current study did not confirm any sex-specific colour dif- species (Colombo-Pallota et al. 2006, Hanelt and Figueroa ferences and did not allow determining whether eventual 2012). As was the case for pigments, there was a signifi- colour difference might be due to the eggs. cant difference in phlorotannin concentrations between Using RGB values to describe the colours, total pig- the sites. However in this case, phlorotannin levels were ments and each individual pigment were found to have an higher in Totoralillo Sur. Increases in phlorotannin con- effect on both the R and G colour values and none on the centrations have been related to changes in PAR and UVR B colour value (Table 1). However, for both R and G, the levels (Swanson and Druehl 2002, Cruces et al. 2013) and effects of pigment concentrations on the colour pattern herbivore pressure (Van Alstyne 1988, Pavia and Toth were relatively minor. It is possible that another pigment 2000). Thus, site-specific environmental conditions cause component could play a more central role in determining variation in pigments and phlorotannin concentrations the colour of D. antarctica. This could particularly be the in D. antarctica and appear to be more important than case for carotenoids, as there are many types of carot- sex-specific differences. This could suggest that kelps enoids and some of them are continuously adjusted as a from some sites are better suited for long-distance rafting result of photoprotective responses (Karsten 2008). As for than those from other sites. In future studies it would be what can be explained by the pigments, it was found that interesting to study other components that could possibly there was a small negative relationship between total pig- have an impact on gamete survival during rafting disper- ments and both the R and G colour values. Visually, this sal, such as carbohydrates or proteins, which might differ means that when an individual has a higher pigment con- between sexes (Jones 1957) and sexual stages (Skriptsova centration, it will be darker. One may therefore assume et al. 2012). that (the darker) females would have higher pigment con- centrations than males. However, this was not the case in our study as reproductive males and females had similar Conclusions and outlook pigment concentrations. Vegetative individuals had a significantly higher con- In this study, it has been demonstrated that reproductive centration of pigments than reproductive individuals, individuals of both sexes of Durvillaea antarctica travel male or female, and both sexes were relatively similar in together in coalesced holdfasts. Furthermore, there is concentration. Though no cause was identified, it is pos- no significant difference between reproductive males sible that vegetative bull kelps, with a full cortex without and females in essential biochemical compounds that colourless conceptacle cells, have higher pigment concen- could result in sex-specific differences in the potential for trations. Phlorotannins are known to serve a protective rafting dispersal. Thus, the phytogeographic pattern of role (Li et al. 2009) and higher concentrations of phlo- D. antarctica along the continental coast of Chile (Fraser rotannins within female tissues may serve to protect the et al. 2010b) does not appear to be due to biological con- gametes of the females. Sexual dimorphism in chemical straints limiting the dispersal potential of this floating concentrations has been demonstrated in other species, kelp. Instead, it seems more likely that the process that such as Gracilaria chilensis (Guillemin et al. 2014). For D. limits the dispersal of species such as D. antarctica is an antarctica, no significant difference was found between ecological or physical one. For instance, density-blocking males and females, which could indicate that both sexes could potentially influence population connectivity in a use phlorotannins equally for protection. Phlorotannins variety of species (Waters et al. 2013, Neiva et al. 2014). are known to act as antioxidants, participate in UVR pro- Also, prevailing winds and currents (e.g. Humboldt tection, and in general defence (Karsten et al. 2009, Wang Current System) can greatly influence the dispersal tra- et al. 2009, Onofrejová et al. 2010, Steinhoff et al. 2012). As jectories of algal rafts (Rothäusler et al. 2015), possibly we found no sex-specific differences in the concentration causing certain areas of the coast to receive less input. of these biochemical compounds it seems that both sexes Future studies should concentrate on studying the impact of D. antarctica are equally fit to survive rafting along the of density-blocking and currents on the population con- coast of central Chile. nectivity of rafting seaweed species.

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 71 48 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller

Acknowledgments: This study was financed by FOND- Friesen, C.R., R.T. Mason, S.J. Arnold and S. Estes. 2014. Patterns ECYT 1131082 to MT and FT, FONDECYT 1131023 to FT and of sperm use in two populations of red-sided garter snake ­(Thamnophis sirtalis parietalis) with long-term female sperm MT, and PhD-fellowship Beca CONICYT-PCHA/Doctorado storage. Can. J. Zoolog. 92: 33–40. Nacional/2014-21140010 to BL. The collaboration of Oscar González, A.V., J. Beltrán, V. Flores and B. Santelices. 2015. Morpho- Pino and Vieia Villalobos in field and laboratory activi- logical convergence in the inter-holdfast coalescence process ties is gratefully acknowledged, as well as that of Julie among kelp and kelp-like seaweeds (Lessonia, Macrocystis, Turgeon, Université Laval, for enabling the internship of Durvillaea). Phycologia 54: 283–291. D L-P. The constructive comments from two anonymous Guillemin, L.M., P. Valenzuela, J.D. Gaitan-Espitia and C. Destombe. 2014. Evidence of reproductive cost in the triphasic life history referees and the editor on the original version of the man- of the red alga Gracilaria chilensis (Gracilariales, Rhodophyta). uscript are also appreciated. J. Appl. Phycol. 26: 569–575. Hanelt, D. and F.L. Figueroa. 2012. Physiological and photomorpho- genic effects of light on marine macrophytes. In: (C. Wiencke and K. Bischof, eds) Seaweed biology. Springer-Verlag Berlin References Heidelberg, Germany. Ecological Studies 219, 1: pp. 3–23. Henley, W.J. and K.H. Dunton. 1995. A seasonal comparison Buschmann, A.H. and A. Bravo. 1990. Intertidal amphipods as of carbon, nitrogen, and pigment content in Laminaria potential dispersal agents of carpospores of Iridaea laminari- solidungula and L. saccharina (Phaeophyta) in the Alaskan oides (Gigartinales, Rhodophyta). J. Phycol. 26: 417–420. arctic. J. Phycol. 31: 325–331. Cheshire, A.C., J.G. Conran and N.D. Hallam. 1995. A cladistics- Hinojosa, I.A., M.M. Rivadeneira and M. Thiel. 2011. Temporal and analysis of the evolution and biogeography of Durvillaea spatial distribution of floating objects in coastal waters of (Phaeophyta). J. Phycol. 31: 644–655. central-southern Chile and Patagonian fjords. Cont. Shelf Res. Collantes, G., R. Riveros and M. Acevedo. 1997. Fenología reproduc- 31: 172–186. tiva de Durvillaea antarctica (Phaeophyta, Durvillaeales) del Inskeep, W.P. and P.R. Bloom. 1985. Extinction coefficients of intermareal de caleta Montemar, Chile central. Rev. Biol. Mar. chlorophyll a and b in N,N-dimethyl-formamide and 80% Oceanogr. 32: 111–116. acetone. Plant Physiol. 77: 483–485. Collantes, G., A. Merino and V. Lagos. 2002. Fenología de la game- Jones, R.F. 1957. Variation of nitrogen and carbohydrate constituents togénesis, madurez de conceptáculos, fertilidad y embriogé- during the development of Himanthalia elongata (L.) S. F. Gray. nesis en Durvillaea antarctica (Chamisso) Hariot (Phaeophyta, Biological Bulletin 112: 81–91. Durvillaeales). Rev. Biol. Mar. Oceanogr. 37: 83–112. Karsten, U. 2008. Defense strategies of algae and cyanobacte- Colombo-Pallota, M., E. García-Mendoza and L. Ladah. 2006. Photo­ ria against solar ultraviolet radiation. In: (C.D. Amsler, ed) synthetic performance, light absorption, and pigment compo- Algal Chemical Ecology. Springer Berlin Heidelberg, Berlin. sition of Macrocystis pyrifera (Laminariales, Phaeophyceae) pp. 273–296. blades from different depths. J. Phycol. 42: 1225–1234. Karsten, U., A. Wulff, M.Y. Roleda, R. Müller, F.S. Steinhoff, Cornelissen, T. and P. Stiling. 2005. Sex-biased herbivory: a meta- J. Fredersdorf and C. Wiencke. 2009. Physiological responses analysis of the effects of gender on plant-herbivore interac- of polar benthic algae to ultraviolet radiation. Bot. Mar. 52: tions. Oikos 111: 488–500. 639–654. Coyer, J.A., G. Hoarau, J.F. Costa, B. Hogerdijk, E.A. Serrao, Kendal, D., C.E. Hauser, G.E. Garrard, S. Jellinek, K.M. Giljohann and E. Billard, M. Valero, G.A. Pearson and J.L. Olsen. 2011. J.L. Moore. 2013. Quantifying plant colour and colour difference Evolution and diversification within the intertidal brown as perceived by humans using digital images. PLoS one 8: macroalgae Fucus spiralis/F. vesiculosus species complex in e72296. the North Atlantic. Mol. Phylogenet. Evol. 58: 283–296. Koivikko, R. 2008. Brown algal phlorotannins: improving and Cruces, E., P. Huovinen and I. Gómez. 2013. Interactive effects of applying chemical methods. Ph. D. thesis, University of Turku, UV radiation and enhanced temperature on photosynthesis, Turku, Finland. phlorotannin induction and antioxidant activities of two sub- Li, Y., Z.J. Qian, B. Ryu, S.H. Lee, M.M. Kim and S.K. Kim. 2009. Antarctic brown algae. Mar. Biol. 160: 1–13. Chemical components and its antioxidant properties in vitro: Delph, L.F. 1999. Sexual dimorphism in life history. In: (M.A. Geber, An edible marine brown alga, Ecklonia cava. Bioorga. Med. T.E. Dawson and L.F. Delph, eds) Gender and sexual dimor- Chem. 17: 1963–1973. phism in flowering plants. Springer-Verlag, Berlin, Germany. Luthringer, R., A. Cormier, S. Ahmed, A.F. Peters, J.M. Cock and pp. 149–173. S.M. Coelho. 2014. Sexual dimorphism in the brown algae. Fraser, C.I., R. Nikula, H.G. Spencer and J.M. Waters. 2009. Kelp Perspect. Phycol. 1: 11–25. genes reveal effects of subantarctic sea ice during the Last Macaya, E.C., S. Boltana, I.A. Hinojosa, J.E. Macchiavello, Glacial Maximum. PNAS 106: 3249–3253. N.A. Valdivia, N.R. Vasquez, A.H. Buschmann, J.A. Vasquez, Fraser, C.I., D.J. Winter, H.G. Spencer and J.M. Waters. 2010a. Multi­ J.M.A. Vega and M. Thiel. 2005. Presence of sporophylls in gene phylogeny of the southern bull-kelp genus Durvillaea floating kelp rafts of Macrocystis spp. (Phaeophyceae) along (Phaeophyceae: Fucales). Mol. Phylogenet. Evol. 57: 1301–1311. the Chilean Pacific coast. J. Phycol. 41: 913–922. Fraser, C.I., M. Thiel, H.G. Spencer and J.M. Waters. 2010b. Con- Malm, T. and L. Kautsky. 2004. Are bladderwrack (Fucus vesiculosus temporary habitat discontinuity and historic glacial ice drive L.) holdfasts that support several fronds composed of one or genetic divergence in Chilean kelp. BMC Evol. Biol. 10:203. several genetic individuals? Aquat. Bot. 80: 221–226.

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 72 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller 49

McKenzie, P.F. and A. Bellgrove. 2008. Dispersal of Hormosira Vandendriessche, S., M. Vincx and S. Degraer. 2007. Floating banksii (Phaeophyceae) via detached fragments: Reproductive seaweed and the influences of temperature, grazing and clump viability and longevity. J. Phycol. 44: 1108–1115. size on raft longevity – A microcosm study. J. Exp. Mar. Biol. Neiva, J., J. Assis, F. Fernandes, G.A. Pearson and E.A. Serrao. 2014. Ecol. 343: 64–73. Species distribution models and mitochondrial DNA phylo- Vergés, A., N.A. Paul and P.D. Steinberg. 2008. Sex and life-history geography suggest an extensive biogeographical shift in the stage alter herbivore responses to a chemically defended red high-intertidal seaweed Pelvetia canaliculata. J. Biogeogr. 41: alga. Ecology 89: 1334–1343. 1137–1148. Vieira, S., M. Biscoito, H. Encarnacao, J. Delgado and T.W. Pietsch. Onofrejová, L., J. Vasickova, B. Klejdus, P. Stratil, L. Misurcova, 2013. Sexual parasitism in the deep-sea ceratioid anglerfish S. Kracmar, J. Kopecky and J. Vacek. 2010. Bioactive phenols Centrophryne spinulosa Regan and Trewavas (Lophiiformes: in algae: The application of pressurized-liquid and solid-phase Centrophrynidae). Copeia 4: 666–669. extraction techniques. J. Pharmaceut. Biomed. 51: 464–470. Wang, T., R. Jonsdottir and G. Olafsdottir. 2009. Total phenolic Pavia, V.J. and G.B. Toth. 2000. Inducible chemical resistance compounds, radical scavenging and metal chelation of extracts to herbivory in the brown seaweed Ascophyllum nodosum. from Icelandic seaweeds. Food Chem. 116: 240–248. Ecology 81: 3212–3225. Waters, J.M., C.I. Fraser and G.M. Hewitt. 2013. Founder takes all: R Development Core Team. 2015. R: A language and environment for density-dependent processes structure biodiversity. Trends statistical computing. R Foundation for Statistical Computing, Ecol. Evol. 28: 78–85. Vienne, Austria. URL: http://www.R-project.org/. Zar, J.H. 2010. Biostatistical analysis. 5th Edition. Pearson Prentice- Rothäusler, E., I. Gomez, I.A. Hinojosa, U. Karsten, F. Tala and Hall, Upper Saddle River (NJ), pp. 944. M. Thiel. 2009. Effect of temperature and grazing on growth Zhang, Y.H., L. Tang, X.J. Lfu, L.L. Liu, W.X. Cao and Y. Zhu. 2014. and reproduction of floating Macrocystis spp. (Phaeophyceae) Modeling dynamics of leaf color based on RGB value in rice. along a latitudinal gradient. J. Phycol. 45: 547–559. Journal Integr. Agr. 13: 749–759. Rothäusler, E., I. Gomez, I.A. Hinojosa, U. Karsten, F. Tala and M. Thiel. 2011. Physiological performance of floating giant kelp Macrocystis pyrifera (Phaeophyceae): Latitudinal variability in the effects of temperature and grazing. J. Phycol. 47: 269–281. Bionotes Rothäusler, E., H. Correll and V. Jormalainen. 2015. Abundance and dispersal trajectories of floating Fucus vesiculosus in the Dominic Lizée-Prynne Northern Baltic Sea. Limnol. Oceanogr. 60: 2173–2184. Facultad Ciencias del Mar, Universidad Santelices, B., J. Cancino, J.C. Castilla and P. Schmiede. 1980. Católica del Norte, Larrondo 1281, Comparative ecology of Lessonia nigrescens and Durvillaea Coquimbo, Chile; and Département de antarctica (Phaeophyta) in central Chile. Mar. Biol. 59: 119–132. Biologie, Université Laval, Québec, 1045, Shine, R., M.J. Elphick, P.S. Harlow, I.T. Moore, M.P. LeMaster and Avenue de la Medecine, Québec, QC G1V R.T. Mason. 2001. Movements, mating, and dispersal of red- 0A6, Canada sided gartersnakes (Thamnophis sirtalis parietalis) from a communal den in Manitoba. Copeia 1: 82–91. Skriptsova, A.V., N.M. Shevchenko, D.V. Tarbeeva and Dominic Lizée-Prynne is a young biologist who finished her Biology T.N. Zvyagintseva. 2012. Comparative study of polysaccharides degree at Université Laval in Canada. This scientific paper is her from reproductive and sterile tissues of five brown seaweeds. first published article and was done as an internship project with Mar. Biotechnol. 14: 304–311. BEDIM at Universidad Católica del Norte in Coquimbo, Chile. She Smith, S.D.A. 2002. Kelp rafts in the Southern Ocean. Global Ecol. plans on continuing a career in research in the near future. Biogeogr. 11: 67–69. Steinhoff, F.S., M. Graeve, K. Bartoszek, K. Bischof and C. Wiencke. Boris López 2012. Phlorotannin production and lipid oxidation as a Facultad Ciencias del Mar, Universidad potential protective function against high photosynthetically Católica del Norte, Larrondo 1281, active and UV radiation in gametophytes of Alaria esculenta Coquimbo, Chile; and Departamento de (Alariales, Phaeophyceae). Photochem. Photobiol. 88: 46–57. Acuicultura y Recursos Agroalimentarios, Swanson, A.K. and L.D. Druehl. 2002. Induction, exudation and the UV Universidad de Los Lagos Avenida protective role of kelp phlorotannins. Aquat. Bot. 73: 241–253. Fuchslocher 1305, Osorno, Chile Tala, F., I. Gomez, G. Luna-Jorquera and M. Thiel. 2013. Morphological, physiological and reproductive conditions of rafting bull kelp (Durvillaea antarctica) in northern-central Boris López is a marine biologist at the Universidad de Los Lagos, Chile (30°S). Mar. Biol. 160: 1339–1351. southern Chile. He is currently a PhD student of the Applied Thornber, C.S. 2006. Functional properties of the isomorphic Biology and Ecology program at the Universidad Católica del biphasic algal life cycle. Integr. Comp. Biol. 46: 605–614. Norte, Chile. His research interests are related to rafting macroal- Van Alstyne, K.L. 1988. Herbivore grazing increases polyphenolic gae and invertebrates along the coast of Chile. Prior experience defenses in the brown alga Fucus distichus. Ecology 69: includes culture of marine invertebrates and technology transfer 655–663. in aquaculture.

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 73 50 D. Lizée-Prynne et al.: Dispersal potential of a dioecious long-distance traveller

Fadia Tala Martin Thiel Facultad Ciencias del Mar, Universidad Facultad Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Católica del Norte, Larrondo 1281, Coquimbo, Chile; and Centro de Coquimbo, Chile; Millennium Nucleus Investigación y Desarrollo Tecnológico en Ecology and Sustainable Management of Algas de la Universidad Católica del Norte Oceanic Island (ESMOI), Larrondo 1281, (CIDTA-UCN), Larrondo 1281, Coquimbo, Coquimbo, Chile; and Centro de Estudios Chile Avanzados en Zonas Áridas (CEAZA), Larrondo 1281,Coquimbo, Chile, Fadia Tala obtained her degree in Marine Biology and her MS [email protected] from the Universidad Católica del Norte (Chile). She continues her research at the same university, including research into brown mac- Martin Thiel is Professor of Marine Biology at Universidad Católica roalgal reproduction and ecophysiology as response to abiotic and del Norte in Coquimbo, Chile, with broad interests in diverse aspects biotic stress under benthic and rafting condition. She also works of marine biology, including behavioural ecology, rafting dispersal on management and culture techniques of seaweeds. Her PhD in and biogeography. Together with his students, collaborators and Botany was obtained at the University of Sao Paulo, Brazil, in 2013. friends, he is directing the national research network “Cientificos de At present she is an Assistant Professor and Director of the Center la Basura” (Litter Scientists), in which schoolkids from all over Chile for Research and Technological Development in Algae (CIDTA-UCN) investigate the problem of marine litter, in a quest to identify causes at the Faculty of Marine Science (UCN). and find solutions (www.cientificosdelabasura.cl).

Brought to you by | De Gruyter / TCS Authenticated | [email protected] Download Date | 2/9/16 2:23 AM 74

Chapter 4

Epibiont communities on stranded kelp rafts of Durvillaea antarctica (Fucales, Phaeophyceae) – do positive interactions facilitate range extensions?

(manuscript published in J. Biogeogr. 45: 1833-1845, 2018)

Boris A. López1,2, Erasmo C. Macaya3,4,5,Marcelo M. Rivadeneira 1,6,7, Fadia Tala1,8, Florence Tellier9,10 and Martin Thiel1,4,6*

1Departamento de Biología Marina, Facultad de Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 2Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos, Avenida Fuchslocher 1305, Osorno, Chile. 3Departamento de Oceanografía, Universidad de Concepción, Casilla 160-C, Concepción, Chile. 4Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile. 5Centro FONDAP de Investigaciones en Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Valdivia, Chile 6Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Coquimbo, Chile. 7Departamento de Biología, Universidad de La Serena, La Serena, Chile. 8Centro de Investigación y Desarrollo Tecnológico en Algas (CIDTA), Facultad de Ciencias del Mar, Universidad Católica del Norte, Larrondo 1281, Coquimbo, Chile. 9Departamento de Ecología, Facultad de Ciencias, Universidad Católica de la Santísima Concepción, Casilla 297, Concepción, Chile. 10Centro de Investigación en Biodiversidad y Ambientes Sustentables (CIBAS), Universidad Católica de la Santísima Concepción, Casilla 297, Concepción, Chile.

*Corresponding author: [email protected]

75

DOI: 10.1111/jbi.13375

RESEARCH PAPER

Epibiont communities on stranded kelp rafts of Durvillaea antarctica (Fucales, Phaeophyceae)—Do positive interactions facilitate range extensions?

Boris A. Lopez 1,2 | Erasmo C. Macaya3,4,5 | Marcelo M. Rivadeneira1,6,7 | Fadia Tala1,8 | Florence Tellier9,10 | Martin Thiel1,4,6

1Departamento de Biologıa Marina, Facultad de Ciencias del Mar, Universidad Abstract Catolica del Norte, Coquimbo, Chile Aim: This study examines how rafting on floating bull kelps can shape the biogeo- 2Departamento de Acuicultura y Recursos graphic patterns of raft-associated species, and analyses the spatio-temporal variabil- Agroalimentarios, Universidad de Los Lagos, Osorno, Chile ity of taxonomic richness and co-occurrences of epibionts on beach-cast rafts of 3Departamento de Oceanografıa, Durvillaea antarctica along a latitudinal gradient. Universidad de Concepcion, Concepcion, ~ Chile Location: Southeast Pacific, along 1,700 km of coastline. 4Millennium Nucleus Ecology and Methods: We examined the epibionts on stranded individuals of D. antarctica on 33 Sustainable Management of Oceanic Island beaches along the continental coast of Chile (28° S–42° S) within four biogeographic (ESMOI), Coquimbo, Chile – 5Centro FONDAP de Investigaciones en districts during the winter and summer of two years (2014/2015 2015/2016). Tax- Dinamica de Ecosistemas Marinos de Altas onomic richness and co-occurrences of epibionts within a holdfast were examined. Latitudes (IDEAL), Valdivia, Chile Known geographic ranges and rafting ranges were compared to determine possible 6Centro de Estudios Avanzados en Zonas Aridas (CEAZA), Coquimbo, Chile range expansions via rafting dispersal. 7Departamento de Biologıa, Universidad de Results: Sessile species were the most frequent epibionts. Taxonomic richness var- La Serena, La Serena, Chile ied among biogeographic zones and seasons, and was higher between 33° S and 8Centro de Investigacion y Desarrollo ° ° ° Tecnologico en Algas (CIDTA), Facultad de 42 S than between 28 S and 33 S, particularly in summer compared to winter. Ciencias del Mar, Universidad Catolica del Taxonomic richness decreased with floating time (indicated by the presence and size Norte, Coquimbo, Chile of Lepas spp.). Habitat-forming epibionts such as mytilid mussels, the polychaete 9Departamento de Ecologıa, Facultad de Ciencias, Universidad Catolica de la Phragmatopoma moerchi and the seaweeds Gelidium lingulatum and Lessonia spicata ı Sant sima Concepcion, Concepcion, Chile favoured co-occurrences of other species within a holdfast, suggesting a habitat cas- 10 Centro de Investigacion en Biodiversidad — — y Ambientes Sustentables (CIBAS), cade (i.e. biogenic holdfast sessile eco-engineers other epibionts), while the bor- Universidad Catolica de la Santısima ing isopod Limnoria chilensis and the excavating limpet Scurria scurra were negatively Concepcion, Concepcion, Chile correlated with many other species. Some rafting epibiont species with low dispersal Correspondence ability were found more than 100–300 km outside of their known geographic Martin Thiel, Departamento de Biologıa Marina, Facultad de Ciencias del Mar, ranges, with more extensive ephemeral range extensions at the southern edge of Universidad Catolica del Norte, Larrondo their respective ranges, probably facilitated by higher availability of rafts in those 1281, Coquimbo, Chile. Email: [email protected] areas. Main Conclusions: These findings confirm that raft-associated species are fre- Editor: Jonathan Waters quently dispersed outside their known geographic ranges, although these range extensions vary strongly depending on the availability and persistence of rafts, and on the biotic interactions within the rafting assemblage.

| Journal of Biogeography. 2018;45:1833–1845. wileyonlinelibrary.com/journal/jbi © 2018 John Wiley & Sons Ltd 1833 76 1834 | LOPEZ ET AL.

KEYWORDS floating seaweeds, geographic range, habitat cascade, marine biogeography, rafting dispersal, range extension

1 | INTRODUCTION (i.e. with direct development or low mobility) but found frequently on floating seaweeds often have wider ranges than can be explained Many molecular studies of epibiont species on floating seaweeds by their intrinsic dispersal capabilities. In line with these observa- have indicated population connectivity through long-distance rafting tions, the current geographic ranges of rafting epibionts are hypoth- (e.g. Nikula, Fraser, Spencer, & Waters, 2010; Nikula, Spencer, & esised to depend on (a) the dispersal ability of the floating Waters, 2013). However, there is little empirical information about substratum and its availability in coastal areas, (b) the biological the rafting process itself and about the organisms capable of endur- interactions among epibionts and (c) their biological traits favouring ing extensive voyages, even though this information is necessary to long-distance rafting journeys (Macaya, Lopez, Tala, Tellier, & Thiel, evaluate the importance of this dispersal mechanism in shaping the 2016; Thiel & Gutow, 2005). geographic ranges of rafting species. Herein we used the southern bull kelp Durvillaea antarctica (Cha- Seaweeds are usually attached to rocks with a holdfast structure. misso) Hariot 1892 as a model-organism to examine the possible Some seaweeds have large holdfasts that can serve as habitat for small effect of rafting dispersal on the geographic ranges of epibiont spe- species (e.g. Thiel & Vasquez, 2000). Strong storms or herbivores can cies along a latitudinal gradient. Durvillaea antarctica has a wide dis- cause seaweeds to detach from the rock and some species with high tribution in the Southern Hemisphere, mainly in sub-Antarctic waters buoyancy, mainly brown algae, can remain afloat for a long time, sup- (Fraser, Thiel, Spencer, & Waters, 2010). In Chile, it is present from porting a diverse associate community (Thiel & Gutow, 2005). After 30° S to Cape Horn (56° S), inhabiting rocky areas, exposed or detachment of floating seaweeds from the primary substratum, the semiexposed to waves, in the lower intertidal and shallow subtidal number of associated organisms can drastically decrease, particularly zone (Hoffmann & Santelices, 1997). It is a species with positive that of mobile ones (e.g. Gutow, Gimenez, Boos, & Saborowski, 2009; buoyancy that has 1–2 m of average length, and can reach up to Miranda & Thiel, 2008), and subsequently epibiont communities tend 10 m and more than 20 kg of wet biomass (Hoffmann & Santelices, to get homogenised while afloat (Gutow, Beermann, Buschbaum, Riva- 1997), with high rafting potential at high latitudes (Tala, Penna-Dıaz, deneira, & Thiel, 2015) because only organisms capable of competing Luna-Jorquera, Roth€ausler, & Thiel, 2017). Many different associate and exploiting resources on the raft (mainly space and food) are cap- species inhabit the coalesced holdfasts of benthic individuals (San- able to persist during long journeys (Thiel & Gutow, 2005). For exam- telices, Castilla, Cancino, & Schmiede, 1980). ple, Stoner and Greening (1984) reported low variability in the Rafts of D. antarctica frequently wash up on sandy beaches epibiont community cohabiting on pelagic Sargassum, according to raft within its geographic range (Duarte, Jaramillo, Contreras, Acuna,~ & age. Interactions among epibionts during rafting journeys are thus Navarro, 2009), and stranded biomass tends to vary strongly among thought to determine the species that survive long trips and can reach biogeographic districts (Lopez, Macaya, Tala, Tellier, & Thiel, 2017), new habitats (Thiel & Gutow, 2005). The size of seaweed rafts also being highest in the northern-central (30° S–33° S) and southern- has a strong influence on the density and species richness of the epi- most districts (37° S–42° S). Stranded specimens are also frequently biont community, because large rafts can support more travellers found outside their northern geographic limit (<30° S), showing indi- (Thiel & Gutow, 2005; Vandendriessche, Vincx, & Degraer, 2007). cations of prolonged floating times (Lopez et al., 2017). This suggests While some studies have examined rafting assemblages at sea, in order that the connectivity of D. antarctica populations (Fraser et al., to assess the contribution of rafting dispersal to population connectiv- 2010) and eventually also of associated epibionts is higher in some ity and structuring of geographic ranges, it is necessary to identify the areas than in others, particularly in zones with high availability of species that return to the coast, which was done herein by using rafts and less stressful conditions at the sea surface. beach-cast rafts as a proxy for successful rafting events on adjacent Herein we tested the following hypotheses: (a) species richness rocky shores. of epibionts on D. antarctica rafts increases along a latitudinal gradi- The extent of geographic ranges of marine benthic species is ent because rafting dispersal opportunities increase in areas where delimited by their physiological tolerances, biological interactions and environmental conditions at the sea surface are less severe (e.g. at their potential dispersal capabilities (e.g. Lester, Ruttenberg, Gaines, high latitudes), (b) species richness of epibionts decreases with & Kinlan, 2007), either due to intrinsic (e.g. the presence of plank- increasing floating time of stranded rafts, (c) range extensions of epi- tonic larvae or adult mobility) and extrinsic factors (e.g. availability of biont species are more common in the southern parts of the study floating substrata, currents and local winds). Therefore, the dispersal area, where supplies of floating seaweeds are higher, (d) epibiont ability of a species should be related to its geographic range. Never- species are more frequently observed outside their geographic theless, epibiont species with low autonomous dispersal potential ranges on rafts with long floating times than on rafts with no

77 LOPEZ ET AL. | 1835 indication of floating or less floating time and (e) epibionts are more A total of 5,219 complete plants were measured during the frequently observed outside of their geographic ranges on rafts with study. For each complete plant of D. antarctica, the following vari- species that favour co-occurrences. In order to test these hypothe- ables were measured: ses, we determined the spatio-temporal variability of species rich- (BH) Biomass of holdfast: the wet weight of the holdfast was ness and co-occurrences of epibionts on holdfasts of recently measured with a portable electronic hanging digital scale of 1 g stranded D. antarctica, according to biogeographic districts of the accuracy. Chilean coast (28° S–42°S) and floating times of stranded specimens. (FT) Floating time: for each plant, we determined whether it This information is used to evaluate the role of rafting dispersal by had been colonised by stalked barnacles of the genus Lepas or floating seaweeds as a modulator of the geographic ranges of ben- not. The presence and size (i.e. capitular length) of these organ- thic species commonly found on rafts. isms can indirectly indicate the floating time of a substratum (Thiel & Gutow, 2005). Cyprid larvae and adult specimens of Lepas spp. were visible to the naked eye and verified from 2 | MATERIALS AND METHODS fronds and holdfasts of stranded bull kelps. According to the presence and size of stalked barnacles, all D. antarctica rafts 2.1 | Characteristics of the study area were categorised in three groups: (a) short floating time (<2days) The continental coast of Chile (from approximately 28° Sto42° S) is —plants without any Lepas; (b) intermediate floating time (2– characterized by a linear topography with a north–south orientation 10 days)—plants with cyprid recruits or small, juvenile Lepas and the absence of distinct geographic barriers to dispersal of marine (<5 mm capitular length); and (c) long floating time (>10 days)— organisms (Camus, 2001). This area has a strong latitudinal gradient plants with large, adult Lepas (≥5 mm capitular length). For some of sea surface temperature within the Humboldt Current System specific analyses (see below in Comparison of geographic ranges), (Camus, 2001). and due to the observed low frequency of rafts with large Lepas This study was conducted on 33 sandy and pebble/cobble bea- spp., the intermediate and long floating time categories were ches (28° S–42° S) across the benthic and pelagic geographic range grouped, and only two categories were considered (absence and of D. antarctica (Figure 1). Distances between adjacent beaches var- presence of Lepas spp.) (for details see Supporting Information ied from 30 to 100 km and the extension of the stretches that were Appendix S2). surveyed on each beach ranged from 0.28 to 11.08 km, depending on beach length and/or amounts of stranded bull kelps (for details 2.3 | Taxonomic richness of epibionts see Lopez et al., 2017). Beaches were distributed across the four biogeographic districts, based on Camus (2001) and a previous study The presence of epibiont species on holdfasts was registered for (Lopez et al., 2017): the southern edge of the Peruvian Province, each stranded bull kelp, checking both sides of each holdfast. Some 28° S–30° S, hereafter termed Coquimbo-Choros District (CCD); species that were difficult to determine in the field were taken Septentrional District (SED), 30° S–33° S; Mediterranean District back to the lab for detailed species identification. Because we (MED), 33° S–37° S; Meridional District (MD), 37° S–42° S (Figure 1, recorded many taxa at different levels of resolution (i.e. species, see also Supporting Information Tables S1.1 and S1.2). A recent genus, order), we herein used the term “taxonomic richness” to study showed that the stranding dynamics of D. antarctica vary represent putative species richness. For each D. antarctica plant, strongly between these biogeographic districts (Lopez et al., 2017); the number of epibiont species was calculated and considered as added to the genetic structure shown for this area by Fraser et al. the taxonomic richness at the plant level. Mean values and stan- (2010) (i.e. high genetic differentiation between nearby populations, dard deviations of taxonomic richness for total epibionts, only ses- 100–200 km), this suggests that rafting-mediated connectivity of sile and only mobile epibionts were calculated according to populations of D. antarctica and its epibionts might also differ biogeographic district, year, season and floating time category, between these districts. using the PRIMER v7 software (Clarke, Gorley, Somerfield, & War- wick, 2014). Because the observed taxonomic richness of epibionts may 2.2 | Sampling and morphological measurements of be influenced by sampling effort and abundance of stranded bull stranded Durvillaea antarctica kelps found on each beach, taxonomic richness was also Recently stranded individuals of D. antarctica were collected on each estimated per beach using the Chao 2 index (hereafter named beach during winters and summers in two consecutive years (2014/ “accumulated taxonomic richness”) (Chao & Lee, 1992), 2015 and 2015/2016). For the purpose of this study, only intact calculating an average value and estimated standard deviation in individuals with both fronds and holdfasts were considered. Like- the case of biogeographic district, year, season and floating time wise, because D. antarctica may present holdfasts with multiple category, using the ‘vegan’ package (Oksanen et al., 2017) in R stipes (each stipe corresponds to a different individual, Gonzalez, 3.4, (R Development Core Team, 2017). Details of the statistical Beltran, Flores, & Santelices, 2015), for practical purposes hereafter analyses can be found in Supporting Information Appendix a stranded specimen will be called a “plant”. S2.

78 1836 | LOPEZ ET AL.

FIGURE 1 Geographic distribution of sampling sites and biogeographic districts described for the coast of Chile described in this study (Coquimbo-Choros District: 28° S–30° S, Septentrional District: 30° S– 33° S, Mediterranean District: 33° S–37° S, Meridional District: 37° S–42° S). The geographic distribution of Durvillaea antarctica within the study area is also indicated. The number of beaches sampled within each biogeographic district is shown. The size of grey circles represents the sampled individuals of D. antarctica in each biogeographic district

presence of others) and random co-occurrences (i.e. unrelated spe- 2.4 | Co-occurrence of epibiont species cies) (Veech, 2014), using the package ‘cooccur’ (Griffith, Veech, & To analyse the co-occurrence of species within a single holdfast of Marsh, 2016) in R 3.4 (R Development Core Team, 2017). stranded individuals of D. antarctica, we worked with a subset of 28 species (see Supporting Information Tables S4.1 and S4.2) that pre- 2.5 | Comparison of geographic ranges sented a minimum of 20 records within all plants sampled through- out the study. With these, the presence/absence matrices of From all epibiont species recorded on stranded bull kelps, 35 species epibiont species were constructed. Of a total of 300 pair combina- were chosen for the range analysis that satisfied the following crite- tions considered, 77 pairs (25.6%) were removed from the analysis ria: (a) identification to species level, (b) more than four records because expected co-occurrence was <1, and thus only 223 pairs among the total of all stranded bull kelp rafts and (c) clear latitudinal were analysed. The species co-occurrence matrix was calculated, delimitation of geographic range described in the specialized litera- showing positive co-occurrences (i.e. species tending to occur ture (Supporting Information Tables S3.1, S4.1 and S4.2). For these together), negative co-occurrences (i.e. species interfering with the 35 species, a literature review was conducted to determine their

79 LOPEZ ET AL. | 1837 currently known geographic ranges along the Chilean coast, using taxonomic richness of all epibionts observed per plant of D. antarc- specialized references for seaweeds and invertebrates (see more tica differed significantly between seasons, and was higher in sum- details in Supporting Information Table S3.1). mer compared to winter (p < 0.05), but no differences were For each species, two types of ranges were considered: (a) the observed for other factors or interactions (Figure 2a and Supporting distribution range based on literature (hereafter called “literature Information Figure S4.1a, Table S4.3). A significant relationship range”) and (b) the range determined by the presence on beach-cast between taxonomic richness of epibionts and the holdfast wet bull kelps along the surveyed beaches (hereafter called “rafting weight of stranded individuals was observed (Supporting Information range”); both ranges (literature and rafting) were expressed as latitu- Figure S4.2). Accumulated taxonomic richness (Chao 2 index) varied dinal bins of 1°. significantly among biogeographic districts, years and seasons For the comparison of geographic ranges, species were grouped (p < 0.05), with higher taxonomic richness in high-latitude districts in three range categories: (a) species that presented rafting ranges (CCD

3 | RESULTS

A total of 89 epibiont taxa were recorded on holdfasts of stranded D. antarctica. Of the species found, 86.5% were sessile and 13.5% were FIGURE 2 Average (mean Æ SD) taxonomic richness of epibionts mobile epibionts. Also, 71.9% were seaweeds and 28.1% were inverte- attached on stranded individuals of Durvillaea antarctica on beaches brates. Within the seaweed group, 48 were Rhodophyta, 10 Phaeo- from the continental coast of Chile (28° S–42° S), according to phyceae and 6 Chlorophyta, and several unidentified crustose marine biogeographic districts (CCD: Coquimbo-Choros District; calcareous algae (Supporting Information Table S4.1). With respect to SED: Septentrional District; MED: Mediterranean District; MD: invertebrates, the most common taxonomic groups were Cnidaria, Meridional District). (a) Taxonomic richness at plant level and accumulated per beach in the different districts (Chao 2 index). (b) Bryozoa, Mollusca, Annelida and Crustacea (Supporting Information Average taxonomic richness of sessile and mobile epibionts. Letters Table S4.2, for more details see Supporting Information Appendix S4). (a-b-c-d) above the columns indicate significant differences between biogeographic districts (p < 0.05). Letters in italics correspond to the 3.1 | Taxonomic richness results of accumulated taxonomic richness (in a) or mobile species (in b). Number of plants (at plant level) and beaches (total accumulated) Taxonomic richness of epibionts ranged from 0 to 8 species per from each biogeographic district are listed at the bottom of each plant of D. antarctica, with an average of 2.6 Æ 1.4 species. The column

80 1838 | LOPEZ ET AL.

Supporting Information Table S4.3). Higher average taxonomic rich- time increased (p < 0.05), while there were no differences in the ness of sessile epibionts per plant was observed in the MED and case of mobile epibionts (Figure 3b). MD compared to CCD and SED (p < 0.05) (Figure 2b), as well as in summer compared to winter (p < 0.05) (Supporting Information Fig- 3.2 | Co-occurrence of epibiont species ure S4.1b). No significant differences were observed in average taxo- nomic richness of mobile epibionts (Figure 2b and Supporting There were 40.4% nonrandom co-occurrences within the total pairs Information Figure S4.1b, Table S4.3). of combinations analysed, of which 56% (22.6% out of the total) Average taxonomic richness of total epibionts per plant varied were classified as positive and 44% (17.8% out of the total) were according to the floating time of beach-cast rafts (Figure 3a and negative co-occurrences. The proportion of positive and negative co- Supporting Information Table S4.4). Taxonomic richness of epibionts occurrences, considering only nonrandom interactions, was not dif- decreased with increasing floating time of rafts (p < 0.05) (Fig- ferent than that expected by chance (v2 = 1.14, p = 0.211). In partic- ure 3a). This same pattern was observed with the accumulated taxo- ular, species such as the bivalves Semimytilus algosus and Perumytilus nomic richness of epibionts (Chao 2 index) (Figure 3a). Average purpuratus, the polychaete Phragmatopoma moerchi, acorn barnacles taxonomic richness of sessile epibionts also varied according to the and the seaweeds Gelidium lingulatum and Lessonia spicata had the floating time of the individuals (Supporting Information Table S4.4), highest positive co-occurrences with other epibiont species on showing that taxonomic richness tended to decrease as flotation stranded individuals of D. antarctica (Figure 4). These species are sessile taxa with complex structural morphology that usually appear in dense aggregations of multiple individuals, and are therefore important habitat engineers. On the other hand, the isopod Limnoria chilensis, the limpet Scurria scurra, crustose calcareous algae and articulated coralline algae had negative co-occurrences with many other epibionts (Figure 4).

3.3 | Comparison of geographic ranges

Twenty-five of the associated species presented rafting ranges over- lapping with their literature ranges (RO). Of these species, most were seaweeds with wide literature ranges (e.g. Macrocystis pyrifera, Cera- mium virgatum and Mazzaella laminarioides), while eight species were small invertebrates, mainly molluscs (e.g. S. scurra, P. purpuratus, Brachidontes granulata, S. algosus, Hiatella solida) (Figure 5a). Ten spe- cies were found outside their literature ranges, with seaweed epi- bionts predominating over invertebrates. Seven species were found on beaches south of their known literature ranges (SE), reporting in some cases >200–300 km range extension, such as the seaweeds Antithamnion densum, Gelidium chilense and Chaetomorpha firma and the gastropod Dendropoma mejillonensis (Figure 5b). On the other hand, three species were detected on beaches further north than their reported literature ranges (NE), most of them with range extensions of ~100–150 km, although the amphipod Parawaldeckia kidderi was found more than 300 km to the north of its literature range (Fig- ure 5c). In general, epibiont species on D. antarctica tended to increase towards the southern zone of the study area (37° S–42° S) for all FIGURE 3 Æ Average (mean SD) taxonomic richness of epibionts range categories, RO, SE and NE (Supporting Information Figure S4.3). attached on stranded individuals of Durvillaea antarctica on beaches The frequencies of D. antarctica rafts with epibionts outside their from the continental coast of Chile (28° S–42° S), according to floating time (short, <2 days; intermediate, 2–10 days and long, benthic ranges differed according to the absence/presence of Lepas >10 days). (a) Taxonomic richness at plant level and accumulated per spp. for SE species (Pseudo-R2 = 0.325; d.f. = 1;12; p = 0.039), floating time (Chao 2 index). (b) Average taxonomic richness of being higher with the presence of Lepas spp. than without them. sessile and mobile epibionts. Letters (a-b-c) above the columns However, in the case of NE species, there were no differences indicate significant differences between floating time categories between rafts with the presence and absence of Lepas spp. (Pseudo- (p < 0.05). Letters in italics correspond to the results of accumulated 2 = = = taxonomic richness (in a) or mobile species (in b). Number of plants R 0.091; d.f. 1;4; p 0.692) (Figure 6). (at plant level) and beaches (total accumulated) from each floating Frequencies of raft-associated species on stranded D. antarctica time category are listed at the bottom of each column found outside of their literature ranges were higher in the presence

81 LOPEZ ET AL. | 1839

FIGURE 4 Species co-occurrence matrix of frequent epibionts attached on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28° S–42° S), according to positive, negative and random species co-occurrences. The numbers of positive/ negative co-occurrences for each species are shown

of epibionts that favour co-occurrences than in the absence of these 4.1 | Taxonomic richness of the epibiont species, for both SE and NE species (Figure 7a,b, Supporting Infor- communities mation Table S4.5). On the other hand, the percentage of rafts with SE species decreased with the presence of epibiont species that lim- The number of epibiont species that were transported by floating ited co-occurrences (Limnoria chilensis and Scurria scurra) but this specimens of D. antarctica varied at the plant-level. In general, after was not the case for NE species (Figure 7c,d, Supporting Information detachment from the primary substratum, the amount of epibionts Table S4.5). tends to decrease drastically because mobile species with low adhe- sion capacity are not able to hold onto floating seaweeds or actively evacuate (Gutow et al., 2009, 2015), and mostly sessile species per- 4 | DISCUSSION sist on rafts (e.g. stalked barnacles, hydrozoans, bryozoans and sea- weeds; Thiel & Gutow, 2005). This coincides with the findings of our Taxonomic richness of the epibiont community on D. antarctica study (i.e. low number of epibionts per holdfast, mainly sessile spe- rafts varied along the southern-central coast of Chile, particularly cies), although diverse groups have been reported in other studies of among biogeographic districts, coinciding with the genetic structure floating seaweeds during the pelagic stage (e.g. large fronds of Asco- of the southern bull kelp described for this area (Fraser et al., phylum, Gutow et al., 2009; Sargassum, Gutow et al., 2015; for an 2010). Epibionts presented positive and negative co-occurrences overview see also Thiel & Fraser, 2016). with other species, suggesting that these interactions might influ- At the plant-level, total taxonomic richness of epibionts was ence the outcome of long-distance journeys. The results also con- higher in summer compared to winter. Abundances of rafting organ- firmed that floating bull kelps can carry epibiont species outside of isms fluctuate more seasonally at high latitudes compared to lower their known literature ranges, apparently with higher dispersal latitudes, increasing in summer compared to winter (Thiel & Gutow, opportunities in areas with more abundant rafts. This suggests that 2005). On the other hand, the CCD (28° S–30° S) is the northern- floating seaweeds and the corresponding probability of rafting dis- most zone that surpasses the northern edge of the geographic range persal can influence the geographic ranges of diverse associated of D. antarctica (~30° S), where lower biomasses and longer floating species. times of specimens have been observed (Lopez et al., 2017; Tala,

82 1840 | LOPEZ ET AL.

FIGURE 5 Comparison of literature and rafting ranges of epibiont species attached on stranded bull kelp Durvillaea antarctica on beaches from the continental coast of Chile (28° S–42° S). (a) Species with rafting ranges within their literature ranges (RO). (b) Species with rafting ranges that surpass the southern edges of their literature ranges, southward extension (SE). (c) Species with rafting ranges that surpass the northern edges of their literature ranges, northward extension (NE). Sessile and mobile species are indicated with triangles and circles respectively. Study area (light grey area) and northern edge of the distribution range of D. antarctica (dark grey line) are also indicated

Gomez, Luna-Jorquera, & Thiel, 2013), which can explain the low number of epibionts at plant level. Sessile epibionts and accumulated richness increased with lati- tude. These latitudinal trends reflect the biogeographic patterns reported for seaweeds (Santelices & Marquet, 1998) and several invertebrate taxa, such as molluscs (Valdovinos, Navarrete, & Mar- quet, 2003), peracarids (Fernandez, Astorga, Navarrete, Valdovinos, & Marquet, 2009; Rivadeneira, Thiel, Gonzalez, & Haye, 2011), and polychaetes (Hernandez, Moreno, & Rozbaczylo, 2005). Hence, the higher taxonomic richness of epibionts on stranded D. antarctica from southern districts can be explained with the increasing taxo- nomic richness of benthic biota (mainly sessile) towards the south of our study area. The decreasing epibiont diversity with increasing floating time of FIGURE 6 Box plot of percentages of stranded individuals of D. antarctica rafts suggests that few species are able to withstand Durvillaea antarctica with epibionts outside their literature ranges according to the absence and presence of Lepas spp. for SE long periods afloat. In general, species inhabiting the lower intertidal (southward extension) and NE species (northward extension) on or shallow subtidal zone (such as D. antarctica and its epibiont com- beaches along the continental coast of Chile (28° S–42° S). munity) tend to have less resistance to environmental stress (e.g. at Horizontal lines represent the median; boxes, the interquartile range; the sea surface) than species that live in the mid-upper intertidal whiskers, 1.5x of interquartile range

83 LOPEZ ET AL. | 1841

FIGURE 7 Box plot of percentages of stranded individuals of Durvillaea antarctica with epibionts outside their literature ranges (% per beach) according to the absence and presence of epibiont species with positive co-occurrences, (a) Gelidium lingulatum and (b) Semimytilus algosus, and epibiont species with negative co-occurrences, (c) Limnoria chilensis and (d) Scurria scurra for SE (southward extension) and NE species (northward extension) on beaches along the continental coast of Chile (28° S–42° S). Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range

zone (e.g. Flores-Molina et al., 2014; Gomez & Huovinen, 2011). complexity, refuges, large size), such as the turf-forming seaweed Also, taxonomic richness of epibionts was lower on rafts with long Gelidium lingulatum (Gonzalez, Stotz, Toledo, Jorquera, & Romero, floating times in the northern (mainly in summer) than in the south- 1991), the canopy-forming kelp Lessonia spicata (Westermeier, Mul- ern districts. This is congruent with the latitudinal and seasonal gra- ler, Gomez, Rivera, & Wenzel, 1994), mussels (Perumytilus purpuratus dient of stressful conditions that suppress the persistence of floating and Semimytilus algosus) that form dense three-dimensional matrices seaweeds at the sea surface (Tala, Velasquez, Mansilla, Macaya, & (Prado & Castilla, 2006), and the reef-building tube worm Phrag- Thiel, 2016; Tala et al., 2013). Hence, our findings support the matopoma moerchi (Sepulveda, Moreno, & Carrasco, 2003) presented hypothesis that some raft-associated species are lost with increasing many positive interactions. On the other hand, mobile species that floating time. Future studies should focus on the functional open cavities in the holdfast, thereby destroying attachment surfaces responses of raft-associated species along a latitudinal gradient, and diminishing the available area for other epibionts during rafting complemented with ecophysiological experiments under controlled journeys, such as the boring isopod Limnoria chilensis (Thiel, 2003) floating conditions. and the excavating limpet Scurria scurra (Vasquez, Veliz, & Pardo, 2001), were negatively correlated with other rafting species. Epibiont species with high number of positive co-occurrences 4.2 | Co-occurrences of epibiont species with other species (e.g. Gelidium lingulatum and Semimytilus algosus) Several positive and negative co-occurrences among epibionts of D. were common on stranded specimens in the southern districts sug- antarctica were observed within single bull kelp rafts. Ecosystem gesting that they contribute to the high richness of rafting species in engineers that generate habitat for other epibionts (e.g. structural these areas and facilitate their range expansions. In the case of

84 1842 | LOPEZ ET AL. epibionts with mostly negative co-occurrences, there was no clear species are more common and wider towards the south of their latitudinal pattern, although they were frequent (e.g. Limnoria chilen- known literature ranges. Moreover, our results suggest that rafting sis) on stranded individuals with indications of longer floating times. dispersal by floating seaweeds might have contributed to the This could indicate that the presence of these species might cause recent range expansion of the introduced seaweed Schottera the disappearance of other epibionts during prolonged rafting jour- nicaeensis (Villasenor-Parada,~ Pauchard, & Macaya, 2014) in the neys. Our results are the first to suggest that biological interactions southern zone (37° S–42° S). within a raft may facilitate or suppress the persistence of other epi- Epibionts found outside their literature ranges showed a slight biont species during long-distance dispersal, thereby potentially tendency to be found on stranded bull kelps with indications of affecting immigration to other areas along a latitudinal gradient. It is prolonged floating times. Lopez et al. (2017) showed that the pres- also possible that our findings are influenced by biotic interactions ence and size of Lepas spp. attached to beach-cast rafts of D. that occurred prior to the detachment of bull kelp from the rocks, antarctica varies strongly along the latitudinal gradient among bio- but there are a number of indications that support our interpreta- geographic districts, being more frequent at 28° S–33° S and tion. Several previous studies had shown that epibiont communities 37° S–42° S, and less frequent from 33° Sto37° S. These geo- on benthic holdfasts of D. antarctica are very different (Edgar & Bur- graphic patterns agree with the range extensions evidenced in our ton, 2000; Santelices et al., 1980) from those observed in the pelagic study for SE and NE species. This suggests that long-distance dis- stage (our study), with a much lower proportion of mobile species persal episodes may be more frequent in certain areas (e.g. 37° S– (e.g. snails, crabs, sea urchins) in floating kelps. Furthermore, rapid 42° S), which probably explains higher range extensions in southern emigration immediately after detachment has been reported in other areas, whereas rafts in northern districts (CCD and SED) usually studies (e.g. Gutow et al., 2009; Miranda & Thiel, 2008) where many have depressed photosynthetic responses and protective mecha- mobile organisms abandon holdfasts during the first minutes after nisms (Tala et al., 2013), which limits their survival at the sea sur- the detachment of buoyant kelps. Indeed, herein we found a high face and would explain the shorter range extensions in those areas. proportion of sessile organisms in the stranded kelps, suggesting that Moreover, higher the presence of epibionts with positive co-occur- important changes had occurred after detachment and that the rences on rafts beyond the literature ranges and lower occurrence observed results are the outcome of interactions during the (possibly of epibiont species with negative co-occurrences in these expan- short) rafting voyages. sions suggest the importance of biotic interactions during long-dis- tance rafting dispersal that facilitate or restrict successful immigration to other areas. Lastly, our results suggest the first 4.3 | Travelling outside of their geographic ranges example of a habitat cascade (sensu Thomsen & Wernberg, 2014) Most rafting epibionts were found within their known literature as a mechanism for range extensions. Holdfasts of floating bull ranges (i.e. RO species). These species are characterized by wide kelps provide habitat to sessile taxa (e.g. mussels) that then gener- geographic ranges and in the case of invertebrates, many of them ate habitat for additional epibiont species, thereby facilitating their have long planktonic larval phases (see Supporting Information long-distance dispersal. Table S3.1). On the other hand, all species that presented exten- sions of their ranges are organisms that have low autonomous dis- persal ability, which suggests that rafting dispersal on floating 5 | CONCLUSION AND OUTLOOK seaweeds could be an effective mechanism of dispersal. Range extensions tend to be wider towards the southern edge of the Our study strongly indicates that some species are frequently being ranges (i.e. SE species), where there are abundant floating kelp sup- dispersed via bull kelp rafts, adding support to phylogeographic plies and environmental conditions at the sea surface are less sev- studies, which had shown long-distance connectivity for peracarid ere (i.e. lower temperature and solar radiation), facilitating raft populations from New Zealand and Chile (e.g. L. chilensis, Nikula persistence and return to the coast (Lopez et al., 2017; Tala et al., et al., 2010; P. kidderi, Haye, Varela, & Thiel, 2012). However, in 2016). In contrast, few species showed extension at the northward other cases, our study actually provides no support for frequent edge of their ranges (i.e. NE species) and these extensions were rafting dispersal—this is, for example, the case for the snail Diloma smaller than in SE species. This is interesting because this zone nigerrima, a species that has genetic similarities between popula- coincides with the biogeographic break at 30° S (Camus, 2001), tions of New Zealand and Chile (Donald, Kennedy, & Spencer, which corresponds to the northern edge of benthic populations of 2005), but that was not observed on any of the stranded kelps D. antarctica (Hoffmann & Santelices, 1997). This area is character- examined in our study. Similar examples are the sea slugs from the ized by oceanographic characteristics (i.e. local current and winds) genus Onchidella (Cumming, Nikula, Spencer, & Waters, 2014) and that affect dispersal and recruitment of several benthic inverte- the seaweed Gracilaria chilensis (Guillemin, Valero, Faugeron, Nel- brates (e.g. Broitman, Navarrete, Smith, & Gaines, 2001), and also son, & Destombe, 2014), both of which were also not found on has consequences for the population connectivity of species with our stranded holdfasts. Finally, their high frequency of occurrence low dispersal ability (e.g. Haye et al., 2014). Therefore, our results on stranded D. antarctica holdfasts suggests that in some species support the hypothesis that range extensions of raft-associated with long-lived autonomous dispersal stages (i.e. planktonic larval

85 LOPEZ ET AL. | 1843 stage for more than 10 days), such as the mytilid mussels S. algosus Duarte, C., Jaramillo, E., Contreras, H., Acuna,~ K., & Navarro, J. M. and P. purpuratus, the limpet S. scurra and the polychaete P. moer- (2009). Importance of macroalgae subsidy on the abundance and population biology of the amphipod Orchestoidea tuberculata (Nicolet) chi (Haye et al., 2014; Trovant, Orensanz, Ruzzante, Stotz, & Basso, in sandy beaches of south central Chile. Revista de Biologia Marina y 2015), rafting on seaweeds might be a complementary dispersal Oceanografia, 44, 691–702. mechanism (in addition to larval dispersal) that could also affect Edgar, G. J., & Burton, H. R. (2000). The biogeography of shallow- their population connectivity. water macrofauna at Heard Island. In M. R. Banks & M. J. Brown (Eds.), Heard Island papers. Papers and proceedings of the Royal Soci- In summary, our study shows that rafting dispersal by floating ety of Tasmania (pp. 23–26). Hobart, Australia: University of Tas- seaweeds is a mechanism that could contribute to the structuring of mania. the geographic ranges of epibiont species, particularly those with Fernandez, M., Astorga, A., Navarrete, S. A., Valdovinos, C., & Marquet, low autonomous dispersal ability. However, these range expansions P. A. (2009). Deconstructing latitudinal species richness patterns in the ocean: Does larval development hold the clue? Ecology Letters, of raft-associated species are strongly influenced by the availability 12, 601–611. https://doi.org/10.1111/j.1461-0248.2009.01315.x of rafts, environmental conditions at the sea surface and local Flores-Molina, M. R., Thomas, D., Lovazzano, C., Nunez,~ A., Zapata, J., oceanographic features that affect successful immigration to other Kumar, M., ... Contreras-Porcia, L. (2014). Desiccation stress in inter- areas. Predicted climate change impacts on floating seaweeds and tidal seaweeds: Effects on morphology, antioxidant responses and – their epibiont communities (i.e. changes in supplies, transport and photosynthetic performance. Aquatic Botany, 113,90 99. https://doi. org/10.1016/j.aquabot.2013.11.004 permanence of rafts; Macreadie, Bishop, & Booth, 2011) should also Fraser, C. I., Thiel, M., Spencer, H. G., & Waters, J. M. (2010). Contempo- be considered in future studies. rary habitat discontinuity and historic glacial ice drive genetic diver- gence in Chilean kelp. BMC Evolutionary Biology, 10, 203. https://doi. org/10.1186/1471-2148-10-203 ACKNOWLEDGEMENTS Gomez, I., & Huovinen, P. (2011). Morpho-functional patterns and zona- tion of South Chilean seaweeds: The importance of photosynthetic This study was financed by CONICYT/FONDECYT 1131082 to and bio-optical traits. Marine Ecology Progress Series, 422,77–91. M.T., F. Tellier and F. Tala and PhD-fellowship Beca CONICYT- https://doi.org/10.3354/meps08937 PCHA/DoctoradoNacional/2014-21140010 to B.L. We wish to Gonzalez, A. V., Beltran, J., Flores, V., & Santelices, B. (2015). Morpholog- ical convergence in the inter-holdfast coalescence process among express our gratitude to Oscar Pino, Vieia Villalobos, Jose Pantoja, kelp and kelp-like seaweeds (Lessonia, Macrocystis, Durvillaea). Phy- Alvaro Gallardo, Ulyces Urtubia, Felipe Saez, Tim Kiessling and Cal- cologia, 54, 283–291. https://doi.org/10.2216/14-105.1 lum Blake for their field and laboratory assistance. Comments from Gonzalez, S., Stotz, W., Toledo, P., Jorquera, M., & Romero, M. (1991). four anonymous reviewers were very helpful in improving the origi- Utilizacion de diferentes microambientes del intermareal como lugares de asentamiento por Fissurella spp (Gastropoda : Proso- nal manuscript. Also, we are grateful to Lucas Eastman for editing branchia) (Palo Colorado, Los Vilos, Chile). Revista de Biologia Marina the final manuscript. y Oceanografia, 26, 325–338. Griffith, D. M., Veech, J. A., & Marsh, C. J. (2016). Cooccur: Probabilistic species co-occurrence analysis in R. Journal of Statistical Software, 69, ORCID 1–17. Guillemin, M. L., Valero, M., Faugeron, S., Nelson, W., & Destombe, C. Martin Thiel http://orcid.org/0000-0001-7535-3888 (2014). Tracing the trans-pacific evolutionary history of a domesti- cated seaweed (Gracilaria chilensis) with archaeological and genetic data. PLoS ONE, 9, e114039. https://doi.org/10.1371/journal.pone. REFERENCES 0114039 Gutow, L., Beermann, J., Buschbaum, C., Rivadeneira, M. M., & Thiel, M. Broitman, B. R., Navarrete, S. A., Smith, F., & Gaines, S. D. (2001). Geo- (2015). Castaways can’t be choosers - Homogenization of rafting graphic variation of southeastern Pacific intertidal communities. Mar- assemblages on floating seaweeds. Journal of Sea Research, 95, 161– ine Ecology Progress Series, 224,21–34. https://doi.org/10.3354/me 171. https://doi.org/10.1016/j.seares.2014.07.005 ps224021 Gutow, L., Gimenez, L., Boos, K., & Saborowski, R. (2009). Rapid changes in Camus, P. A. (2001). Biogeografıa marina de Chile continental. Revista the epifaunal community after detachment of buoyant benthic macroal- Chilena de Historia Natural, 74, 587–617. gae. Journal of the Marine Biological Association of the United Kingdom, Chao, A., & Lee, S. M. (1992). Estimating the number of classes via sam- 89,323–328. https://doi.org/10.1017/S0025315408002658 ple coverage. Journal of the American Statistical Association, 87, 210– Haye, P. A., Segovia, N. I., Munoz-Herrera,~ N. C., Galvez, F. E., Martınez, 217. https://doi.org/10.1080/01621459.1992.10475194 A., Meynard, A., ... Faugeron, S. (2014). Phylogeographic structure in Clarke, K. R., Gorley, R. N., Somerfield, P. J., & Warwick, R. M. (2014). benthic marine invertebrates of the southeast Pacific coast of Chile Change in marine communities: An approach to statistical analysis and with differing dispersal potential. PLoS ONE, 9, e88613. https://doi. interpretation (3rd ed.). Plymouth, UK: PRIMER-E. org/10.1371/journal.pone.0088613 Cumming, R. A., Nikula, R., Spencer, H. G., & Waters, J. M. (2014). Trans- Haye, P. A., Varela, A. I., & Thiel, M. (2012). Genetic signatures of rafting oceanic genetic similarities of kelp-associated sea slug populations: dispersal in algal-dwelling brooders Limnoria spp. (Isopoda) along the Long-distance dispersal via rafting? Journal of Biogeography, 41, SE Pacific (Chile). Marine Ecology Progress Series, 455, 111–122. 2357–2370. https://doi.org/10.1111/jbi.12376 https://doi.org/10.3354/meps09673 Donald, K. M., Kennedy, M., & Spencer, H. G. (2005). Cladogenesis as Hernandez, C. E., Moreno, R. A., & Rozbaczylo, N. (2005). Biogeographi- the result of long-distance rafting events in South Pacific topshells cal patterns and Rapoport’s rule in southeastern Pacific benthic poly- (Gastropoda, Trochidae). Evolution, 59, 1701–1711. https://doi.org/ chaetes of the Chilean coast. Ecography, 28, 363–373. https://doi. 10.1111/j.0014-3820.2005.tb01819.x org/10.1111/j.0906-7590.2005.04013.x

86 1844 | LOPEZ ET AL.

Hoffmann, A., & Santelices, B. (1997). Flora marina de Chile Central. Santi- Tala, F., Gomez, I., Luna-Jorquera, G., & Thiel, M. (2013). Morpholog- ago de Chile, Chile: Ediciones Universidad Catolica de Chile. ical, physiological and reproductive conditions of rafting bull kelp Lester, S. E., Ruttenberg, B. I., Gaines, S. D., & Kinlan, B. P. (2007). The (Durvillaea antarctica) in northern-central Chile (30°S). Marine relationship between dispersal ability and geographic range size. Ecol- Biology, 160, 1339–1351. https://doi.org/10.1007/s00227-013- ogy Letters, 10, 745–758. https://doi.org/10.1111/j.1461-0248.2007. 2186-8 01070.x Tala, F., Penna-Dıaz, M. A., Luna-Jorquera, G., Roth€ausler, E., & Thiel, M. Lopez, B. A., Macaya, E. C., Tala, F., Tellier, F., & Thiel, M. (2017). The (2017). Daily and seasonal changes of photobiological responses in variable routes of rafting: Stranding dynamics of floating bull kelp floating bull kelp Durvillaea antarctica (Chamisso) Hariot (Fucales: Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Phaeophyceae). Phycologia, 56, 271–283. https://doi.org/10.2216/ Pacific. Journal of Phycology, 53,70–84. https://doi.org/10.1111/jpy. 16-93.1 12479 Tala, F., Velasquez, M., Mansilla, A., Macaya, E. C., & Thiel, M. (2016). Macaya, E. C., Lopez, B., Tala, F., Tellier, F., & Thiel, M. (2016). Float Latitudinal and seasonal effects on short-term acclimation of floating and raft: Role of buoyant seaweeds in the phylogeography and kelp species from the South-East Pacific. Journal of Experimental Mar- genetic structure of non-buoyant associated flora. In Z. M. Hu & C. ine Biology and Ecology, 483,31–41. https://doi.org/10.1016/j.jembe. I. Fraser (Eds.), Seaweed phylogeography (pp. 97–130). Dordrecht, 2016.06.003 The Netherlands: Springer. https://doi.org/10.1007/978-94-017- Thiel, M. (2003). Reproductive biology of Limnoria chilensis: Another bor- 7534-2 ing peracarid species with extended parental care. Journal of Natural Macreadie, P. I., Bishop, M. J., & Booth, D. J. (2011). Implications of cli- History, 37, 1713–1726. https://doi.org/10.1080/ mate change for macrophytic rafts and their hitchhikers. Marine Ecol- 00222930210125416 ogy Progress Series, 443, 285–292. https://doi.org/10.3354/me Thiel, M., & Fraser, C. I. (2016). The role of floating plants in dispersal of ps09529 biota across habitats and ecosystems. In E. Olafsson (Ed.), Marine Miranda, L., & Thiel, M. (2008). Active and passive migration in boring macrophytes as foundation species (pp. 76–99). Boca Raton, FL: CRC isopods Limnoria spp. (Crustacea, Peracarida) from kelp holdfasts. Press, Taylor & Francis Group. https://doi.org/10.4324/ Journal of Sea Research, 60, 176–183. https://doi.org/10.1016/j.seare 9781315370781 s.2008.06.002 Thiel, M., & Gutow, L. (2005). The ecology of rafting in the marine envi- Nikula, R., Fraser, C. I., Spencer, H. G., & Waters, J. M. (2010). Circumpo- ronment. II. The rafting organisms and community. Oceanography and lar dispersal by rafting in two subantarctic kelp-dwelling crustaceans. Marine Biology: An Annual Review, 43, 279–418. https://doi.org/10. Marine Ecology Progress Series, 405, 221–230. https://doi.org/10. 1201/CRCOCEMARBIO 3354/meps08523 Thiel, M., & Vasquez, J. A. (2000). Are kelp holdfasts islands on the ocean Nikula, R., Spencer, H. G., & Waters, J. M. (2013). Passive rafting is a floor? Indication for temporarily closed aggregations of peracarid powerful driver of transoceanic gene flow. Biology Letters, 9, crustaceans. Hydrobiologia, 440,45–54. https://doi.org/10.1023/A: 20120821. 1004188826443 Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P., McGlinn, Thomsen, M. S., & Wernberg, T. (2014). On the generality of cascading D., ... Wagner, H. (2017). Vegan: Community ecology package. R habitat-formation. Proceedings of the Royal Society of London B, Biolog- package version 2.4-4. Retrieved from http://CRAN.R-project.org/pac ical Sciences, 281, 20131994. https://doi.org/10.1098/rspb.2013. kage=vegan. 1994 Prado, L., & Castilla, J. C. (2006). The bioengineer Perumytilus purpuratus Trovant, B., Orensanz, J. M., Ruzzante, D. E., Stotz, W., & Basso, N. G. (Mollusca: Bivalvia) in central Chile: Biodiversity, habitat structural (2015). Scorched mussels (Bivalvia: Mytilidae: Brachidontinae) from complexity and environmental heterogeneity. Journal of the Marine the temperate coasts of South America: Phylogenetic relationships, Biological Association of the United Kingdom, 86, 417–421. https://doi. trans-Pacific connections and the footprints of Quaternary glacia- org/10.1017/S0025315406013282 tions. Molecular Phylogenetics and Evolution, 82,60–74. https://doi. R Development Core Team. (2017). R: A language and environment for org/10.1016/j.ympev.2014.10.002 statistical computing. Vienna, Austria: R Foundation for Statistical Valdovinos, C., Navarrete, S. A., & Marquet, P. A. (2003). Mollusk species Computing. Retrieved from http://www.r-project.org/. diversity in the Southeastern Pacific: Why are there more species Rivadeneira, M. M., Thiel, M., Gonzalez, E. R., & Haye, P. A. (2011). An towards the pole? Ecography, 26, 139–144. https://doi.org/10.1034/j. inverse latitudinal gradient of diversity of peracarid crustaceans along 1600-0587.2003.03349.x the Pacific Coast of South America: Out of the deep south. Global Vandendriessche, S., Vincx, M., & Degraer, S. (2007). Floating seaweed Ecology and Biogeography, 20, 437–448. https://doi.org/10.1111/j. and the influences of temperature, grazing and clump size on raft 1466-8238.2010.00610.x longevity - A microcosm study. Journal of Experimental Marine Biology Santelices, B., Castilla, J. C., Cancino, J., & Schmiede, P. (1980). Compara- and Ecology, 343,64–73. https://doi.org/10.1016/j.jembe.2006.11. tive ecology of Lessonia nigrescens and Durvillaea antarctica (Phaeo- 010 phyta) in central Chile. Marine Biology, 59, 119–132. https://doi.org/ Vasquez, J. A., Veliz, D., & Pardo, L. M. (2001). Biodiversidad de macroin- 10.1007/BF00405461 vertebrados bajo las grandes algas. In K. Alveal & T. Antezana (Eds.), Santelices, B., & Marquet, P. A. (1998). Seaweeds, latitudinal diversity Sustentabilidad de la biodiversidad. Un problema actual: Bases cientıfico patterns, and Rapoport’s rule. Diversity and Distributions, 4,71–75. tecnicas, teorizaciones y proyecciones (pp. 293–308). Concepcion: https://doi.org/10.1046/j.1472-4642.1998.00005.x Universidad de Concepcion. Sepulveda, R. D., Moreno, R. A., & Carrasco, F. D. (2003). Diversidad de Veech, J. A. (2014). The pairwise approach to analysing species co-occur- macroinvertebrados asociados a arrecıfes de Phragmatopoma moerchi rence. Journal of Biogeography, 41, 1029–1035. https://doi.org/10. Kinberg, 1867 (Polychaeta: Sabellariidae) en el intermareal rocoso de 1111/jbi.12318 Cocholgue,€ Chile. Gayana (Concepcion) , 67,45–54. Villasenor-Parada,~ C., Pauchard, A., & Macaya, E. (2014). Expansion del Stoner, A. W., & Greening, H. S. (1984). Geographic variation in the area de distribucion de la macroalga introducida Schottera nicaeensis macrofaunal associates of pelagic Sargassum and some biogeographic (Rhodophyta: Gigartinales) en la costa chilena: ¿evidencia de una implications. Marine Ecology Progress Series, 20, 185–192. https://doi. invasion? Boletın de la Red Latinoamericana para el Estudio de Especies org/10.3354/meps020185 Invasoras, 4,19–27.

87 LOPEZ ET AL. | 1845

Westermeier, R., Muller, D. G., Gomez, I., Rivera, P., & Wenzel, H. SUPPORTING INFORMATION (1994). Population biology of Durvillaea antarctica and Lessonia nigrescens (Phaeophyta) on the rocky shores of southern Chile. Mar- Additional supporting information may be found online in the ine Ecology Progress Series, 110, 187–194. https://doi.org/10.3354/ Supporting Information section at the end of the article. meps110187

How to cite this article: Lopez BA, Macaya EC, Rivadeneira BIOSKETCHES MM, Tala F, Tellier F, Thiel M. Epibiont communities on stranded kelp rafts of Durvillaea antarctica (Fucales, Boris A. Lopez is a PhD student of the Applied Biology and Ecol- Phaeophyceae)—Do positive interactions facilitate range ogy program at the Universidad Catolica del Norte (UCN), Chile. extensions?. J Biogeogr. 2018;45:1833–1845. https://doi.org/ He is a marine biologist interested in ecology, biogeography and 10.1111/jbi.13375 genetic connectivity of marine populations. He and the other authors collaborate in ecological and biogeographic studies that help understand the patterns of marine biodiversity across differ- ent taxa (mainly seaweeds and invertebrates) in the Southeast Pacific.

Author contributions: B.A.L., F. Tala and M.T. conceived the idea; B.A.L., E.C.M. and M.T. collected the data; analyses were con- ducted by B.A.L. and M.M.R.; B.A.L. and M.T. led the writing with assistance from M.M.R., E.C.M., F. Tala and F. Tellier.

88

Chapter 5

Phylogeography of two intertidal seaweeds, Gelidium lingulatum and G. rex (Rhodophyta: Gelidiales), along the South East Pacific – patterns explained by rafting dispersal?

(manuscript published in Mar. Biol. 164: 188, 2017)

Boris A. López1,2, Florence Tellier3,4*, Juan C. Retamal-Alarcón3, Karla Pérez-Araneda3,4, Ariel O. Fierro3; Erasmo C. Macaya5,6,7, Fadia Tala1,8 & Martin Thiel1,6,9

1Facultad de Ciencias del Mar, Universidad Católica del Norte, Coquimbo, Chile. 2Departamento de Acuicultura y Recursos Agroalimentarios, Universidad de Los Lagos, Osorno, Chile. 3Departamento de Ecología, Facultad de Ciencias, Universidad Católica de la Santísima Concepción, Concepción, Chile. 4Centro de Investigación en Biodiversidad y Ambientes Sustentables (CIBAS), Universidad Católica de la Santísima Concepción, Concepción, Chile. 5Laboratorio de Estudios Algales ALGALAB, Departamento de Oceanografía, Facultad de Ciencias Naturales y Oceanográficas, Universidad de Concepción, Concepción, Chile. 6Millennium Nucleus Ecology and Sustainable Management of Oceanic Island (ESMOI), Coquimbo, Chile. 7Centro FONDAP de Investigaciones en Dinámica de Ecosistemas Marinos de Altas Latitudes (IDEAL), Valdivia, Chile. 8Centro de Investigación y Desarrollo Tecnológico en Algas. Universidad Católica del Norte, Coquimbo, Chile. 9Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Coquimbo, Chile.

*Corresponding author: [email protected]

89

Mar Biol (2017) 164:188 DOI 10.1007/s00227-017-3219-5

ORIGINAL PAPER

Phylogeography of two intertidal seaweeds, Gelidium lingulatum and G. rex (Rhodophyta: Gelidiales), along the South East Pacifc: patterns explained by rafting dispersal?

Boris A. López1,2 · Florence Tellier3,4 · Juan C. Retamal‑Alarcón3 · Karla Pérez‑Araneda3,4 · Ariel O. Fierro3 · Erasmo C. Macaya5,6,7 · Fadia Tala8,9 · Martin Thiel6,8,10

Received: 31 October 2016 / Accepted: 4 August 2017 © Springer-Verlag GmbH Germany 2017

Abstract Rafting on foating seaweeds facilitates disper- respectively) were characterized using a mitochondrial sal of associated organisms, but there is little information marker (COI) and, for a subset, using a chloroplastic marker on how rafting afects the genetic structure of epiphytic (rbcL). Gelidium lingulatum had higher genetic diversity, seaweeds. Previous studies indicate a high presence of sea- but its genetic structure did not follow a clear geographic weeds from the genus Gelidium attached to foating bull kelp pattern, while G. rex had less genetic diversity with a shal- Durvillaea antarctica (Chamisso) Hariot. Herein, we ana- low genetic structure and a phylogeographic break coincid- lyzed the phylogeographic patterns of Gelidium lingulatum ing with the phylogeographic discontinuity described for (Kützing 1868) and G. rex (Santelices and Abbott 1985), this region (29°S–33°S). In G. lingulatum, no isolation- species that are partially co-distributed along the Chilean by-distance was observed, in contrast to G. rex. The phy- coast (28°S–42°S). A total of 319 individuals from G. lin- logeographic pattern of G. lingulatum could be explained gulatum and 179 from G. rex (20 and 11 benthic localities, mainly by rafting dispersal as an epiphyte of D. antarctica, although other mechanisms cannot be completely ruled out (e.g., human-mediated dispersal). The contrasting pattern Responsibile Editor: O. Puebla. observed in G. rex could be attributed to other factors such as intertidal distribution (i.e., G. rex occurs in the lower Reviewed by Undisclosed experts. zone compared to G. lingulatum) or diferential efciency of recruitment after long-distance dispersal. This study indi- Electronic supplementary material The online version of this article (doi:10.1007/s00227-017-3219-5) contains supplementary cates that rafting dispersal, in conjunction with the intertidal material, which is available to authorized users.

6 * Florence Tellier Millennium Nucleus Ecology and Sustainable Management [email protected] of Oceanic Island (ESMOI), Coquimbo, Chile 7 1 Centro FONDAP de Investigaciones en Dinámica de Doctorado en Biología y Ecología Aplicada, Universidad Ecosistemas Marinos de Altas Latitudes (IDEAL), Valdivia, Católica del Norte, Coquimbo, Chile Chile 2 Departamento de Acuicultura y Recursos Agroalimentarios, 8 Departamento de Biología Marina, Facultad de Ciencias del Universidad de Los Lagos, Osorno, Chile Mar, Universidad Católica del Norte, Coquimbo, Chile 3 Departamento de Ecología, Facultad de Ciencias, 9 Centro de Investigación y Desarrollo Tecnológico en Algas Universidad Católica de la Santísima Concepción, (CIDTA), Universidad Católica del Norte, Coquimbo, Chile Concepción, Chile 10 4 Centro de Estudios Avanzados en Zonas Áridas (CEAZA), Centro de Investigación en Biodiversidad y Ambientes Coquimbo, Chile Sustentables (CIBAS), Universidad Católica de la Santísima Concepción, Concepción, Chile 5 Laboratorio de Estudios Algales ALGALAB, Departamento de Oceanografía, Facultad de Ciencias Naturales y Oceanográfcas, Universidad de Concepción, Concepción, Chile

Vol.:(0123456789)1 3

90 188 Page 2 of 19 Mar Biol (2017) 164:188 distribution, can modulate the phylogeographic patterns of of foating species, but also of their epibiont communities seaweeds. (Thiel and Haye 2006). Only a few studies have evaluated the efects of rafting on the genetic diversity and structure of epibionts (also called Introduction secondary rafters), focusing mostly on animals associated with foating kelps (see Nikula et al. 2010, 2011a, b, 2013; The dispersal ability of marine species is a major trait deter- Cumming et al. 2014). In their recent review of phylogeo- mining the genetic structure of their benthic populations graphic studies on non-buoyant seaweeds associated with (Weersing and Toonen 2009; Selkoe and Toonen 2011; Haye foating substrata, Macaya et al. (2016) have shown that most et al. 2014). In general, species with high dispersal ability of these epibionts present low genetic structure and high (i.e., presence of planktonic larvae, swimming or crawling genetic connectivity among populations. Nevertheless, in structures in adults) tend to have lower genetic structure due most cases the authors of the genetic studies only suggested to higher gene fow between geographically distant popula- this connectivity via rafting of foating seaweeds and could tions compared to species with direct development (absence not completely exclude other vectors of dispersal (e.g., foat- of larvae) or low mobility (e.g., Dawson et al. 2014; Haye ing marine litter, see Kiessling et al. 2015). et al. 2014). However, other factors such as oceanographic, A good choice to study the phylogeography of epibionts geological, geographical and ecological features can also is conducting research in areas where there is extensive prior afect connectivity, and therefore, the distribution of genetic information about abundances and environmental factors diversity (Palumbi 1994). In particular, on rocky shores, the that could afect the persistence of foating substrata, espe- tidal height where the organisms are distributed might infu- cially detached seaweeds. In particular, one of the oceans ence the genetic structure of local populations, with species where there have been several studies on rafting and phylo- from medium and high tidal levels having greater genetic geography of seaweeds is the South East Pacifc coast (SEP, structure than species from the low intertidal or subtidal ~14°S to 56°S) (Thiel and Gutow 2005b; Fraser et al. 2010; zone (Kelly and Palumbi 2010). This is frequently assumed Macaya and Zuccarello 2010a, b; see also for review Guil- to be due to the patchiness and greater variety of environ- lemin et al. 2016a). In this zone, phylogeographic studies of mental stresses in the high- to mid-intertidal zones that may benthic species (invertebrates and seaweeds) have focused generate diferential natural selection than in lower zones on testing the concordance between the proposed biogeo- where the conditions tend to be more homogeneous. Sev- graphic boundaries (at 30°S and 42°S) and phylogeographic eral studies have reported this pattern, which tends to be breaks (for recent reviews see Haye et al. 2014; Guillemin more prevalent in seaweeds and sessile invertebrates (Engel et al. 2016a). In particular, seaweed species with low dis- et al. 2004; Billard et al. 2005; Valero et al. 2011; Krueger- persal ability presented notorious phylogeographic breaks, Hadfeld et al. 2013; Robuchon et al. 2014). suggesting that evolutionary lineages constitute distinct Seaweeds from intertidal or shallow subtidal habitats are phylogenetic species, as in the intertidal macroalgae Les- considered good models for phylogeographic studies (Hu sonia nigrescens (now separated into L. berteroana and L. et al. 2016). This is due to the complex reproductive cycles spicata; Tellier et al. 2009; González et al. 2012) and Maz- (alternation of haploid and diploid phases) of numerous spe- zaella laminarioides (Montecinos et al. 2012). On the other cies from all seaweed divisions that may afect the genetic hand, seaweeds with high dispersal ability have shallow structure of their populations (Krueger-Hadfeld and Hoban phylogeographic breaks and a low genetic structure, such 2016), coupled with the low dispersal capacity of spores as the foating kelp Macrocystis pyrifera (Macaya and Zuc- (Santelices 1980; Destombe et al. 1992). However, other carello 2010a, b). A distinct phylogeographic pattern (i.e., mechanisms such as rafting permit dispersal over long dis- strong genetic structure and high values of genetic diversity) tances (Thiel and Haye 2006; Muhlin et al. 2008; Fraser has been reported for the bull kelp Durvillaea antarctica, a et al. 2009a; 2010; Coyer et al. 2011a, b). For example, species with a high dispersal potential by rafting, along the some buoyant seaweeds, such as the bull kelp Durvillaea continental coast of Chile (Fraser et al. 2010). This has been antarctica (Fraser et al. 2010) and the giant kelp Macro- attributed to inefective long-distance dispersal, either due cystis pyrifera (Macaya and Zuccarello 2010a, b) can foat to low efectiveness in recruitment of new individuals in over extensive distances (>1000 km) after detachment from resident populations or because bull kelp supplies are highly the primary substratum, and occasionally even cross entire variable in certain areas (Fraser et al. 2010). More than 40 ocean basins (e.g., over 5000 km, between the coasts of New species of seaweeds, mainly Rhodophyta, have been found Zealand and Chile), disrupting the potential for genetic dif- attached to holdfasts of stranded specimens of D. antarctica ferentiation among distant populations (Thiel and Gutow along the continental coast of Chile (Macaya et al. 2016). 2005a; Fraser et al. 2010; Coyer et al. 2011b). Rafting may Therefore, dispersal as secondary rafter could also modulate not only facilitate gene fow among benthic populations the genetic structure of these epiphytic algae as suggested

1 3

91 Mar Biol (2017) 164:188 Page 3 of 19 188 for other non-buoyant seaweeds with low genetic diferentia- lingulatum is present at 1–2 m above MLLW (mean lower tion among distant populations (Boo et al. 2014a; Guillemin low water) and G. rex is most often found at lower inter- et al. 2014). tidal heights, about 0–0.5 m above MLLW (Santelices Gelidium lingulatum (Kützing 1868) and G. rex (San- 1986). The reported geographical distribution of G. lin- telices and Abbott 1985) are two endemic red seaweeds gulatum extends from Antofagasta (23°S) to Tierra del from the SEP (Santelices 1990). They form monospecifc Fuego (56°S) (Ramírez and Santelices 1991; Hofmann beds at wave-exposed sites (Ortega et al. 2001), mainly and Santelices 1997; Fig. 1). However, the current distri- in the intertidal zone, where they grow attached to rocks bution of the species is not clearly established, because and calcareous shells, but also to holdfasts of large kelps identifcation was based solely on morphological charac- (Santelices 1990; Macaya et al. 2016). Particularly, G. ters (Santelices 1990). Since G. lingulatum features high

Fig. 1 Geographic location of the coast of Chile, show- ing the main biogeographic zones (provinces and districts) and breaks (30°S–33°S and 41°S–42°S) (modifed from Camus 2001). The sampled distributions of Gelidium lingulatum and G. rex along the Chilean coast are also indicated (shaded bars), as well as the geographic distribution accord- ing to the literature (lines). Also, local strandings (stranded biomass and length) of foating bull kelp Durvillaea antarc- tica in diferent biogeographic districts are shown (correspond- ence between circle sizes and stranded biomasses, and kelp sizes and lengths are indicated) (extracted from López et al. 2017)

1 3

92 188 Page 4 of 19 Mar Biol (2017) 164:188 phenotypic plasticity and juveniles are morphologically Materials and methods very similar to other species of the genus, there is a cer- tain risk of erroneous identifcations. On the other hand, Biogeographical features of the study area G. rex has morphological features facilitating identifca- tion in the feld. This species is distributed more narrowly, The SEP coast (~14°S to 56°S) is characterized by a lin- between Coquimbo (30°S) and Concepción (36°S) (Hof- ear topography and no major topographical discontinuities mann and Santelices 1997), although it has been suggested between 14°S and 42°S, south of where it becomes a coast that its northern and southern distribution could extend to characterized by the presence of channels and fords (Camus 16°S and 39°S, respectively (Santelices and Abbott 1985; 2001; Thiel et al. 2007; Försterra 2009). Ocean circulation Fig. 1). in this area is mainly determined by the Humboldt Current Both species of Gelidium have a reproductive cycle with with south–north orientation, and by the southward Cape an alternation of haploid and diploid phases (Hernández Horn Current in the southernmost area (Thiel et al. 2007). 1997), thalli that can re-attach to the substratum (Rojas Also, it is characterized by a latitudinal temperature gradient et al. 1996), and their spores only survive for short time in surface waters (Tapia et al. 2014) where the occurrence periods in the water so that the dispersal potential via spores of seasonally persistent upwelling events afect the bioge- is limited (Bobadilla and Santelices 2005). Also, these turf- ographic structure of the coastal zone (Lachkar and Gru- forming seaweeds are ecologically important as settlement ber 2012; Aravena et al. 2014). Two major biogeographic and nursery area for small invertebrates (González et al. provinces have been described for the continental coast of 1991). Moreover, they are economically important for agar Chile: the Peruvian Province (18°S–30°S) and the Magel- extraction (Matsuhiro and Urzúa 1990, 1991; Melo 1998). lanic Province (42°S–56°S), which are separated by a broad Within their geographic ranges individuals of both species transition zone, the Intermediate Area, between 30°S and are frequently found in holdfasts of foating bull kelp Durvil- 42°S (Camus 2001) (Fig. 1). A recent study had reported laea antarctica (i.e., >10% in the case of G. lingulatum and a strong pattern of stranded biomass and length of beach- 1–10% in G. rex, Macaya et al. 2016), and while intrinsic cast bull kelps (Durvillaea antarctica) in diferent biogeo- dispersal ability is strongly limited in these species, dispersal graphical districts (i.e., subdivisions of the biogeographic on foating bull kelps could potentially enhance connectivity provinces), particularly within the Intermediate Area (López between their populations. However, this could also be mod- et al. 2017, Fig. 1), suggesting areas where the connectivity ulated by the distribution across the tidal gradient, where of their populations and that of their secondary rafters could more structure would be expected in G. lingulatum from the be greater than in others. mid-intertidal zone compared to G. rex, which grows in the low intertidal zone. Sampling of Gelidium lingulatum and G. rex Genetic studies in Gelidiales have revealed high species diversity within the group and also important limitations Species identifcation of the morphological identifcation of species (e.g., Nelson et al. 2006; Boo et al. 2013, 2014b, 2016). Currently, few Both species were identifed using morphological traits as phylogeographic studies are available for species from the described by Santelices and Montalva (1983), Santelices genus Gelidium. For example, both G. canariense in the and Stewart (1985), Vargas and Collado-Vides (1996), and Canary Islands (Bouza et al. 2006) and G. elegans on the Hofmann and Santelices (1997). For G. lingulatum some coast of Korea, China and Japan (Kim et al. 2012) have difculties in visual species identifcation were encoun- high genetic variability between populations, numerous tered due to its close morphological similarity with other private haplotypes, and low genetic connectivity. This high co-occurring Gelidium species (e.g., G. chilense), par- level of genetic structure among populations has also been ticularly in the northern part of the described distribution observed for species with a wide geographical range, such range of the species. Fully developed individuals consist of as G. vagum (Yoon et al. 2014), G. crinale and G. pusillum a crawling and an erect portion. The creeping axes adhere to (Kim and Boo 2012). the substratum by short discoidal rhizoids, while the erect Using two molecular markers (COI and rbcL), the present axes are cylindrical with tongue-like blades and sparsely study aimed to determine the geographical distribution of branched at the base (Hoffmann and Santelices 1997) genetic diversity for two species from the genus Gelidium, (Fig. 2). While G. lingulatum is supposed to occur as far which are partially co-distributed along the Chilean coast north as 23°S (Ramírez and Santelices 1981), no individu- and occur at distinct tidal levels. Based on these results, this als with the morphological characteristics of G. lingula- study also aimed to evaluate whether the observed phylogeo- tum were found in eight locations from the northern part graphic patterns might be infuenced by rafting dispersal via of our study area (~20°S to 28°S, Fig. 2, Online Resource foating seaweeds. 1). In locations at 28°S (i.e., BURR and APOL), we found

1 3

93 Mar Biol (2017) 164:188 Page 5 of 19 188 individuals with typical traits of the genus Gelidium, but not Germany) at 240 rpm for 5 min. The subsequent DNA of G. lingulatum, and molecular characterization confrmed extraction was performed using the EZNA­ ® Tissue DNA Kit the distinctiveness from G. lingulatum (unpubl. data). This (Bio-Tek OMEGA, Atlanta, USA), according to the manu- suggests that G. lingulatum does not occur north of 29°S, facturer’s specifcations. where other, morphologically similar species have been Polymerase chain reaction (PCR) amplifcation of the found; consequently, the distribution range of G. lingulatum partial Cytochrome Oxidase c subunit I gene (COI) was per- seems to be more restricted than reported by Ramírez and formed using primers designed by Saunders (2005) for red Santelices (1991) and Hofmann and Santelices (1997). On seaweeds (GazF1: 5′ TCAACA​ AAT​ CAT​ AAA​ GAT​ ATTGG​ the other hand, a clear morphological distinction of G. rex is 3′ and GazR2: 5′ ACTTCT​ GGA​ TGT​ CCA​ AAA​ AAYCA​ 3′) possible because of the cylindrical axes at the base and fat- and using the same conditions for PCR concentrations and tened middle and upper parts with toothed margins. Also, program as Fraser et al. (2009b). Reactions were done using this species lacks branchlets along the main axis and has a dNTPs and DNA polymerase GoTaq, Fermelo Biotec (Pro- rigid, crispate and cartilaginous thallus (Fig. 3). In addition, mega, Madison, USA), and PCR reactions were performed G. rex is the largest species from the genus Gelidium present in a thermocycler Veriti (Applied Biosystems, Foster City, in Chile (Santelices and Abbott 1985). No individuals of G. USA). rex were found in surveys north of 28°S (Online Resource In order to compare the results of the COI marker with a 1) and south of 34°S (Fig. 3). marker with a slower evolving rate, a subset of 22 individu- als (10 G. lingulatum and 12 G. rex) was selected for the Sampling locations rbcL sequencing, considering primarily the specimens hav- ing diferent COI haplotypes and trying to cover the maxi- Individuals of G. lingulatum and G. rex were collected in mum of the species distribution range. It is important to note winters and summers of 2012–2015 from natural popula- that a single marker may not be representative of the species tions in the mid–lower intertidal zone (0.5–1 m) of wave- history, and therefore, the combination of multiple mark- exposed rocky shores. Sampling was performed in a total of ers facilitates the detection of diferent processes occurring 11 locations (28°S–34°S; 790 km of coastline) for G. rex, at diferent time scales (Ballard and Whitlock 2004). The covering 68% of the described geographic range, and in a chloroplast-encoded rbcL corresponds to the large subunit total of 20 locations (29°S–42°S; 1770 km of coastline) for G. of the ribulose-1,5-bisphosphate carboxylase/oxygenase lingulatum (45% of the initially described geographic range) (RuBisCo). PCR amplifcations of rbcL were performed (Table 1; Figs. 2, 3). For both G. lingulatum and G. rex, we using two primer combinations, F7–R753 (F7: 5′AAC​TCT​ collected at least 15 individuals per locality (for this study, GTA​GAA​CGNACAAG 3′; R753: 5′ GCT​CTT​TCA​TAC​ an individual was composed of one or several erect axes that ATATCT​ TCC​ 3′; Freshwater and Rueness 1994; Gavio and arise from stoloniferous thalli, Santelices 1986), except in Fredericq 2002) and F645–RrbcSstart (F645: 5′ ATG​CGT​ those locations of low species abundance, such as in the north TGG AAA​GAA​AGA​TTC​T 3′ and RrbcSstart: 5′ TGT​GTT​ of the study range (Table 1). A total of 319 and 179 specimens GCGGCC​ GCC​ CTT​ GTG​ TTA​ GTC​ TCA​ C​ 3′; Freshwater and were analyzed for G. lingulatum and G. rex, respectively. Rueness 1994; Lin et al. 2001). The conditions for PCR con- centrations and program were identical to Boo et al. (2013). Sample manipulation The PCR reagents used were similar to those described for the COI marker. For each sample, several branches of small and well-iden- PCR products were purifed and then sequenced using tifed patches of G. lingulatum or G. rex were collected. the reverse amplifcation primers (GazR2 for COI, R753 Samples were only taken from patches that had a minimum and RrbcSstart for rbcL) by Macrogen Inc. (Seoul, South distance of 1 m apart, since vegetative propagation occurs Korea: http://www.macrogen.com). Sequences were visual- by prostrate stoloniferous thalli (Santelices 1986). Sam- ized and edited in Chromas v2.5.1 (Technelysium Pty Ltd ples were carefully cleaned from epibionts, then stored in 2016) and multiple sequence alignment was performed using individual plastic bags flled with silica gel beads for rapid CLUSTALW function of BioEdit 7.2.5 (Hall 1999) for each dehydration, and transported to the laboratory for further species and marker dataset. Final alignments were checked genetic analysis. visually. The resulting datasets for G. lingulatum consisted of a 622-base pair (bp) alignment for the mitochondrial DNA extraction, PCR amplifcation, sequencing DNA region and of a 1484-bp alignment for the chloroplas- and sequences alignment tic DNA region, while G. rex datasets consisted of a 628-bp alignment and a 1543-bp alignment, respectively. All hap- For each sample, a small piece of dry tissue (50 mg) was lotype sequences were deposited in GenBank (Accession finely ground using the Tissue ­Lyser® (Rotsch, Hilden, Numbers KX961986–KX962024) and analyzed by BLAST

1 3

94 188 Page 6 of 19 Mar Biol (2017) 164:188 analysis to identify matches with other sequences (Altschul locations with less than 14 samples were excluded and each et al. 1990). combination of groups was tested using 500 permutations.

Genetic diversity and genetic diferentiation Isolation by distance

Estimations of standard genetic diversity indices The isolation-by-distance model (Slatkin 1993) was tested per location and per species using a Mantel test in Arlequin with 1000 permutations, test- ing for a positive correlation between pairwise geographic The following molecular diversity indices were computed distance (in km) and raw (D) average pairwise diferences at the species-level only for the rbcL dataset and at the spe- for COI datasets, excluding locations with less than 14 sam- cies- and location-levels for COI, using Arlequin v 3.5.2.2 ples. Linearized population pairwise φST values could not (Excofer and Lischer 2010): the number of haplotypes (h), be used in the Mantel test because several locations were the number of private haplotypes (i.e., haplotypes found at genetically monomorphic for diferent haplotypes, and pair- a single sampled location, hpriv), the number of polymor- wise comparisons between such fxed populations gave an phic sites (S), haplotype diversity (H, based on haplotype φST of 1.0. The geographical distance between location pairs frequency, the probability that two randomly chosen haplo- was measured as distance along the coast for continental types are diferent; Excofer and Lischer 2010) and nucleo- locations and taken as the straight-line distance for the island tide diversity (π, the probability that two randomly chosen locations (Chiloé Island: MBRA, CUCA and SBA), using homologous nucleotide sites are diferent, expressed as %π; the ‘path ruler’ tool in Google Earth (http://earth.google. Excofer and Lischer 2010). For each COI dataset, consider- com/). ing the diferent sample sizes, a rarefaction method was used with the Contrib program (Petit et al. 1998) to calculate the Haplotype network reconstruction and historical standardized haplotype diversity at location (excluding the demography locations with less than 14 individuals) and overall (species- level). Since sample size of G. lingulatum (n = 319 samples) To represent the genealogical relationship between haplo- is about twice as large as that of G. rex (n = 179), we con- types, a network of COI haplotypes was constructed for sidered a sample size of rarefaction of 179 individuals for each species, using the median-joining algorithm imple- the case of G. lingulatum in order to compare between the mented in NETWORK v5.0 (Bandelt et al. 1999). This two species. method is based on a maximum parsimony algorithm to simplify the complex branching pattern and to represent Estimations of pairwise and overall φST the most parsimonious intraspecifc phylogenies (Polzin and Daneshmand 2003). Population diferentiation between populations of G. lingu- To infer the historical demography of G. lingulatum latum and G. rex, and within species was inferred by cal- and G. rex, we frst calculated neutrality tests, Tajima’s D culating pairwise and overall (species-level) φST-statistics (Tajima 1989), and Fu’s Fs (Fu 1997) statistics for each (FST-like taking into account haplotype frequencies and COI dataset, in order to detect signifcant past changes in amount of diferences among haplotype pairs). Only loca- population size. Signifcant departure from selection-drift tions with a minimum of 14 individuals were included in this equilibrium was tested by 1000 bootstrap replicates in Arle- analysis. Computing values and tests for signifcance were quin. Under the assumption of neutrality, negative values done using non-parametric permutation tests (1000 permuta- characterize populations in expansion while positive values, tions, with Arlequin). Sequential Bonferroni correction was associated to the loss of rare haplotypes, are considered as a used for multiple comparisons. signature of recent bottlenecks. As a complementary approach to infer the historical Geographic structure demography of each species, we compared the observed mismatch distributions of the number of diferences between We evaluated whether locations of G. lingulatum and G. rex pairs of sequences to estimated values under a model of sud- were geographically structured through spatial analysis of den pure demographic expansion (Rogers and Harpending molecular variance (SAMOVA) test using SAMOVA v2.0 1992) and a model of spatial expansion (Excofer 2004) software (Dupanloup et al. 2002). Genetic diferentiation using Arlequin. For each expansion model and each spe- was investigated using a hierarchical analysis of the genetic cies, the ft between observed and estimated mismatch dis- variance by partitioning FST into FSC and FCT indicating tributions was calculated through a generalized least squares the genetic diferentiation of populations within groups approach and tested by 1000 permutations. A multimodal and between groups, respectively. For each COI dataset, distribution generally indicates a population in demographic

1 3

95 Mar Biol (2017) 164:188 Page 7 of 19 188 0.039 0.574 0.095 0.737 0.416 N.A. 0.794 0.508 0.935 0.082 0.273 N.A. N.A. 0.978 N.A. 0.659 0.384 0.355 N.A. 0.021 N.A. 0.323 N.A. N.A. N.A. N.A. 0.476 0.062 N.A. N.A. 0.013 N.A. P value s − 6.287 0.580 −0.838 0.999 −0.384 N.A. 1.345 0.834 3.792 −1.094 −0.388 0.000 0.000 4.866 N.A. 0.635 0.463 −0.523 0.000 − 1.546 0.000 −0.598 0.000 N.A. 0.000 N.A. 0.217 −1.420 0.000 N.A. − 2.070 F N.A. 0.223 0.061 0.146 0.413 0.207 N.A. 0.805 0.034 0.917 0.118 0.293 1.000 1.000 0.972 N.A. 0.290 0.724 0.246 1.000 0.055 1.000 0.190 1.000 N.A. 1.000 N.A. 0.685 0.165 1.000 N.A. 0.059 P value N.A. −0.853 −1.424 −1.165 −0.341 −0.908 N.A. 0.791 − 1.685 1.333 −1.156 −0.438 0.000 0.000 1.750 N.A. −0.743 0.022 −0.742 0.000 −1.491 0.000 −0.941 0.000 N.A. 0.000 N.A. 0.210 −1.010 0.000 N.A. −1.422 D N.A. 0.352 (±0.216) 0.150 (±0.106) 0.118 (±0.102) 0.017 (±0.033) 0.243 (±0.170) 0.267 (±0.187) N.A. 0.292 (±0.197) 0.064 (±0.071) 0.213 (±0.155) 0.012 (±0.027) 0.084 (±0.027) N.A. N.A. 0.293 (±0.197) N.A. 0.201 (±0.147) 0.047 (±0.058) 0.222 (±0.161) N.A. 0.043 (±0.056) N.A. 0.167 (±0.128) N.A. N.A. N.A. N.A. 0.094 (±0.087) 0.098 (±0.092) N.A. N.A. 0.065 (±0.070) % π (±SD) N.A. 0.781 (±0.013) 0.348 (±0.203) 0.186 (±0.110) 0.105 (±0.092) 0.558 (±0.113) 0.593 (±0.144) N.A. 0.601 (±0.111) 0.133 (±0.112) 0.442 (±0.088) 0.077 (±0.070) 0.484 (±0.142) N.A. N.A. 0.456 (±0.085) N.A. 0.463 (±0.120) 0.426 (±0.110) 0.507 (±0.140) N.A. 0.257 (±0.142) N.A. 0.384 (±0.113) N.A. N.A. N.A. N.A. 0.534 (±0.097) 0.467 (±0.148) N.A. N.A. 0.380 (±0.134) H (±SD) N.A. 5 1 6 7 5 5 3 3 1 2 0 0 4 0 6 1 6 0 2 0 6 0 0 0 2 2 3 0 1 3 0 21 S 0 0 1 1 0 1 0 1 0 2 0 0 0 0 0 1 2 0 1 0 2 0 0 1 2 2 3 0 1 3 0 priv 15 h 3 2 4 5 4 4 2 2 2 3 1 1 2 1 4 2 5 1 3 1 5 1 1 1 3 3 4 1 2 4 1 24 h 2 7 9 4 1 21 25 20 14 13 18 15 20 26 14 16 19 19 22 18 17 15 15 17 28 18 14 23 15 20 19 319 N 72°47 ′ W 74º07 ′ W 74º01 ′ W 73º50 ′ W 73°44 ′ W 73°42 ′ W 73º24 ′ W 73º39 ′ W 73°11 ′ W 72°02 ′ W 72°49 ′ W 72°02 ′ W 72°38 ′ W 72°25 ′ W 72°25 ′ W 71°42 ′ W 72°00 ′ W 71°42 ′ W 71°50 ′ W 71°30 ′ W 71°33 ′ W 71º31 ′ W 71º25 ′ W 71°34 ′ W 71°28 ′ W 71°41 ′ W 71°21 ′ W 71°29 ′ W 71°19 ′ W Longitude 71°31 ′ W 71°27 ′ W 42°51 ′ S 42º40 ′ S 41º56 ′ S 41°18 ′ S 40°34 ′ S 40°31 ′ S 39º49 ′ S 37º22 ′ S 37°09 ′ S 34°38 ′ S 36°10 ′ S 34°25 ′ S 35°50 ′ S 33°54 ′ S 35°19 ′ S 33°24 ′ S 34°22 ′ S 33°11 ′ S 33°54 ′ S 31°51 ′ S 32°57 ′ S 31º45 ′ S 32º37 ′ S 31°30 ′ S 32°33 ′ S 30°31 ′ S 29°57 ′ S 29°10 ′ S 29°41 ′ S Latitude Coordinates 28°54 ′ S 29°13 ′ S SBA CUCA MBRA HUAR BAM PUCA CRNC QICO LOT BUCA COB PCH CUR LBO CON PTTR PCH QTAY LBO PAMA MONT CHLO MAIT FUAD ZAP SAUC PAM APOL ARRA BURR PTCH Code . The code used to identify each sampling site, coordinates (latitude and longitude), indices of genetic diversity diversity (latitude and longitude), indices of genetic site, coordinates sampling identify each . The code used to sites of benthicSampling populations of Gelidium lingulatum and G. rex Los Burros Overall COI Overall Mean COI Santa Bárbara Cucao Mar Brava Hua-Huar Bahía Mansa Pucatrihue Curiñanco Quidico Lota Bucalemu Cobquecura Pichilemu Curanipe La Boca Constitución Punta Tralca Pichilemu Quintay La Boca Playa Amarilla Playa Montemar Chigualoco Maitencillo Fundo Agua Dulce Agua Fundo Zapallar El Sauce La Pampilla El Apolillado El Arrayán Punta de Choros Gelidium rex Gelidium lingulatum 1 Table and their indicated tests signifcance are of neutrality results COI, the mitochondrial marker for found site Sampling

1 3

96 188 Page 8 of 19 Mar Biol (2017) 164:188

equilibrium, while a unimodal distribution is associated with a recent pure demographic expansion or a range expansion. 0.004 0.005 0.048 0.027 P value In the particular case of G. rex, in both approaches, the COI dataset was separated into three groups, according to the results of the geographic distribution of haplotypes (see “Results”), s − 3.514 − 4.090 − 3.295 − 5.085 F those locations from 28°S to 31°S (including FUAD named “G. rex north”), the same locations without FUAD (i.e., 28°S to 30°S named “G. rex north-f”), and those locations of the south- ern range (between ~31°S and 34°S, called “G. rex south”). 0.047 0.037 0.119 0.103 P value

Results − 1.368 − 1.509 −1.179 −1.208 D Sequence characteristics

A 622-bp portion of COI was analyzed from 319 individual G. lingulatum, detecting 24 haplotypes with 21 polymorphic sites (Table 1). On the other hand, the 628-bp portion of COI 0.022 (±0.036) 0.059 (±0.064) 0.097 (±0.086) 0.136 (±0.107) 0.029 (±0.039) % π (±SD) sequenced for 179 individual G. rex revealed 11 haplotypes, with 10 polymorphic sites (Table 1). From the 10 individuals of G. lingulatum sequenced also for the rbcL marker (1484- bp alignment), three haplotypes were detected, difering by 2 polymorphic sites (Online Resource 2), while sequencing G. rex rbc 0.134 (±0.046) 0.344 (±0.076) 0.533 (±0.059) 0.628 (±0.029) 0.165 (±0.225) H (±SD) of 12 individuals (1543-bp portion of L) revealed only one single haplotype. G. rex 3 5 6 A query of sequences for the COI haplotypes of , 10 S using a BLAST search, revealed a 100% identity between two of our haplotypes (GR4 and GR9, query cover: 516 bp) 3 5 6 9 priv G. rex

h and a reference sequence identifed as , from Tongoy Bay, Chile (30°15′S; 71°29′W; GenBank Accession Num- ber: HM629875; Kim et al. 2011). Similarly, the unique 4 6 7 11 h rbcL haplotype recovered for G. rex (GR701) presented a 94–100% match with the two reference sequences for the 62 76

103 G. rex 179 N species, both identifed as from Tongoy Bay (Gen- standardized haplotype diversity and SD (after rarefaction); and SD (after rarefaction); diversity S number of polymorphic sites, H standardized haplotype haplotypes, number of private

Fu’s test. Mean and overall values correspond to the average values of the locations and the total for each species, respectively. species, respectively. each of the locations and the total for values correspond values the to average Mean and overall test. Fu’s Bank Accession Number: AF305801, query cover: 1430 bp, s priv Thomas and Freshwater 2001; GenBank Accession Number: HM629835, query cover: 1353 bp, Kim et al. 2011). The taxonomic unit G. lingulatum was absent from GenBank, Longitude but some of our haplotypes presented a 100% identity with sequences registered as Gelidium sp., from Chile (Chun- gungo: 29°26′S; 71°18′W and Caleta Horcón: 32°42′S; ′ Latitude Coordinates 71°29 W). The sequences from Chungungo matched with GL5 and GL702 haplotypes, for COI and rbcL, respectively, while sequences from Caleta Horcón matched with GL3 and GL703 haplotypes, respectively (GenBank accession num- Code bers: COI, JX891593–JX891594; rbcL, JX89619–JX891622; query cover: COI, 511 bp; rbcL: 1354 bp; Boo et al. 2013). Overall, for COI the nucleotide diversity (%π) was 0.352 ± 0.216 and 0.131 ± 0.112 in G. lingulatum and G. rex, respectively, while standardized haplotype diversity was 0.781 ± 0.013 and 0.628 ± 0.029 after rarefaction, being in both cases greater in G. lingulatum than in G. rex (see Overall COI south COI Overall Overall COI north-f COI Overall Overall COI north COI Overall Overall COI Overall Mean COI 1 Table (continued) or uncalculated (locations withN < 14) available in bold. N.A. = not shown ( P < 0.05) are Signifcant values Sampling site Sampling h h number of haplotypes, N number of individuals sequenced, F test, and SD, D Tajima’s diversity of nucleotide % π percentage Table 1). On the other hand, for the rbcL marker, nucleotide

1 3

97 Mar Biol (2017) 164:188 Page 9 of 19 188 diversity and standardized haplotype diversity in G. lingu- them (12 haplotypes) being unique (i.e., haplotypes found latum were %π = 0.059 ± 0.051 and H = 0.711 ± 0.086, only in one single individual), while for G. rex, 9 of 11 hap- respectively. In the case of G. rex, no genetic diversity was lotypes were private (81.5%), but only two of them were observed for this marker. unique. A contrasting pattern among species was observed regarding the distribution of the frequent haplotypes. Phylogeographical patterns The three most frequent COI haplotypes of G. lingulatum were widespread and shared among geographically distant For G. lingulatum, 15 of 24 COI haplotypes were private locations (GL3: 11 locations distributed along the com- (62.5%) (i.e., haplotypes found at a single location), most of plete study range ~29°S–42°S, 2000 km distance; GL5:

Fig. 2 Geographic distribution of haplotypes and haplotype networks of Gelidium lingula- tum for mitochondrial COI and chloroplastic rbcL markers. Sampling locations where no individuals of the species were found from the northern part of the study area are also indicated. Photographs of a specimen and intertidal patches of species are shown in the lower right. The within-location diversity and the geographical extent of each haplotype are shown. On the map each circle represents a location and the proportion of pie chart indicates the frequency of individuals for each haplotype. The pie chart color-code corresponds to the one used in haplotype networks of each marker. In the networks, each circle represents a haplo- type and its size is proportional to the frequency in which the haplotype was encountered (correspondence between circle sizes and numbers of individu- als is indicated). Perpendicular bars between each haplotype pair correspond to the number of mutational steps among them. Abbreviations for location codes are as in Table 1

1 3

98 188 Page 10 of 19 Mar Biol (2017) 164:188

Fig. 3 Geographic distribution of haplotypes and haplotype networks of Gelidium rex for mitochondrial COI and chloro- plastic rbcL markers. Sampling locations where no individu- als of the species were found from the northern and southern sites of the study area are also indicated. Photographs of a specimen and intertidal patches of species are shown in the lower right. See legend of Fig. 2 for details

10 locations between ~29°S and 41°S, 1800 km distance; In the case of the rbcL marker, the three haplotypes in GL2: 9 locations between ~33°S and 42°S, 1550 km dis- G. lingulatum were distributed from ~33°S to 40°S, co- tance; Table 1; Fig. 2). On the other hand, the two most occurring in some locations (CUR and QICO, Fig. 2, see frequent COI haplotypes of G. rex presented disjunct geo- Online Resource 2), whereas for G. rex the single haplo- graphic distributions, with GR4 exclusively found at the type was observed from ~28°S to 34°S (Fig. 3; see Online three northernmost locations (~28°S to 30°S) and GR1 Resource 2). only at the seven southernmost locations (~31°S to 34°S; For the G. lingulatum COI dataset, average values per Fig. 3). In between, the location FUAD presented a sin- sampled location of standardized haplotype diversity (H) gular pattern, as all sampled individuals (i.e., 14) shared and nucleotide diversity (%π) were 0.348 ± 0.203 and the GR2 haplotype, which is private from this location. 0.150 ± 0.106, respectively (Table 1). Of the 20 sampled

1 3

99 Mar Biol (2017) 164:188 Page 11 of 19 188

which are distributed interspersed along the latitudinal gradient covered by the study (i.e., group 1: PAM, LOT, CRNC, CUCA, 1770 km distant between the most extreme sites; group 8: LBO, PCH, CON, MBRA, SBA, 1550 km distant between the most extreme sites; see Fig. 2 and Online Resource 3). In contrast, in G. rex the three detected groups coincide completely with the disjunct distribu- tion of the three most frequent haplotypes, as described above (Online Resource 3). The groups corresponded to (1) northern locations (BURR, APOL and SAUC), (2) the single site FUAD, and (3) southern locations (from CHLO to BUCA). According to the Mantel test, in G. lingulatum the corre- lation between genetic distance and geographic distance was 2 −5 not signifcant (r = 9.658 e , F1,103 = 0.009, P = 0.514; Fig. 4a), while in G. rex the correlation was signifcant 2 (r = 0.103, F1,43 = 4.939, P = 0.047), indicating an iso- lation-by-distance pattern for the latter species (Fig. 4b). However, no isolation-by-distance was observed when the three geographic groups were considered separately, G. rex 2 north (r = 0.199, F1,6 = 2.339, P = 0.177), G. rex north- 2 f (r = 0.835, F1,4 = 2.861, P = 0.166), and G. rex south 2 −4 (r = 1.441 e , F1,37 = 0.283, P = 0.598). The overall φST (at species-level) for G. lingulatum was Fig. 4 Scatter plot of genetic diferentiation and geographic distance 0.629. Most pairwise φST-values were signifcant, indicat- of pairwise locations for COI marker. a Gelidium lingulatum and b G. ing diferentiation among locations. Interestingly, similari- rex D . Pairwise genetic distances, represented as , are plotted against ties between geographically distant (over 1000 km) loca- pairwise geographic distances (km). Each point corresponds to a pairwise comparison of locations. The results of the statistical analy- tions were observed (e.g., LBO, PCH, CON, MBRA and ses and the regression line for signifcant relationship (G. rex) are also SBA) (Fig. 2, Online Resource 4). On the other hand, the shown. Locations with N < 14 were excluded from analyses overall φST for G. rex was 0.859. The φST-values were sig- nifcant for all pairwise comparisons among locations from distinct groups. Likewise, genetic diferentiation was evi- locations, 17 were polymorphic with up to 5 haplotypes per denced among all location pairs from the group G. rex north location (Fig. 2; Table 1). The haplotype network showed (i.e., BURR, APOL, SAUC and FUAD), whereas within the that private haplotypes difered from one of the three most group G. rex south, only the location BUCA was signif- frequent haplotypes mostly by 1 (and up to 3) mutational cantly diferent from the other sampled locations of that zone steps and the maximum pairwise diference among G. lin- (Fig. 3; Online Resource 5). gulatum haplotypes is 7 steps (Fig. 2). A lower diversity was observed for the G. rex COI data- set, with average values per sampled location of standard- Historical demography (COI datasets) ized haplotype diversity (H) and nucleotide diversity (%π) of 0.165 ± 0.225 and 0.029 ± 0.039, respectively (Table 1). The mismatch distribution for the G. lingulatum COI data- Most of the locations were monomorphic (7 of 11) and set was ftted to the sudden demographic expansion model polymorphic locations showed up to 4 haplotypes (Table 1; (SSD = 0.017, P = 0.398), and the spatial expansion model Fig. 3). The haplotypes difered by 1–3 mutational steps in (SSD = 0.015, P = 0.496; Fig. 5a). Neutrality tests on the haplotype network, with all private haplotypes found at overall G. lingulatum COI data also supported partially a 1 single step from one of the two most frequent haplotypes demographic expansion, with a negative Tajima’s D index (Fig. 3). (although not signifcant: D = −0.853, P = 0.223) and a neg- ative and signifcant Fu’s Fs index (Fs = −6.287, P = 0.039; Within‑species genetic structure (COI datasets) Table 1). For G. rex north and G. rex north-f, a demographic The SAMOVA revealed eight diferent groups for G. lin- population expansion was not or only poorly supported gulatum, with only two groups formed by several locations by both tested models (G. rex north: sudden demographic

1 3

100 188 Page 12 of 19 Mar Biol (2017) 164:188

Fig. 5 Mismatch distribution for COI datasets for Gelidium lingulatum (a), G. rex north (b), G. rex north-f (c) and G. rex south (d), according to spatial expansion models. The observed distributions of the number of pairwise diferences (bars) are contrasted to their expected distributions (solid lines) under a model of spatial expansion

expansion, SSD = 0.013, P = 0.045 and spatial expansion, mitochondrial marker and partially sharing the same geo- SSD = 0.013, P = 0.001; G. rex north-f: sudden demo- graphic area of study (Montecinos et al. 2012; Guillemin graphic expansion, SSD = 0.002, P = 0.025 and spatial et al. 2016b). On the other hand, the G. rex pattern has expansion, SSD = 0.002, P = 0.089; Fig. 5b, c). Tajima’s similarities to the shallow genetic structure of M. pyrifera D was not signifcant for G. rex north, while Fu’s Fs indi- (Macaya and Zuccarello 2010a). cated a demographic expansion (D = −1.179, P = 0.119; Fs = −3.295, P = 0.048). In the case of G. rex north-f, both Contrasts in genetic diversity and structure indices showed a demographic expansion (D = −1.509, P = 0.037; Fs = −4.090, P = 0.005). In contrast, for G. rex Our results for genetic diversity in G. lingulatum and G. south stronger evidence was found for population expansion rex are within the observed range of previous studies done both from mismatch analysis (with a higher support for the for other Gelidium species, using the COI marker (e.g., G. sudden demographic expansion: SSD = 0.014, P = 0.309; elegans: h = 34, H = 0.711, %π = 0.734, Kim et al. 2012; G. compared to the spatial expansion model: SSD = 0.014, vagum: h = 17, S = 16, H = 0.844, %π = 0.173, Yoon et al. P = 0.162; Fig. 5d), and from neutrality tests, both signif- 2014; H values not standardized by rarefaction), revealing icant (D = −1.368, P = 0.047; Fs = −3.514, P = 0.004; a high genetic diversity at species-level, particularly in the Table 1). case of G. lingulatum, and after standardizing the H values to the smaller sample size of G. rex. Another red seaweed (Mazzaella laminarioides) from the coast of Chile showed Discussion higher genetic diversity indices (h = 24, S = 62, H = 0.871, %π = 3.42, H value not standardized) compared with G. lin- The two seaweed species presented contrasting genetic gulatum and G. rex, although that study covered a larger geo- diversity and structure. Gelidium lingulatum had higher graphic area (29°S–54°S) (Montecinos et al. 2012). These genetic diversity, but genetic structure did not follow a clear three red seaweed species (G. lingulatum, G. rex, M. lami- geographic pattern, while G. rex had low genetic diversity, narioides) share some characteristics of their habitat (cohab- a phylogeographic break, but shallow genetic structure. In iting in the rocky intertidal shore, partially co-distributed particular, the phylogeographic pattern of G. lingulatum along the Chilean coast) and of their life history (low auton- is not consistent with that observed for other intertidal red omous dispersal capacity, triphasic isomorphic life cycle). seaweeds described for the coast of Chile, using the same Nevertheless, they present contrasting genetic diversity;

1 3

101 Mar Biol (2017) 164:188 Page 13 of 19 188 these diferences could be associated with the distribution dispersal mechanism, which cannot be explained by intrinsic in the intertidal zone (i.e., M. laminarioides lives higher in dispersal abilities alone due to limited autonomous dispersal the intertidal zone than both Gelidium species) or with the potential via spores/gametes. type of vegetative reproduction found in the genus Gelidium Gelidium rex is a species found in the very low intertidal (i.e., fragmentation and re-attachment to the substratum, zone compared to G. lingulatum, which grows closer to the Santelices and Varela 1994; Rojas et al. 1996; Perrone et al. mid-intertidal zone (Santelices 1986). In general, species 2006). In particular G. lingulatum tends to monopolize the that are distributed in an area with less environmental vari- rocky substratum, which could suppress local genetic diver- ability (e.g., longer immersion times below the tidal gradi- sity, in contrast to the individual and unconnected thalli of ent as for G. rex) tend to have less genetic structure (Kelly Mazzaella (Gómez and Westermeier 1991). and Palumbi 2010). This might be due to less patchy dis- The amplitude of the latitudinal range could also have tributions and larger population sizes in the low intertidal implications in the genetic diversity, because a wider geo- or subtidal zones (i.e., distribution width efect, Robuchon graphic range is related to a larger efective population et al. 2014), which would reduce the selection pressure and size and a higher gradient of environmental variability, the action of genetic drift observed in upper intertidal zone which can lead to selection and local adaptation (Alberto (although the distribution of G. rex in the lower intertidal et al. 2010). This could explain the diferences in genetic zone tends to be very patchy rather than being a continu- diversity between the two species studied, considering the ous fringe; Santelices and Abbott 1985). For example, for wider geographic range of G. lingulatum compared to G. two sister species of laminarian kelps co-distributed along rex. Therefore, our results suggest that the amplitude of the the coast of France, Robuchon et al. (2014) showed that geographic range contribute to the diferences in genetic populations of the species inhabiting the shallow subtidal diversity observed for both species, and other red seaweeds zone (Laminaria hyperborea) had less genetic structure than from the Chilean coast. those of the intertidal species (L. digitata). In addition, this In the case of G. lingulatum, genetic structure was is congruent with the observed pattern of M. laminarioides evidenced throughout its range, but without a clear geo- from the mid-intertidal zone, which has a much stronger graphical pattern (i.e., haplotypes disappear and reappear genetic structure than both Gelidium species, showing two repeatedly throughout its geographic range and no phylo- strongly diferentiated haplogroups (separated by 15–45 bp geographic break was detected), and there was no genetic for COI) between 29°S and 37°S, and up to 3 haplogroups isolation-by-distance. On the other hand, G. rex compared considering locations up to 42°S (Montecinos et al. 2012). to G. lingulatum showed a diferent pattern with a disjunct However, if both species of Gelidium are compared, genetic haplotype distribution where a separation occurs at ~31°S structure of G. lingulatum did not follow a clear geographic between the northern, FUAD, and southern populations of pattern in contrast to G. rex and overall (species-level) φST- its geographic range. This coincides with the biogeographic value is lower for G. lingulatum (0.629) than G. rex (0.859). break at 30°S (Camus 2001) and is also consistent with the Therefore, our results do not support the hypothesis that sea- phylogeographic breaks described for that region for many weed species from the mid-intertidal zone have more genetic intertidal species of invertebrates and macroalgae with lim- structure compared with organisms from lower zones and it ited dispersal abilities (i.e., 29°S–33°S, Tellier et al. 2009; suggests that other factors may be important. A trend of less Sánchez et al. 2011; Montecinos et al. 2012; Haye et al. genetic structure in species from the upper intertidal zone 2014; Guillemin et al. 2016a). However, unlike other inter- has also been observed in two intertidal barnacles (Jehlius tidal seaweeds (e.g., Lessonia nigrescens complex, Tellier cirratus and Notochthamalus scabrosus) from the Chilean et al. 2011; M. laminarioides, Montecinos et al. 2012), this coast (18°S–54°S) (Zakas et al. 2009; Ewers-Saucedo et al. geographical subdivision is not based on a strong genetic 2016; Guo and Wares 2017). Future studies should also use diference, since the separation between populations is only complementary nuclear markers to improve understanding one mutational step, similar to the shallow genetic structure of the genetic structure of these species within their tidal described for M. pyrifera, a kelp species with high dispersal distribution. potential via rafting (Macaya and Zuccarello 2010a). Indeed, the rbcL marker, despite the low sample sizes, Phylogeographic patterns as a result of rafting dispersal revealed no indication for a phylogeographic break due to the complete absence and lower polymorphism for this In G. lingulatum, the lack of a phylogeographic break, marker in G. rex and G. lingulatum, respectively. This is patchy distribution of haplotypes within its geographic consistent with the generally lower mutation rate of this range and no isolation-by-distance are indications of marker, compared to COI (Engel et al. 2008; Grant 2016). long-distance dispersal events. Human-mediated trans- Therefore, all these results suggest that G. lingulatum and port (Banks et al. 2015) or rafting dispersal (e.g., wood G. rex (more evident in G. lingulatum) have a long-distance and foating seaweeds, Thiel and Gutow 2005a) can move

1 3

102 188 Page 14 of 19 Mar Biol (2017) 164:188 organisms over long distances. In the case of human-medi- Long-distance dispersal could also be consistent with the ated dispersal, transport through maritime trafc (ballast historical patterns observed. For G. lingulatum and G. rex a waters and ship hulls) has been shown to afect the phy- recent population expansion was detected, although in the logeographic patterns of other seaweeds (e.g., Undaria latter this was only observed in southern populations, while pinnatifda, Voisin et al. 2005; Caulerpa cylindracea, in northern populations (particularly, those from BURR to Piazzi et al. 2016). Although G. lingulatum germlings SAUC) this pattern was not so clear. This coincides with the have a high tolerance to total darkness (Santelices et al. high presence of private haplotypes in both species, which 2002), conditions in ballast waters are strongly adverse in the case of G. rex were detected only at the northernmost (i.e., anoxia) and maritime trafc occurs ofshore and in and southernmost sampling sites (i.e., BURR and BUCA). ports, not in rocky areas, so that this transport mechanism Nikula et al. (2010) reported genetic signatures of popula- is much less likely than through rafting dispersal by foat- tion expansion in epifaunal invertebrates (i.e., peracarids) ing seaweeds. In addition, at least one species from the associated with holdfasts of foating bull kelp D. antarc- genus Gelidium has been detected on derelict aquaculture tica in subantarctic areas. This suggests that rapid histori- buoys in Coquimbo Bay (30°S) (Astudillo et al. 2009), cal population growth might have been favored by frequent and so other foating substrata cannot be completely ruled rafting events. out as dispersal vehicle. Another long-distance dispersal Successful immigration after rafting journeys is likely mechanism through drifting fronds has been also reported also infuenced by other factors such as substratum avail- in G. versicolor on the south coast of England (Dixon and ability, settlement capacity of immigrant propagules, and Irvine 1977). the density of the resident population (i.e., density block- Detached seaweeds are one of the most common foating ing, Waters et al. 2013; Neiva et al. 2014). In dense local substrata along the coast of Chile (Hinojosa et al. 2010; populations, new haplotypes that arrive with few immigrant 2011; Wichmann et al. 2012). Rafting transport could individuals have a high probability of being outcompeted increase the gene fow between distant populations and because of their rarity, which leads to rapid elimination of thus modify the genetic structure, as had been described these new haplotypes by genetic drift. For example, this for some invertebrates inhabiting holdfasts of D. antarctica could be happening for G. rex in locations such as FUAD, (Nikula et al. 2010; Haye et al. 2012). In Chile, both G. where a single private haplotype was very frequent among lingulatum and G. rex are often found attached to holdfasts the sampled individuals. This is congruent with records in of foating and recently stranded bull kelps D. antarctica locations adjacent to FUAD (30°S–31°S), where higher (higher frequencies in G. lingulatum than G. rex, Macaya population abundances of this species have been observed et al. 2016), but the continental clade (30°S–44°S) of this in comparison to northern and southern sites (Broitman et al. bull kelp presents a very diferent phylogeographic pattern 2001, Vásquez and Vega 2004). This could be because, as (Fraser et al. 2010) than the two red seaweeds. For exam- observed in other species of Gelidium (i.e., G. arbuscula, ple, D. antarctica has a much more genetically structured Sosa and García-Reina 1992; Sosa et al. 1998), stoloniferous pattern (i.e., more mutational steps among pairs of haplo- outgrowths of creeping axes is a common way of propa- types) compared to G. rex and its geographical haplotype gation; therefore, locally adapted clones could propagate distribution is not similar to the patchy pattern of G. lin- asexually and became predominant through competitive gulatum. This suggests that other factors during, or after, advantage, thereby minimizing the availability of unoccu- along-shore rafting journeys could be afecting connectivity pied substratum and limiting opportunities for recruitment among distant populations. Moreover, Macaya et al. (2016) of new genotypes (via sexual reproduction). In addition, the suggested that the physiological capacity to tolerate new ability of thallus reattachment of these species (Rojas et al. environmental conditions at the sea surface during rafting 1996) would favor the monopolization of the substratum. For might be directly related to the bathymetric distribution example, Alberto et al. (1999) suggested that populations of pattern of seaweeds in their benthic habitats. Particularly G. sesquipedale from northern France to Morocco maintain in turf algae, changes in solar radiation levels during trans- the gene fow among populations (<500 km) through occa- fer from the benthic to the pelagic environment (rafting at sional transport of detached fronds by local currents during the sea surface) could afect performance and persistence storm events and subsequent reattachment to new substrata. of these algae. Given the intertidal distribution of the two Conversely, strong disturbances with massive local mor- study species, these shifts in light regime should be more talities (e.g., coastal uplifts after earthquakes) could change critical in G. rex than in G. lingulatum. In addition, the dif- this pattern (Castilla et al. 2010; Jaramillo et al. 2012), ference in latitudinal distribution between the two species enhancing the possibility of successful immigration to (i.e., wider in G. lingulatum compared to G. rex) could also uncolonized habitats or those with lower population den- suggest that there are diferent tolerance capacities to harsh sity. Habitat heterogeneity could also be an important fac- conditions between them. tor infuencing phylogeography and population connectivity

1 3

103 Mar Biol (2017) 164:188 Page 15 of 19 188 in intertidal seaweeds. For example, the extent of sandy depending on seaweed biology (e.g., functional and repro- beaches (particularly from 36°S to 40°S on the coast of ductive characteristics of these epiphytic non-buoyant sea- Chile, Thiel et al. 2007) could reduce the availability of pri- weeds) and population ecology (e.g., density blocking). mary substratum for intertidal seaweeds inhabiting rocky Further studies should in particular focus on rafting routes, shores and thus, genetic drift and small efective popula- via genetic characterization of the source populations of tion sizes probably contribute strongly to the divergence stranded D. antarctica, particularly those holdfasts car- between their populations (Fraser et al. 2010). Therefore, rying G. rex or G. lingulatum individuals. Recent stud- our results suggest that the phylogeographic patterns of these ies indicate that supplies of bull kelp rafts to the shore intertidal algae are afected by rafting dispersal via foat- vary strongly along the coast of Chile (López et al. 2017, ing seaweeds, although there may be diferential functional Fig. 1), which could afect connectivity among the popula- capabilities during rafting journeys and/or diferential ef- tions of D. antarctica and of associated epibionts. ciency of recruitment after long-distance dispersal that could explain the divergent patterns between both species. Future Acknowledgements This study was fnanced by the following grants: CONICYT/FONDECYT 1131082 to MT, F. Tala and F. Tellier, CONI- studies should also focus on phenology and the relationships CYT/FONDECYT 1110437 to EM, and CONICYT/FONDECYT between diferent phases of the life cycle in these species. 11121504 to F. Tellier. BL received fnancial support by PhD-fellow- ship Beca CONICYT-PCHA/Doctorado Nacional/2014-21140010. Additional support came from International Research Network “Diversity, Evolution and Biotechnology of Marine Algae” (GDRI N ̊ Conclusions and outlook 0803). The collaboration of Óscar Pino, José Pantoja, Alvaro Gallardo, Solange Pacheco, Ricardo Jeldres, María Fabiola Monsalvez, Ariel Our phylogeographic study confrms the presence of G. lin- Cáceres, Ulyces Urtubia, Vieia Villalobos and Tim Kiessling in feld gulatum along the Chilean coast at least from 29°S to 42°S activities is gratefully acknowledged. The valuable comments from two anonymous referees were very helpful in improving the original manu- (no recent records are available for the south, 42°S–56°S, script. We are grateful to Lucas Eastman for checking the language of John et al. 2003; Soto et al. 2012), but our surveys suggest the fnal manuscript. that this species does not occur north of 29°S. Similarly, we only found individuals of G. rex between 28°S and 34°S, Compliance with ethical standards despite a reported distribution ranging from 16°S to 39°S (Santelices and Abbott 1985), thus suggesting a previous Confict of interest All authors declare that they have no confict of interests. overestimation of the geographical range in both study spe- cies (Fig. 1). Human and animals rights This article does not contain any studies Gelidium lingulatum had some genetic structure (i.e., φST with human participants or animals performed by any of the authors. values are highly signifcant among several locations), but did not follow a clear geographic pattern (i.e., no phylo- geographic break, and haplotypes disappear and reappear repeatedly along its geographical range), contrasting with References fndings for other red seaweeds with similar life histories and distribution ranges (e.g., M. laminarioides, Montecinos Alberto F, Santos R, Leitão JM (1999) Assessing patterns of geo- et al. 2012; Nothogenia chilensis, Lindstrom et al. 2015). graphic dispersal of Gelidium sesquipedale (Rhodophyta) A shallow genetic structure was observed in G. rex, with a through RAPD diferentiation of populations. Mar Ecol Prog Ser 191:101–108. doi:10.3354/Meps191101 phylogeographic break coinciding with the phylogeographic Alberto F, Raimondi PT, Reed DC, Coelho NC, Leblois R, Whitmer discontinuity described for other species between 29°S and A, Serrão EA (2010) Habitat continuity and geographic dis- 33°S (Tellier et al. 2009; Sánchez et al. 2011; Montecinos tance predict population genetic diferentiation in giant kelp. et al. 2012). We propose that these contrasting patterns of Ecology 91:49–56. doi:10.1890/09-0050.1 G. lingulatum G. rex Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) and might be due to (1) diferences in Basic local alignment search tool. J Mol Biol 215:403–410. tidal level and species-specifc adaptations in physiology and doi:10.1016/S0022-2836(05)80360-2 reproductive biology (e.g., temperatures below 10 °C are Aravena G, Broitman B, Stenseth NC (2014) Twelve years of change limiting for growth of these species, Oliger and Santelices in coastal upwelling along the central-northern coast of Chile: spatially heterogeneous responses to climatic variability. PLoS 1981) and (2) diferences in extrinsic dispersal capacities, One 9:e90276. doi:10.1371/journal.pone.0090276 with more efective rafting dispersal for G. lingulatum than Astudillo JC, Bravo M, Dumont CP, Thiel M (2009) Detached aqua- for G. rex. culture buoys in the SE Pacifc: potential dispersal vehicles As previously suggested (Macaya et al. 2016), while for associated organisms. Aquat Biol 5:219–231. doi:10.3354/ ab00151 our study provides support for efcient rafting dispersal, Ballard JW, Whitlock MC (2004) The incomplete natu- it also indicates that the relative contribution of rafting ral history of mitochondria. Mol Ecol 13:729–744. to contemporaneous population connectivity may vary, doi:10.1046/j.1365-294X.2003.02063.x

1 3

104 188 Page 16 of 19 Mar Biol (2017) 164:188

Bandelt HJ, Forster P, Rohl A (1999) Median-joining networks for Gracilaria verrucosa (Gracilariales, Rhodophyta). Bot Mar inferring intraspecifc phylogenies. Mol Biol Evol 16:37–48 35:93–98. doi:10.1515/botm.1992.35.2.93 Banks NC, Paini DR, Bayliss KL, Hodda M (2015) The role of global Dixon PS, Irvine LM (1977) Seaweeds of the British Isles. Volume 1. trade and transport network topology in the human-mediated Rhodophyta. Part 1. Introduction, Nemaliales, Gigartinales. The dispersal of alien species. Ecol Lett 18:188–199. doi:10.1111/ Natural History Museum, London, p 252 ele.12397 Dupanloup I, Schneider S, Excofer L (2002) A simulated annealing Billard E, Daguin C, Pearson G, Serrão E, Engel C, Valero M (2005) approach to defne the genetic structure of populations. Mol Ecol Genetic isolation between three closely related taxa: Fucus vesic- 11:2571–2581. doi:10.1046/j.1365-294X.2002.01650.x ulosus, F. spiralis, and F. ceranoides (Phaeophyceae). J Phycol Engel CR, Destombe C, Valero M (2004) Mating system and gene 41:900–905. doi:10.1111/j.0022-3646.2005.04221.x fow in the red seaweed Gracilaria gracilis: efect of haploid- Bobadilla M, Santelices B (2005) Variations in the dispersal curves diploid life history and intertidal rocky shore landscape on of macroalgal propagules from a source. J Exp Mar Biol Ecol fne-scale genetic structure. Heredity 92:289–298. doi:10.1038/ 327:47–57. doi:10.1016/j.jembe.2005.06.006 sj.hdy.6800407 Boo GH, Park JK, Boo SM (2013) Gelidiophycus (Rhodophyta: Engel CR, Billard E, Voisin M, Viard F (2008) Conservation and Gelidiales): a new genus of marine algae from East Asia. Taxon polymorphism of mitochondrial intergenic sequences in 62:1105–1116. doi:10.12705/626.7 brown algae (Phaeophyceae). Eur J Phycol 43:195–205. Boo GH, Mansilla A, Nelson W, Bellgrove A, Boo SM (2014a) Genetic doi:10.1080/09670260701823437 connectivity between trans-oceanic populations of Capreolia Ewers-Saucedo C, Pringle JM, Sepúlveda HH, Byers JE, Navarrete SA, implexa (Gelidiales, Rhodophyta) in cool temperate waters of Wares JP (2016) The oceanic concordance of phylogeography Australasia and Chile. Aquat Bot 119:73–79. doi:10.1016/j. and biogeography: a case study in Notochthamalus. Ecol Evol aquabot.2014.08.004 6:4403–4420. doi:10.1002/ece3.2205 Boo GH, Kim KM, Nelson WA, Riosmena-Rodríguez R, Yoon KJ, Excoffier L (2004) Patterns of DNA sequence diversity Boo SM (2014b) Taxonomy and distribution of selected species and genetic structure after a range expansion: lessons of the agarophyte genus Gelidium (Gelidiales, Rhodophyta). J from the infinite-island model. Mol Ecol 13:853–864. Appl Phycol 26:1243–1251. doi:10.1007/s10811-013-0111-7 doi:10.1046/j.1365-294X.2003.02004.x Boo GH, Le Gall L, Miller KA, Freshwater DW, Wernberg T, Terada Excofer L, Lischer HEL (2010) Arlequin (version 3.5): arlequin suite R, Yoon KJ, Boo SM (2016) A novel phylogeny of the Gelidiales ver 3.5: A new series of programs to perform population genetics (Rhodophyta) based on fve genes including the nuclear CesA, analyses under Linux and Windows. Mol Ecol Resour 10:564– with descriptions of Orthogonacladia gen. nov and Orthogona- 567. doi:10.1111/j.1755-0998.2010.02847.x cladiaceae fam. nov. Mol Phylogenet Evol 101:359–372. Försterra G (2009) Ecological and biogeographical aspects of the Chil- doi:10.1016/j.ympev.2016.05.018 ean Fjord region. In: Häussermann V, Försterra G (eds) Marine Bouza N, Caujape-Castells J, González-Pérez MA, Sosa PA (2006) benthic fauna of chilean patagonia. Nature in Focus, Puerto Genetic structure of natural populations in the red algae Gelid- Montt, pp 61–76 ium canariense (Gelidiales, Rhodophyta) investigated by ran- Fraser CI, Nikula R, Spencer HG, Waters JM (2009a) Kelp genes reveal dom amplifed polymorphic DNA (RAPD) markers. J Phycol efects of subantarctic sea ice during the Last Glacial Maxi- 42:304–311. doi:10.1111/j.1529-8817.2006.00201.x mum. Proc Natl Acad Sci USA 106:3249–3253. doi:10.1073/ Broitman BR, Navarrete SA, Smith F, Gaines SD (2001) Geographic pnas.0810635106 variation of southeastern Pacifc intertidal communities. Mar Fraser CI, Hay CH, Spencer HG, Waters JM (2009b) Genetic Ecol Prog Ser 224:21–34. doi:10.3354/Meps224021 and morphological analyses of the southern bull kelp Dur- Camus PA (2001) Biogeografía marina de Chile continental. Rev Chil villaea antarctica (Phaeophyceae: Durvillaeales) in New Hist Nat 74:587–617 Zealand reveal cryptic species. J Phycol 45:436–443. Castilla JC, Manríquez PH, Camaño A (2010) Efects of rocky shore doi:10.1111/j.1529-8817.2009.00658.x coseismic uplift and the 2010 Chilean mega-earthquake on Fraser CI, Thiel M, Spencer HG, Waters JM (2010) Contem- intertidal biomarker species. Mar Ecol Prog Ser 418:17–23. porary habitat discontinuity and historic glacial ice drive doi:10.3354/meps08830 genetic divergence in Chilean kelp. BMC Evol Biol 10:203. Coyer JA, Hoarau G, Costa JF, Hogerdijk B, Serrão EA, Billard E, doi:10.1186/1471-2148-10-203 Valero M, Pearson GA, Olsen JL (2011a) Evolution and diversi- Freshwater DW, Rueness J (1994) Phylogenetic relationships of some fcation within the intertidal brown macroalgae Fucus spiralis/F. European Gelidium (Gelidiales, Rhodophyta) species, based vesiculosus species complex in the North Atlantic. Mol Phylo- on rbcL nucleotide sequence analysis. Phycologia 33:187–194. genet Evol 58:283–296. doi:10.1016/j.ympev.2010.11.015 doi:10.2216/i0031-8884-33-3-187.1 Coyer JA, Hoarau G, Van Schaik J, Luijckx P, Olsen JL (2011b) Trans- Fu YX (1997) Statistical tests of neutrality of mutations against popu- Pacifc and trans-Arctic pathways of the intertidal macroalga lation growth, hitchhiking and background selection. Genetics Fucus distichus L. reveal multiple glacial refugia and coloniza- 147:915–925 tions from the North Pacifc to the North Atlantic. J Biogeogr Gavio B, Fredericq S (2002) Grateloupia turuturu (Halymeniaceae, 38:756–771. doi:10.1111/j.1365-2699.2010.02437.x Rhodophyta) is the correct name of the non-native species in Cumming RA, Nikula R, Spencer HG, Waters JM (2014) Transoce- the Atlantic known as Grateloupia doryphora. Eur J Phycol anic genetic similarities of kelp-associated sea slug populations: 37:349–359. doi:10.1017/s0967026202003839 long-distance dispersal via rafting? J Biogeogr 41:2357–2370. Gómez IM, Westermeier RC (1991) Frond regrowth from basal disc doi:10.1111/jbi.12376 in Iridaea laminarioides (Rhodophyta, Gigartinales) at Mehuín, Dawson MN, Hays CG, Grosberg RK, Raimondi PT (2014) Dispersal southern Chile. Mar Ecol Prog Ser 73:83–91. doi:10.3354/ potential and population genetic structure in the marine inter- meps073083 tidal of the eastern North Pacifc. Ecol Monogr 84:435–456. González S, Stotz W, Toledo P, Jorquera M, Romero M (1991) Uti- doi:10.1890/13-0871.1 lización de diferentes microambientes del intermareal como Destombe C, Godin J, Lefebvre C, Dehorter O, Vernet P (1992) Dif- lugares de asentamiento por Fissurella spp (Gastropoda: Proso- ferences in dispersal abilities of haploid and diploid spores of branchia) (Palo Colorado, Los Vilos, Chile). Rev Biol Mar

1 3

105 Mar Biol (2017) 164:188 Page 17 of 19 188

Oceanogr 26: 325–338. http://www.revbiolmar.cl/escanea- National Park, Estero Elefantes, Chonos Archipelago and Kata- dos/262-325.pdf. Accessed 29 Aug 2016 lalixar). The Natural History Museum, London González A, Beltrán J, Hiriart-Bertrand L, Flores V, de Reviers B, Cor- Kelly RP, Palumbi SR (2010) Genetic structure among 50 species rea JA, Santelices B (2012) Identifcation of cryptic species in of the Northeastern Pacifc rocky intertidal community. PLoS the Lessonia nigrescens complex (Phaeophyceae, Laminariales). One 5:13. doi:10.1371/journal.pone.0008594 J Phycol 48:1153–1165. doi:10.1111/j.1529-8817.2012.01200.x Kiessling T, Gutow L, Thiel M (2015) Marine litter as habitat and Grant SW (2016) Paradigm shifts in the phylogeographic dispersal vector. In: Bergmann M, Gutow L, Klages K (eds) analysis of seaweeds. In: Hu ZM, Fraser CI (eds) Sea- Marine anthropogenic litter, Springer, Berlin, pp 141–184. weed phylogeography. Springer, Dordrecht, pp 23–62. doi:10.1007/978-3-319-16510-3_6 doi:10.1007/978-94-017-7534-2_2 Kim KM, Boo SM (2012) Phylogenetic relationships and distribution Guillemin ML, Valero M, Faugeron S, Nelson W, Destombe C (2014) of Gelidium crinale and G. pusillum (Gelidiales, Rhodophyta) Tracing the trans-pacifc evolutionary history of a domesticated using cox1 and rbcL sequences. Algae 27:83–94. doi:10.4490/ seaweed (Gracilaria chilensis) with archaeological and genetic algae.2012.27.2.083 data. PLoS One 9:e114039. doi:10.1371/journal.pone.0114039 Kim KM, Hwang IK, Park JK, Boo SM (2011) A new agarophyte Guillemin ML, Valero M, Tellier F, Macaya EC, Destombe C, Faugeron species, Gelidium eucorneum sp. nov. (Gelidiales, Rhodo- S (2016a) Phylogeography of seaweeds in the South East Pacifc: phyta), based on molecular and morphological data. J Phycol complex evolutionary processes along a latitudinal gradient In: 47:904–910. doi:10.1111/j.1529-8817.2011.01005.x Hu ZM, Fraser CI (eds) Seaweed Phylogeography. Springer, Kim KM, Hoarau GG, Boo SM (2012) Genetic structure and dis- Dordrecht, pp 251–278. doi:10.1007/978-94-017-7534-2_10 tribution of Gelidium elegans (Gelidiales, Rhodophyta) in Guillemin ML, Contreras-Porcia L, Ramírez ME, Macaya EC, Bulboa- Korea based on mitochondrial cox1 sequence data. Aquat Bot Contador C, Woods H, Wyatt C, Brodie J (2016b) The bladed 98:27–33. doi:10.1016/j.aquabot.2011.12.005 Bangiales (Rhodophyta) of the South Eastern Pacifc: molecular Krueger-Hadfeld SA, Hoban SM (2016) The importance of efective species delimitation reveals extensive diversity. Mol Phylogenet sampling for exploring the population dynamics of haploid– Evol 94:814–826. doi:10.1016/j.ympev.2015.09.027 diploid seaweeds. J Phycol 52:1–9. doi:10.1111/jpy.12366 Guo B, Wares JP (2017) Large-scale gene fow in the barnacle Jehlius Krueger-Hadfeld SA, Roze D, Mauger S, Valero M (2013) Inter- cirratus and contrasts with other broadly-distributed taxa along gametophytic selfng and microgeographic genetic structure the Chilean coast. PeerJ 5:e2971. doi:10.7717/peerj.2971 shape populations of the intertidal red seaweed Chondrus Hall TA (1999) BioEdit: a user-friendly biological sequence alignment crispus. Mol Ecol 22:3242–3260. doi:10.1111/mec.12191 editor and analysis program for Windows 95/98/NT. Nucl Acids Kützing FT (1868) Tabulae phycologicae; oder, Abbildungen der Symp Ser 41:95–98 Tange. Vol. XVIII pp. [i–iii], 1–35, 100 pls. Nordhausen: Haye PA, Varela AI, Thiel M (2012) Genetic signatures of rafting Gedruckt auf kosten des Verfassers (in commission bei W. dispersal in algal-dwelling brooders Limnoria spp. (Isopoda) Köhne) along the SE Pacifc (Chile). Mar Ecol Prog Ser 455:111–122. Lachkar Z, Gruber N (2012) A comparative study of biological pro- doi:10.3354/meps09673 duction in eastern boundary upwelling systems using an arti- Haye PA, Segovia NI, Muñoz-Herrera NC, Gálvez FE, Martínez A, fcial neural network. Biogeosciences 9:293–308. doi:10.5194/ Meynard A, Pardo-Gandarillas MC, Poulin E, Faugeron S (2014) bg-9-293-2012 Phylogeographic structure in benthic marine invertebrates of the Lin SM, Fredericq S, Hommersand MH (2001) Systematics southeast Pacifc coast of Chile with difering dispersal potential. of the Delesserlaceae (Ceramiales, Rhodophyta) based PLoS One 9:e88613. doi:10.1371/journal.pone.0088613 on large subunit rDNA and rbcL sequences, including Hernández CJ (1997) Análisis de la variación estacional e interanual de the Phycodryoideae, subfam. nov. J Phycol 37:881–899. la cosecha de Gelidium robustum en Baja California Sur, México. doi:10.1046/j.1529-8817.2001.01012.x Master thesis, Centro Interdisciplinario de Ciencias Marinas, La Lindstrom SC, Gabrielson PW, Hughey JR, Macaya EC, Nelson WA Paz, Baja California Sur, México (2015) Sequencing of historic and modern specimens reveals Hinojosa IA, Pizarro M, Ramos M, Thiel M (2010) Spatial and cryptic diversity in Nothogenia (Scinaiaceae, Rhodophyta). temporal distribution of floating kelp in the channels and Phycologia 54:97–108. doi:10.2216/14-077.1 fords of southern Chile. Estuar Coast Shelf Sci 87:367–377. López B, Macaya EC, Tellier F, Tala F, Thiel M (2017) The variable doi:10.1016/j.ecss.2009.12.010 routes of rafting: stranding dynamics of foating bull-kelp Dur- Hinojosa IA, Rivadeneira MM, Thiel M (2011) Temporal and spatial villaea antarctica (Fucales, Phaeophyceae) on beaches in the distribution of foating objects in coastal waters of central-south- SE Pacifc. J Phycol 54:70–84. doi:10.1111/jpy.12479 ern Chile and Patagonian fords. Cont Shelf Res 31:172–186. Macaya EC, Zuccarello GC (2010a) Genetic structure of the giant doi:10.1016/j.csr.2010.04.013 kelp Macrocystis pyrifera along the southeastern Pacifc. Mar Hofmann A, Santelices B (1997) Flora marina de Chile Central. Edi- Ecol Prog Ser 420:103–112. doi:10.3354/meps08893 ciones Universidad Católica de Chile, Santiago de Chile Macaya EC, Zuccarello GC (2010b) DNA barcoding and genetic Hu ZM, De-Lin D, López-Baptista J (2016) Seaweed phylogeog- divergence in the giant kelp Macrocystis (Laminariales). J raphy from 1994 to 2014: an overview. In: Hu ZM, Fraser CI Phycol 46:736–742. doi:10.1111/j.1529-8817.2010.00845.x (eds) Seaweed Phylogeography. Springer, Dordrecht, pp 3–22. Macaya EC, López B, Tala F, Tellier F, Thiel M (2016) Float and doi:10.1007/978-94-017-7534-2_1 raft: role of buoyant seaweeds in the phylogeography and Jaramillo E, Dugan JE, Hubbard DM, Melnick D, Manzano M, Duarte genetic structure of non-buoyant associated fora. In: Hu ZM, C, Campos C, Sánchez R (2012) Ecological implications of Fraser CI (eds) Seaweed phylogeography, Springer, Dordrecht, extreme events: footprints of the 2010 earthquake along the pp 97–130. doi:10.1007/978-94-017-7534-2_4 Chilean coast. PLoS One 7:8. doi:10.1371/journal.pone.0035348 Matsuhiro B, Urzúa CC (1990) Agars from Gelidium rex (Gelidi- John D, Paterson G, Evans N, Ramírez M, Spencer J, Báez P, Ferrero ales, Rhodophyta). Hydrobiologia 204:545–549. doi:10.1007/ T, Valentine C, Reid D (2003) Manual de biotopos marinos de bf00040284 la región de Aysén, Sur de Chile (A manual of marine biotopes Matsuhiro B, Urzúa CC (1991) Agars from Chilean Gelidiaceae. of Region Aysén, Southern Chile. The Laguna San Raphael Hydrobiologia 221:149–156. doi:10.1007/bf00028371

1 3

106 188 Page 18 of 19 Mar Biol (2017) 164:188

Melo RA (1998) Gelidium commercial exploitation: natural resources co-distributed along the coast of Brittany, France. Mol Ecol and cultivation. J Appl Phycol 10:303–314. doi:10.102 23:2669–2685. doi:10.1111/mec.12774 3/A:1008070419158 Rogers AR, Harpending H (1992) Population growth makes waves Montecinos A, Broitman BR, Faugeron S, Haye PA, Tellier F, Guil- in the distribution of pairwise genetic diferences. Mol Biol lemin ML (2012) Species replacement along a linear coastal Evol 9:552–569 habitat: phylogeography and speciation in the red alga Mazzaella Rojas R, León N, Rojas R (1996) Practical and descriptive techniques laminarioides along the south east Pacifc. BMC Evol Biol 12:97. for Gelidium rex (Gelidiales, Rhodophyta) culture. Hydrobio- doi:10.1186/1471-2148-12-97 logia 327:367–370 Muhlin JF, Engel CR, Stessel R, Weatherbee RA, Brawley SH (2008) Sánchez R, Sepúlveda RD, Brante A, Cárdenas L (2011) Spatial The infuence of coastal topography, circulation patterns, and pattern of genetic and morphological diversity in the direct rafting in structuring populations of an intertidal alga. Mol Ecol developer Acanthina monodon (Gastropoda: Mollusca). Mar 17:1198–1210. doi:10.1111/j.1365-294X.2007.03624.x Ecol Prog Ser 434:121–131. doi:10.3354/meps09184 Neiva J, Assis J, Fernandes F, Pearson GA, Serrão EA, Maggs C Santelices B (1980) Phytogeographic characterization of the tem- (2014) Species distribution models and mitochondrial DNA perate coast of Pacifc South-America. Phycologia 19:1–12. phylogeography suggest an extensive biogeographical shift in doi:10.2216/i0031-8884-19-1-1.1 the high-intertidal seaweed Pelvetia canaliculata. J Biogeogr Santelices B (1986) The wild harvest and culture of the economically 41:1137–1148. doi:10.1111/jbi.12278 important species of Gelidium in Chile. In: Doty MS, Caddy Nelson WA, Farr TJ, Broom JES (2006) Phylogenetic diversity of New JS, Santelices B (eds) Case of studies of seven commercial Zealand gelidiales as revealed by rbcL sequence data. J Appl seaweed resources. FAO Fisheries Technical Paper, vol 281, Phycol 18:653–661. doi:10.1007/s10811-006-9068-0 FAO, Rome, pp 165–192 Nikula R, Fraser CI, Spencer HG, Waters JM (2010) Circumpolar dis- Santelices B (1990) New and old problems in the taxonomy of persal by rafting in two subantarctic kelp-dwelling crustaceans. the Gelidiales (Rhodophyta). Hydrobiologia 204:125–135. Mar Ecol Prog Ser 405:221–230. doi:10.3354/meps08523 doi:10.1007/bf00040224 Nikula R, Spencer HG, Waters JM (2011a) Evolutionary conse- Santelices B, Abbott IA (1985) Gelidium rex sp. nov. (Gelidiales. quences of microhabitat: population-genetic structuring in Rhodophyta) from central Chile. In: Abbott IA, Norris JN kelp- vs. rock-associated chitons. Mol Ecol 20:4915–4924. (eds) Taxonomy of economic seaweeds with reference to some doi:10.1111/j.1365-294X.2011.05332.x Pacifc and Caribbean species. California Sea Grant College Nikula R, Spencer HG, Waters JM (2011b) Comparison of population- Program, La Jolla, pp 33–36 genetic structuring in congeneric kelp- versus rock-associated Santelices B, Montalva S (1983) Taxonomic studies on Gelidiaceae snails: a test of a dispersal-by-rafting hypothesis. Ecol Evol (Rhodophyta) from central Chile. Phycologia 22:185–196. 1:169–180. doi:10.1002/ece3.16 doi:10.2216/i0031-8884-22-2-185.1 Nikula R, Spencer HG, Waters JM (2013) Passive rafting is a pow- Santelices B, Stewart JG (1985) Pacifc species of Gelidium Lamour- erful driver of transoceanic gene fow. Biol Lett 9:20120821. oux and other Gelidiales (Rhodophyta), with keys and descrip- doi:10.1098/rsbl.2012.0821 tions to the common or economically important species. In: Oliger P, Santelices B (1981) Physiological ecology stud- Abbott IA, Norris JN (eds) Taxonomy of economic seaweeds ies on Chilean Gelidiales. J Exp Mar Biol Ecol 53:65–75. with reference to some Pacifc and Caribbean species. Califor- doi:10.1016/0022-0981(81)90084-8 nia Sea Grant College Program, California, La Jolla, pp 17–31 Ortega M, Godínez-Ortega J, Garduño G (2001) Catálogo de algas Santelices B, Varela D (1994) Abiotic control of reattachment in bénticas de las costas mexicanas del Golfo de México y Mar Gelidium chilense (Montagne) Santelices and Montalva Caribe. Instituto de Biología, Universidad Nacional Autónoma (Gelidiales, Rhodophyta). J Exp Mar Biol Ecol 177:145–155. de México, México doi:10.1016/0022-0981(94)90233-x Palumbi SR (1994) Genetic divergence, reproductive isolation, and Santelices B, Aedo D, Hofmann A (2002) Banks of microscopic marine speciation. Annu Rev Ecol Syst 25:547–572. doi:10.1146/ forms and survival to darkness of propagules and micro- annurev.ecolsys.25.1.547 scopic stages of macroalgae. Rev Chil Hist Nat 75:547–555. Perrone C, Felicini GP, Bottalico A (2006) The prostrate system of doi:10.4067/S0716-078X2002000300006 the Gelidiales: diagnostic and taxonomic importance. Bot Mar Saunders GW (2005) Applying DNA barcoding to red macroalgae: 49:23–33. doi:10.1515/Bot2006.003 a preliminary appraisal holds promise for future applica- Petit RJ, El Mousadik A, Pons O (1998) Identifying populations for tions. Philos Trans R Soc Lond B Biol Sci 360:1879–1888. conservation on the basis of genetic markers. Conserv Biol doi:10.1098/rstb.2005.1719 12:844–855. doi:10.1046/j.1523-1739.1998.96489.x Selkoe KA, Toonen RJ (2011) Marine connectivity: a new look at Piazzi L, Balata D, Bulleri F, Gennaro P, Ceccherelli G (2016) The pelagic larval duration and genetic metrics of dispersal. Mar invasion of Caulerpa cylindracea in the Mediterranean: the Ecol Prog Ser 436:291–305. doi:10.3354/meps09238 known, the unknown and the knowable. Mar Biol 163:161–174. Slatkin M (1993) Isolation by distance in equilibrium and nonequilib- doi:10.1007/S00227-016-2937-4 rium populations. Evolution 47:264–279. doi:10.2307/2410134 Polzin T, Daneshmand SV (2003) On Steiner trees and minimum span- Sosa P, García-Reina G (1992) Genetic variability and diferentiation ning trees in hypergraphs. Oper Res Lett 31:12–20. doi:10.1016/ of sporophytes and gametophytes in populations of Gelidium s0167-6377(02)00185-2 arbuscula (Gelidiaceae: Rhodophyta) determined by isozyme Ramírez M, Santelices B (1981) Análisis biogeográfco de la fora electrophoresis. Mar Biol 113:679–688 algológica de Antofagasta (norte de Chile). Bol Mus Nac Hist Sosa PA, Valero M, Batista F, González-Pérez MA (1998) Genetic Nat Santiago de Chile 38:5–20 structure of natural populations of Gelidium species: a re- Ramírez ME, Santelices B (1991) Catálogo de las algas marinas evaluation of results. J Appl Phycol 10:279–284. doi:10.102 bentónicas de la costa temperada del Pacífco de Sudamérica. 3/A:1008092023549 Monografías Biológicas 5, Ediciones Universidad Católica de Soto EH, Báez P, Ramírez ME, Letelier S, Naretto J, Rebolledo Chile, Santiago de Chile A (2012) Biotopos marinos intermareales entre Canal Trini- Robuchon M, Le Gall L, Mauger S, Valero M (2014) Contrast- dad y Canal Smyth, Sur de Chile. Rev Biol Mar Oceanogr ing genetic diversity patterns in two sister kelp species 47:177–191

1 3

107 Mar Biol (2017) 164:188 Page 19 of 19 188

Tajima F (1989) Statistical method for testing the neutral mutation Thomas DT, Freshwater DW (2001) Studies of Costa Rican Gelidi- hypothesis by DNA polymorphism. Genetics 123:585–595 ales (Rhodophyta): four Caribbean taxa including Pterocla- Tapia FJ, Largier JL, Castillo M, Wieters EA, Navarrete SA (2014) Lat- diella beachii sp nov. Phycologia 40:340–350. doi:10.2216/ itudinal discontinuity in thermal conditions along the nearshore i0031-8884-40-4-340.1 of central-northern Chile. PLoS One 9:e110841. doi:10.1371/ Valero M, Destombe C, Mauger S, Ribout C, Engel CR, Daguin- journal.pone.0110841 Thiebaut C, Tellier F (2011) Using genetic tools for sustainable Technelysium Pty Ltd (2016) CHROMAS LITE. Available at: http:// management of kelps: a literature review and the example of www.technelysium.com.au/chromas_lite.html. Accessed 12 Laminaria digitata. Cah Biol Mar 52:467–483 April 2016 Vargas DR, Collado-Vides L (1996) Architectural models for apical Tellier F, Meynard AP, Correa JA, Faugeron S, Valero M (2009) patterns in Gelidium (Gelidiales, Rhodophyta): hypothesis of Phylogeographic analyses of the 30°S south-east Pacifc bio- growth. Phycol Res 44:95–100 geographic transition zone establish the occurrence of a sharp Vásquez JA, Vega JMA (2004) Ecosistemas marinos costeros del genetic discontinuity in the kelp Lessonia nigrescens: vicariance Parque Nacional Bosque Fray Jorge. In: Squeo F, Gutiérrez J, or parapatry? Mol Phylogenet Evol 53:679–693. doi:10.1016/j. Hernández I (eds) Historia Natural del Parque Nacional Bosque ympev.2009.07.030 Fray Jorge. Ediciones Universidad de La Serena, La Serena, pp Tellier F, Vega JMA, Broitman BR, Vásquez JA, Valero M, Faugeron 235–252 S (2011) The importance of having two species instead of one Voisin M, Engel CR, Viard F (2005) Diferential shufing of native in kelp management: the Lessonia nigrescens species complex. genetic diversity across introduced regions in a brown alga: Cah Biol Mar 52:455–465 aquaculture vs. maritime trafc efects. Proc Natl Acad Sci USA Thiel M, Gutow L (2005a) The ecology of rafting in the marine envi- 102:5432–5437. doi:10.1073/pnas.0501754102 ronment. I. The foating substrata. Oceanogr Mar Biol Ann Rev Waters JM, Fraser CI, Hewitt GM (2013) Founder takes all: density- 42:181–263. doi:10.1201/9780203507810 dependent processes structure biodiversity. Trends Ecol Evol Thiel M, Gutow L (2005b) The ecology of rafting in the marine envi- 28:78–85. doi:10.1016/j.tree.2012.08.024 ronment. II. The rafting organisms and community. Oceanogr Weersing K, Toonen RJ (2009) Population genetics, larval dispersal, Mar Biol Ann Rev 43:279–418. doi:10.1201/9781420037449 and connectivity in marine systems. Mar Ecol Prog Ser 393:1– Thiel M, Haye PA (2006) The ecology of rafting in the marine 12. doi:10.3354/meps08287 environment. III. Biogeographical and evolutionary con- Wichmann CS, Hinojosa IA, Thiel M (2012) Floating kelps in Patago- sequences. Oceanogr Mar Biol Ann Rev 44:323–429. nian Fjords: an important vehicle for rafting invertebrates and doi:10.1201/9781420006391.ch7 its relevance for biogeography. Mar Biol 159:2035–2049. Thiel M, Macaya EC, Acuña E, Arntz WE, Bastias H, Brokordt K, doi:10.1007/s00227-012-1990-x Camus PA, Castilla JC, Castro LR, Cortés M, Dumont CP, Yoon KJ, Kim KM, Boo GH, Miller KA, Boo SM (2014) Mito- Escribano R, Fernández M, Gajardo JA, Gaymer CF, Gómez chondrial cox1 and cob sequence diversities in Gelidium I, González AE, González HE, Haye PA, Illanes JE, Iriarte JL, vagum (Gelidiales, Rhodophyta) in Korea. Algae 29:15–25. Lancellotti DA, Luna-Jorquera G, Luxoroi C, Manríquez PH, doi:10.4490/algae.2014.29.1.015 Marín V, Muñoz P, Navarrete SA, Pérez E, Poulin E, Sellanes J, Zakas C, Binford J, Navarrete SA, Wares JP (2009) Restricted gene Sepúlveda HH, Stotz W, Tala F, Thomas A, Vargas CA, Vásquez fow in Chilean barnacles refects an oceanographic and bio- JA, Vega JMA (2007) The Humboldt Current system of north- geographic transition zone. Mar Ecol Prog Ser 394:165–177. ern and central Chile. Oceanogr Mar Biol Ann Rev 45:195–344. doi:10.3354/meps08265 doi:10.1201/9781420050943

1 3

108

II. General Discussion

In the present study it was shown that dispersal capacity of floating individuals of D. antarctica, using beach-cast raft strandings as proxy, can vary strongly over several spatial (biogeographic districts) and temporal (inter and intra-annually) scales, suggesting that along the Chilean coast there would be areas of retention, of variable supplies and other factors that favor the arrival of rafts (see Hinojosa et al. 2011). The reproductive and physiological characteristics of stranded specimens did not vary in the study area, suggesting that these are not the cause of the observed low dispersal and the lack of genetic connectivity between bull kelp populations. On the other hand, a high capacity of floating D. antarctica as transport vehicle for epibionts (mainly sessile taxa) was evidenced, suggesting that successful dispersal depends not only on the floating seaweed capacity, but also on the biological interactions (facilitation or competition) between epibiont species during transport and arrival to other areas. Likewise, the phylogeographic patterns of the epiphytic seaweeds of genus Gelidium (particularly G. lingulatum) support the effective dispersal capacity of floating D. antarctica as vector. Therefore, this study shows empirical evidence of the role of rafting dispersal via floating seaweeds in the population connectivity, gene flow and the structuring of geographic ranges of species. Moreover, the results suggest that oceanographic factors (i.e. local currents and winds, occurrence of local storms) and the availability of floating rafts due to demographic fluctuations of the benthic populations would be key to contribute successful immigration from distant sites.

Dispersal patterns of floating bull kelps along Chilean coast - implications for its population connectivity Dynamic strandings of floating bull kelps (both semiannually and bimonthly) indicated that along the continental coast of Chile (28°S-42°S) general dispersal patterns are contrasting between zones. These zones coincide with the biogeographic districts described by Camus (2001) for the Intermediate Area (30°S-42°S). In the northern zone (i.e. Septentrional District, SED, 30°S-33°S), high values of stranded biomass and prolonged floating times of

109 the rafts were observed (Fig. 6.1). This area is close to the northern limit of the species (~ 30°S), where it is expected that the benthic populations are smaller and more fragmented than central populations (Sagarin et al. 2006). Furthermore, this area is characterized by several oceanographic discontinuities (i.e. local upwelling, eddy kinetic energy) (Hormazábal 2004, Yuras et al. 2005, Tapia et al. 2014). Also, this area coincides with the biogeographic break at 30°S (Camus 2001) and with the phylogeographic breaks described for many intertidal species of invertebrates and macroalgae with limited dispersal abilities (i.e. 29°S-33°S, Tellier et al. 2009, Sánchez et al. 2011, Montecinos et al. 2012, Haye et al. 2014, Guillemin et al. 2016). Moreover, harsh enviromental conditions at the sea surface (i.e. high water temperature or intense solar radiation) suppress the survival of floating seaweeds (Hobday 2000b, Vandendriessche et al. 2007, Tala et al. 2013, 2016, 2017) and can also cause loss of reproductive tissues of seaweed rafts (Macaya et al. 2005, McKenzie and Bellgrove 2008, Rothäusler et al. 2009), which may affect the generation of gametes and decrease the possibility of successful immigration to other sites. However, the evidence of long floating time (i.e. large sizes of Lepas spp. attached) in the stranded individuals suggests that the supplies come from more distant populations, probably from the southern populations, due to the south-north orientation of the Humboldt Current (Thiel et al. 2007) or nearshore currents and eddies (Hawes et al. 2017), which would explain the high accumulations of beach-cast rafts in this area. Also, large-scale oceanographic factors (e.g. El Niño) might affect the availability of floating individuals, which has been observed in benthic populations of other floating seaweeds (e. g. Macrocystis pyrifera, Ladah et al. 1999, Vega et al. 2005). From this area there would also be supplies to areas further north (i.e. Coquimbo-Choros District, CCD, 28°S-30°S), beyond the geographic range of the species (~200 km), which indicates that D. antarctica has some long-distance dispersal capacity, even at lower latitudes. A recent study reported beach-cast rafts of D. antarctica on the southern coast of Brazil and Uruguay (i.e. 27°S-34°S) (Batista et al. 2018), which suggests that along the Atlantic basin long-distance dispersal events may also be occurring. A similar pattern was observed in the southernmost area of this study (i.e. Meridional District, MD, 37°S-42°S) where high stranded biomass, large sizes and long floating times of stranded specimens were observed (Fig. 6.1). In this zone the environmental conditions at sea surface are less severe (i.e. temperature, solar radiation),

110 which favors the persistence of the floating specimens (Rothäusler et al. 2009). Additionally, coastal winds tend to push the floating rafts back to the shore (Hinojosa et al. 2010). The presence and sizes of attached specimens of Lepas spp. also suggest that a fraction of the observed biomasses in this area comes from more distant sources, as reported in SED. The results in these two zones (SED and MD) suggest that environmental variables operate differentially on dispersal capacities of floating individuals between both edges. Towards the northern edge of the geographic range of the species (~30°S), physiological restrictions associated with high temperature and solar radiation could affect the availability of floating specimens (i.e. low benthic abundances have been observed in this area, Fig. 6.2, Schreiber 2018) and survival of rafts at the sea surface (Graiff et al., 2013, Tala et al., 2013, 2016). Moreover, bull kelp rafts transported outside of their physiological range (beyond the north of 30°S), would have little chance of surviving. This has also been reported for floating Sargassum natans/fluitans, which are occasionally transported via the Gulf Stream towards the coasts in Northwest Europe (Schneider and Searles 1991), but which clearly would not be able to survive the cold winters there. Franke et al. (1998) and Hoeksema et al. (2012) reported similar results of warm-water animal species being transported by rafting to high latitude with no chance of survival during winter. On the other hand, at the southern edge of this study area, the physiological limitations on dispersal capacity appear to be low due to the favorable environmental conditions during the pelagic stage that are less severe. Consequently the availability of floating rafts is higher than in other areas (Hinojosa et al. 2011), which favors rafting opportunities. These differences between both edges of the geographic range (particularly between low and high latitudes) have been reported in other organims (see review Sagarin and Gaines 2006, Sagarin et al., 2006), because these areas may have different geological histories or ecological regimes. Moreover, along a latitudinal gradient, the ACH is expected to be fulfilled in species that have their distribution center close to the equator, where the effects of abiotic and biotic factors tend to be symmetrical with respect to the species distribution. Conversely, in the case of species whose geographical range is generally at higher latitudes (the case of D. antarctica), the effect of environmental factors on the dispersal capacity should be asymmetrical, with differential results at high and low latitude

111 distribution limits. Also, it has been reported that ACH does not apply to all species equally (Rivadeneira et al. 2010, Dallas et al. 2017). On the other hand, in the central zone (i.e. Mediterranean District, MED, 33°S- 37°S), apparently the connectivity between populations is much lower than in other areas, because the stranded biomasses were consistently low, the stranded individual sizes were small and floating times (absence or small sizes of Lepas spp) indicate that supplies come almost exclusively from nearby sources (Fig. 6.1). Oceanographic features with a strong seasonal component, such as local upwelling (Sobarzo et al. 2007) can form retention zones for floating substrata (Hinojosa et al. 2011), and river plumes (e.g. Maipo and Itata rivers, Piñones et al. 2005, Saldías et al. 2012, 2016) represent barriers to dispersal and may affect the return of floating rafts to the shore. Also, the wide extension of sandy beaches in this area (Thiel et al. 2007), limits the availability of potential substratum for the species (Fig. 6.2). Moreover, this area concentrates the most intense exploitation of natural beds of D. antarctica in Chile (Gelcich et al. 2006, SERNAPESCA 2017), which affects the abundances and sizes of benthic specimens, and subsequently availability of floating rafts (Castilla and Bustamante 1989, Bustamante and Castilla 1990, Castilla et al. 2007). Also, previous genetic studies in seaweeds indicated phylogeographic discontinuities in this area (e.g. 33°S and 38°S, Montecinos et al. 2012), which suggests that the dispersal and population connectivity can vary strongly, being higher in certain areas than others. In this zone, oceanographic factors seem to be key in limiting the effective dispersal. Nevertheless, ecological factors (i.e. density blocking, Waters et al. 2013, Neiva et al. 2014, Fraser et al. 2017) can not be discarded. Upon return to coasts, rafters might be confronted with well- established communities of competitors and/or consumers, leaving little opportunity for successful colonization (e.g. Hoffmann 1987, Bellgrove et al. 2010, Waters 2011,), because growth and survival of recruits might be suppressed (Worm and Chapman 1996, Taylor and Schiel 2005, 2010, Aguilera et al. 2015). However, in central Chile there would be some release of patches of primary substratum due to more intense exploitation in certain areas and seasons (although some extraction practices leave the adhesion holdfast of specimens, E. Macaya com. pers.), albeit for example, Aburto (2016) found no temporal variations in genetic diversity and haplotypes in benthic populations of D. antarctica from southern- central Chile (between 36°S-39°S) that were recolonized post-disturbance (i.e. earthquake

112 of 2010). This suggests that adjacent population supplied propagules that repopulated these disturbed sites, which also supports the observed results of this study (i.e. low dispersal in this zone).

Fig. 6.1. Map of the continental coast of Chile, showing the main results of this study. The main biogeographic zones (provinces and districts) and breaks (30°S and 41°S–42°S) (modified from Camus 2001) are indicated. Benthic abundance pattern of D. antarctica extracted from Schreiber (2018).

113

Although there was a high temporal variability of strandings with marked periods of increase (summer-autumn) and decrease (spring) in the availability of floating specimens, due to the annual growth cycles of the benthic populations. The results of this study indicate that dispersal by floating bull kelps is high in SED and MD, while it is low in MED, suggesting that limited population connectivity in central Chile might be due to lack of rafting opportunities. Also, factors related to the functional and/or reproductive capacities of the floating specimens do not seem to be decisive in the dispersal potential (at least in the southern-central zone), while oceanographic factors and fluctuations in benthic abundances appear to determine the limited connectivity between central populations.

Fig. 6.2. Stranded biomass, extension of potentially suitable habitat (PSH) and benthic biomass of Durvillaea antarctica along the continental coast of Chile (30°S – 43°S). Mean stranded biomass values for 28 beaches were obtained from Chapter 1 (Extracted from Schreiber 2018).

Durvillaea antarctica as dispersal agent for associated species This study shows (through empirical evidence based on epibiont analysis on stranded bull kelps, as well as genetic analysis of benthic populations of two epiphytic seaweeds frequent on holdfasts of beach-cast rafts) that D. antarctica can transport a wide variety of species (mainly invertebrates and non-buoyant seaweeds) along the continental coast of Chile. The

114 results indicated that rafting dispersal by floating bull kelps contributes to the structuring of the geographic ranges of epibionts (particularly in those with low dispersal ability), and to the geographical distribution of the genetic diversity of their benthic populations. Previous studies had only examined the associated fauna on floating seaweeds in the Chilean coast (i.e M. pyrifera and D. antarctica, Thiel 2003, Hinojosa et al. 2007, Wichmann et al. 2012), even though evaluating the epibionts that return to the coast (as done herein) better allows understanding the effectiveness of this dispersal mechanism. Several genetic and ecological studies have suggested the vector role of floating seaweeds (see Thiel and Haye 2006, Macaya et al 2016) in the population connectivity and gene flow of associated species. In the case of D. antarctica there is evidence of transoceanic rafting between the coasts of New Zealand, Chile and subantarctic islands (e.g. Helmuth et al. 1994, Donald et al. 2005, Nikula et al. 2010, 2011a, 2011b, Fraser et al. 2011, Nikula et al. 2013, Cumming et al. 2014, González-Wevar et al. 2018, Waters et al. 2018a), and along the continental coast of Chile (Haye et al. 2012). Also, seaweed pathogens might use rafting dispersal via floating bull kelps to extend their infection prevalence (Blake et al. 2017). Nonetheless, the results of this thesis suggest that biological interactions among epibionts, particularly positive, during transport are crucial to facilitate survival/presence of other epibionts during long-distance dispersal and promote range extensions. The reported range extensions were more frequent and extensive in the southern zone of study area (> 37°S), most likely because there the conditions at the sea surface are more benign, facilitating long-term survival of rafts and of rafting organisms (Fig. 6.1). On the other hand, the phylogeographic patterns of Gelidium species suggest that these can be explained by long-distance dispersal of floating bull kelps, particularly in the case of G. lingulatum. In G. lingulatum there was no clear geographic pattern (i.e. haplotypes that disappear and reappear along the coast), which supports that there are areas where the dispersal of species is higher than in others, as also observed in other chapters of this thesis. Also, the pattern observed for G. rex indicates that there could be short-distance dispersal, within its northern (28°S-31°S) and southern (31°S-34°S) populations along its geographic range. These results suggest that some epibiont species could withstand long- distance transport at the sea surface more than others, allowing their successful immigration to other sites. Likewise, other dispersal mechanisms (e.g. human-mediated dispersal) that

115 affect phylogeographic patterns of seaweeds (Vousin et al. 2005, Piazzi et al. 2016) can not be completely discarded (transport through maritime traffic, such as ballast waters and ship hulls), although this transport mechanism in Gelidium species is much less likely than through rafting dispersal by floating seaweeds (see Chapter 5). Future studies on functional and/or reproductive parameters of epibiont species, simulating floating conditions could help to understand their dispersal capacity along the coast. Also, considering other floating seaweeds that have high pelagic abundances, high buoyancy, large sizes and a wide geographic distribution, such as M. pyrifera (Hobday 2000a, 2000b, 2000c, Macaya 2010a, 2010b, Hinojosa et al. 2011, Tala et al. 2016), could be useful to contrast dispersal and transport capabilities of other organisms.

Conclusions The results of this study suggest long-distance dispersal events by rafting of floating bull kelp D. antarctica and its associated epibionts along the continental coast of Chile, particularly between 30°S-33°S and 37°S-42°S, as indicated by the high stranded biomasses observed in those zones. In general, the observed spatial patterns of stranded biomass and floating times of stranded individuals indicate that most of the floating raft supplies come from adjacent benthic populations (i.e. within a radius of ~10 km near each beach, Schreiber 2018), which helps to understand the phylogeographic pattern of D. antarctica described for this study area (i.e. low genetic connectivity among nearby populations, < 200 km, Fraser et al. 2010). Likewise, the results indicated that local oceanographic features (i.e. upwellings, river plumes), the latitudinal gradient of abiotic factors along the Chilean coast, especially at the sea surface (i.e. temperature, solar radiation), and seasonal variations in the abundances of the benthic populations are the main factors that explain the availability of floating bull kelps, transport, survival and return of rafts to the shore. These factors differentially affect the dispersal capacity of floating individuals and raft-asociated species along the geographic range of the species. Future studies evaluating the effect of antropogenic drivers, such as coastal artificial infrastructure (i.e. presence of breackwaters and jetties, Bulleri and Chapman, 2010, Aguilera 2018) and coastal pollution (Rech et al. 2018) on the transport and arrival of

116 floating seaweeds to coastal areas would be complementary to verify changes in spatio- temporal patterns, particularly in zones with high levels of coastal urbanization. In summary, this study shows that the effectiveness of rafting dispersal via floating seaweeds in temperate regions varies strongly along a latitudinal gradient, as well as on time scales. Local oceanographic features, and demographic variations of benthic populations affect the dispersal capacity of floating specimens. The presence of local upwellings generates retention zones for floating specimens that limit their return to the coast. At the low latitude edge of the geographic distribution of floating seaweed species, physiological constraints suppress their dispersal potential. On the other hand, at the high latitude edge the variations in raft availability due to changes in benthic abundances, and ecological interactions with established communities affect successful immigration and population connectivity (Fig. 6.1). Future studies that consider genetic analysis of stranded specimens (particularly those with indications of long-distance dispersal, such as presence of Lepas spp.), and epibiont species frequently associated with floating rafts would be useful to confirm the observed patterns in this study, and would help to understand the dispersal trajectories of floating seaweeds. Also, changes in species distributions under future climate scenarios might affect the dispersal capabilities of floating seaweeds and their associated species (i.e. variations in supplies, transport and permanence of rafts, Macreadie et al. 2011), which should be incorporated in future studies.

117

III. General References

Aburto, C. E. 2016. Diversidad genética y conectividad de propágulos vía rafting de Durvillaea antarctica (Heterokonthophyta, Fucales) post-terremoto 2010 en el centro-sur de Chile. Master Dissertation, Universidad Católica de la Santísima Concepción, Concepción. 53 pp. Aguilera, M. A. 2018. Artificial defences in coastal marine ecosystems in Chile: Opportunities for spatial planning to mitigate habitat loss and alteration of the marine community structure. Ecol Eng 120:601-10. Aguilera, M. A., Valdivia, N. & Broitman, B. R. 2015. Herbivore-alga interaction strength influences spatial heterogeneity in a kelp-dominated intertidal community. PLoS One 10:e0137287. Araújo, R., Serrão, E. A., Sousa-Pinto, I. & Aberg, P. 2011. Phenotypic differentiation at southern limit borders: The case study of two fucoid macroalgal species with different life-history traits. J. Phycol. 47:451-62. Araújo, R., Serrão, E. A., Sousa-Pinto, I., Arenas, F., Monteiro, C. A., Toth, G., Pavia, H. & Aberg, P. 2015. Trade-offs between life-history traits at range-edge and central locations. J. Phycol. 51:808-18. Batista, M. B., Batista, A., Franzan, P., Simionatto, P., Lima, T. C., Velez-Rubio, G. M., Scarabino, F., Camacho, O., Schmitz, C., Martinez, A., Ortega, L., Fabiano, G., Rothman, M. D., Liu, G., Ojeda, J., Mansilla, A., Barreto, L. M., Assis, J., Serrão, E. A., Santos, R. & Horta, P. A. 2018. Kelps’ long-distance dispersal: Role of ecological/oceanographic processes and implications to marine forest conservation. Diversity 10:11. Bellgrove, A., McKenzie, P. F., McKenzie, J. L. & Sfiligoj, B. J. 2010. Restoration of the habitat-forming fucoid alga Hormosira banksii at effluent-affected sites: competitive exclusion by coralline turfs. Mar. Ecol. Prog. Ser. 419:47-56. Blake, C., Thiel, M., López, B. A. & Fraser, C. I. 2017. Gall-forming protistan parasites infect southern bull- kelp across the Southern Ocean, with prevalence increasing to the south. Mar. Ecol. Prog. Ser. 583:95-106. Bobadilla, M. & Santelices, B. 2005. Variations in the dispersal curves of macroalgal propagules from a source. J. Exp. Mar. Biol. Ecol. 327:47-57. Brown, J. H. 1984. On the relationship between abundance and distribution of species. Am Nat 124:255-79. Buchanan, J. & Zuccarello, G. C. 2012. Decoupling of short- and long-distance dispersal pathways in the endemic new zealand seaweed Carpophyllum maschalocarpum (Phaeophyceae, Fucales). J Phycol 48:518-29. Bulleri, F. & Chapman, M. G. 2010. The introduction of coastal infrastructure as a driver of change in marine environments. J Appl Ecol 47:26-35. Bussolini, L. T. & Waters, J. M. 2015. Genetic analyses of rafted macroalgae reveal regional oceanographic connectivity patterns. J. Biogeogr. 42:1319-26. Bustamante, R. H. & Castilla, J. C. 1990. Impact of human exploitation on populations of the intertidal Southern bull kelp Durvillaea antarctica (Phaeophyta, Durvilleales) in central Chile. Biological Conservation 52:205-20.

118

Cahill, A. E. & Levinton, J. S. 2016. Genetic differentiation and reduced genetic diversity at the northern range edge of two species with different dispersal modes. Mol Ecol 25:515-26. Camus, P. A. 2001. Biogeografía marina de Chile continental. Rev. Chil. Hist. Nat. 74:587- 617. Castilla, J. C. & Bustamante, R. H. 1989. Human exclusion from rocky intertidal of Las Cruces, central Chile: effects on Durvillaea antarctica (Phaeophyta, Durvilleales). Mar. Ecol. Prog. Ser. 50:203-14. Castilla, J. C., Campo, M. A. & Bustamante, R. H. 2007. Recovery of Durvillaea antarctica (Durvilleales) inside and outside Las Cruces Marine Reserve, Chile. Ecol. Appl. 17:1511-22. Collantes, G., Riveros, R. & Acevedo, M. 1997. Fenología reproductiva de Durvillaea antarctica (Phaeophyta, Durvillaeales) del intermareal de caleta Montemar, Chile central. Rev. Biol. Mar. Oceanogr. 32:111-16. Collantes, G., Merino, A. & Lagos, V. 2002. Fenología de la gametogénesis, madurez de conceptáculos, fertilidad y embriogénesis en Durvillaea antarctica (Chamisso) Hariot (Phaeophyta, Durvillaeales). Rev. Biol. Mar. Oceanogr. 37:83-112. Collins, C. J., Fraser, C. I., Ashcroft, A. & Waters, J. M. 2010. Asymmetric dispersal of southern bull-kelp (Durvillaea antarctica) adults in coastal New Zealand: testing an oceanographic hypothesis. Mol. Ecol. 19:4572-80. Coyer, J. A., Hoarau, G., Van Schaik, J., Luijckx, P. & Olsen, J. L. 2011a. Trans-Pacific and trans-Arctic pathways of the intertidal macroalga Fucus distichus L. reveal multiple glacial refugia and colonizations from the North Pacific to the North Atlantic. J. Biogeogr. 38:756-71. Coyer, J. A., Hoarau, G., Costa, J. F., Hogerdijk, B., Serrão, E. A., Billard, E., Valero, M., Pearson, G. A. & Olsen, J. L. 2011b. Evolution and diversification within the intertidal brown macroalgae Fucus spiralis/F. vesiculosus species complex in the North Atlantic. Mol. Phylogenet. Evol. 58:283-96. Craw, D. & Waters, J. 2018. Long distance kelp-rafting of rocks around southern New Zealand. N Z J Geol Geophys:1-16. Cumming, R. A., Nikula, R., Spencer, H. G. & Waters, J. M. 2014. Transoceanic genetic similarities of kelp-associated sea slug populations: long-distance dispersal via rafting? J. Biogeogr. 41:2357-70. Dallas, T., Decker, R. R. & Hastings, A. 2017. Species are not most abundant in the centre of their geographic range or climatic niche. Ecol Lett 20:1526-33. Donald, K. M., Kennedy, M. & Spencer, H. G. 2005. Cladogenesis as the result of long- distance rafting events in South Pacific topshells (Gastropoda, Trochidae). Evolution 59:1701-11. Duarte, C., Jaramillo, E. & Contreras, H. 2008. Stranded algal wracks on a sandy beach of south central Chile: feeding and habitat preferences of juveniles and adults of Orchestoidea tuberculata (Nicolet), (Amphipoda, Talitridae). Rev. Chil. Hist. Nat. 81:69-81. Duarte, C., Jaramillo, E., Contreras, H., Acuña, K. & Navarro, J. M. 2009. Importance of macroalgae subsidy on the abundance and population biology of the amphipod Orchestoidea tuberculata (Nicolet) in sandy beaches of south central Chile. Rev. Biol. Mar. Oceanogr. 44:691-702.

119

Fenberg, P. B. & Rivadeneira, M. M. 2011. Range limits and geographic patterns of abundance of the rocky intertidal owl limpet, Lottia gigantea. J. Biogeogr. 38:2286- 98. Franke, H.-D., Gutow, L. & Janke, M. 1998. The recent arrival of the oceanic isopod Idotea metallica Bosc off Helgoland (German Bight, North Sea): an indication of a warming trend in the North Sea? Helgol. Meeresunters. 52:347-357. Fraser, C. I., Nikula, R., Spencer, H. G. & Waters, J. M. 2009a. Kelp genes reveal effects of subantarctic sea ice during the Last Glacial Maximum. Proc. Natl. Acad. Sci. U. S. A. 106:3249-53. Fraser, C. I., Spencer, H. G. & Waters, J. M. 2009b. Glacial oceanographic contrasts explain phylogeography of Australian bull kelp. Mol. Ecol. 18:2287-96. Fraser, C. I., Hay, C. H., Spencer, H. G. & Waters, J. M. 2009c. Genetic and morphological analyses of the southern bull kelp Durvillaea antarctica (Phaeophyceae: Durvillaeales) in New Zealand reveal cryptic species. J. Phycol. 45:436-43. Fraser, C. I., Thiel, M., Spencer, H. G. & Waters, J. M. 2010. Contemporary habitat discontinuity and historic glacial ice drive genetic divergence in Chilean kelp. BMC Evol Biol 10:203. Fraser, C. I., Nikula, R. & Waters, J. M. 2011. Oceanic rafting by a coastal community. Proc. R. Soc. B-Biol. Sci. 278:649-55. Fraser, C. I., Davies, I. D., Bryant, D. & Waters, J. M. 2017. How disturbance and dispersal influence intraspecific structure. J. Ecol. 106:1298-306. Fraser, C. I., Morrison, A. K., Hogg, A. M., Macaya, E. C., van Sebille, E., Ryan, P. G., Padovan, A., Jack, C., Valdivia, N. & Waters, J. M. 2018. Antarctica’s ecological isolation will be broken by storm-driven dispersal and warming. Nat. Clim. Change. 8:704-708. Garden, C. J., Craw, D., Waters, J. M. & Smith, A. 2011. Rafting rocks reveal marine biological dispersal: A case study using clasts from beach-cast macroalgal holdfasts. Estuar. Coast. Shelf Sci. 95:388-94. Garden, C. J., Currie, K., Fraser, C. I. & Waters, J. M. 2014. Rafting dispersal constrained by an oceanographic boundary. Mar. Ecol. Prog. Ser. 501:297-302. Gelcich, S., Edwards-Jones, G., Kaiser, M. J. & Castilla, J. C. 2006. Co-management policy can reduce resilience in traditionally managed marine ecosystems. Ecosystems 9:951-66. González, A. V., Beltrán, J., Flores, V. & Santelices, B. 2015. Morphological convergence in the inter-holdfast coalescence process among kelp and kelp-like seaweeds (Lessonia, Macrocystis, Durvillaea). Phycologia 54:283-91. González-Wevar, C. A., Segovia, N. I., Rosenfeld, S., Ojeda, J., Hüne, M., Naretto, J., Saucede, T., Brickle, P., Morley, S., Féral, J. P., Spencer, H. G. & Poulin, E. 2018. Unexpected absence of island endemics: Long-distance dispersal in higher latitude sub-Antarctic Siphonaria (Gastropoda: Euthyneura) species. J. Biogeogr. 45:874- 84. Graiff, A., Karsten, U., Meyer, S., Pfender, D., Tala, F. & Thiel, M. 2013. Seasonal variation in floating persistence of detached Durvillaea antarctica (Chamisso) Hariot thalli. Bot. Mar. 56:3-14. Graiff, A., Pantoja, J. F., Tala, F. & Thiel, M. 2016. Epibiont load causes sinking of viable kelp rafts: seasonal variation in floating persistence of giant kelp Macrocystis pyrifera. Mar. Biol. (Berl.) 163:190-204. 120

Guillemin, M. L., Valero, M., Tellier, F., Macaya, E. C., Destombe, C. & Faugeron, S. 2016. Phylogeography of seaweeds in the South East Pacific: complex evolutionary processes along a latitudinal gradient In: Hu, Z. M. & Fraser, C. I. [Eds.] Seaweed Phylogeography. Springer, Dordrecht, The Netherlands, pp. 251-78. Gutow, L., Beermann, J., Buschbaum, C., Rivadeneira, M. M. & Thiel, M. 2015. Castaways can't be choosers - Homogenization of rafting assemblages on floating seaweeds. J. Sea Res. 95:161-71. Hawes, N.A., Taylor, D.I. & Schiel, D.R. 2017. Transport of drifting fucoid algae: nearshore transport and potential for long distance dispersal. J Exp Mar Biol Ecol 490:34-41 Haye, P. A., Varela, A. I. & Thiel, M. 2012. Genetic signatures of rafting dispersal in algal- dwelling brooders Limnoria spp. (Isopoda) along the SE Pacific (Chile). Mar. Ecol. Prog. Ser. 455:111-22. Haye, P. A., Segovia, N. I., Muñoz-Herrera, N. C., Gálvez, F. E., Martínez, A., Meynard, A., Pardo-Gandarillas, M. C., Poulin, E. & Faugeron, S. 2014. Phylogeographic structure in benthic marine invertebrates of the southeast Pacific coast of Chile with differing dispersal potential. PLoS One 9:e88613. Helmuth, B., Veit, R. R. & Holberton, R. 1994. Long distance dispersal of a subantarctic brooding bivalve (Gaimardia trapesina) by kelp rafting. Mar. Biol. 120:421-26. Hernández, C. J. 1997. Análisis de la variación estacional e interanual de la cosecha de Gelidium robustum en Baja California Sur, México. M. Sc. Dissertation M. Sc. Dissertation, Centro Interdisciplinario de Ciencias Marinas, 60 pp. Hinojosa, I., González, E., Ugalde, P., Valdivia, N., Macaya, E. & Thiel, M. 2007. Distribución y abundancia de macroalgas flotando a la deriva y su fauna peracárida asociada a en los canales de la XI Región, Chile. Revista Ciencia y Tecnología del Mar 30:37-50. Hinojosa, I. A., Pizarro, M., Ramos, M. & Thiel, M. 2010. Spatial and temporal distribution of floating kelp in the channels and fjords of southern Chile. Estuar. Coast. Shelf Sci. 87:367-77. Hinojosa, I. A., Rivadeneira, M. M. & Thiel, M. 2011. Temporal and spatial distribution of floating objects in coastal waters of central-southern Chile and Patagonian fjords. Cont. Shelf Res. 31:172-86. Hobday, A. J. 2000a. Abundance and dispersal of drifting kelp Macrocystis pyrifera rafts in the Southern California Bight. Mar. Ecol. Prog. Ser. 195:101-16. Hobday, A. J. 2000b. Age of drifting Macrocystis pyrifera (L.) C. Agardh rafts in the Southern California Bight. J. Exp. Mar. Biol. Ecol. 253:97-114. Hobday, A. J. 2000c. Persistence and transport of fauna on drifting kelp (Macrocystis pyrifera (L.) C. Agardh) rafts in the Southern California Bight. J. Exp. Mar. Biol. Ecol. 253:75-96. Hoeksema, P. J., Roos, B. W. & Cadée, G. C. 2012. Trans-Atlantic rafting by the brooding reef coral Favia fragum on man-made flotsam. Mar. Ecol. Prog. Ser. 445:209-18. Hoffmann, A. 1987. The arrival of seaweed propagules at the shore: a review. Bot. Mar. 30:151-66. Hoffmann, A. & Santelices, B. 1997. Flora marina de Chile Central. Ediciones Universidad Católica de Chile, Santiago de Chile, Chile, 434 pp. Hormazábal, S., Shaffer, G. & Leth, O. 2004. Coastal transition zone off Chile. J. Geophys. Res. Oceans 109:C01021. 121

Iacchei, M., Gaither, M. R., Bowen, B. W. & Toonen, R. J. 2016. Testing dispersal limits in the sea: range-wide phylogeography of the pronghorn spiny lobster Panulirus penicillatus. J. Biogeogr. 43:1032-44. Ingólfsson, A. 1998. Dynamics of macrofaunal communities of floating seaweed clumps off western Iceland: a study of patches on the surface of the sea. J. Exp. Mar. Biol. Ecol. 231:119-37. Ladah, L. B., Zertuche‐González, J. A. & Hernández‐Carmona, G. 1999. Giant kelp (Macrocystis pyrifera, Phaeophyceae) recruitment near its southern limit in Baja California after mass disappearance during ENSO 1997–1998. J. Phycol. 35:1106- 12. Macaya, E. C., Boltana, S., Hinojosa, I. A., Macchiavello, J. E., Valdivia, N. A., Vásquez, N. R., Buschmann, A. H., Vásquez, J. A., Vega, J. M. A. & Thiel, M. 2005. Presence of sporophylls in floating kelp rafts of Macrocystis spp. (Phaeophyceae) along the Chilean Pacific coast. J. Phycol. 41:913-22. Macaya, E. C. & Zuccarello, G. C. 2010a. Genetic structure of the giant kelp Macrocystis pyrifera along the southeastern Pacific. Mar. Ecol. Prog. Ser. 420:103-12. Macaya, E. C. & Zuccarello, G. C. 2010b. DNA barcoding and genetic divergence in the giant kelp Macrocystis (Laminariales). J. Phycol. 46:736-42. Macaya, E. C., López, B., Tala, F., Tellier, F. & Thiel, M. 2016. Float and raft: role of buoyant seaweeds in the phylogeography and genetic structure of non-buoyant associated flora. In: Hu, Z. M. & Fraser, C. I. [Eds.] Seaweed Phylogeography. Springer, Dordrecht, The Netherlands, pp. 97-130. Macreadie, P. I., Bishop, M. J. & Booth, D. J. 2011. Implications of climate change for macrophytic rafts and their hitchhikers. Mar. Ecol. Prog. Ser. 443:285-92. Maggs, C. A., Castilho, R., Foltz, D., Henzler, C., Jolly, M. T., Kelly, J., Olsen, J., Pérez, K. E., Stam, W., Vainola, R., Viard, F. & Wares, J. 2008. Evaluating signatures of glacial refugia for North Atlantic benthic marine taxa. Ecology 89:S108-S22. Mansilla, A. O., Avila, M. & Cáceres, J. 2017. Reproductive biology of Durvillaea antarctica (Chamisso) Hariot in the sub-Antartic ecoregion of Magallanes (51–56° S). J. Appl. Phycol. 29: 2567–2574. McKenzie, P. F. & Bellgrove, A. 2008. Dispersal of Hormosira banksii (Phaeophyceae) via detached fragments: Reproductive viability and longevity. J Phycol 44:1108-15. Moon, K.L., Chown, S.L. & Fraser, C.I. 2017. Reconsidering connectivity in the sub- Antarctic. Biol Rev Camb Philos Soc 92:2164-2181. Montecinos, A., Broitman, B. R., Faugeron, S., Haye, P. A., Tellier, F. & Guillemin, M. L. 2012. Species replacement along a linear coastal habitat: phylogeography and speciation in the red alga Mazzaella laminarioides along the south east Pacific. BMC Evol Biol 12:97. Neiva, J., Pearson, G. A., Valero, M. & Serrão, E. A. 2012. Drifting fronds and drifting alleles: range dynamics, local dispersal and habitat isolation shape the population structure of the estuarine seaweed Fucus ceranoides. J. Biogeogr. 39:1167-78. Neiva, J., Assis, J., Fernandes, F., Pearson, G. A., Serrão, E. A. & Maggs, C. 2014. Species distribution models and mitochondrial DNA phylogeography suggest an extensive biogeographical shift in the high-intertidal seaweed Pelvetia canaliculata. J. Biogeogr. 41:1137-48.

122

Nikula, R., Fraser, C. I., Spencer, H. G. & Waters, J. M. 2010. Circumpolar dispersal by rafting in two subantarctic kelp-dwelling crustaceans. Mar. Ecol. Prog. Ser. 405:221-30. Nikula, R., Spencer, H. G. & Waters, J. M. 2011a. Evolutionary consequences of microhabitat: population-genetic structuring in kelp- vs. rock-associated chitons. Mol. Ecol. 20:4915-24. Nikula, R., Spencer, H. G. & Waters, J. M. 2011b. Comparison of population-genetic structuring in congeneric kelp- versus rock-associated snails: a test of a dispersal- by-rafting hypothesis. Ecol. Evol. 1:169-80. Nikula, R., Spencer, H. G. & Waters, J. M. 2013. Passive rafting is a powerful driver of transoceanic gene flow. Biol Lett 9:20120821. Ortega, M., Godínez-Ortega, J. & Garduño, G. 2001. Catálogo de algas bénticas de las costas mexicanas del Golfo de México y Mar Caribe. UNAM, CONABIO, México, 594 pp. Otaíza, R. D., Rodríguez, C. Y., Cáceres, J. H. & Sanhueza, A. G. 2018. Fragmentation of thalli and secondary attachment of fragments of the agarophyte Gelidium lingulatum (Rhodophyta, Gelidiales). J. Appl. Phycol. 30:1921-31. Piazzi, L., Balata, D., Bulleri, F., Gennaro, P. & Ceccherelli, G. 2016. The invasion of Caulerpa cylindracea in the Mediterranean: the known, the unknown and the knowable. Mar. Biol. 163:161–74. Piñones, A., Valle-Levinson, A., Narváez, D., Vargas, C., Navarrete, S., Yuras, G. & Castilla, J. 2005. Wind-induced diurnal variability in river plume motion. Estuar. Coast. Shelf Sci. 65:513-25. Pironon, S., Papuga, G., Villellas, J., Angert, A. L., Garcia, M. B. & Thompson, J. D. 2017. Geographic variation in genetic and demographic performance: new insights from an old biogeographical paradigm. Biol. Rev. Camb. Philos. Soc. 92:1877-909. Quintanilla-Ahumada, D., Quijón, P. A., Navarro, J. M., Pulgar, J. & Duarte, C. 2018. Living on a trophic subsidy: Algal quality drives an upper-shore herbivore’s consumption, preference and absorption but not growth rates. PloS one 13:e0196121. Ramírez, M. E. & Santelices, B. 1991. Catálogo de las algas marinas bentónicas de la costa temperada del Pacífico de Sudamérica. Monografías Biológicas. pp. 1-437. Rech, S., Pichs, Y. J. B. & García-Vazquez, E. 2018. Anthropogenic marine litter composition in coastal areas may be a predictor of potentially invasive rafting fauna. PloS One 13:e0191859. Rivadeneira, M. M., Hernáez, P., Baeza, J. A., Boltana, S., Cifuentes, M., Correa, C., Cuevas, A., del Valle, E., Hinojosa, I., Ulrich, N., Valdivia, N., Vásquez, N., Zander, A. & Thiel, M. 2010. Testing the abundant-centre hypothesis using intertidal porcelain crabs along the Chilean coast: linking abundance and life-history variation. J. Biogeogr. 37:486-98. Rodríguez, S. R. 2003. Consumption of drift kelp by intertidal populations of the sea urchin Tetrapygus niger on the central Chilean coast: possible consequences at different ecological levels. Mar. Ecol. Prog. Ser. 251:141-51. Rothäusler, E., Gómez, I., Hinojosa, I. A., Karsten, U., Tala, F. & Thiel, M. 2009. Effect of temperature and grazing on growth and reproduction of floating Macrocystis spp. (Phaeophyceae) along a latitudinal gradient. J. Phycol. 45:547-59.

123

Rothäusler, E., Gutow, L. & Thiel, M. 2012. Floating seaweed and their communities. In: Wiencke, C. & Bischof, K. [Eds.] Seaweed Biology. Springer, Berlin, Germany, pp. 109-34. Rothäusler, E., Corell, H. & Jormalainen, V. 2015. Abundance and dispersal trajectories of floating Fucus vesiculosus in the Northern Baltic Sea. Limnol. Oceanogr. 60:2173- 84. Sagarin, R. D. & Gaines, S. D. 2002a. Geographical abundance distributions of coastal invertebrates: using one-dimensional ranges to test biogeographic hypotheses. J. Biogeogr. 29:985-97. Sagarin, R. D. & Gaines, S. D. 2002b. The 'abundant centre' distribution: to what extent is it a biogeographical rule? Ecology Letters 5:137-47. Sagarin, R. D. & Gaines, S. D. 2006. Recent studies improve understanding of population dynamics across species ranges. Oikos 115:386-88. Sagarin, R. D., Gaines, S. D. & Gaylord, B. 2006. Moving beyond assumptions to understand abundance distributions across the ranges of species. Trends Ecol. Evol. 21:524-30. Saldías, G. S., Sobarzo, M., Largier, J., Moffat, C. & Letelier, R. 2012. Seasonal variability of turbid river plumes off central Chile based on high-resolution MODIS imagery. Remote Sens. Environ. 123:220-33. Saldías, G. S., Largier, J. L., Mendes, R., Pérez-Santos, I., Vargas, C. A. & Sobarzo, M. 2016. Satellite-measured interannual variability of turbid river plumes off central- southern Chile: Spatial patterns and the influence of climate variability. Prog. Oceanogr. 146:212-22. Santelices, B. 1980. Phytogeographic characterization of the temperate coast of Pacific South-America. Phycologia 19:1-12. Santelices, B., Castilla, J. C., Cancino, J. & Schmiede, P. 1980. Comparative ecology of Lessonia nigrescens and Durvillaea antarctica (Phaeophyta) in central Chile. Mar. Biol. 59:119-32. Santelices, B. & Abbott, I. A. 1985. Gelidium rex sp. nov. (Gelidiales. Rhodophyta) from central Chile. In: Abbott, I. A. & Norris, J. [Eds.] Taxonomy of economic seaweeds with reference to some Pacific and Caribbean species. California Sea Grant College Program, La Jolla, California, USA, pp. 33-36. Santelices, B. 1986. The wild harvest and culture of the economically important species of Gelidium in Chile. In: Doty, M. S., Caddy, J. S. & Santelices, B. [Eds.] Case of studies of seven commercial seaweed resources. FAO Fisheries Technical Paper, pp. 165-92. Santelices, B. 1990a. Patterns of reproduction, dispersal and recruitment in seaweeds. Oceanography and Marine Biology 28:177-276. Santelices, B. 1990b. New and old problems in the taxonomy of the Gelidiales (Rhodophyta). Hydrobiologia 204:125-35. Santelices, B. 1991. Production ecology of Gelidium. Hydrobiologia 221:31-44. Saunders, G. W. 2014. Long distance kelp rafting impacts seaweed biogeography in the Northeast Pacific: the kelp conveyor hypothesis. J Phycol 50:968-74. Schneider, C. W. & Searles, R. B. 1991. Seaweeds of the Southeastern United States: Cape Hatteras to Cape Canaveral. Duke University Press, Durham, Norht Carolina.

124

Schreiber, L. 2018. Connecting benthic and pelagic populations of the bull kelp Durvillaea antarctica along the continental coast of Chile. Bachelor Thesis, Humboldt- Universität zu Berlin, Berlin, Germany. 36 pp. SERNAPESCA 2017. Anuario Estadístico de Pesca Servicio Nacional de Pesca y Acuicultura, Ministerio de Economía, Fomento y Turismo, Gobierno de Chile, Chile. Smith, S. D. A. 2002. Kelp rafts in the Southern Ocean. Global Ecol. Biogeogr. 11:67-69. Sobarzo, M., Bravo, L., Donoso, D., Garcés-Vargas, J. & Schneider, W. 2007. Coastal upwelling and seasonal cycles that influence the water column over the continental shelf off central Chile. Prog. Oceanogr. 75:363-82. Tala, F., Gómez, I., Luna-Jorquera, G. & Thiel, M. 2013. Morphological, physiological and reproductive conditions of rafting bull kelp (Durvillaea antarctica) in northern- central Chile (30 degrees S). Mar. Biol. 160:1339-51. Tala, F., Velásquez, M., Mansilla, A., Macaya, E. C. & Thiel, M. 2016. Latitudinal and seasonal effects on short-term acclimation of floating kelp species from the South- East Pacific. J. Exp. Mar. Biol. Ecol. 483:31-41. Tala, F, Penna-Díaz M,A,, Luna-Jorquera, G., Rothäusler, E. & Thiel, M. 2017. Daily and seasonal changes of photobiological responses in floating bull kelp Durvillaea antarctica (Chamisso) Hariot (Fucales: Phaeophyceae). Phycologia 56:271-283. Tapia, F. J., Largier, J. L., Castillo, M., Wieters, E. A. & Navarrete, S. A. 2014. Latitudinal discontinuity in thermal conditions along the nearshore of central-northern Chile. PLoS One 9:e110841. Taylor, D. I. & Schiel, D. R. 2005. Self-replacement and community modification by the southern bull kelp Durvillaea antarctica. Mar. Ecol. Prog. Ser. 288:87-102. Taylor, D. I. & Schiel, D. R. 2010. Algal populations controlled by fish herbivory across a wave exposure gradient on southern temperate shores. Ecology 91:201-11. Tellier, F., Meynard, A. P., Correa, J. A., Faugeron, S. & Valero, M. 2009. Phylogeographic analyses of the 30°S south-east Pacific biogeographic transition zone establish the occurrence of a sharp genetic discontinuity in the kelp Lessonia nigrescens: vicariance or parapatry? Mol Phylogenet Evol 53:679-93. Thiel, M. 2003. Rafting of benthic macrofauna: important factors determining the temporal succession of the assemblage on detached macroalgae. Hydrobiologia 503:49-57. Thiel, M. & Gutow, L. 2005a. The ecology of rafting in the marine environment. I. The floating substrata. Oceanogr. Mar. Biol. Ann. Rev. 42:181-263. Thiel, M. & Gutow, L. 2005b. The ecology of rafting in the marine environment. II. The rafting organisms and community. Oceanogr. Mar. Biol. Ann. Rev. 43:279-418. Thiel, M. & Haye, P. A. 2006. The ecology of rafting in the marine environment. III. Biogeographical and evolutionary consequences. Oceanogr. Mar. Biol. Ann. Rev. 44:323-429. Thiel, M., Macaya, E. C., Acuña, E., Arntz, W. E., Bastias, H., Brokordt, K., Camus, P. A., Castilla, J. C., Castro, L. R., Cortés, M., Dumont, C. P., Escribano, R., Fernández, M., Gajardo, J. A., Gaymer, C. F., Gómez, I., González, A. E., González, H. E., Haye, P. A., Illanes, J. E., Iriarte, J. L., Lancellotti, D. A., Luna-Jorquera, G., Luxoroi, C., Manríquez, P. H., Marín, V., Muñoz, P., Navarrete, S. A., Pérez, E., Poulin, E., Sellanes, J., Sepúlveda, H. H., Stotz, W., Tala, F., Thomas, A., Vargas, C. A., Vásquez, J. A. & Vega, J. M. A. 2007. The Humboldt Current System of northern and central Chile. Oceanogr Mar Biol 45:195-344. 125

Thiel, M., & Fraser, C.I. 2016. The role of floating plants in dispersal of biota across habitats and ecosystems. In: Olafsson E (ed) Marine Macrophytes as Foundation Species. CRC Press, Taylor & Francis Group, USA, pp 76-99. Vandendriessche, S., Vincx, M. & Degraer, S. 2007. Floating seaweed and the influences of temperature, grazing and clump size on raft longevity - A microcosm study. J. Exp. Mar. Biol. Ecol. 343:64-73. Vega, J. M. A., Vásquez, J. A. & Buschmann, A. H. 2005. Population biology of the subtidal kelps Macrocystis integrifolia and Lessonia trabeculata (Laminariales, Phaeophyceae) in an upwelling ecosystem of northern Chile: interannual variability and El Niño 1997–1998. Rev. Chil. Hist. Nat. 78:33-50. Voisin, M., Engel, C. R. & Viard, F. 2005. Differential shuffling of native genetic diversity across introduced regions in a brown alga: aquaculture vs. maritime traffic effects. Proc Natl Acad Sci USA 102:5432-7. Waters, J. M. 2011. Competitive exclusion: phylogeography’s ‘elephant in the room’? Mol. Ecol. 20:4388-94. Waters, J. M. & Craw, D. 2018. Cyclone-driven marine rafting: storms drive rapid dispersal of buoyant kelp rafts. Mar. Ecol. Prog. Ser. 602:77-85. Waters, J. M., Fraser, C. I. & Hewitt, G. M. 2013. Founder takes all: density-dependent processes structure biodversity. Trends Ecol. Evol. 28:78-85. Waters, J. M., King, T.M., Fraser, C.I. & Garden, C. 2018a. Rafting dispersal in a brooding southern sea star (Asteroidea: Anasterias). Invertebr. Syst. 32:253-258. Waters J.M., King, T.M., Fraser, C.I. & Craw, D. 2018b. Crossing the front: contrasting storm-forced dispersal dynamics revealed by biological, geological and genetic analysis of beach-cast kelp. J R Soc Interface 15:20180046. Westermeier, R., Muller, D. G., Gomez, I., Rivera, P. & Wenzel, H. 1994. Population biology of Durvillaea antarctica and Lessonia nigrescens (Phaeophyta) on the rocky shores of southern Chile. Mar. Ecol. Prog. Ser. 110:187-94. Wichmann, C. S., Hinojosa, I. A. & Thiel, M. 2012. Floating kelps in Patagonian Fjords: an important vehicle for rafting invertebrates and its relevance for biogeography. Mar. Biol. 159:2035-49. Worm, B. & Chapman, A. R. 1996. Interference competition among two intertidal seaweeds: Chondrus crispus strongly affects survival of Fucus evanescens recruits. Mar. Ecol. Prog. Ser. 145:297-301. Yuras, G., Ulloa, O. & Hormazábal, S. 2005. On the annual cycle of coastal and open ocean satellite chlorophyll off Chile (18 degrees-40 degrees S). Geophys. Res. Lett. 32:L23604.

126

Appendices

Supplementary Material 1

Journal of Phycology

The variable routes of rafting: stranding dynamics of floating bull-kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches in the SE Pacific

Boris A. López, Erasmo C. Macaya, Fadia Tala, Florence Tellier and Martin Thiel

127

Supplementary Figures

Figure S1.1. Box plot of stranded biomass of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer and years (2013-2015). Different letters above the box plot indicate differences between biogeographic districts and years (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

128

Figure S1.2. Box plot of maximum length of stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

129

Figure S1.3. Box plot of maximum wet weight of stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

130

Figure S1.4. Box plot of maximum number of stipes per plant of stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

131

Fig. S1.5. Box plot of maximum size of Lepas spp. attached in stranded individuals of Durvillaea antarctica on beaches to the continental coast of Chile (28°S-42°S), according to marine biogeographic districts during winter and summer (2013-2015). Different letters above the box plot indicate differences between biogeographic districts (P < 0.05). CCD: Coquimbo-Choros District; SED: Septentrional District; MED: Mediterranean District; MD: Meridional District. Horizontal lines represent the median; boxes, the interquartile range; whiskers, 1.5x of interquartile range; circles, the outlier values.

132

Supplementary Tables

Table S1.1. Beaches sampled in the study, according to biogeographic districts (Coquimbo-Choros District and Septentrional District) of continental coast of Chile (28°S-33°S). Coordinates and distances surveyed are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth.

Coordinates Distance surveyed (km) District Beach type Latitude Longitude winter summer winter summer winter summer (S) (W) 2013 2013/14 2014 2014/15 2015 2015/16 Bahía Sarco boulder 28°48’04’’ 71°24’31’’ 2.83 2.83 2.83 2.84 2.75 2.72 Mamani boulder 29°06’35’’ 71°28’42’’ 3.56 3.50 3.54 4.75 3.32 3.34 Punta de sandy 29°16’44’’ 71°26’21’’ 5.25 5.08 6.59 7.05 6.61 6.57 Coquimbo- Choros Choros Caleta boulder 29°37’14’’ 71°17’42’’ 2.67 2.67 2.68 2.68 2.72 2.71 District Hornos (CCD) Punta sandy 29°52’57’’ 71°17’25’’ 6.83 2.40 2.15 2.57 2.34 2.39 Teatinos Lagunillas sandy 30°09’50’’ 71°22’30’’ 6.78 5.41 6.45 6.33 5.74 5.84 El Sauce boulder 30°32’10’’ 71°41’35’’ 1.68 1.33 2.58 2.76 1.68 1.68 Fundo Agua sandy 31°30’19’’ 71°34’27’’ 1.84 0.28 0.48 1.94 0.73 0.47 Dulce Playa sandy 31°52’54’’ 71°30’39’’ 2.82 2.82 2.82 2.75 2.76 2.63 Septentrional Amarilla District (SED) Pichicuy sandy 32°21’15’’ 71°27’28’’ 2.25 1.52 2.24 1.85 2.02 2.22 Maitencillo sandy 32°38’17’’ 71°25’50’’ 1.39 1.39 1.83 1.89 2.27 2.31 Ritoque sandy 32°53’39’’ 71°30’45’’ 3.06 1.90 3.74 3.71 3.42 3.22 Quintay sandy 33°10’58’’ 71°41’02’’ 1.15 0.84 1.15 1.18 1.11 1.13 Punta Tralca sandy 33°25’23’’ 71°42’15’’ 0.56 0.53 1.03 1.48 1.06 1.05

133

Table S1.2. Beaches sampled in the study, according to biogeographic districts (Mediterranean District and Meridional District) of continental coast of Chile (33°S-42°S). Coordinates and distances surveyed are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth.

Coordinates Distance surveyed (km) District Beach type Latitude Longitude winter summer winter summer winter summer (S) (W) 2013 2013/14 2014 2014/15 2015 2015/16 Santo boulder 33°39’31’’ 71°38’05’’ 2.48 2.63 3.15 2.79 3.87 3.17 Domingo Matanzas sandy 33°58’50’’ 71°53’00’’ 1.78 2.13 2.49 2.47 2.37 2.34 Pichilemu sandy 34°25’37’’ 72°01’33’’ 4.19 4.21 4.12 4.69 4.61 3.88 Mediterranean Duao sandy 34°53’06’’ 72°08’53’’ 6.46 6.42 5.17 3.31 3.26 6.34 District Las Cañas sandy 35°27’50’’ 72°29’30’’ 5.74 6.14 5.93 6.36 6.18 6.42 (MED) Curanipe sandy 35°50’43’’ 72°37’28’’ 2.19 2.50 2.13 2.73 2.67 2.48 Itata Norte sandy 36°18’00’’ 72°49’47’’ 2.77 2.25 2.96 2.71 2.44 2.42 San Pedro de sandy 36°56’10’’ 73°09’13’’ 11.08 10.89 6.22 9.49 9.10 4.11 la Paz Quidico sandy 37°25’28’’ 73°38’32’’ 5.49 4.04 3.93 0.62 4.35 4.45 Quidico- sandy 38°14’42’’ 73°28’14’’ 2.43 2.43 2.44 2.43 2.33 2.29 Tirua Playa Piedra sandy 38°58’27’’ 73°20’15’’ 2.81 3.22 2.83 2.11 3.49 1.94 Queule sandy 39°20’29’’ 73°12’24’’ 4.46 2.55 3.59 4.58 5.62 2.87 Meridional Curiñanco sandy 39°44’30’’ 73°23’18’’ 2.05 1.25 0.83 2.01 0.98 1.08 District (MD) Chaihuin sandy 39°56’28’’ 73°35’34’’ 2.12 2.02 1.92 1.99 1.94 1.90

Pucatrihue sandy 40°31’34’’ 73°43’09’’ 2.26 0.94 1.04 1.17 1.93 1.43 Huar-Huar sandy 41°18’56’’ 73°49’51’’ 2.02 2.62 2.54 2.12 2.44 2.39 Carelmapu sandy 41°43’39’’ 73°44’31’’ 2.59 1.94 1.92 2.55 2.84 1.35 Mar Brava sandy 41°54’23’’ 74°00’51’’ 3.11 3.16 2.52 1.22 0.54 0.72 Cucao boulder 42°38’43’’ 74°07’13’’ 1.96 1.98 2.60 2.39 2.12 1.72 Journal of Applied Phycology

134

Supplementary Material 2

Journal of Applied Phycology

Spatio-temporal variability of strandings of the southern bull kelp Durvillaea antarctica (Fucales, Phaeophyceae) on beaches on the coast of Chile - linking with local storms.

Boris A. López, Erasmo C. Macaya, Ricardo Jeldres, Nelson Valdivia, César C. Bonta, Fadia Tala and Martin Thiel

135

Table S2.1. Summary of GLM (only source of variations and P-values) for stranded biomass and number of rafts of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Pairwise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Nonsignificant pairwise tests are not shown. Pi: Pichicuy; It: Itata Norte; Cu: Curiñanco; Y1: year 1 (2014/15), Y2: year 2 (2015/16); Y3: year 3 (2016/17); Ju: July; Se: September; No: November; Ja: January; Ma: March; May: May.

Variable Source of variation P-value Significant pairwise comparison Stranded Beach, (B) < 0.001 Pi > It < Cu Biomass Year, (Y) < 0.050 Y1 = Y2 > Y3 Month, (M) < 0.001 Ju > Se = No < Ja = Ma = May Douglas, (D) < 0.050 B x Y < 0.001 Pi: Y1 = Y2 > Y3 It: Y1 < Y2 > Y3 Cu: Y1 < Y2 > Y3 B x M < 0.010 Pi: Ju > Se = No < Ja = Ma = May It: Ju > Se = No = Ja = Ma = May Cu: Ju > Se = No < Ja = Ma = May Y x M < 0.010 Y1: Ju > Se = No < Ja = Ma = May Y2: Ju > Se = No < Ja < Ma = May Y3: Ju = Se = No = Ja = Ma = May B x D < 0.050 Y x D 0.133 M x D 0.337 B x Y x D 0.065 B x M x D 0.196 Y x M x D 0.203 Number Beach, (B) < 0.001 Pi > It < Cu of rafts Year, (Y) < 0.050 Y1 < Y2 > Y3

136

Month, (M) < 0.001 Ju > Se = No < Ja = Ma = May Douglas, (D) < 0.050 B x Y < 0.010 Pi: Y1 < Y2 > Y3 It: Y1 < Y2 > Y3 Cu: Y1 = Y2 > Y3 B x M < 0.010 Pi: Ju > Se = No < Ja = Ma = May It: Ju = Se = No = Ja = Ma = May Cu: Ju > Se = No < Ja = Ma = May Y x M < 0.010 Y1: Ju > Se = No < Ja = Ma = May Y2: Ju > Se = No < Ja < Ma = May Y3: Ju = Se = No = Ja = Ma = May B x D < 0.050 Y x D 0.209 M x D 0.228 B x Y x D 0.092 B x M x D 0.478 Y x B x D 0.512

137

Table S2.2. Summary of GLM (only source of variations and P-values) for length and wet weight of stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (32°S-39°S), according to beaches, years, months as factors, and Douglas sea scale (weekly maximum value) as covariate. Pairwise post hoc comparisons were done on significant terms. Significant values (P < 0.05) are shown in bold. Nonsignificant pairwise tests are not shown. Pi: Pichicuy; It: Itata Norte; Cu: Curiñanco; Y1: year 1 (2014/15), Y2: year 2 (2015/16); Y3: year 3 (2016/17); Ju: July; Se: September; No: November; Ja: January; Ma: March; May: May.

Variable Source of variation P-value Significant pairwise comparison Length Beach, (B) 0.107 Year, (Y) 0.073 Month, (M) 0.420 Douglas, (D) 0.213 B x Y < 0.001 Pi: Y1 = Y2 > Y3 It: Y1 = Y2 < Y3 Cu: Y1 < Y2 = Y3 B x M < 0.001 Pi: Ju = Se = No = Ja = Ma = May It: Ju > Se = No = Ja = Ma = May Cu: Ju < Se > No = Ja = Ma = May Y x M < 0.001 Y1: Ju = Se = No = Ja = Ma = May Y2: Ju = Se = No = Ja = Ma = May Y3: Ju > Se > No = Ja = Ma = May B x Y x M 0.096 B x D 0.186 Y x D 0.320 M x D 0.563 B x Y x D 0.238 B x M x D 0.609 Y x M x D 0.581 B x Y x M x D 0.318 Wet Beach, (B) < 0.010 Pi > It < Cu Weight Year, (Y) < 0.010 Y1 = Y2 < Y3 Month, (M) 0.256

138

Douglas, (D) 0.182 B x Y < 0.001 Pi: Y1 > Y2 > Y3 It: Y1 = Y2 < Y3 Cu: Y1 < Y2 < Y3 B x M < 0.001 Pi: Ju < Se = No = Ja = Ma > May It: Ju = Se > No = Ja = Ma = May Cu: Ju < Se > No = Ja < Ma > May Y x M < 0.001 Y1: Ju > Se = No < Ja = Ma = May Y2: Ju > Se = No < Ja = Ma = May Y3: Ju = Se > No = Ja = Ma = May B x Y x M 0.563 B x D 0.172 Y x D 0.491 M x D 0.163 B x Y x D 0.417 B x M x D 0.288 Y x M x D 0.195 B x Y x M x D 0.374

139

Fig. S2.1. Summary of main change tendencies on studied responses of bull kelp Durvillaea antarctica strandings. Letters (a-b-c-d) in the right column indicate significant differences between factor categories (P < 0.05). Letters on the right side correspond to the results of pairwise comparisons for each response variable. Horizontal solid gray lines correspond to reference for significant groups.

140

Fig. S2.2. Relationships between stranded biomass of Durvillaea antarctica (kg per km of shoreline) and Douglas sea scale (mean and maximum values of biweekly lag according to the date of each survey) on three beaches from the continental coast of Chile (28°S- 42°S). A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated.

141

Fig. S2.3. Relationships between stranded biomass of Durvillaea antarctica (kg per km of shoreline) and Douglas sea scale (mean and maximum values of monthly lag according to the date of the each survey) on three beaches from the continental coast of Chile (28°S- 42°S). A-B: Pichicuy (32°S), C-D: Itata Norte (36°S) and E-F: Curiñanco (39°S). Summary of simple linear regression between both variables for each case is also indicated.

142

Fig. S2.4. Percentages of stranded biomass of bull kelps of Durvillaea antarctica, according to complete plants (i.e. individuals with intact fronds that are coalesced within a single holdfast) and fragments (i.e. parts of a frond without holdfasts or only holdfasts) on three beaches from the continental coast of Chile (28°S-42°S) during bimonthly surveys on three years (2014/15-2016/17). A-B-C: Pichicuy (32°S), D-E-F: Itata Norte (36°S) and G-H-I: Curiñanco (39°S).

143

Fig. S2.5. Bimonthly accumulated landings of artisanal fisheries of bull kelps (Durvillaea antarctica) on three political-administrative regions of Chile which include the surveyed beaches, between July 2014 to June 2017. A: Vth Region (for Pichicuy, 32°S), B: VIIIth Region (for Itata Norte, 36°S) and C: XIVth Region (for Curiñanco, 39°S). Source: Fisheries Statistical Yearbook 2014-2017. National Fisheries and Aquaculture Service (Anuario Estadístico de Pesca. Servicio Nacional de Pesca y Acuicultura). Gobierno de Chile. (http://www.sernapesca.cl/informes/estadisticas).

144

Supplementary Material 3

Journal of Biogeography

SUPPORTING INFORMATION

Epibiont communities on stranded kelp rafts of Durvillaea antarctica (Fucales, Phaeophyceae) – do positive interactions facilitate range extensions?

Boris A. López, Erasmo C. Macaya, Marcelo M. Rivadeneira, Fadia Tala, Florence Tellier and Martin Thiel

145

Appendix S3.1: Sampling sites

Table S3.1.1. Beaches sampled in the study, according to biogeographic districts (Coquimbo-Choros District, CCD and Septentrional District, SED) along the continental coast of Chile (28°S-33°S). Coordinates, distances surveyed and number of sampled individuals of Durvillaea antarctica are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth. Number of sampled specimens of D. antarctica are shown in parentheses.

Coordinates Distance surveyed (km) District Beach type Latitude Longitude winter summer winter summer (S) (W) 2014 2014/15 2015 2015/16 Bahía Sarco boulder 28°48’04’’ 71°24’31’’ 2.83 (0) 2.84 (0) 2.75 (0) 2.72 (0) Coquimbo- Mamani boulder 29°06’35’’ 71°28’42’’ 3.54 (1) 4.75 (0) 3.32 (0) 3.34 (0) Choros Punta de Choros sandy 29°16’44’’ 71°26’21’’ 6.59 (52) 7.05 (25) 6.61 (1) 6.57 (6) District Caleta Hornos boulder 29°37’14’’ 71°17’42’’ 2.68 (3) 2.68 (12) 2.72 (0) 2.71 (1) (CCD) Punta Teatinos sandy 29°52’57’’ 71°17’25’’ 2.15 (21) 2.57 (25) 2.34 (20) 2.39 (1) Lagunillas sandy 30°09’50’’ 71°22’30’’ 6.45 (34) 6.33 (10) 5.74 (60) 5.84 (1) El Sauce boulder 30°32’10’’ 71°41’35’’ 2.58 (14) 2.76 (26) 1.68 (10) 1.68 (5) Fundo Agua Dulce sandy 31°30’19’’ 71°34’27’’ 0.48 (88) 1.94 (55) 0.73 (140) 0.47 (56) Playa Amarilla sandy 31°52’54’’ 71°30’39’’ 2.82 (99) 2.75 (39) 2.76 (65) 2.63 (11) Septentrional Pichicuy sandy 32°21’15’’ 71°27’28’’ 2.24 (48) 1.85 (117) 2.02 (138) 2.22 (7) District (SED) Maitencillo sandy 32°38’17’’ 71°25’50’’ 1.83 (25) 1.89 (13) 2.27 (2) 2.31 (46) Ritoque sandy 32°53’39’’ 71°30’45’’ 3.74 (14) 3.71 (10) 3.42 (24) 3.22 (20) Quintay sandy 33°10’58’’ 71°41’02’’ 1.15 (12) 1.18 (11) 1.11 (8) 1.13 (16) Punta Tralca sandy 33°25’23’’ 71°42’15’’ 1.03 (24) 1.48 (18) 1.06 (53) 1.05 (56)

146

Table S3.1.2. Beaches sampled in the study, according to biogeographic districts (Mediterranean District, MED and Meridional District, MD) along the continental coast of Chile (33°S-42°S). Coordinates, distances surveyed and number of sampled individuals of Durvillaea antarctica are given, according to season and year. Distances were corrected by the start and end points of each survey on Google Earth. Number of sampled specimens of D. antarctica are shown in parentheses.

Coordinates Distance surveyed (km) District Beach type Latitude Longitude winter summer winter summer (S) (W) 2014 2014/15 2015 2015/16 Santo Domingo boulder 33°39’31’’ 71°38’05’’ 3.15 (0) 2.79 (27) 3.87 (6) 3.17 (28) Matanzas sandy 33°58’50’’ 71°53’00’’ 2.49 (17) 2.47 (30) 2.37 (43) 2.34 (30) Pichilemu sandy 34°25’37’’ 72°01’33’’ 4.12 (30) 4.69 (8) 4.61 (114) 3.88 (19) Mediterranean Duao sandy 34°53’06’’ 72°08’53’’ 5.17 (56) 3.31 (27) 3.26 (39) 6.34 (42) District Las Cañas sandy 35°27’50’’ 72°29’30’’ 5.93 (27) 6.36 (17) 6.18 (7) 6.42 (52) (MED) Curanipe sandy 35°50’43’’ 72°37’28’’ 2.13 (9) 2.73 (27) 2.67 (6) 2.48 (10) Itata Norte sandy 36°18’00’’ 72°49’47’’ 2.96 (37) 2.71 (33) 2.44 (77) 2.42 (134) San Pedro de la Paz sandy 36°56’10’’ 73°09’13’’ 6.22 (0) 9.49 (2) 9.10 (0) 4.11 (19) Quidico sandy 37°25’28’’ 73°38’32’’ 3.93 (18) 0.62 (97) 4.35 (112) 4.45 (139) Quidico-Tirua sandy 38°14’42’’ 73°28’14’’ 2.44 (79) 2.43 (63) 2.33 (112) 2.29 (139) Playa Piedra sandy 38°58’27’’ 73°20’15’’ 2.83 (2) 2.11 (21) 3.49 (30) 1.94 (13) Queule sandy 39°20’29’’ 73°12’24’’ 3.59 (18) 4.58 (19) 5.62 (65) 2.87 (3) Meridional Curiñanco sandy 39°44’30’’ 73°23’18’’ 0.83 (92) 2.01 (16) 0.98 (95) 1.08 (68) District (MD) Chaihuin sandy 39°56’28’’ 73°35’34’’ 1.92 (32) 1.99 (13) 1.94 (56) 1.90 (35) Pucatrihue sandy 40°31’34’’ 73°43’09’’ 1.04 (19) 1.17 (29) 1.93 (71) 1.43 (105) Huar-Huar sandy 41°18’56’’ 73°49’51’’ 2.54 (32) 2.12 (75) 2.44 (137) 2.39 (40) Carelmapu sandy 41°43’39’’ 73°44’31’’ 1.92 (25) 2.55 (38) 2.84 (76) 1.35 (79) Mar Brava sandy 41°54’23’’ 74°00’51’’ 2.52 (74) 1.22 (85) 0.54 (196) 0.72 (107) Cucao boulder 42°38’43’’ 74°07’13’’ 2.60 (100) 2.39 (83) 2.12 (22) 1.72 (69)

147

Appendix S3.2: Expanded Materials and Methods

Sampling and morphological measurements of stranded individuals of Durvillaea antarctica

Recently stranded individuals of D. antarctica were collected on each beach during winters and summers in two consecutive years (2014/15 and 2015/16). Surveys were done on foot following the coastline, collecting entire individuals of D. antarctica, along the most recent flotsam lines (from the last 2 - 3 high tides). Only recently stranded specimens were collected (i.e. bull kelps with greenish or dark-brown color that maintained some flexibility as an indicator of freshness), because during the summer bull kelps rapidly dried out and the possibility of finding attached organisms is lower in old, stranded individuals. Likewise, because the stranded individuals of D. antarctica may present coalesced holdfasts with multiple stipes (each stipe corresponds to a different individual, González et al., 2015), hereafter a stranded specimen that has one or multiple stipes within a single holdfast will be called a “plant”. For the purpose of this study, only intact individuals with both fronds and holdfast were considered.

A total of 5,219 complete plants were measured during the study. For each complete plant of D. antarctica, the following variables were measured:

(BH) Biomass of holdfast: the wet weight of the holdfast of each plant was measured with a portable electronic hanging digital scale of 1 g accuracy. For this, all stipes were separated from the holdfast at the base of the stipe (< 1 cm).

(FT) Floating time: for each plant we determined whether it had been colonized by stalked barnacles of the genus Lepas or not. Any item floating at the sea surface (including buoyant seaweeds) is rapidly colonized by these organisms. The availability of its competent larvae may have some seasonality (i.e. higher in spring), but generally they can be found throughout the year (Anderson 1994). Therefore, presence and size can indirectly indicate the floating time of a substratum (Thiel & Gutow, 2005). In this case, for each stranded bull kelp the presence of cyprids (settled larvae) or adult specimens of Lepas spp. was verified in the fronds and the holdfast. In the case of cypris larvae, their presence was noted in the field, while for metamorphosed specimens of Lepas spp. the 10 to 20

148 largest individuals from each plant were collected for further measurements. Subsequently, the Lepas individuals were photographed with a scale and measured, considering as a size estimator the capitular length (rectilinear distance between the distal angle of the carina plate and the beginning of the peduncle) using Image Pro Plus v6 (Media Cybernetics Inc., Rockville, USA).

According to the presence and size of stalked barnacles, all D. antarctica rafts were categorized in three groups: (i) short floating time (< 2 days) – plants without any Lepas; (ii) intermediate floating time (2 - 10 days) – plants with cyprid recruits or small, juvenile Lepas (< 5 mm capitular length); and (iii) long floating time (> 10 days) – plants with large, adult Lepas (≥ 5 mm capitular length). According to Thiel and Gutow (2005), the growth rates of L. anatifera and L. australis range from 0.22 to 0.46 mm ꞏ day-1. Therefore, plants of D. antarctica with Lepas> 5 mm are equivalent to more than 10 days of floating times.

Taxonomic richness of epibionts (statistical analyses)

For the purpose of the quantitative analyses, only those species that had more than one record on all stranded bull kelps sampled throughout the entire study were considered. Three-way ANCOVA tests were run to compare the taxonomic richness of epibionts (total, only sessile, and only mobile) considering biogeographic district (fixed factor: CCD, SED, MED, MD), year (fixed factor: Year 1 and 2), and season (fixed factor: winter and summer) as factors, and the holdfast wet weight as covariate, following analysis of slope homogeneity and square root x+1 transformation of the dependent variable (Zar, 2010). The holdfast wet weight was used because there exists a strong relationship between biomass of kelp holdfasts and epibiont taxonomic richness (Thiel & Vásquez, 2000).

A separate one-way ANCOVA was done in the case of floating time (fixed factor: short, intermediate and long), because it was not orthogonal to the other factors, applying the same procedure as described above. Lastly, if there were significant effects (P < 0.05), the differences between categories of factors were examined a posteriori with Tukey HSD tests (Zar, 2010). All statistical analyses of taxonomic richness were done using “lme4” and “multcomp” packages in R 3.4 (R Development Core Team, 2017).

149

Comparison of geographical ranges (statistical analyses)

In order to examine whether the floating time of stranded D. antarctica (i.e. absence and presence of Lepas spp. attached) was associated with the probability of finding rafting epibionts outside their literature ranges, generalized linear models (GLMs) with a negative binomial distribution with a logarithm link function (Lawless, 1987) were performed for SE and NE species, separately. Modelling was based on the number of beach-cast rafts of D. antarctica with epibionts outside their respective literature ranges according to floating time (i.e. absence and presence of Lepas spp. attached), offset by the logarithm of total number of sampled rafts for each beach in the study area. The model used was: model = glm.nb (number of specimens outside their range~Lepas spp. + offset (log (total number of rafts)), data=data).

In order to examine if positive and negative co-occurrences correlate with the observed range expansions, for each chosen species, i.e. positive co-occurrences (Gelidium lingulatum and Semimytilus algosus) and negative co-occurrences (Limnoria chilensis and Scurria scurra), generalized linear models (GLMs) with a negative binomial distribution were performed for SE and NE species, separately. In each case, modelling was based on the number of beach-cast rafts of D. antarctica with epibionts outside their respective literature ranges, expressed as percentage per beach, according to absence and presence of each of four selected epibiont species, offset by the logarithm of total number of sampled rafts for each beach in the study area. The model used was: model = glm.nb (number of specimens outside their range~epibiont species with co-occurrences + offset (log (total number of rafts)), data=data). All these statistical tests (GLMs) were done with R 3.4., using the “MASS” package (R Development Core Team, 2017).

References

150

Anderson, D.T. (1994) Barnacles: Structure, Function, Development and Evolution. Chapman and Hall, London, UK.

Lawless, J.F. (1987) Negative binomial and mixed Poisson regression. Canadian Journal of Statistics, 15, 209-225.

González, A.V., Beltrán, J., Flores, V. & Santelices, B. (2015) Morphological convergence in the inter-holdfast coalescence process among kelp and kelp-like seaweeds (Lessonia, Macrocystis, Durvillaea). Phycologia, 54, 283-291.

R Develoment Core Team (2017) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL: http://www.r-project.org/.

Thiel, M. & Vásquez, J.A. (2000) Are kelp holdfasts islands on the ocean floor? indication for temporarily closed aggregations of peracarid crustaceans. Hydrobiologia, 440, 45-54.

Thiel, M. & Gutow, L. (2005) The ecology of rafting in the marine environment. II. The rafting organisms and community. Oceanography and Marine Biology: An Annual Review, 43, 279-418.

Zar, J.H. (2010) Biostatistical analysis, 5th Edition. Prentice-Hall, Englewood Cliffs, NJ, USA.

151

Appendix S3.3: Geographic ranges and biological features of epibiont species.

Table S3.3.1: Epibiont species attached on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S) selected for analysis of comparison of geographical ranges. Type of epibiont (seaweed or animal), motility (sessile or mobile), geographic range limits (northern and southern edges), percentage of individuals of D. antarctica with Lepas spp. attached (total records of each species are also indicated in parentheses), autonomous dispersal ability (high or low, according to the criterion based on the presence of planktonic larvae or propagules in the water column for more than 7 days, a period previously described in Haye et al. 2014 as important for the genetic differentiation of marine invertebrate populations across the phylogeographic break at 30°S on the coast of Chile), reproduction type (DD: direct development; PL: pelagic larval; PP: propagules such as spores, gametes, zygotes, papillae or sporophylls; VR: vegetative reproduction, such as fragmentation or prostrate thalli; pelagic larval duration are also mentioned) and references for each species are indicated.

Species Seaweed Sessile Benthic Percent of Autonomous Reproduction References (S) or (Sess) ranges stranded dispersal type Animal or limits bull kelps ability (A) Mobile (northern with Lepas (Mob) and spp. southern) (number of specimens) 1: Scurria scurra A Mob 18°S- 12.1 High PL: 7-10 Brattström & Johanssen, 1983; 56°S (1573) days Thiel & Ullrich, 2002; Aldea & Valdovinos, 2005; Kolbin & Kulikova, 2011; Soto et al., 2015. 2: Perumytilus A Sess 18°S- 10.2 (137) High PL: 14 days Soot-Ryen, 1955; Ramorino & purpuratus 56°S Campos, 1979; Prado &

152

Castilla, 2006; Pérez et al., 2008; Briones et al., 2013; Trovant et al., 2015; Velásquez et al., 2016. 3: Macrocystis S Sess 18°S- 23.5 (17) High PP Ramírez & Santelices, 1991; pyrifera 56°S Hoffmann & Santelices, 1997. Macaya & Zuccarello, 2010. 4: Ceramium virgatum S Sess 18°S- 0 (5) Low PP Ramírez & Santelices, 1991; 56°S Hoffmann & Santelices, 1997; Leonardi et al., 2006. 5: Phragmatopoma A Sess 18°S- 6.4 (717) High PD: > 90 Rozbaczylo, 1985; Sepúlveda moerchi 53°S days et al., 2003; Rozbaczylo et al., 2017. 6: Phymactis papillosa A Sess 18°S- 17.6 (17) Low PD: < 1 day Excoffon & Zamponi, 1991; 45°S Häussermann, 2004; 2006; Gomes et al., 2012. 7: Brachidontes A Sess 18°S- 75.0 (4) High PD: 16-24 Soot-Ryen, 1959; Sepúlveda et granulata 42°S days al., 2003; Aldea & Valdovinos, 2005; Kelaher et al., 2007; Lagos et al., 2007; Trovant et al., 2015. 8: Semimytilus algosus A Sess 18°S- 6.7 (713) High PD: 14-20 Brattström & Johanssen, 1983; 42°S days Tokeshi & Romero, 1995; Cerda & Castilla, 2001; Caro & Castilla, 2004; Villegas et al., 2005; Caro et al., 2008. 9: Asterfilopsis S Sess 21.9°S- 8.6 (46) Low PP Ramírez & Santelices, 1991; disciplinalis 42°S Hoffmann & Santelices, 1997. Calderon & Boo, 2016. 10: Symphyocladia S Sess 18°S- 0 (8) Low PP Ramírez & Santelices, 1991; dendroidea 42°S Hoffmann & Santelices, 1997; Ramírez et al., 2008;

153

Bustamante et al., 2016. 11: Griffithsia S Sess 18°S- 20.0 (5) Low PP Ramírez & Santelices, 1991; chilensis 42°S Hoffmann & Santelices, 1997; Uribe et al., 2015. 12: Antithamnionella S Sess 23°S- 10.1 (59) Low PP Ramírez & Santelices, 1991; ternifolia 56°S Athanasiadis, 1996; Hoffmann & Santelices, 1997.Vásquez et al., 2001; Ramírez et al., 2008. 13: Schottera S Sess 23°S- 7.5 (40) Low PP Santelices et al., 1989; Ramírez nicaeensis 42°S & Santelices, 1991; Hoffmann & Santelices, 1997; Villaseñor- Parada et al., 2014; 2017; 2018. 14: Mazzaella S Sess 28°S- 10.0 (30) Low PP Ramírez & Santelices, 1991; laminarioides 56°S Hoffmann & Santelices, 1997; Faugeron et al., 2001; Palma et al., 2007; Montecinos et al., 2012; Velásquez et al., 2016. 15: Limnoria chilensis A Mob 29°S- 13.3 Low DD Paternoster & Elias, 1980, 50°S (1207) Thiel & Vásquez, 2000; Thiel, 2003; Miranda & Thiel, 2008; Hinojosa et al., 2010; Haye et al., 2012. 16: Gelidium rex S Sess 28°S- 16.0 (100) Low PP, VR Santelices & Abbott, 1985; 36.5°S Ramírez & Santelices, 1991; Rodríguez & Santelices, 1995; Hoffmann & Santelices, 1997; Vásquez & Vega 2004; Ramírez et al., 2008; López et al., 2017. 17: Gelidium S Sess 28°S- 4.7 (313) Low PP, VR Santelices & Stewart, 1985; lingulatum 43°S Rodríguez & Santelices, 1995;

154

Ramírez & Santelices, 1991; Hoffmann & Santelices, 1997; López et al., 2017. 18: Hiatella solida A Sess 33°S- 22.2 (9) High PL: 30-60 Osorio & Reid, 2004; Aldea & 56°S days Valdovinos, 2005; Osorio et al., 2005; Ríos et al., 2007; Díaz & Campos, 2014; Laakkonen et al., 2015; Campos & Landaeta, 2016. 19: Sarcothalia S Sess 33°S- 4.2 (47) Low PP Ramírez & Santelices, 1991; crispata 56°S Ávila et al., 1996; Hoffmann & Santelices, 1997; Opazo & Otaíza, 2007; Werlinger et al., 2008. 20: Desmarestia S Sess 32.7°S- 75 (4) Low PP Ramírez & Santelices, 1991; ligulata 56°S Hoffmann & Santelices, 1997. 21: Mariaramirezia S Sess 34°S- 0 (4) Low PP Ramírez & Santelices, 1991; osornoensis 56°S León et al., 1995; Hoffmann & Santelices, 1997; Calderon et al., 2014; 2016. 22: Paraglossum S Sess 33.9°S- 0 (11) Low PP Ramírez & Santelices, 1991; crassinervium 56°S Hoffmann & Santelices, 1997. 23: Chondria S Sess 34°S- 16.6 (6) Low PP Ramírez & Santelices, 1991; secundata 42°S Hoffmann & Santelices, 1997; Ramírez et al., 2008. 24: Mastocarpus S Sess 33.9°S- 0 (4) Low PP Ávila & Alveal, 1987; Ramírez latissimus 41.9°S & Santelices, 1991; Hoffmann & Santelices, 1997; Castilla et al., 2005; Oróstica et al., 2012; Macaya et al., 2013; Villaseñor-Parada et al., 2018. 25: Camontagnea S Sess 42°S- 60.0 (5) Low PP Ramírez & Santelices, 1991;

155 oxyclada 56°S Mansilla et al., 2013. 26: Rhodymenia S Sess 18°S- 10.0 (21) Low PP Ramírez & Santelices, 1991; skottsbergii 41.8°S León et al., 1995; Hoffmann & Santelices, 1997; Ramírez et al., 2008; Gómez & Huovinen, 2011; Huovinen & Gómez, 2011; Villaseñor-Parada et al., 2014. 27: Antithamnion S Sess 21.9°S- 18.5 (27) Low PP Ramírez & Santelices, 1991; densum 32.2°S Hoffmann & Santelices, 1997; Valdivia et al., 2005. 28: Dendropoma A Sess 23.5°S- 40.0 (20) Low DD Vásquez & Vega, 2004; mejillonensis 30.3°S Pacheco & Laudien, 2008; Pacheco et al., 2011. 29: Petalonia S Sess 23.6°S- 25.0 (4) Low PP Ramírez & Santelices, 1991; binghamiae 33°S Hoffmann & Santelices, 1997; Santelices et al., 2002. 30: Gelidium chilense S Sess 23.6°S- 3.6 (55) Low PP, VR Santelices & Stewart, 1985; 36.5°S Ramírez & Santelices, 1991; Santelices & Varela, 1994; Rodríguez & Santelices, 1995; Hoffmann & Santelices, 1997; Santelices, 1999; Thomas & Freshwater, 2001; Medina et al., 2005; Wieters, 2005; Wieters et al., 2013; Soler et al., 2016. 31: Chaetomorpha S Sess 28.3°S- 20.0 (20) Low PP Ramírez & Santelices, 1991; firma 33.5°S Hoffmann & Santelices, 1997; Santelices et al., 2002, Navarrete et al., 2008; Ramírez et al., 2008.

156

32: Schimmelmannia S Sess 32.9°S- 42.8 (21) Low PP Ramírez & Santelices, 1991; plumosa 41°S Hoffmann & Santelices, 1997; Castilla & Neill, 2009; Villaseñor-Parada et al., 2018. 33: Lessonia spicata S Sess 29.0°S- 13.8 (579) Low PP Ramírez & Santelices, 1991; 41°S Hoffmann & Santelices, 1997; Tellier et al., 2009, 2011a, 2011b, 2011c; González et al., 2012; López-Cristoffanini et al., 2013; Flores-Molina et al., 2014; González et al., 2014; Ortega et al., 2014; Poore et al., 2014; Aguilera et al., 2015; Koch et al., 2015; Parada et al., 2016; Velásquez et al., 2016; Ritter et al., 2017. 34: Ballia callitricha S Sess 32.2°S- 75 (4) Low PP Ramírez & Santelices, 1991; 56°S Hoffmann & Santelices, 1997; Ramírez et al., 2008; Mansilla et al., 2013. 35: Parawaldeckia A Mob 41°S- 25 (4) Low DD Menzies, 1962; Hinojosa et al., kidderi 56°S 2007; González et al., 2008.

References

Aguilera, M.A., Valdivia, N. & Broitman, B.R. (2015) Herbivore-alga interaction strength influences spatial heterogeneity in a kelp- dominated intertidal community. PLoS One, 10, e0137287.

157

Aldea, C. & Valdovinos, C. (2005) Moluscos del intermareal rocoso del centro-sur de Chile (36-38 S): taxonomía y clave de identificación. Gayana (Concepción), 69, 364-396.

Athanasiadis, A. (1996) Morphology and classification of the Ceramioideae (Rhodophyta) based on phylogenetic principles. Opera Botanica Belgica, 127, 1-221.

Ávila, M. & Alveal, K. (1987) Ciclo de vida de Mastocarpus papillatus en el área de Concepción, Chile (Petrocelidaceae, Rhodophyta). Investigación Pesquera (Chile), 34, 129-138.

Ávila, M., Otaíza, R., Norambuena, R. & Nuñez, M. (1996) Biological basis for the management of ‘luga negra’ (Sarcothalia crispata Gigartinales, Rhodophyta) in southern Chile. Hydrobiologia, 326/327, 245-252.

Brattström, H. & Johanssen, A. (1983) Ecological and regional zoogeography of the marine benthic fauna of Chile - Report no 49 of the Lund University Chile Expedition 1948-49. Sarsia, 68, 289-334.

Briones, C., Presa, P., Pérez, M., Pita, A. & Guiñez, R. (2013) Genetic connectivity of the ecosystem engineer Perumytilus purpuratus north to the 32°S southeast Pacific ecological discontinuity. Marine Biology, 160, 3143-3156.

Bustamante, D.E., Won, B.Y. & Cho, T.O. (2016) Morphology and phylogeny of Pterosiphonia dendroidea (Rhodomelaceae, Ceramiales) described as Pterosiphonia tanakae from Japan. Botanica Marina, 59, 353-361.

Calderon, M.S. & Boo, S.M. (2016) Phylogeny of Phyllophoraceae (Rhodophyta, Gigartinales) reveals Asterfilopsis gen. nov. from the Southern Hemisphere. Phycologia, 55, 543-554.

Calderon, M.S., Boo, G.H. & Boo, S.M. (2014) Morphology and phylogeny of Ramirezia osornoensis gen. & sp. nov. and Phyllymenia acletoi sp. nov. (Halymeniales, Rhodophyta) from South America. Phycologia, 53, 23-36.

158

Calderon, M.S., Boo, G.H. & Boo, S.M. (2016) Correction to the paper "Morphology and phylogeny of Ramirezia osornoensis gen. & sp. nov. and Phyllymenia acletoi sp. nov. (Halymeniales, Rhodophyta) from South America". Phycologia, 55: 610.

Campos, B. & Landaeta, M.F. (2016) Moluscos planctónicos entre el fiordo Reloncaví y el golfo Corcovado, sur de Chile: ocurrencia, distribución y abundancia en invierno. Revista de Biología Marina y Oceanografía, 51, 527-539.

Caro, A.U. & Castilla, J.C. (2004) Predator-inducible defences and local intrapopulation variability of the intertidal mussel Semimytilus algosus in central Chile. Marine Ecology Progress Series, 276, 115-123.

Caro, A.U., Escobar, J., Bozinovic, F., Navarrete, S.A. & Castilla, J.C. (2008) Phenotypic variability in byssus thread production of intertidal mussels induced by predators with different feeding strategies. Marine Ecology Progress Series, 372, 127-134.

Castilla, J.C., Uribe, M., Bahamonde, N., Clarke, M., Desqueyroux-Faúndez, R., Kong, I., Moyano, H., Rozbaczylo, N., Santelices, B., Valdovinos, C. & Zavala, P. (2005) Down under the southeastern Pacific: marine non-indigenous species in Chile. Biological Invasions, 7, 213-232.

Castilla, J.C. & Neill, P.E. (2009) Marine bioinvasions in the southeastern Pacific: status, ecology, economic impacts, conservation and management. Biological invasions in marine ecosystems (ed. by G. Rilov and J.A. Crooks), pp. 439-457. Springer-Verlag, Berlin, Germany.

Cerda, M. & Castilla, J.C. (2001) Diversidad y biomasa de macro-invertebrados en matrices intermareales del tunicado Pyura praeputialis (Heller, 1878) en la Bahía de Antofagasta, Chile. Revista Chilena de Historia Natural, 74, 841-853.

Díaz, P. & Campos, B. (2014) Ontogenia de la concha larval y postlarval de cuatro especies de bivalvos de la costa del Pacífico sureste. Revista de Biología Marina y Oceanografía, 49, 175-191.

159

Excoffon, A. & Zamponi, M. (1991) La biología reproductiva de Phymactis clematis Dana, 1849 (Actiniaria: Actiniidae); gametogénesis, períodos reproductivos, desarrollo embrionario y larval. Spheniscus, 9, 25-39.

Faugeron, S., Valero, M., Destombe, C., Martínez, E.A. & Correa, J.A. (2001) Hierarchical spatial structure and discriminant analysis of genetic diversity in the red alga Mazzaella laminarioides (Gigartinales, Rhodophyta). Journal of Phycology, 37, 705-716.

Flores-Molina, M.R., Thomas, D., Lovazzano, C., Núñez, A., Zapata, J., Kumar, M., Correa, J.A. & Contreras-Porcia, L. (2014) Desiccation stress in intertidal seaweeds: Effects on morphology, antioxidant responses and photosynthetic performance. Aquatic Botany, 113, 90-99.

Gomes, P.B., Schama, R. & Solé-Cava, A.M. (2012) Molecular and morphological evidence that Phymactis papillosa from Argentina is, in fact, a new species of the genus Bunodosoma (Cnidaria: Actiniidae). Journal of the Marine Biological Association of the United Kingdom, 92, 895-910.

Gómez, I. & Huovinen, P. (2011) Morpho-functional patterns and zonation of South Chilean seaweeds: the importance of photosynthetic and bio-optical traits. Marine Ecology Progress Series, 422, 77-91.

González, A.V., Borras-Chavez, R., Beltrán, J., Flores, V., Vásquez, J.A. & Santelices, B. (2014) Morphological, ultrastructural, and genetic characterization of coalescence in the intertidal and shallow subtidal kelps Lessonia spicata and L. berteroana (Laminariales, Heterokontophyta). Journal of Applied Phycology, 26, 1107-1113.

González, A., Beltrán, J., Hiriart-Bertrand, L., Flores, V., de Reviers, B., Correa, J.A. & Santelices, B. (2012) Identification of cryptic species in the Lessonia nigrescens complex (Phaeophyceae, Laminariales). Journal of Phycology, 48, 1153-65.

González, E.R., Haye, P.A., Balanda, M.J. & Thiel, M. (2008) Lista sistemática de especies de peracáridos de Chile (Crustacea, Eumalacostraca). Gayana (Concepción), 72, 157-177.

160

Häussermann, V. (2004) Identification and taxonomy of soft-bodied hexacorals exemplified by Chilean sea anemones; including guidelines for sampling, preservation and examination. Journal of the Marine Biological Association of the United Kingdom, 84, 931-936.

Häussermann, V. (2006) Biodiversity of Chilean sea anemones (Cnidaria: Anthozoa): distribution patterns and zoogeographic implications, including new records for the fjord region. Investigaciones Marinas. Valparaiso, 34, 23-35.

Haye, P.A., Varela, A.I. & Thiel, M. (2012) Genetic signatures of rafting dispersal in algal-dwelling brooders Limnoria spp. (Isopoda) along the SE Pacific (Chile). Marine Ecology Progress Series, 455, 111-122.

Haye, P.A., Segovia, N.I., Muñoz-Herrera, N.C., Gálvez, F.E., Martínez, A., Meynard, A., Pardo-Gandarillas, M.C., Poulin, E. & Faugeron, S. (2014) Phylogeographic structure in benthic marine invertebrates of the southeast Pacific coast of Chile with differing dispersal potential. PLoS One, 9, e88613.

Hinojosa, I., González, E., Ugalde, P., Valdivia, N., Macaya, E. & Thiel, M. (2007) Distribución y abundancia de macroalgas flotando a la deriva y su fauna peracárida asociada a en los canales de la XI Región, Chile. Revista Ciencia y Tecnología del Mar, 30, 37-50.

Hinojosa, I.A., Pizarro, M., Ramos, M. & Thiel, M. (2010) Spatial and temporal distribution of floating kelp in the channels and fjords of southern Chile. Estuarine Coastal and Shelf Science, 87, 367-377.

Hoffmann, A. & Santelices, B. (1997) Flora marina de Chile Central. Ediciones Universidad Católica de Chile, Santiago de Chile, Chile.

Huovinen, P. & Gómez, I. (2011) Spectral attenuation of solar radiation in Patagonian fjord and coastal waters and implications for algal photobiology. Continental Shelf Research, 31, 254-259.

161

Kelaher, B.P., Castilla, J.C. & Prado, L. (2007) Is there redundancy in bioengineering for molluscan assemblages on the rocky shores of central Chile? Revista Chilena de Historia Natural, 80, 173-186.

Koch, K., Thiel, M., Tellier, F., Hagen, W., Graeve, M., Tala, F., Laeseke, P. & Bischof, K. (2015) Species separation within the Lessonia nigrescens complex (Phaeophyceae, Laminariales) is mirrored by ecophysiological traits. Botanica Marina, 58, 81- 92.

Kolbin, K. & Kulikova, V. (2011) Reproduction and larval development of the limpet Lottia persona (Rathke, 1833) (Gastropoda: Lottiidae). Russian Journal of Marine Biology, 37, 239-242.

Laakkonen, H.M., Strelkov, P. & Väinölä, R. (2015) Molecular lineage diversity and inter‐oceanic biogeographical history in Hiatella (Mollusca, Bivalvia). Zoologica Scripta, 44, 383-402.

Lagos, N.A., Tapia, F.J., Navarrete, S.A. & Castilla, J.C. (2007) Spatial synchrony in the recruitment of intertidal invertebrates along the coast of central Chile. Marine Ecology Progress Series, 350, 29-39.

León, J.F., Osorio, C.A., Quispe, V.C. & Quispe, L.S. (1995) Estudio de las algas rhodophyta del litoral de Tacna (Boca del Río, Vila- Vila y Cánepa). Ciencia & Desarrollo, 1, 5-11.

Leonardi, P.I., Miravalles, A.B., Faugeron, S., Flores, V., Beltran, J. & Correa, J.A. (2006) Diversity, phenomenology and epidemiology of epiphytism in farmed Gracilaria chilensis (Rhodophyta) in northern Chile. European Journal of Phycology, 41, 247-257.

López, B.A., Tellier, F., Retamal-Alarcón, J.C., Pérez-Araneda, K., Fierro, A.O., Macaya, E.C., Tala, F. & Thiel, M. (2017) Phylogeography of two intertidal seaweeds, Gelidium lingulatum and G. rex (Rhodophyta: Gelidiales), along the South East Pacific: patterns explained by rafting dispersal? Marine Biology, 164, 188.

162

López-Cristoffanini, C., Tellier, F., Otaíza, R., Correa, J.A. & Contreras-Porcia, L. (2013) Tolerance to air exposure: a feature driving the latitudinal distribution of two sibling kelp species. Botanica Marina, 56, 431-440.

Macaya, E.C. & Zuccarello, G.C. (2010) Genetic structure of the giant kelp Macrocystis pyrifera along the southeastern Pacific. Marine Ecology Progress Series, 420, 103-112.

Macaya, E.C., Pacheco, S., Cáceres, A. & Musleh, S. (2013) Range extension of the non-indigenous alga Mastocarpus sp. along the Southeastern Pacific coast. Revista de Biología Marina y Oceanografía, 48, 661-665.

Mansilla, A., Ávila, M., Ramírez, M.E., Rodríguez, J.P., Rosenfeld, S., Ojeda, J. & Marambio, J. (2013) Macroalgas marinas bentónicas del submareal somero de la ecorregión subantártica de Magallanes, Chile. Anales del Instituto de la Patagonia, 41, 51-64.

Medina, M., Andrade, S., Faugeron, S., Lagos, N., Mella, D. & Correa, J.A. (2005) Biodiversity of rocky intertidal benthic communities associated with copper mine tailing discharges in northern Chile. Marine Pollution Bulletin, 50, 396-409.

Menzies, R.J. (1962) The zoogeography, ecology and systematics of the Chilean marine isopods. Lunds Universitets Årsskrift NF, 57, 1-162.

Miranda, L. & Thiel, M. (2008) Active and passive migration in boring isopods Limnoria spp. (Crustacea, Peracarida) from kelp holdfasts. Journal of Sea Research, 60, 176-183.

Montecinos, A., Broitman, B.R., Faugeron, S., Haye, P.A., Tellier, F. & Guillemin, M.L. (2012) Species replacement along a linear coastal habitat: phylogeography and speciation in the red alga Mazzaella laminarioides along the south east Pacific. BMC Evolutionary Biology, 12, 97.

163

Navarrete, A.H., Camus, P.A. & Opazo, L.F. (2008) Variación ambiental y patrones dietarios del erizo negro Tetrapygus niger en costas intermareales rocosas del norte de Chile. Revista Chilena de Historia Natural, 81, 305-319.

Opazo, L.F. & Otaíza, R.D. (2007) Vertical distribution of spores of blade-forming Sarcothalia crispata (Gigartinaceae) and crustose corallines (Corallinaceae) in the water column. Botanica Marina, 50, 97-105.

Oróstica, M.H., Otaíza, R.D. & Neill, P.E. (2012) Blades and papillae as likely dispersing propagules in Chilean populations of Mastocarpus sp. (Rhodophyta, Gigartinales). Revista de Biología Marina y Oceanografía, 47, 109-119.

Ortega, K.J., Sáez, C.A. & Macaya, E.C. (2014) Changes in invertebrate assemblages inhabiting Lessonia spicata (Phaeophyceae) holdfasts after the 2010 earthquake-mediated coastal uplift in Chile. Revista de Biología Marina y Oceanografía, 49, 129-134.

Osorio, C. & Reid, D.G. (2004) Moluscos marinos intermareales y submareales entre la Boca del Guafo y el Estero Elefantes, sur de Chile. Investigaciones Marinas, 32, 71-89.

Osorio, C., Reid, D.G. & Ramajo, L. (2005) Moluscos en los canales del Sur de Chile entre boca del Guafo y estero Elefantes (CIMAR 7 Fiordos). Ciencia y Tecnología del Mar, 28, 91-98.

Pacheco, A. & Laudien, J. (2008) Dendropoma mejillonensis sp. nov. a new species of Vermetid (Mollusca: Caenogastropoda,) from Northern Chile. The Veliger, 50, 219-224.

Pacheco, A.S., Laudien, J., Thiel, M., Oliva, M. & Heilmayer, O. (2011) Succession and seasonal onset of colonization in subtidal hard-bottom communities off northern Chile. Marine Ecology, 32, 75-87.

Palma, M., González, F., Romo, H., Ruiz, E. & Fuentealba, C. (2007) Efectos de la contaminación sobre la diversidad genética de Mazzaella laminarioides (Bory) Fredericq (Gigartinales, Rhodophyta) en bahías. Gayana Botánica, 64, 24-32.

164

Parada, G.M., Tellier, F. & Martínez, E.A. (2016) Spore dispersal in the intertidal kelp Lessonia spicata: macrochallenges for the harvested Lessonia species complex at microscales of space and time. Botanica Marina, 59, 283-289.

Paternoster, I.K. & Elias, I. (1980) Redescripción de Limnoria (Phycolimnoria) chilensis Menzies, 1962. Neotropica, 26, 35-41.

Pérez, M., Guínez, R., Llavona, A., Toro, J.E., Astorga, M. & Presa, P. (2008) Development of microsatellite markers for the ecosystem bioengineer mussel Perumytilus purpuratus and cross-priming testing in six Mytilinae genera. Molecular Ecology Resources, 8, 449-451.

Poore, A.G., Gutow, L., Pantoja, J.F., Tala, F., Jofré, D. & Thiel, M. (2014) Major consequences of minor damage: impacts of small grazers on fast-growing kelps. Oecologia, 174, 789-801.

Prado, L. & Castilla, A.C. (2006) The bioengineer Perumytilus purpuratus (Mollusca: Bivalvia) in central Chile: biodiversity, habitat structural complexity and environmental heterogeneity. Journal of the Marine Biological Association of the United Kingdom, 86, 417-421.

Ramírez, M. & Santelices, B. (1981) Análisis biogeográfico de la flora algológica de Antofagasta (norte de Chile). Boletín del Museo Nacional de Historia Natural, 38, 5-20.

Ramírez, M.E., García-Huidobro, M.R. & Orellana, N. (2008) Flora marina bentónica de Caleta Quintay (Región de Valparaíso), litoral central de Chile. Boletín del Museo Nacional de Historia Natural (Chile), 57, 9-20.

Ramorino, L. & Campos, B. (1979) Desarrollo larval y postlarval de Perumytilus purpuratus (Lamarck, 1819) Bivalvia: Mytilidae. Anales del Museo de Historia Natural, Valparaíso, 12, 207-218.

Ríos, C., Arntz, W.E., Gerdes, D., Mutschke, E. & Montiel, A. (2007) Spatial and temporal variability of the benthic assemblages associated to the holdfasts of the kelp Macrocystis pyrifera in the Straits of Magellan, Chile. Polar Biology, 31, 89-100.

165

Ritter, A., Cabioch, L., Brillet-Guéguen, L., Corre, E., Cosse, A., Dartevelle, L., Duruflé, H., Fasshauer, C., Goulitquer, S., Thomas, F., Correa, J.A., Potin, P., Faugeron, S. & Leblanc, C. (2017) Herbivore-induced chemical and molecular responses of the kelps Laminaria digitata and Lessonia spicata. PLoS One, 12, e0173315.

Rodríguez, D. & Santelices, B. (1995) Temporal variation in the apical architecture of two species of Gelidium from central Chile. Taxonomy of economic seaweeds with reference to some Pacific species (ed. by I.A. Abbott), pp. 147-161. California Sea Grant College Program, La Jolla, California.

Rozbaczylo, N. (1985) Los anélidos poliquetos de Chile: Índice sinonímico y distribución geográfica de especies. Facultad de Ciencias Biológicas, Pontificia Universidad Católica de Chile.

Rozbaczylo, N., Moreno, R. & Díaz-Díaz, O. (2017) Poliquetos bentónicos en Chile. Poliquetos de Sudamérica (ed. by O. Díaz-Díaz, D. Bone, C.T. Rodríguez and V.H. Delgado-Blas), pp. 51-70. Instituto Oceanográfico de Venezuela. Cumaná, Venezuela.

Santelices, B. (1999) Patterns of carposporangial production among species of Gelidium (Gelidiales, Rhodophyta). Taxonomy of economic seaweeds with reference to some Pacific species (ed. by I.A. Abbott), pp. 55-69. California Sea Grant College System, La Jolla, California.

Santelices, B. & Abbott, I.A. (1985) Gelidium rex sp. nov. (Gelidiales. Rhodophyta) from central Chile. Taxonomy of economic seaweeds with reference to some Pacific and Caribbean species (ed. by I.A. Abbott and J. Norris), pp. 33-36. California Sea Grant College Program, La Jolla, California, USA.

Santelices, B. & Stewart, J.G. (1985) Pacific species of Gelidium Lamouroux and other Gelidiales (Rhodophyta), with keys and descriptions to the common or economically important species. Taxonomy of economic seaweeds with reference to some

166

Pacific and Caribbean species. (ed. by I.A. Abbott and J.N. Norris), pp. 17-31. California Sea Grant College Program, La Jolla, California, USA.

Santelices, B. & Varela, D. (1994) Abiotic control of reattachment in Gelidium chilense (Montagne) Santelices and Montalva (Gelidiales, Rhodophyta). Journal of Experimental Marine Biology and Ecology, 177, 145-155.

Santelices, B., Ramírez, M.E. & Abbott, I.A. (1989) A new species and new records of marine-algae from Chile. British Phycological Journal, 24, 73-82.

Santelices, B., Aedo, D. & Hoffmann, A. (2002) Banks of microscopic forms and survival to darkness of propagules and microscopic stages of macroalgae. Revista Chilena de Historia Natural, 75, 547-555.

Sepúlveda, R.D., Moreno, R.A. & Carrasco, F.D. (2003) Diversidad de macroinvertebrados asociados a arrecifes de Phragmatopoma moerchi Kinberg, 1867 (Polychaeta: Sabellariidae) en el intermareal rocoso de Cocholgüe, Chile. Gayana (Concepción), 67, 45-54.

Soler, N.S., Macaya, E.C., Lluch, J.R. & Garreta, A.G. (2016) Nuclear DNA content in Gelidium chilense (Gelidiales, Rhodophyta) from the Chilean coast. Revista de Biología Marina y Oceanografía, 51, 113-122.

Soot-Ryen, T. (1955) A report on the family Mytilidae (Pelecypoda). Allan Hancock Pacific Expeditions, 20, 1-154.

Soot-Ryen, T. (1959) Pelecypoda. Reports of the Lund University Chile Expedition 1948-49 N°35. Lunds Universitets Årsskrift. Ny Foljd. Avdelningen 2, pp. 1-86. C.W.K. Gleerup.

Soto, E.H., Báez, P., Ramírez, M.E., Letelier, S., Naretto, J. & Rebolledo, A. (2015) Biotopos marinos intermareales y sublitorales someros entre Canal Trinidad y Canal Smyth, XII Región, Chile. Revista Ciencia y Tecnología del Mar, 36, 91-103.

167

Tellier, F., Meynard, A.P., Correa, J.A., Faugeron, S. & Valero, M. (2009) Phylogeographic analyses of the 30°S south-east Pacific biogeographic transition zone establish the occurrence of a sharp genetic discontinuity in the kelp Lessonia nigrescens: vicariance or parapatry? Molecular Phylogenetics and Evolution, 53, 679-93.

Tellier, F., Tapia, J., Faugeron, S., Destombe, C. & Valero, M. (2011a) The Lessonia nigrescens species complex (Laminariales, Phaeophyceae) shows strict parapatry and complete reproductive isolation in a secondary contact zone. Journal of Phycology, 47, 894-903.

Tellier, F., Faugeron, S. & Valero, M. (2011b) Possible role of a mitochondrial genome derangement in maintaining the spatial segregation of two cryptic species of the Lessonia nigrescens species complex. Cahiers de Biologie Marine, 52, 371-383.

Tellier, F., Vega, J.M.A., Broitman, B.R., Vásquez, J.A., Valero, M. & Faugeron, S. (2011c) The importance of having two species instead of one in kelp management: the Lessonia nigrescens species complex. Cahiers de Biologie Marine, 52, 455-465.

Thiel, M. (2003) Reproductive biology of Limnoria chilensis: another boring peracarid species with extended parental care. Journal of Natural History, 37, 1713-1726.

Thiel, M. & Vásquez, J.A. (2000) Are kelp holdfasts islands on the ocean floor? - indication for temporarily closed aggregations of peracarid crustaceans. Hydrobiologia, 440, 45-54.

Thiel, M. & Ullrich, N. (2002) Hard rock versus soft bottom: the fauna associated with intertidal mussel beds on hard bottoms along the coast of Chile, and considerations on the functional role of mussel beds. Helgoland Marine Research, 56, 21-30.

Thomas, D.T. & Freshwater, D.W. (2001) Studies of Costa Rican Gelidiales (Rhodophyta): four Caribbean taxa including Pterocladiella beachii sp. nov. Phycologia, 40, 340-350.

168

Tokeshi, M. & Romero, L. (1995) Filling a gap - dynamics of space occupancy on a mussel-dominated subtropical rocky shore. Marine Ecology Progress Series, 119, 167-176.

Trovant, B., Orensanz, J.M., Ruzzante, D.E., Stotz, W. & Basso, N.G. (2015) Scorched mussels (Bivalvia: Mytilidae: Brachidontinae) from the temperate coasts of South America: phylogenetic relationships, trans-Pacific connections and the footprints of Quaternary glaciations. Molecular Phylogenetics and Evolution, 82 60-74.

Uribe, R.A., Ortiz, M., Macaya, E.C. & Pacheco, A.S. (2015) Successional patterns of hard-bottom macrobenthic communities at kelp bed (Lessonia trabeculata) and barren ground sublittoral systems. Journal of Experimental Marine Biology and Ecology, 472, 180-188.

Valdivia, N., Heidemann, A., Thiel, M., Molis, M. & Wahl, M. (2005) Effects of disturbance on the diversity of hard-bottom macrobenthic communities on the coast of Chile. Marine Ecology Progress Series, 299, 45-54.

Vásquez, J., Fonck, E. & Vega, M. (2001) Comunidades submareales rocosas dominadas por macroalgas en el norte de Chile: diversidad, abundancia y variabilidad temporal. Sustentabilidad de la biodiversidad (ed. by K. Alveal and T. Antezana), pp. 351-366. Universidad de Concepción, Concepción, Chile.

Vásquez, J.A. & Vega, J.A.M. (2004) Ecosistemas marinos costeros del Parque Nacional Bosque Fray Jorge. Historia Natural del Parque Nacional Bosque Fray Jorge (ed. by F. Squeo, J. Gutiérrez and I. Hernández), pp. 235-252. Ediciones Universidad de La Serena, La Serena, Chile.

Velásquez, C., Jaramillo, E., Camus, P.A., Manzano, M. & Sánchez, R. (2016) Biota del intermareal rocoso expuesto de la Isla Grande de Chiloé, Archipiélago de Chiloé, Chile: Patrones de diversidad e implicancias ecológicas y biogeográficas. Revista de Biología Marina y Oceanografía, 51, 33-50.

169

Villaseñor-Parada, C., Pauchard, A. & Macaya, E. C. (2014) Expansión del área de distribución de la macroalga introducida Schottera nicaeensis (Rhodophyta: Gigartinales) en la costa chilena: ¿evidencia de una invasión? Boletín de la Red Latinoamericana para el Estudio de Especies Invasoras, 4, 19-27.

Villaseñor-Parada, C., Pauchard, A. & Macaya, E.C. (2017) Ecología de invasiones marinas en Chile continental: ¿Qué sabemos y que nos falta por saber? Revista de Biología Marina y Oceanografía, 52, 1-17.

Villaseñor-Parada, C., Pauchard, A., Ramírez, M.E. & Macaya, E.C. (2018) Macroalgas exóticas en la costa de Chile: patrones espaciales y temporales en el proceso de invasión. Latin American Journal of Aquatic Research, 46, 147-165.

Villegas, M.J., Stotz, W. & Laudien, J. (2005) First record of an epibiosis between the sand crab Emerita analoga (Stimpson, 1857) (Decapoda: Hippidae) and the mussel Semimytilus algosus (Gould, 1850) (Bivalvia, Mytilidae) in southern Peru. Helgoland Marine Research, 60, 25.

Werlinger, C., Mansilla, A., Villarroel, A. & Palacios, M. (2008) Effects of photon flux density and agricultural fertilizers on the development of Sarcothalia crispata tetraspores (Rhodophyta, Gigartinales) from the Strait of Magellan, Chile. Journal of Applied Phycology, 20, 757-765.

Wieters, E.A. (2005) Upwelling control of positive interactions over mesoscales: a new link between bottom-up and top-down processes on rocky shores. Marine Ecology Progress Series, 301, 43-54.

Wieters, E.A., Medrano, A. & Quiroga, G. (2013) Spatial variation in photosynthetic recovery of intertidal turf algae from acute UVB and temperature stress associated with low tides along the central coast of Chile. Journal of Experimental Marine Biology and Ecology, 449, 340-348.

170

Appendix S3.4: Expanded Results

A total of 89 epibiont taxa were recorded on holdfasts of stranded D. antarctica. Only 7.3% of the stranded D. antarctica had no epibiont species. Of the species found, 86.5% were sessile and 13.5% were mobile epibionts. Also, 71.9% were seaweeds and 28.1% were invertebrates. Within the seaweed group, 48 were Rhodophyta, 10 Phaeophyceae and 6 Chlorophyta, and several unidentified crustose calcareous algae (Table S3.4.1). The species of algae most frequently found were articulated coralline algae (26.1%), Lessonia spicata (11.1%), Gelidium lingulatum (5.9%), Gelidium sp. (2.3%) and G. rex (1.9%). With respect to invertebrates, the most representative taxonomic groups were Cnidaria, Bryozoa, Mollusca, Annelida and Crustacea (Table S3.4.2). The most common invertebrate taxa were acorn barnacles (62.3%), the limpet Scurria scurra (21.3%), the isopod Limnoria chilensis (23.1%), the bivalve Semimytilus algosus (13.7%), and the polychaete Phragmatopoma moerchi (13.7%) (Table S3.4.2). Of the total species found, 29 species were reported from a single stranded specimen throughout the study (Tables S3.4.1 and S3.4.2).

171

Table S3.4.1. Seaweed species found on holdfasts of stranded bull kelp Durvillaea antarctica on sandy beaches from the continental coast of Chile (28°S - 42°S) during winters and summers 2014/15-2015/16. The percentage is based on a total of 5,219 stranded individuals of D. antarctica. Only species with more than one record were included in statistical analyses. (*) Species that were incorporated in co-occurrence analysis. () Species that were incorporated in geographical range analysis. The unidentified species corresponded to individuals that due to their small size, absence of reproductive structures and/or tissue deterioration, it was not possible to reach a more specific taxonomic level.

Presence Species Number of stranded Percentage individuals

Chlorophyta Chaetomorpha firma* # 20 0.4 Cladophora sp. 1 < 0.1 Codium sp. 5 < 0.1 Rhizoclonium ambiguum 1 < 0.1 Ulva sp.* 60 1.2 Unidentified species 5 < 0.1

Phaeophyceae Adenocystis utricularis 1 < 0.1 Colpomenia sp. 1 < 0.1 Desmarestia ligulata # 4 < 0.1 Ectocarpus sp. 1 < 0.1 Halopteris sp. 1 < 0.1

172

Lessonia spicata* # 579 11.1 Macrocystis pyrifera # 17 0.3 Petalonia binghamiae # 4 < 0.1 Petalonia fascia 2 < 0.1 Unidentified species 1 < 0.1

Rhodophyta Ahnfeltiopsis durvillaei 1 < 0.1 Antithamnion densum* # 27 0.5 Antithamnionella ternifolia* # 59 1.1 Articulated coralline algae* 1361 26.1 Asparagopsis armata 1 < 0.1 Asterfilopsis disciplinalis* # 46 0.8 Ballia callitricha # 4 < 0.1 Bossiella orbigniana 1 < 0.1 Bostrychia intricata 1 < 0.1 Branchioglossum bipinnatifidum 1 < 0.1 Callithamnion sp. 1 < 0.1 Callophyllis concepcionensis 1 < 0.1 Camontagnea oxyclada # 5 < 0.1 Centroceras clavulatum 1 < 0.1 Ceramiales unidentified 12 0.2 Ceramium virgatum # 5 < 0.1 Chondria secundata # 19 0.3 Chondrus canaliculatus 1 < 0.1 Corallinaceae unidentified 1 < 0.1 Crustose calcareous algae* 2886 55.3 Florideophyceae unidentified 2 < 0.1 Gelidiales unidentified 4 < 0.1 Gelidium chilense* # 55 1.1 Gelidium lingulatum* # 313 5.9 Gelidium rex* # 100 1.9

173

Gelidium sp.* 118 2.3 Gigartinales unidentified 18 0.2 Griffithsia chilensis # 5 < 0.1 Halymeniales unidentified 4 < 0.1 Mariaramirezia osornoensis # 4 < 0.1 Mastocarpus latissimus # 4 < 0.1 Mazzaella laminarioides* # 30 0.5 Mazzaella membranacea 2 < 0.1 Montemaria horridula 1 < 0.1 Nothogenia sp. 1 < 0.1 Paraglossum crassinervium # 11 0.2 Plocamium cartilagineum 2 < 0.1 Polysiphonia pacifica 1 < 0.1 Polysiphonia sp.* 22 0.4 Prionitis sp. 1 < 0.1 Pyropia (Porphyra) sp.* 56 1.1 Rhodoglossum sp. 1 < 0.1 Rhodymenia skottsbergii * # 21 0.4 Sarcothalia crispata* # 47 0.9 Schimmelmannia plumosa* # 21 0.4 Schottera nicaeensis* # 40 0.7 Symphyocladia dendroidea # 8 0.2 Unidentified species 160 2.2

174

Table S3.4.2. Invertebrate species found on holdfasts of stranded bull kelp Durvillaea antarctica on sandy beaches from the continental coast of Chile (28°S - 42°S) during winters and summers 2014/15-2015/16. The percentage is based on a total of 5,219 stranded individuals of D. antarctica. Only species with more than one record were included in statistical analyses. (*) Species that were incorporated in co-occurrence analysis. (#) Species that were incorporated in geographical ranges analysis. (+) Mobile taxa. The unidentified species corresponded to individuals that due to their small size, absence of reproductive structures and/or tissue deterioration, it was not possible to reach a more specific taxonomic level.

Presence Species Number of stranded Percentage individuals

Porifera 2 < 0.1

Cnidaria Hydrozoa* 336 6.4 Phymactis papillosa # 17 0.2

Nemertea + 11 0.2

Bryozoa* 125 2.3

Mollusca Acanthina monodon + 1 < 0.1 Aulacomya ater 2 < 0.1 Brachidontes granulata # 4 < 0.1

175

Crepidula sp. 1 < 0.1 Dendropoma mejillonensis* # 20 0.4 Fissurella sp. + 1 < 0.1 Hiatella solida # 9 0.2 Perumytilus purpuratus* # 137 2.6 Polyplacophora + 1 < 0.1 Scurria scurra* # + 1573 21.3 Semimytilus algosus* # 713 13.7

Annelida Nereididae + 3 < 0.1 Phragmatopoma moerchi* # 717 13.7 Serpulidae 1 < 0.1 Spirorbidae 1 < 0.1

Crustacea Acorn barnacles* 3334 63.9 Limnoria chilensis* # + 1207 23.1 Megalopa larvae 4 < 0.1 Parawaldeckia kidderi # + 4 < 0.1 Sphaeromatidae + 2 < 0.1

Pycnogonida + 1 < 0.1

Tunicata Pyura sp. 3 < 0.1

176

Table S3.4.3. Summary of three-way ANCOVA for taxonomic richness of epibionts attached (total, sessile and mobile) on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to biogeographic district (CCD, SED, MED and MD), year (2014 and 2015), season (winter and summer) and holdfast wet weight as covariate. Significant values (P < 0.05) are shown in bold.

Variable Source of variation df F P-value Taxonomic District (D) 3;5187 1.89 0.127 Richness (Total) Year (Y) 1;5187 1.29 0.267 Season (S) 1;5187 18.61 < 0.001 Weight (W) 1;5187 75.78 < 0.001 D x Y 3;5187 0.97 0.403 D x S 3;5187 4.68 < 0.01 Y x S 1;5187 2.76 0.096 D x W 3;5187 4.71 < 0.01 Y x W 1;5187 7.35 < 0.01 S x W 1;5187 0.98 0.320 D x Y x S 3;5187 1.67 0.381 D x Y x W 3;5187 0.83 0.473 D x S x W 3;5187 1.99 0.111 Y x S x W 1;5187 2.04 0.152 D x Y x S x W 3;5187 0.80 0.492 Taxonomic District (D) 3;5187 9.90 < 0.001 Richness Year (Y) 1;5187 2.75 0.097 (Sessile) Season (S) 1;5187 17.79 < 0.001 Weight (W) 1;5187 4.65 < 0.050 D x Y 3;5187 0.90 0.440

177

D x S 3;5187 1.53 0.202 Y x S 1;5187 1.12 0.288 D x W 3;5187 1.61 0.184 Y x W 1;5187 0.33 0.564 S x W 1;5187 0.01 0.962 D x Y x S 3;5187 0.92 0.432 D x Y x W 3;5187 0.63 0.592 D x S x W 3;5187 0.10 0.957 Y x S x W 1;5187 2.67 0.101 D x Y x S x W 3;5187 0.87 0.451 Taxonomic District (D) 3;5187 1.41 0.061 Richness Year (Y) 1;5187 0.61 0.443 (Mobile) Season (S) 1;5187 0.11 0.731 Weight (W) 1;5187 11.43 < 0.001 D x Y 3;5187 1.37 0.067 D x S 3;5187 1.88 0.159 Y x S 1;5187 1.68 0.194 D x W 3;5187 0.04 0.987 Y x W 1;5187 0.46 0.498 S x W 1;5187 1.78 0.181 D x Y x S 3;5187 0.86 0.457 D x Y x W 3;5187 1.48 0.216 D x S x W 3;5187 1.66 0.172 Y x S x W 1;5187 0.18 0.669 D x Y x S x W 3;5187 0.86 0.456

178

Table S3.4.4. Summary of one-way ANCOVA for taxonomic richness of epibionts attached (total, sessile and mobile) on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to floating time (short, intermediate and long) and holdfast wet weight as covariate. Significant values (P < 0.05) are shown in bold.

Variable Source of variation df F P-value Taxonomic Floating Time (FT) 2;5213 5.17 < 0.010 Richness (Total) Weight (W) 1;5213 17.62 < 0.001 FT x W 2;5213 0.90 0.404 Taxonomic Floating Time (FT) 2;5213 3.37 < 0.050 Richness Weight (W) 1;5213 3.00 0.083 (Sessile) FT x W 2;5213 0.54 0.580 Taxonomic Floating Time (FT) 2;5213 1.18 0.307 Richness Weight (W) 1;5213 5.16 < 0.050 (Mobile) FT x W 2;5213 0.14 0.861

179

Fig. S3.4.1. Average (mean ± SD) taxonomic richness of epibionts attached on stranded individuals of Durvillaea antarctica on beaches from the continental coast of Chile (28°S-42°S), according to years (2014 and 2015) and seasons (winter and summer). (A) Taxonomic richness at plant-level and accumulated per season and year (Chao 2 index). (B) Average taxonomic richness of sessile and mobile epibionts. Letters (a-b-c-d) above the columns indicate significant differences between season and year (P < 0.05). Letters in italics correspond to the results of accumulated taxonomic richness (in A) or mobile species (in B). Number of plants (at plant-level) and beaches (total accumulated) from each year and season are listed at the bottom of each column.

180

Fig. S3.4.2. Relationship between taxonomic richness of associated epibionts and holdfast wet weight (kg) on stranded individuals of Durvillaea antarctica on beaches along the continental coast of Chile (28°S-42°S). Summary of ANOVA of simple linear regression between both variables is indicated.

181

Table S3.4.5. Summary of GLMs with a negative binomial distribution of the frequencies of raft-associated species of stranded D. antarctica found outside their known benthic ranges according to the presence/absence of two species with high positive co- occurrences (Gelidium lingulatum and Semimytilus algosus) and two species with high negative co-occurrences (Limnoria chilensis and Scurria scurra) for SE and NE species from the continental coast of Chile (28°S-42°S). Significant values (P< 0.05) are shown in bold.

Species Expansion df Pseudo-R2 P-value range Positive co-occurrences Gelidium lingulatum SE Species 1;12 0.345 0.042 NE Species 1;4 0.356 0.039 Semimytilus algosus SE Species 1;12 0.279 0.048 NE Species 1;4 0.382 0.024 Negative co-occurrences Limnoria chilensis SE Species 1;12 0.301 0.031 NE Species 1;4 0.153 0.059 Scurria scurra SE Species 1;12 0.316 0.023 NE Species 1;4 0.189 0.066

182

Fig. S3.4.3. Number of stranded rafts of Durvillaea antarctica per km surveyed shoreline (white bars) and their epibiont species attached on beaches along the latitudinal gradient (bins of 1°) of the Chilean coast (28°S-42°S), according to literature (light gray bars) and rafting ranges (dark gray bars). Epibiont species are shown in three categories, RO: rafting occurs within their literature ranges, SE: rafting surpasses the southern edges of their literature ranges, NE: rafting surpasses the northern edges of their literature ranges.

183

Supplementary Material 4

Marine Biology

SUPPORTING INFORMATION

Phylogeography of two intertidal seaweeds, Gelidium lingulatum and G. rex (Rhodophyta: Gelidiales), along the South East Pacific – patterns explained by rafting dispersal?

Boris A. López, Florence Tellier, Juan C. Retamal-Alarcón, Karla Pérez-Araneda, Ariel O. Fierro; Erasmo C. Macaya, Fadia Tala & Martin Thiel

184

Table S4.1. Sampling sites from the northern part of the study area, where no individuals of Gelidium lingulatum and G. rex (*) were found. The code used to identify each sampling site and coordinates (latitude and longitude) are indicated.

Coordinates Sampling site Code Latitude Longitude Huayquique* HUQ 20°16'S 70°07'W Playa Blanca* PBLA 20°19'S 70°08'W Los Verdes* LVER 20°25'S 70°09'W Caleta CERR 23°26'S 70°35'W Errázuriz* U. Antofagasta* UANT 23°42'S 70°25'W La Poza* LPOZ 23°43'S 70°26'W Los Burros BURR 28°54'S 71°31'W El Apolillado APOL 29°10'S 71°29'W

185

Table S4.2. Variable sites and sequence frequencies for rbcL of Gelidium lingulatum and G. rex.

rbcL nucleotide Haplotype N Sampling site 22 406 Gelidium lingulatum GL701 2 QICO, BAM C T GL702 4 LBO, PCH, CUR, QICO C C GL703 4 CUR, COB (x2), QICO T C

Gelidium rex GR701 12 BURR (x3), APOL (x3), - - PTTR (x2), LBO (x2), PCH (x2)

186

Table S4.3. Spatial analysis of molecular variance (SAMOVA) for 2 to 10 groups of locations of Gelidium lingulatum, and for 2 to 5 groups of locations of G. rex. The codes represent the sampling sites. Locations with N ˂ 14 were excluded from analysis. G1, G2 and Gn corresponding to the groups of locations formed after analysis, according to the number of groups required. In bold, the best fit of groups is shown.

Number of groups Spatial structure tested F-statistic P-value Gelidium lingulatum 2 G1:(PAM;MONT;CUR;COB;LOT;QICO;CRNC;PUCA;HUAR;CUCA) FCT = 0.564 P < 0.001 G2:(LBO;PCH;CON;MBRA;SBA) FSC = 0.147 P < 0.001 FST = 0.289 P < 0.001 3 G1:(PAM;CUR;COB;LOT;QICO;CRNC;PUCA;HUAR;CUCA) FCT = 0.594 P < 0.001 G2:(MONT) FSC = 0.095 P < 0.001 G3:(LBO;PCH;CON;MBRA;SBA) FST = 0.312 P < 0.001 4 G1:(PAM;LOT;QICO;CRNC;PUCA;CUCA) FCT = 0.620 P < 0.001 G2:(MONT) FSC = 0.034 P < 0.001 G3:(CUR;COB; HUAR) FST = 0.347 P < 0.001 G4:(LBO;PCH;CON;MBRA;SBA) 5 G1:(PAM;LOT;QICO;CRNC;PUCA;CUCA) FCT = 0.623 P < 0.001 G2:(MONT) FSC = 0.029 P < 0.001 G3:(CUR;HUAR) FST = 0.347 P < 0.001 G4:(COB) G5:(LBO;PCH;CON;MBRA;SBA) 6 G1:(PAM;LOT;CRNC;CUCA) FCT = 0.625 P < 0.001 G2:(MONT) FSC = 0.083 P = 0.108 G3:(CUR;HUAR) FST = 0.367 P < 0.001 G4:(COB) G5:(QICO;PUCA) G6:(LBO;PCH;CON;MBRA;SBA)

187

7 G1:(PAM;LOT;CRNC;CUCA) FCT = 0.624 P < 0.001 G2:(MONT) FSC = 0.072 P = 0.101 G3:(CUR) FST = 0.368 P < 0.001 G4:(HUAR) G5:(COB) G6:(QICO;PUCA) G7:(LBO;PCH;CON;MBRA;SBA) 8 G1:(PAM;LOT;CRNC;CUCA) FCT = 0.632 P < 0.001 G2:(MONT) FSC = -0.024 P = 0.319 G3:(CUR) FST = 0.371 P < 0.001 G4:(HUAR) G5:(COB) G6:(QICO) G7:(PUCA) G8:(LBO;PCH;CON;MBRA;SBA) 9 G1:(PAM;LOT;CRNC;CUCA) FCT = 0.627 P < 0.001 G2:(MONT) FSC = -0.066 P = 0.534 G3:(CUR) FST = 0.379 P < 0.001 G4:(HUAR) G5:(COB) G6:(QICO) G7:(PUCA) G8:(PCH;CON;MBRA;SBA) G9:(LBO)

Gelidium rex 2 G1:(BURR;APOL;SAUC;FUAD) FCT = 0.721 P < 0.001 G2:(CHLO;PAMA;QTAY;PTTR;PCH;BUCA) FSC = 0.138 P < 0.001 FST = 0.140 P < 0.001 3 G1:(BURR;APOL;SAUC) FCT = 0.844 P < 0.001 G2:(FUAD) FSC = 0.015 P < 0.001 G3:(CHLO;PAMA;QTAY;PTTR;PCH;BUCA) FST = 0.140 P < 0.001

188

4 G1:(BURR;APOL) FCT = 0.841 P < 0.001 G2:(SAUC) FSC = 0.007 P < 0.001 G3:(FUAD) FST = 0.151 P < 0.001 G4:(CHLO;PAMA;QTAY;PTTR;PCH;BUCA)

189

Table S4.4. Population pairwise φST values for mitochondrial marker COI in Gelidium lingulatum. The codes represent 15 sampling

sites (locations with N < 14 were excluded from analysis). Below: φST / Above: P-values. Significant φST values (P-values after Bonferroni correction) are shown in bold.

PAM MONT LBO PCH CON CUR COB LOT QICO CRNC PUCA HUAR MBRA CUCA SBA PAM 0 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.038 0.004 0.158 0.006 < 0.001 < 0.001 0.058 < 0.001 MONT 0.626 0 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 LBO 0.874 0.852 0 0.188 0.135 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.147 < 0.001 0.523 PCH 0.666 0.738 0.074 0 0.815 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.003 < 0.001 0.670 < 0.001 0.500 CON 0.694 0.754 0.050 -0.036 0 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 0.580 < 0.001 0.435 CUR 0.364 0.420 0.613 0.385 0.429 0 0.007 < 0.001 < 0.001 0.002 0.012 0.382 < 0.001 < 0.001 < 0.001 COB 0.778 0.503 0.969 0.796 0.802 0.256 0 < 0.001 < 0.001 < 0.001 < 0.001 0.017 < 0.001 < 0.001 < 0.001 LOT 0.119 0.709 0.954 0.775 0.785 0.457 0.950 0 0.002 1.000 < 0.001 < 0.001 < 0.001 1.000 < 0.001 QICO 0.202 0.597 0.698 0.483 0.528 0.269 0.624 0.284 0 0.026 0.054 < 0.001 < 0.001 0.018 < 0.001 CRNC 0.054 0.640 0.886 0.645 0.674 0.327 0.838 0.024 0.138 0 0.074 < 0.001 < 0.001 1.000 < 0.001 PUCA 0.209 0.573 0.539 0.298 0.350 0.181 0.575 0.296 0.089 0.128 0 0.006 < 0.001 0.005 < 0.001 HUAR 0.334 0.374 0.740 0.526 0.563 -0.002 0.122 0.447 0.290 0.320 0.233 0 < 0.001 < 0.001 < 0.001 MBRA 0.652 0.728 0.061 -0.031 -0.019 0.382 0.767 0.750 0.480 0.628 0.310 0.519 0 < 0.001 0.345 CUCA 0.090 0.688 0.945 0.741 0.756 0.427 0.946 0.002 0.243 0.009 0.252 0.418 0.717 0 < 0.001 SBA 0.784 0.805 0.018 -0.020 -0.021 0.526 0.889 0.869 0.609 0.777 0.436 0.661 0.001 0.849 0

190

Table S4.5. Population pairwise φST values for mitochondrial marker COI in Gelidium rex. The codes represent 10 sampling sites

(locations with N < 14 were excluded from analysis). Below φST /Above: P-values. Significant φST values (P-values after Bonferroni correction) are shown in bold.

BURR APOL SAUC FUAD CHLO PAMA QTAY PTTR PCH BUCA BURR 0 0.035 0.012 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 APOL 0.031 0 0.015 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 SAUC 0.099 0.134 0 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 FUAD 0.808 1.000 0.737 0 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 < 0.001 CHLO 0.827 1.000 0.758 1.000 0 0.999 0.999 0.241 0.999 0.017 PAMA 0.823 1.000 0.753 1.000 0.000 0 0.999 0.211 0.999 0.033 QTAY 0.813 1.000 0.742 1.000 0.000 0.000 0 0.257 0.999 0.028 PTTR 0.744 0.880 0.698 0.924 0.118 0.112 0.100 0 0.103 0.027 PCH 0.831 1.000 0.763 1.000 0.000 0.000 0.000 0.123 0 0.021 BUCA 0.692 0.827 0.654 0.885 0.103 0.097 0.084 0.099 0.109 0

191