UNIVERSITÉ DE ÉCOLE DOCTORALE « Sciences de la société : Territoires, Économie, Droit »

UMR CNRS 7324-CITERES UNIVERSIDAD AUTÓNOMA DE FACULTAD DE BIOLOGIA

THÈSE présentée par : Joaquín SOLER GIRBES

soutenue le : 20 décembre 2018

pour obtenir le grade de : Docteur de l’université de Tours Discipline/ Spécialité : Sciences de l’Environnement/Ecologie

et

el título de Doctor en BIOLOGÍA por la Universidad Autónoma de Madrid Conservation ecology of Margaritifera auricularia (Spengler, 1793) in

THÈSE co-dirigée par : M. WANTZEN Karl Matthias Professeur et Chaire UNESCO, Université de Tours M. ARAUJO Rafael Docteur, Museo de Ciencias Naturales-CSIC, Madrid, Espagne

RAPPORTEURS : M GEIST Jürgen Professeur, Technische Universität München, Allemagne Mme. ONDINA Paz Professeur, Universidad de Santiago de Compostela, Espagne

JURY : M. ARAUJO Rafael Docteur, Museo de Ciencias Naturales-CSIC, Madrid, Espagne Mme. BAUDRIMONT, Magali Professeur, Université de Bordeaux M. GEIST Jürgen Professeur, Technische Universität München, Allemagne Mme. ONDINA Paz Professeur, Universidad de Santiago de Compostela, Espagne M. PINCEBOURDE Sylvain Professeur, Université de Tours M. WANTZEN Karl Matthias Professeur et Chaire UNESCO, Université de Tours

Acknowledgments/Remerciements

This work would not have been possible without the help of countless people to whom I would like to express my sincere gratitude here.

To Rafael Araujo for co-directing this thesis and for supporting me during all phases of this work. Thank you for everything you have taught me about freshwater and for giving me the opportunity to resume research. Thank you for hosting me at the Museum and, above all, thank you for your friendship.

To Matthias Wantzen for co-directing this thesis and giving me the opportunity to do it. Thank you very much for sharing your extensive knowledge in fluvial ecology. Thank you for your support, for the discussions and suggestions and for your contagious enthusiasm.

To Ángel Luque, for being my thesis tutor at the Universidad Autónoma de Madrid. For all your efforts and advice.

This thesis has been sponsored by the LIFE Project 'LIFE13 BIO / FR / 001162 Conservation of the Giant Pearl in Europe'. The technical and personal resources were financed by this project and the thesis corresponds to a large extent to the research outlined in the project proposal. Thank you to Nina Richard, Laure Morisseau, Yann Guerez and all the people who, despite the difficulties, have made it possible. Especially many thanks to Philippe Jugé. Without him I would not have been able to survive the troubled waters. Thank you for your moral support and for your friendship, and also thanks for your commitment in the work and for all the technical solutions and great ideas, of which this thesis is largely indebted. Many thanks also to Carmen Jugé for her hospitality and sympathy.

This thesis has been provided under the auspices of the UNESCO Chair "Fleuves et Patrimoine (River Culture)" by Karl M. Wantzen. Thanks also to UMR CITERES for hosting the project and supporting this thesis. Thanks to the University of Tours and especially to Veronica Serrano, and Stephanie Gosset for their support. To the Museo Nacional de Ciencias Naturales de Madrid to welcome me and especially to all the friends who have made me feel at home: Bea, Javi, Lola, Annie, Iván, Violeta, David, "la prima", Paula, Óscar, Jesús , Pepe, Marian, Mario, Anna, Victoria, Carolina, Nacho and many others.

3 I also want to show my gratitude to the researchers that I have met throughout this thesis and from whom I have learned many interesting things. Thanks to Catherine Boisneau of the University of Tours for sharing with me her extensive knowledge of freshwater fish and Michele de Monte for your availability and help. To Stephane Rodrigues and Coraline Wintenberger for their guidance on fluvial sedimentology and for their hospitality. Thanks to Jürgen Geist, Bernhard Stoeckle and the rest of the members of the Molecular Biology laboratory of the Aquatische Systembiologie chair of the Technical University of Munich for their wonderful welcome and enthusiastic collaboration.

Thanks to all the freshwater mussel conservation programs that I had the good fortune to visit in Lugo, Zaragoza, Banyoles and Brittany for opening their doors and teaching me the ins and outs of captive breeding. To Paz, Catuxa, Keiko, Marie and Pierre Yves for sharing your experience. Thanks to Miquel, Carles, Gerard, Ramón and the rest of the team for how much I learned and for making my stay in Banyoles one of the most beautiful moments of this time.

Thanks to all the people who have participated in the fieldwork. To the members of Gemosclera and the Club Plongé of Chinon for having collaborated selflessly in the samplings, most of the time under harsh and dangerous conditions. To the team of Logrami, Biotope and of the Fishing Federations that collaborated in the electrofishing surveys and to Vienne Nature and Valentin Viennot that helped me in the Vienne and Creuse surveys. Thanks to Ana, Yannick and Paula for reviewing parts of the manuscripts.

And to all the friends that have made my stay in Tours an unforgettable experience: "el berberecho", Eli, Pili, Jean Baptiste, Serena, Lillo, Edu, Laura, Edu, Maria Angeles, Romeo, Frederic, Helena, Tierry, Kity, Dom, Paloma, Diego, Alex, los Mata and many other Spanish and Italian expatriate friends in Tours. Thanks also to Sam, los Flakis and the Berthom brothers.

To my family, because without them I would not know how to explain myself and especially my sister and my parents, who have taught me to persevere and believe that a better and fairer world is possible.

The last words of thanks are for Bea. Without your love I could not have gotten here. Thank you for having accompanied me during a good part of my life, and for all your support and patience during this thesis period.

4 Abstract

Among the many victims of global biodiversity loss in the Anthropocene, the giant freshwater pearl mussel Margaritifera auricularia is one of the most endangered species. Originally occurring in many European rivers, M. auricularia is a relict species now restricted to a few ageing population in France and Spain, in which natural reproduction is almost absent. The general goal of this thesis was to improve the knowledge of the conservation status of the M. auricularia French populations, their reproductive biology and the early life stages of the species. Despite harboring approximately 90% of the world population, these French populations have been little studied before. To achieve these objectives, the bibliography and museum collections were revised and the populations of the Charente, Vienne and Creuse rivers, considered the most important ones in France, were studied.

On historical times, Margaritifera auricularia was present in the Rhine in France and Germany, the Seine and the Rhône in France, the Po in Italy and the Tagus in Spain, where the species is now believed to be extirpated. Today, M. auricularia is considered restricted to five watersheds in France and Spain (Loire, Charente, Garonne, Adour and Ebro), representing a range contraction of about a 90% in the last two centuries. Recruitment is very scarce in all populations although evidences of an important recent recruitment were found during the present study in the Vienne and Creuse rivers, constituting the most notable event of M. auricularia recent reproduction reported so far. The current lack of recruitment does not seem to be related to the infertility of the specimens. Ova and developing embryos were found throughout the month of March in the Charente and Creuse populations, and the beginning of glochidial release occurred in early to mid-April. Compared with the Spanish populations, the reproductive period begins some weeks later, likely due to differences in water temperature. Development from first cleavage to glochidial maturity took 25-37 days and the estimated number of larvae per gravid mussel was around 2,000,000.

As the vast majority of freshwater mussel species, Margaritifera auricularia depends on fishes to host larvae until their metamorphosis into juveniles. The discovery of three new host species (Gasterosteus aculeatus, Petromyzon marinus and the exotic Silurus glanis) increases the number of known physiological hosts to eight species from five different families, indicating that M. auricularia has a more host generalist behavior than previously expected. This is the

5 first time that a lamprey has been tested and used as a possible host for a freshwater bivalve belonging to the Unionoida Order. The potential use of Petromyzon marinus to reinforce M. auricularia populations along its entire geographic range may help to counteract the decline of the species. All currently known Margaritifera auricularia hosts have a notable tolerance to salinity. This pattern could have arisen early in its evolution because of the potential advantage for dispersion that diadromous species confer. However, anthropogenic causes, such as dam construction and degradation of water quality and habitat, have led to a severe decline of most migratory fish species in Europe during the last two centuries. Although other causes may have contributed to the overall decline of M. auricularia, the loss of hosts seems to be a main factor. The utilization of M. auricularia by Rhodeus amarus, a cyprinid fish that spawns in the mantle cavity of freshwater mussels resulting in a fitness cost for the mussels, was also discovered in the present Thesis. The extensive expansion of this bitterling species, and the reduction in fitness that it could cause on mussels, may be problematic for species facing extinction risk especially in areas inhabited by spatially restricted mussel species.

The juvenile breeding of Margaritifera auricularia is a difficult task and although the survival results are still very low, it seems that the use of an initial laboratory phase with the newborn juveniles reared in detritus boxes can offer more successful results than the direct cultivation of juveniles into raceways. As it seems to happen in other species of freshwater mussels, the high mortality rates of juveniles of M. auricularia found in the first months of culture may be related to their ontogenetic development, characterized by the initial small size of the juveniles and the slow growth until reaching the first millimeter.

Keywords

Margaritifera auricularia; freshwater mussel; brooding; glochidia release; fecundity; conservation status; host fish; juveniles; captive rearing.

6 French Abstract/Résumé

Parmi les nombreuses victimes de la perte de biodiversité mondiale dans l'Anthropocène, la grande mulette Margaritifera auricularia est l'une des plus menacées. Présente à l'origine dans de nombreuses rivières européennes, la grande mulette est une relique désormais limitée à des populations vieillissantes en France et en Espagne, dans lesquelles la reproduction naturelle est presque inexistante. L'objectif général de cette thèse était d'améliorer nos connaissances sur le statut, la biologie de la reproduction et les premiers stades de développement de M. auricularia, en particulier chez les populations françaises. Bien qu'elles abritent environ 90% de la population mondiale, ces populations françaises ont été peu étudiées auparavant. Pour atteindre ces objectifs, la bibliographie et les collections des musées ont été révisées, et les populations des fleuves Charente, Vienne et Creuse, considérées comme les plus importantes de France, ont été étudiées.

Aux époques historiques, Margaritifera auricularia était présente dans le Rhin en France ou en Allemagne, dans la Seine et le Rhône en France, dans le Pô en Italie et dans le Tajo en Espagne, où l’espèce serait à présent disparue. Aujourd'hui, M. auricularia est considéré comme limité à cinq bassins versants en France et en Espagne (Loire, Charente, Garonne, Adour et Ébre), ce qui représente une contraction d'environ 90% de son aire de répartition au cours des deux derniers siècles. Le recrutement est très rare dans toutes les populations, bien que des preuves d'un recrutement récent important aient été découvertes au cours de la présente étude dans les rivières Vienne et Creuse, constituant l'événement le plus notable de la reproduction récente de M. auricularia signalé à ce jour. Le manque actuel de recrutement ne semble pas être lié à la stérilité des spécimens. Des ovules et des embryons en développement ont été trouvés tout au long du mois de mars dans les populations de la Charente et de la Creuse et le début de la libération des glochidies a eu lieu du début à la mi-avril. Par rapport aux populations espagnoles, la période de reproduction commence quelques semaines plus tard, probablement en raison des différences de température de l'eau. Le développement du premier clivage à la maturité des glochidies a pris 25-37 jours, et le nombre estimé de larves par moule gravide était d'environ 2.000.000.

Comme la grande majorité des espèces de moules d'eau douce, Margaritifera auricularia dépend des poissons pour héberger les larves jusqu'à leur métamorphose en juvéniles. La

7 découverte de trois nouvelles espèces hôtes (Gasterosteus aculeatus, Petromyzon marinus et l'exotique Silurus glanis) augmente le nombre d'hôtes physiologiques connus à huit espèces appartenant à cinq familles différentes, ce qui indique que M. auricularia a un comportement plus généraliste que prévu. L’utilisation potentielle de Petromyzon marinus pour renforcer les populations de M. auricularia sur l’ensemble de son aire de répartition géographique peut aider à lutter contre le déclin de l'espèce. Tous les hôtes de Margaritifera auricularia actuellement connus ont une tolérance notable à la salinité. Cette préférence aurait pu apparaître tôt dans son évolution en raison des avantages potentiels en termes de dispersion que les espèces diadromes confèrent. Cependant, des causes anthropiques, telles que la construction de barrages et la dégradation de la qualité de l'eau et de l'habitat, ont entraîné un grave déclin de la plupart des espèces de poissons migrateurs en Europe au cours des deux derniers siècles. Bien que d’autres causes aient pu contribuer au déclin général de M. auricularia, la perte d’hôtes semble être un facteur principal. L'utilisation de M. auricularia par Rhodeus amarus, un poisson cyprinidé qui se reproduit dans la cavité du manteau des moules d'eau douce, entraînant un coût de fitness pour les moules, a également été découverte dans la présente thèse. La vaste extension de cette espèce et la perte fitness qu'elle pourrait causer aux moules, sauraient poser problèmes aux espèces menacées d'extinction, en particulier dans les zones habitées par des espèces avec une distribution spatiale restreinte.

L'élevage juvénile de Margaritifera auricularia est une tâche difficile, et bien que les résultats en termes de survie restent très faibles, il semble que le recours à une phase initiale de laboratoire (boîtes à détritus) puisse donner de meilleurs résultats que la culture directe de juvéniles nouveau-nés dans des auges d'élevage. Comme cela semble se produire chez d'autres espèces de moules d'eau douce, les taux de mortalité élevés des juvéniles de M. auricularia trouvés au cours des premiers mois de culture peuvent être liés à leur développement ontogénétique, caractérisé par la petite taille initiale des juvéniles et leur croissance lente jusqu'au premier millimètre.

Mots clés

Margaritifera auricularia; moules d'eau douce; incubation; libération de glochidies; fécondité; statut de conservation; poisson hôte; juvéniles; élevage en captivité.

8 Spanish Abstract/Resumen

Entre las muchas víctimas de la pérdida global de biodiversidad en el Antropoceno, Margaritifera auricularia es una de las especies más amenazadas. Aunque originalmente distribuida por muchos ríos europeos, hoy en día M. auricularia ha quedado restringida a unas pocas poblaciones envejecidas en Francia y España en las que apenas hay reproducción natural. El objetivo general de esta tesis ha sido mejorar el conocimiento sobre el estado de conservación, la biología reproductiva y las primeras fases de vida de M. auricularia, especialmente en las poblaciones de Francia. A pesar de que las poblaciones francesas albergan aproximadamente el 90% de la población mundial de la especie, han sido muy poco estudiadas. Para lograr estos objetivos, se revisaron la bibliografía y las colecciones de museos y se estudiaron las poblaciones de los ríos Charente, Vienne y Creuse, consideradas las más importantes de Francia.

En épocas históricas, Margaritifera auricularia estuvo presente en el Rhin en Francia y Alemania, el Sena y el Ródano en Francia, el Po en Italia y el Tajo en España, donde se cree que la especie está actualmente extinguida. Hoy en día, M. auricularia se considera restringida a cinco cuencas hidrográficas en Francia y España (Loira, Charente, Garona, Adur y Ebro), lo que representa una contracción de su distribución de aproximadamente un 90% en los dos últimos siglos. El reclutamiento es muy escaso en todas las poblaciones, aunque durante el presente estudio se encontraron evidencias de un importante reclutamiento reciente en los ríos Vienne y Creuse, lo que constituye el evento de reproducción reciente de M. auricularia más notable conocido hasta la fecha. La actual falta de reclutamiento no parece estar relacionada con la infertilidad de los especímenes. Se encontraron óvulos y embriones en desarrollo durante todo el mes de marzo en las poblaciones de Charente y Creuse y el inicio de la liberación de gloquidios ocurrió de principios a mediados de abril. Comparado con las poblaciones españolas, el período reproductivo comienza unas semanas más tarde, probablemente debido a las diferencias en la temperatura del agua. El desarrollo desde la primera división embrionaria hasta la madurez de los gloquidios duró entre 25 y 37 días y el número estimado de larvas por ejemplar grávido fue de alrededor de 2.000.000.

Como la gran mayoría de las especies de náyades, Margaritifera auricularia necesita de peces para que sus larvas o gloquidios se fijen antes de pasar una metamorfosis que los convertirá en

9 juveniles. El descubrimiento de tres nuevas especies de hospedadores (Gasterosteus aculeatus, Petromyzon marinus y el exótico Silurus glanis) aumenta el número de hospedadores fisiológicos conocidos a ocho especies de cinco familias diferentes, lo que indica que M. auricularia tiene un comportamiento más generalista de lo que se esperaba. Es la primera vez que se experimenta y se utiliza una lamprea como posible hospedador de un bivalvo de agua dulce del Orden Unionoida. El uso potencial de Petromyzon marinus para reforzar las poblaciones de M. auricularia en todo su rango geográfico puede ayudar a contrarrestar el declive de la especie.

Todos los hospedadores de Margaritifera auricularia conocidos actualmente tienen una notable tolerancia a la salinidad. Esta relación podría haber surgido gracias a la ventaja potencial que confieren las especies diádromas para la dispersión de los moluscos, lo que podría haberse mantenido en esta especie como un constraint evolutivo. Sin embargo, nuevas amenazas como la construcción de presas y la degradación de la calidad del agua y el hábitat, han provocado un grave declive de la mayoría de las especies de peces migradores en Europa durante los últimos dos siglos. Aunque el declive general de M. auricularia puede deberse a numerosas causas, la pérdida de especies hospedadoras parece haber sido un factor principal. En esta tesis también se descubrió la utilización de M. auricularia por Rhodeus amarus, un pez ciprínido que pone sus huevos en la cavidad del manto de las náyades provocando un coste en fitness para las náyades. La gran expansión de R. amarus y la reducción de fitness que podría causar en las almejas de agua dulce, puede poner en peligro la supervivencia de algunas especies en riesgo de extinción, especialmente las que tienen una distribución más restringida.

La cría en cautividad de juveniles de Margaritifera auricularia es una tarea difícil y, aunque los resultados de supervivencia son todavía muy bajos, parece que el uso de una fase inicial de laboratorio donde los juveniles recién nacidos se críen en cajas de detritus, puede ofrecer resultados más exitosos que el cultivo directo de juveniles en diferentes tipos de canales de cultivo. Como parece ocurrir en otras especies de almejas de agua dulce, las altas tasas de mortalidad de los juveniles de M. auricularia encontradas en los primeros meses de cultivo pueden estar relacionadas con su desarrollo ontogenético, caracterizado por el pequeño tamaño inicial de los juveniles y su lento crecimiento hasta alcanzar el primer milímetro.

Palabras clave: Margaritifera auricularia; almeja de agua dulce; incubación; liberación de gloquidios; fecundidad; estado de conservación; pez hospedador; juveniles; cría en cautividad.

10 French long abstract/Résumé significatif

Les moules d'eau douce constituent un élément remarquable et important des écosystèmes aquatiques. Elles peuvent former des agrégats de plus de 100 animaux par m2, appelés lits de moules, où ils peuvent dominer la biomasse benthique. Les assemblages de moules denses ont la capacité de filtrer un volume d'eau énorme, pouvant dépasser les débits journaliers d'un ruisseau (Haag, 2012). En tant que bivalves filtreurs, les moules jouent un rôle important dans le cycle des nutriments, car elles influencent la chimie et la clarté de l'eau (par exemple, Kryger & Riisgard, 1988; Strayer et al., 1999; Vaughn & Hakenkamp, 2001). Leurs déchets donnent la possibilité de renforcer les populations locales d'algues (Vaughn et al., 2007) et de macroinvertébrés (Vaughn & Spooner, 2006), et leurs coquilles fournissent un habitat à d'autres organismes benthiques (Spooner & Vaughn, 2006; Vaughn & Spooner, 2006).

Les moules d'eau douce fournissent également d'importants services directs aux humains, tels que la purification de l'eau (Lopes-Lima et al., 2017). En outre, elles ont été récoltées comme source de perles, de nacre et d’alimentation pour l’humain depuis la préhistoire (par exemple, Kunz, 1898, Morrison, 1942, Claassen, 1994; Anthony & Downing, 2001; Walker et al., 2001).

Malgré avoir survécu depuis le Triasique (Haas, 1969; Watters, 2001), elles sont maintenant décimées à l'échelle mondiale en raison d'activités humaines et constituent aujourd'hui l'un des groupes d'animaux les plus menacés de la planète (Bogan, 1993; Young et al., 2001; Lydeard et al. 2004; Strayer et al., 2004; Strayer, 2008). Répertoriées dans la Liste Rouge 2015 des Espèces Menacées de l'UICN, 44% des espèces de moules d'eau douce sont classées dans la catégorie quasi menacée ou menacée (Lopes-Lima et al. 2017).

Margaritifera auricularia (Spengler, 1793) est considérée comme l'une des espèces de bivalves les plus rares et les plus menacées en Europe. Elle est classée en danger critique d'extinction par l'UICN (Araujo & Ramos, 2001a; Prié, 2010a; Prié et al., 2018). Bien que probablement répandue dans la plupart des rivières d'Europe occidentale au début du XXe siècle, il ne reste aujourd'hui seulement quelques populations de cette espèce en Espagne et en France. La quasi- absence de moules juvéniles indique que la reproduction est fortement réduite au sein de ces populations, ce qui représente le problème le plus critique pour la survie de l'espèce.

Bien que les connaissances scientifiques sur M. auricularia se soient considérablement améliorées au cours des deux dernières décennies, d'importantes lacunes empêchent encore une préservation efficace de l'espèce. En outre, malgré que près de 90% des populations vivantes

11 existent dans des rivières françaises, la plupart des connaissances biologiques de M. auricularia ne proviennent que d'études sur les populations espagnoles, y compris les études sur le sex-ratio et la gamétogenèse (Grande et al., 2001), la morphologie des larves ou glochidies (Araujo & Ramos, 1998), la saison d’émission des glochidies (Araujo et al., 2000), les poissons hôtes (Araujo et al., 2001), la gamme de degrés-jours pour compléter la métamorphose des glochidies (Araujo et al., 2003), les menaces nationales (Araujo & Ramos, 2000b; Araujo & Álvarez- Cobelas, 2016), et les modèles de croissance (Nakamura et al., 2018).

L'objectif principal du travail de recherche décrit dans cette thèse, est de fournir les connaissances essentielles pour la mise en œuvre des programmes de conservation de M. auricularia en France. Par conséquent, les recherches et études présentées ici, se concentrent principalement sur l’étude des populations françaises de M. auricularia. La majorité des populations françaises vivantes actuellement connues de cette espèce sont réduites en nombre de spécimens, ce qui les rendent plus vulnérables. Par conséquent, la plupart des travaux décrits dans cette thèse ont été réalisés dans les populations considérées comme les plus importantes de France, situées dans deux bassins hydrographiques différents, la Loire (comprenant la Vienne et la Creuse) et la Charente.

Compte tenu des connaissances limitées sur les populations françaises et de la variété des facteurs qui menacent la survie de M. auricularia, cette thèse inclut l’étude des questions suivantes :

1. Clarifier la distribution historique de M. auricularia en France

Bien que les anciens textes montrent que M. auricularia était répandue en Europe avant le vingtième siècle, la répartition précise de cette espèce avant l'effondrement actuel des populations reste mal connue, notamment en France. Pour pallier ce manque de connaissances sur la biogéographie de l'espèce, une revue exhaustive de la bibliographie et des collections des musées a été réalisée sur les populations françaises.

La littérature disponible a fourni des données précieuses, bien que généralement sans localisation ni date précise. Margaritifera auricularia est connue aux Pays-Bas, en Angleterre et en Allemagne d'après des archives fossiles uniquement. Cependant, certains coquillages recueillis dans la rivière Unstrut en Allemagne sont très bien conservés et pourraient peut-être remonter à des époques historiques, du moins jusqu'au début du Moyen Âge (Bössneck et al, 2006). Les données sur les fossiles en Espagne incluent un fleuve méditerranéen du Quaternaire

12 à Yecla (Murcie) avec des spécimens âgés de 129.000 à 140.000 ans (Andrés & Ortuño, 2014) et de nombreux autres cours d'eau atlantiques avec des spécimens de 5.000 ans (Araujo & Moreno, 1999). En France, nous avons trouvé des données sur les fossiles près de (provenant de fouilles archéologiques), et dans le Massif Central (procédant des fossiles rassemblés dans une grotte) qui étaient vraisemblablement le résultat d'un transport humain.

Selon les données historiques recueillies, Margaritifera auricularia n'a été découverte que dans les grandes rivières avec substrat calcaire, en France, en Espagne et en Italie. En France, les données historiques proviennent principalement des bassins versants de l'Atlantique et de la Manche, avec une seule occurrence dans les bassins versants de la Méditerranée, dans la Saône (affluent du Rhône). En Italie et en Espagne, l'espèce est historiquement connue dans deux bassins versants de la Méditerranée, le Pô et l'Èbre réciproquement (Araujo & Ramos, 2000a). En Espagne, M. auricularia vivait dans deux canaux historiques de l'Èbre, le Canal Imperial et le Canal de Tauste, où il y avait environ 5.000 spécimens vivants. Les données les plus récentes publiées sur ces populations espagnoles ont été enregistrées dans Araujo & Ramos (2000b), Gómez & Araujo (2008) et Araujo & Álvarez-Cobelas (2016).

Les spécimens de M. auricularia conservés dans les collections des musées ont d'abord été examinés à grande échelle par Araujo & Ramos (2001), principalement dans les musées nationaux du monde entier. Notre étude s'est concentrée uniquement sur les collections régionales françaises. Sur les 58 collections identifiées, 25 avaient au moins un spécimen de M. auricularia. Une partie des données des collections des musées étaient des spécimens fossiles. Nous avons trouvé un total de 400 spécimens non fossiles dans les collections des musées, y compris les 37 spécimens déjà trouvés par Araujo et Ramos (2001). Parmi eux, 332 ont été localisés à une échelle de bassin versant. Un tiers de ces spécimens provenait du bassin versant de la Garonne, 19% de la Saône (la moitié d'entre eux, provenant d'un seul lot collecté par Coutagne en 1879), et 17% de l'Èbre. Les autres bassins versants représentent moins de 30% des spécimens des collections des musées. La plupart des données provenant de ces collections correspondaient aux données de la littérature, à l'exception de celles des fleuves Arros et Vézère en France. Environ 80% des spécimens datés ont été collectés avant le début du vingtième siècle.

En résumé, l'analyse historique a confirmé que M. auricularia était autrefois présente jusqu’à la Tamise en Angleterre, aux Pays-Bas et en Allemagne, où des spécimens de fossiles ont été découverts et étudiés (Araujo & Ramos, 2001). Aux époques historiques, nous avons retrouvé

13 des archives de musée (coquillages récents) provenant du Rhin en France ou en Allemagne (leur localisation exacte reste inconnue), de la Seine et du Rhône en France, du Pô en Italie et du Tajo en Espagne, où nous pensons á ce jour que l'espèce aurait disparu (Araujo & Ramos, 2001).

2. Mettre à jour les informations disponibles sur le statut des populations vivantes

Depuis la redécouverte des populations françaises actuelles de M. auricularia, de nombreuses enquêtes de terrain ont été réalisées en France et en Espagne. Mais la plupart de ces résultats sont non publiés ou disponibles uniquement en tant que littérature grise. Une revue sur l'ensemble de cette documentation sur M. auricularia en France et en Espagne a été réalisée. Elle résume les informations sur la répartition actuelle de l'espèce, la taille de ses populations et leur état par rapport à la reproduction.

Contrairement à Margaritifera margaritifera L, 1758, qui est généralement limitée aux habitats en amont, M. auricularia vit généralement dans des écosystèmes en aval. L'étude de cet habitat est difficile car il est souvent profond, turbide, avec beaucoup de courant et navigable. Dans ces cas, les prospections sont effectuées en plongée bouteille. Pour les populations plus accessibles, la plongée en apnée a permis des prospections efficaces. Les populations de Margaritifera auricularia ont été estimées à l'aide de différentes méthodes, notamment des dénombrements exhaustifs d'individus vivants observés, des analyses statistiques et une appréciation subjective basée sur la densité des spécimens observés.

Avant 2007, seules trois populations (Ebro, Charente et Vienne-Creuse) étaient connues à l'échelle mondiale. Des prospections intensives menées au cours de la dernière décennie sur 2.500 km de rivières en France et en Espagne, ont permis de redécouvrir neuf autres populations. Aujourd'hui, Margaritifera auricularia est considérée comme limité à cinq bassins versants: du nord au sud, le bassin versant de la Loire (deux populations proches dans les rivières de la Vienne et la Creuse), le bassin versant de la Charente, le bassin versant de la Garonne (deux populations très isolées, dans les rivières de la Dronne et la Save), le bassin versant de l'Adour (au moins trois populations isolées dans les rivières Adour, Luy et Arros) et l'Èbre (quatre populations, trois dans des canaux et une petite dans l'Èbre même). Vu l'ampleur des efforts déployés pour étudier l'espèce dans son aire de répartition historique, nous pensons maintenant qu'il y a très peu de chances de retrouver des populations non remarquées (sauf peut-être dans le nord-est de la France). Cette distribution actuelle représente une contraction

14 dans sa distribution d’environ 90% au cours des deux derniers siècles, comme précédemment estimée par Prié et al. (2014).

En Espagne, la principale population, avec 5.000 spécimens vivants, vit dans le Canal Imperial de Aragón, bien qu'il y ait eu des mortalités récentes (Nakamura, Guerrero et al., 2018). Une nouvelle population a récemment été découverte dans le fossé d'irrigation de Quinto, avec 25 spécimens vivants. Le Canal de Tauste héberge encore plusieurs spécimens vivants, mais l’espèce est aujourd'hui disparue de la partie basse de l'Èbre.

En France, la Creuse et la Vienne abritent environ 250 individus. Dans le bassin versant de la Garonne, trois sites contenant quelques dizaines de spécimens vivants ont été découverts dans la rivière de la Dronne, et seuls cinq spécimens vivants ont été observés dans la rivière de la Save. Dans le bassin versant de l'Adour, la population totale est estimée à environ 300 spécimens sur toute la longueur de l'Adour, à environ 150 spécimens sur l'affluent Luy et à environ 200 individus sur les 54 km d'habitat favorable de l’Arros. La plus grande population mondiale vit dans la Charente avec environ 100.000 individus.

Le recrutement est très faible dans toutes les populations, mais des spécimens vivants âgés de moins de 10 à 15 ans ont été trouvés dans l'Èbre en Espagne et dans les rivières Vienne, Charente, Dronne et Adour en France. Néanmoins, au terme de cette étude, certaines populations ont déjà disparu de l'Èbre et de l'Adour. Le statut de l'espèce reste donc préoccupant. Les populations prioritaires pour la conservation sont : celle de la Charente, car elle est de loin la plus grande du monde; la population de la Vienne et de la Creuse, puisqu’elle présente le plus haut taux de recrutement naturel; les populations du bassin versant de l'Adour, parce qu'elles forment une métapopulation importante composée de sous-populations très éloignées; et la population de l'Èbre, car c’est la seule population restante dans les bassins versants de la Méditerranée.

3. Décrire le statut des populations de la Charente et de la Creuse-Vienne

Des données détaillées sur la taille et la structure des populations de M. auricularia sont nécessaires pour élaborer des mesures de conservation, et fournir des données de base pour des investigations ultérieures, permettant de mettre en évidence les changements survenus dans la population étudiée. Ces informations sont rares pour la plupart des populations françaises de M. auricularia. La structure de la population dans les rivières Charente et Vienne-Creuse a déjà été analysée, mais elle était basée sur une taille réduite des échantillons ou sur des estimations

15 uniquement à partir des coquilles d'individus morts, ce qui peut transmettre une interprétation inexacte de la situation réelle. Cependant, cette information nous a permis de comparer nos données. En outre, l'évolution temporelle de ces populations a été insuffisamment étudiée depuis sa redécouverte, et il existe peu d'informations pour appuyer le choix d'une méthodologie appropriée pour réaliser des études comparatives.

L’objectif de cette étude était 1) d’évaluer les caractéristiques écologiques de base des populations de la Charente et de la Vienne-Creuse telles que la répartition, l’abondance et la structure de la population, 2) de comparer ces résultats avec des études antérieures afin de détecter des changements temporels, 3) de décrire la capacité de déplacement de l'espèce et 4) discuter de l'adéquation de différentes approches méthodologiques pour détecter des changements temporels dans les populations de M. auricularia.

Les populations de la Charente et de la Vienne-Creuse ont été choisies pour cette étude car ceux sont les plus importantes en France, et car il existe des données de base pour les comparer avec nos résultats. Afin d'établir ces comparaisons, les méthodes d'échantillonnages utilisées dans les études précédentes ont été suivies telles quelles. Dans la Vienne-Creuse, des prospections ont été effectuées dans 14 stations réparties sur un tronçon de 54 km de long avec des bathiscopes, et par la plongée en apnée. Dans la Charente, 12 stations réparties sur un tronçon de 25 km de long ont été étudiées. A chaque station, des transects de 20 m de long ont été étudiés en plongée bouteille.

La caractérisation de la structure des populations était basée sur des mesures biométriques. La longueur de la coquille des spécimens vivants et morts de M. auricularia a été mesurée à 0,1 mm près avec des pieds à coulisse. Pour déduire la structure de la population et détecter un recrutement récent possible, une distribution de fréquence de taille de 10 mm d'intervalle a été utilisée.

Afin d'identifier les spécimens mesurés biométriquement et de permettre des actions de suivi futures, une petite étiquette numérotée a été collée sur l'une des valves. De plus, dans la population Vienne-Creuse, l'emplacement précis des spécimens marqués a été enregistré à l'aide d'un GPS à précision centimétrique. Un an plus tard, les individus géoréférencés ont été à nouveau examinés afin d'évaluer le taux de mortalité, et son déplacement horizontal.

Les plus fortes densités dans ces rivières ont été trouvées dans les zones urbaines ou près des rejets d'eaux usées, suggérant que M. auricularia (même les individus jeunes) semblent

16 capables de tolérer un certain degré de pollution de l'eau. Ceci est conforme aux observations précédentes de Prié (2010), mais des études écotoxicologiques complémentaires sont nécessaires pour étudier les limites de tolérance de M. auricularia à différentes concentrations de polluants, et en particulier, aux premiers stades de leur vie. Nous apportons également des données sur la qualité de l’eau des rivières étudiées, indiquant une tolérance plus grande de M. auricularia à différentes concentrations de calcium et de conductivité de l’eau que nous pensions auparavant.

La distribution des fréquences de taille des coquilles indique une population vieillissante dans la Charente et un recrutement presque inexistant, ce qui est en accord avec les résultats précédemment rapportés par Nienhuis (2003), Prié et al. (2008) et Prié (2010). Dans cette rivière, bien que la comparaison avec les études précédentes nécessite être interprétée avec prudence, les résultats suggèrent une population de M. auricularia semblant être restée relativement stable au cours des 6 à 9 dernières années.

Inversement, dans les rivières de la Vienne et de la Creuse, les résultats obtenus dans cette étude diffèrent nettement des résultats précédents de Cochet (2001) et de Nienhuis (2003). Des évidences d'un important recrutement récent ont été trouvées, constituant l'événement le plus notable de reproduction récente de M. auricularia signalée à ce jour à l'échelle mondiale. Nous avons observé que la taille des spécimens vivants variait entre 8 et 18 centimètres, avec 34% des individus appartenant à des classes de taille inférieures à 12 cm. Une étude préliminaire de l'âge des individus basée sur le comptage des anneaux de croissance internes et externes des coquilles de cette population, suggère que la classe de taille 10-11 cm correspond à un âge de 15-20 ans (J. Soler, données non publiées). Cela implique qu’après une période sans reproduction effective, un événement de recrutement important s’est produit entre la fin des années 90 et le début des années 2000. Néanmoins, la comparaison des abondances actuelles avec celles trouvées par Cochet (2002) suggère une baisse de 45% au cours des 14 dernières années, ce qui est conforme à un taux de mortalité annuel estimé compris entre 1 et 13%. Par conséquent, malgré l'important recrutement récent observé, les résultats obtenus ici montrent que si les mesures de conservation ne sont pas prises en compte de manière urgente, l'espèce pourrait disparaître dans un avenir proche.

Une évaluation du déplacement horizontal effectuée par des spécimens marqués dans les rivières Vienne et Creuse a montré, que 71% des moules étudiées se sont déplacées de moins de 20 cm en un an, et seulement 7% ont été trouvées à une distance supérieure à 50 cm. Ces

17 résultats confirment que M. auricularia est essentiellement sédentaire, ce qui indique que la prospection du même emplacement au cours des campagnes de suivi successives, peut constituer une approche efficace pour évaluer les variations temporelles des populations. Compte tenu de la profondeur et de la densité relativement élevée de M. auricularia dans la Charente, une étude quantitative des parcelles permanentes dans cette rivière est recommandée. Selon les caractéristiques de la population des rivières de la Vienne et de la Creuse, nous proposons que les futures actions de surveillance de ces populations soient basées sur le suivi d'individus précisément géoréférencés et marqués avec des pit-tags.

4. Augmenter l'information sur les facteurs de menace pour M. auricularia en France

De nombreux facteurs ont été identifiés comme responsables du déclin global de M. auricularia, y compris la destruction de son habitat, la surexploitation commerciale, la pollution de l'eau et des sédiments, la perte de poissons hôtes en raison de poissons exotiques et des modifications du débit naturel par la construction de barrages. L'introduction d'espèces envahissantes est une menace relativement récente pour les moules d'eau douce, mais pourrait devenir l'une des préoccupations majeures pour leur conservation. Les effets d’espèces bivalves invasives telles que la moule zébrée, Dreissena polymorpha (Pallas, 1771) et la corbicule, Corbicula fluminea (Müller, 1774), sur les populations de M. auricularia n’ont pas été bien étudiés, mais ils pourraient être très importants étant donné les capacités de filtration très élevées de ces petits bivalves (Bogan, 1993; Parker et al., 1998; Ricciardi et al., 1998; Neves, 1999; Yeager et al., 1999; Burlakova et al., 2000; Lydeard et al., 2004; Strayer, 2006). Les espèces de poissons exotiques peuvent également constituer une menace importante pour les moules d'eau douce, en contribuant de manière significative à la disparition des faunes de poissons indigènes, qui sont des hôtes essentiels pour les unionoïdes.

L'objectif de cette étude était d'alerter la communauté scientifique et les gestionnaires de la biodiversité sur une nouvelle menace potentielle pour M. auricularia, dérivé de la colonisation récente de leurs habitats par une espèce de poisson exotique. La présence de la Bouvière, Rhodeus amarus (Bloch, 1782), a été détectée par la pêche électrique lors d’une étude visant à identifier les poissons hôtes de M. auricularia dans les rivières Vienne, Creuse et Charente. Ce poisson cyprinidé est connu pour son cycle de vie très particulier, caractérisé par sa symbiose obligatoire avec les moules d'eau douce, à l'intérieur desquelles la femelle de la bouvière pond ses œufs. (Smith et al., 2004).

18 La présence d’embryons de la bouvière dans les moules entraîne un coût de fitness pour les hôtes: ils rivalisent avec l’hôte pour l’oxygène (Smith et al. 2001) et réduisent la circulation de l’eau au niveau des branchies des moules, ce qui peut affecter leur capacité de filtration (Mills et al., 2005). Ces effets pourraient expliquer les taux de croissance considérablement réduits des moules infestées par des embryons de R. amarus (Reichard et al., 2006). Dans les moules d'eau douce, la taille est positivement corrélée à la fécondité (Bauer, 1994); par conséquent, une réduction de la croissance représente un coût de fitness pour les moules.

Malgré le fait que la bouvière est considérée comme une espèce indigène sur une grande partie de son aire de répartition actuelle en Europe, des études récentes ont montré qu’elle s’est étendue relativement récemment (des siècles à des millénaires avant notre époque) à partir de la région de la mer Noire à l'Europe centrale et occidentale (Bohlen et al., 2006, Van Damme et al., 2007, Bryja et al., 2010). Au cours de son expansion, R. amarus a utilisé avec succès des espèces de moules présentes dans son aire de répartition historique comme hôtes, telles que Anodonta anatina (L, 1758), A. cygnea (L, 1758), Pseudanodonta complanata (Rossmässler, 1835), pictorum (L, 1758), U. tumidus Retzius, 1788 (Wiepkema, 1961; Balon, 1962, Reynolds et al., 1997; Smith et al., 2000; Smith et al., 2004) et U. crassus Philipsson, 1788 (Reichard et al., 2010; Tatoj et al., 2017). Comparées aux populations d'unionidés de la région pontique, les populations de moules indigènes d'Europe continentale sont moins adaptés pour éviter ou éjecter les œufs de R. amarus, probablement en raison d'une durée plus courte de sympatrie (Reichard et al., 2010).

Après son déclin spectaculaire en abondance de 1960 à 1980, R. amarus a été déclarée espèce en voie de disparition en Europe centrale et occidentale et fut inscrite à l'Annexe II de la Directive Habitat (92/43/CEE). Néanmoins, depuis 1980, la distribution de R. amarus s'est étendue dans de nombreuses régions d'Europe, notamment en Europe orientale, où une augmentation considérable et rapide de l'abondance a pu être observée (Kozhara et al., 2007). En France, cette espèce était limitée aux bassins du nord-est au dix-neuvième siècle (Valenciennes, 1848; Gehin, 1868; Gensoul, 1908). Cependant, de nos jours, elle s'est répandue dans presque tout le pays, en particulier dans le sud-ouest, où elle est considérée comme envahissante (Kottelat & Freyhof, 2007). Deux raisons ont été proposées pour expliquer la récente expansion de R. amarus: les activités anthropiques, y compris son introduction par les pêcheurs à la ligne et les connexions artificielles de systèmes de voies navigables, et le changement climatique (Kozhara et al., 2007; Van Damme et al., 2007).

19 L'expansion récente de R. amarus dans de nouveaux habitats et de nouvelles zones géographiques a entraîné son contact avec différentes espèces de moules d'eau douce. À la suite de sa récente expansion en Europe occidentale, il a été observé que R. amarus ovoposait dans Unio mancus Lamarck, 1819 et Potomida littoralis (Cuvier, 1798) dans des rivières françaises (Prié, 2017). La vaste expansion de cette espèce et la perte de fitness qu'elle peut entraîner, pourraient être problématiques pour les espèces de moules d’eau douce menacées de disparition, comme dans le cas de la Grande Mulette, Margaritifera auricularia.

Les observations sur le terrain et en laboratoire ont permis de vérifier pour la première fois que R. amarus utilise M. auricularia comme hôte de leurs œufs et de leurs embryons. Étant donné que l'anatomie des branchies des margaritiféridés est différente de celle des unionidés (voir section 3.2.1 de cette thèse), cette découverte soutient l'idée que R. amarus peut parasiter toutes les espèces de moules d’eau douce européennes indigènes, indépendamment de l’anatomie de leurs branchies.

Dans les rivières d’Europe centrale, les moules indigènes ont coexisté avec R. amarus pendant des centaines, voire des milliers d'années, sans une diminution apparente de leur population malgré la présence de ces derniers. Cependant, les moules d'eau douce européennes sont en déclin actuellement et, 12 des 16 espèces actuellement reconnues sont classées comme menacées ou quasi menacées par l'UICN (Lopes-Lima et al., 2017). Les populations de moules qui ont connu un déclin drastique de la densité sauraient particulièrement impactées par la présence de R. amarus, car les moules individuelles pourraient accueillir un plus grand nombre d'embryons de R. amarus, ce qui représenterait un stress supplémentaire pour ces moules (Van Damme et al., 2007, Prié 2017, Tatoj et al., 2017). Ainsi, l’expansion de R. amarus et le coût de fitness qu’elle pourrait causer sur les moules, sauraient problématiques pour des espèces telles que M. auricularia, faisant face à un risque d’extinction.

5. Améliorer les connaissances sur la biologie de la reproduction de M. auricularia

Étant donné que l’absence de reproduction des populations naturelles est l’un des problèmes les plus importants pour la conservation de l’espèce, élargir les connaissances sur ce sujet était l’un des objectifs principaux de cette thèse. La biologie de la reproduction des populations françaises n'a jamais été étudiée et, quelques questions fondamentales, comme les suivantes, n'ont pas encore été abordées. Le manque actuel de recrutement est-il dû à la stérilité des spécimens? Si elles sont toujours fertiles, quelle est la saison de reproduction et quelles

20 variables abiotiques peuvent la déterminer? Quelles espèces de poissons hôtes sont utilisées? La disponibilité des poissons hôtes limite-t-elle la reproduction? Dans quelle mesure le déclin global de l'espèce est-il lié à la disponibilité des poissons hôtes? La réponse à ces questions est l’objectif principal de cette section, détaillée dans les objectifs spécifiques suivants:

5.1 Identification de la période de reproduction des populations françaises de M. auricularia

Bien que la période de reproduction des populations espagnoles ait été identifiée auparavant, cette information est encore inconnue pour les populations françaises. Ces connaissances de base sont essentielles pour établir de futures mesures de conservation. En fait, la collecte de femelles gravides au bon moment de l’année est primordiale pour mener à bien les programmes d’élevage en captivité. Pour éviter de longues périodes de captivité des moules adultes en laboratoire, la collecte doit avoir lieu lorsque les larves sont matures. En outre, les larves des espèces avec une gravidité de courte durée (ou tachytictiques, comme on le suppose pour tous les margaritiferidés) sont généralement prêtes pendant une très courte période, par conséquent, manquer le moment adéquate peut entraîner l'échec des activités du programme de reproduction artificielle de tout l’année. De plus, le moment de la libération des glochidies revêt une importance cruciale par rapport à la disponibilité des poissons hôtes, et peut donner des pistes pour identifier des espèces hôtes utilisées dans la nature.

Sur la base des informations disponibles sur la saison de reproduction de l'espèce en Espagne (Araujo et al., 2000), les populations de M. auricularia de la Creuse et de la Charente ont été étudiées entre mars et mai des années 2015-2017. Les moules ont été collectées en plongée bouteille et régulièrement inspectées pour en déterminer la gravidité et le stade de développement des embryons, à la fois sur le terrain et en laboratoire. Les évaluations de la gravidité sur le terrain ont été réalisées en observant des échantillons du contenu des branchies, obtenus à l'aide d'une seringue jetable et observées au microscope optique. Les moules gravides ont été transportées au laboratoire et maintenues dans des aquariums remplis d'eau de rivière constamment aérée. Le matériel larvaire libéré par chaque moule était inspecté quotidiennement au microscope binoculaire et les dates d'émission des glochidies enregistrées.

Les connaissances antérieures sur la saison de reproduction de M. auricularia sont basées sur les populations du bassin de l'Èbre, qui sont gravides en février et libèrent des glochidies en mars (Araujo et al., 2000). Les observations de terrain et de laboratoire des deux populations

21 françaises réalisées de 2015 à 2017 ont dévoilé que les moules présentaient des embryons en développement jusqu'en mars, et que le début de la libération de glochidies tenait lieu du début à la mi-avril. Pris ensemble, cela indique une différence dans le moment de la libération des glochidies entre les populations espagnoles et françaises, et même entre les deux populations françaises, ou la libération des glochidies dans la population plus septentrionale de la rivière Creuse étant retardée de plusieurs jours.

Bien que la plupart des informations disponibles sur la saison de libération des glochidies soient basées sur des observations en laboratoire, les évidences suggèrent qu’il existe un gradient latitudinal qui influence le moment de la saison de reproduction chez cette espèce, similaire à celui proposé pour d’autres espèces de moules d'eau douce. Howard (1915) a suggéré que les différences dans le moment de la reproduction pourraient être liées aux différences latitudinales de la température de l'eau. Les différences thermiques entre les rivières semblent être responsables des dissimilitudes de temps de reproduction chez d'autres margaritiferids tels que M. laevis (Awakura, 1968; Naito, 1988), M. falcata (Meyers & Millemann, 1977) et M. margaritifera (Hastie & Young, 2003). Dans ce dernier cas, les auteurs ont suggéré que les différences annuelles en degrés-jours accumulés (somme de la température moyenne quotidienne) pourraient constituer un indicateur plus fiable des différences thermiques. Les degrés-jours cumulés de six années consécutives (2011-2016) des trois rivières comparées dans notre étude (Èbre, Charente et Creuse) indiquent un gradient de température latitudinal qui semble corroborer cette hypothèse.

Malgré des variations dans la période de reproduction entre les différentes populations, M. auricualaria commence à libérer des glochidies de la fin de l'hiver au début du printemps, ce qui est plus précoce que chez les autres margaritiferids. Le moment de la libération des glochidies est d’une importance cruciale par rapport à la disponibilité des poissons hôtes.

5.2 Description du processus d'incubation chez M. auricularia

Bien que des études antérieures aient considérablement amélioré les connaissances sur la biologie de la reproduction de M. auricularia, à l'exception de la brève description fournie par Haas (1914, 1924), aucune étude n'a caractérisé l'anatomie des branchies et les modifications du marsupium au cours de l’incubation, la période de développement et la fécondité chez cette espèce. Cette information est fondamentale pour mieux comprendre d’autres traits biologiques, et elle est aussi très importante pour les études taxonomiques et phylogénétiques.

22 Les observations des branchies de M. auricularia ont été effectuées sur des spécimens recueillis au Canal Imperial (bassin de l'Èbre, Espagne) et conservées dans la collection malacologique du Musée National des Sciences (Museo Nacional de Ciencias Naturales) de Madrid. En raison de la rareté de cette espèce, nous n'avons pas sacrifié de spécimens supplémentaires pour cette étude. Trois spécimens (deux gravides et un non gravide) ont été traités pour des analyses histologiques. Des portions de la partie centrale des démibranches externes et internes ont été excisées, progressivement déshydratées au moyen d'une série d'éthanol, incluses dans du Paraplast® et sectionnées en série (5 à 10 µm) avec un microtome. Les lames ont ensuite été colorées à l’hématoxyline-éosine et à l’azan de Heidenhain et observées au microscope optique équipé d’une caméra.

Afin d'estimer la durée du processus d’incubation et le nombre de larves hébergées par une seule moule, des individus des populations de la Creuse et de la Charente ont été collectées par plongée bouteille pendant la période de reproduction identifiée auparavant. La durée de la maturation des œufs a été estimée en comptant le nombre de jours entre la première observation d'œufs non divisés et la présence de glochidies matures chez deux spécimens de la Creuse. Pour estimer le nombre total de glochidies hébergées par une seule moule, nous avons recueilli tout le matériel larvaire libéré individuellement par trois spécimens des deux populations. Nous avons dilué et homogénéisé ce matériau dans un volume connu d'eau de rivière, puis compté le nombre de glochidies dans trois échantillons de 1 ml dans une chambre de comptage Sedgewick Rafter pour chaque échantillon.

Les ctenidia de Margaritifera auricularia étaient composées de démibranches internes et externes, chacune avec des lamelles ascendantes et descendantes. Le démibranche intérieur était toujours plus grand que l’extérieur. Presque toute la longueur du démibranche externe était attachée dorsalement au manteau, 1/9 à 1/4 de sa longueur étant libre postérieurement. Le démibranche interne était relié dorsalement au sac viscéral, à l'exception des parties centrale et postérieure. Comme dans le reste des espèces de , les extrémités postérieures libres des deux ctenidia étaient en contact l'une avec l'autre, et avec la paroi interne du manteau, formant le pseudodiaphragme.

Les lamelles de chaque démibranche étaient unies par intermittence par des jonctions irrégulièrement dispersées formant légèrement un motif diagonal. Ces jonctions interlamellaires (ILJ) consistent généralement en des extensions de tissu entre deux ostia adjacentes dans la direction du filament. Dans le tissu lamellaire des démibranches, les

23 vaisseaux sanguins, que nous pouvons observer de manière macroscopique, courent et se ramifient dorso-ventralement. Des canaux d'eau, reliant la lumière intérieure des démibranches à la cavité du manteau à travers l'ostia, étaient toujours présents.

En ce qui concerne les modifications anatomiques associées à l’incubation chez M. auricularia, nous avons observé que, pendant la gravidité, les démibranches augmentent le volume de leur lumière interne pour héberger les larves en réduisant la largeur du tissu conjonctif lamellaire. Apparemment, ceci est contrôlé par la relaxation des fibres musculaires des ILJ, qui se dilatent au maximum à partir d'un état contracté, de sorte que les épithéliums de la lumière interne à la base des ILJ s'étendent vers les filaments branchiaux, réduisant ainsi la largeur du tissu conjonctif lamellaire et en allongeant des ILJ. Par conséquent, ceux sont les ajustements dans les dimensions linéaires du tissu, plutôt que l’action d’étirement, qui sont responsables de la légère augmentation de la largeur du démibranche pendant l’incubation. Ce manque d'élasticité, qui a également été signalé pour M. margaritifera (Smith, 1979), est contraire aux résultats rapportés pour les (Ortmann, 1911; Fuller, 1972, 1973; Richard et al., 1991; Tankersley & Dimock, 1992), et peut s’avérer être une contrainte phylogénétique de la famille Margaritiferidae (Smith, 1979).

Comme décrit pour M. margaritifera par Smith (1979), les cellules épithéliales des ILJ et du reste de la lumière interne subissent des changements morphologiques au cours du processus d’incubation, probablement pour assumer une fonction de sécrétion. La fonction de ces sécrétions ne semble pas claire. D'après des études effectuées sur plusieurs Unionidés (espèces d'Anodontinae et de Lampsilini), certains auteurs ont suggéré que les concrétions de nutriments et de calcium pourraient être transférées des femelles aux larves en développement via du mucus sécrété par des cellules situées dans les septa interlamellaires (Wood, 1974, Tankersley & Dimock, 1992; Tankersley, 1996; Schwartz & Dimock, 2001; McElwain & Bullard, 2014).

Comme d'autres Margaritiferidae (Harms, 1907; Howard, 1915; Haas, 1916; Murphy, 1942; Smith, 1980), M. auricularia est une espèce tétragène. Dans les spécimens analysés de M. auricularia, les démibranches externes présentaient un nombre d'embryons considérablement supérieur à celui des démibranches internes. À l'heure actuelle, nous ne pouvons pas déterminer s'il s'agit d'une caractéristique de cette espèce ou bien si les spécimens que nous avons observés n'étaient pas encore complètement gravides avant la fixation

24 A partir du matériel prélevé sur trois spécimens, nous avons estimé la fécondité de M. auricularia à environ 2.000.000 de glochidies par adulte. Bien que cette estimation de la fécondité est élevée par rapport à celle des autres membres d'Unionoida (Haag, 2013), elle est inférieure à celle observée pour les autres margaritiferids. Vraisemblablement, pour des moules de taille similaire, plus de glochidies peuvent être produites avec le même investissement énergétique si les glochidies sont plus petites (Bauer, 1994). La fécondité apparemment inférieure de M. auricularia chez les margaritiféridés pourrait être une conséquence de la taille des glochidies. Margaritifera auricularia produit des glochidies d’une longueur comprise entre 127 et 155 μm (Araujo & Ramos, 1998; cette étude), les plus grandes parmi tous les Margaritiferidae. Le glochidium de M. auricularia, décrit précédemment par Araujo & Ramos (1998) et confirmé dans cette étude, est du type sans crochet et dépourvu de fils larvaires. Il a plutôt des dents minuscules dans la marge de la coquille (Araujo & Ramos, 1998), qui sont probablement utilisées pour se fixer aux branchies des poissons hôtes.

D’après nos observations du développement embryonnaire, réunissant autant les études sur le terrain qu’en laboratoire, le temps de maturation des embryons (du premier clivage jusqu’au stade glochidial mature) dans la Creuse à une température moyenne de 11°C est de 31 à 37 jours. Cette période requise pour la maturation est similaire à celle observée dans plusieurs populations européennes de M. margaritifera (Harms, 1907; Scheder, Gumpinger & Csar, 2011). Par conséquent, chez M. auricularia, la ponte et la fécondation ont probablement lieu du début à la mi-mars dans les populations françaises et un peu plus tôt, entre la fin de janvier au début du mois de février en Espagne, le moment précis dépendant de la température (Wellmann, 1943; Meyers & Millemann, 1977; Young & Williams, 1984; Ross, 1992; Hastie & Young, 2003).

5.3 Identification des espèces de poissons hôtes des populations françaises de M. auricularia

Le déclin général de M. auricularia au cours des périodes historiques semble être étroitement lié à la disparition de l’un de ses poissons hôtes: Acipenser sturio L, 1758 (Altaba, 1990; Araujo & Ramos, 2000, 2001; López et al., 2007). Deux faits appuient cette hypothèse. La répartition historique de A. sturio correspond à l'aire géographique d'origine de M. auricularia (Araujo & Ramos 2000a, c). Les deux espèces ont également connu un déclin considérable de leur population à la fin du dix-neuvième siècle et au début du vingtième siècle. Aujourd'hui, A.

25 sturio est pratiquement éteinte et on ne connaît plus qu'une seule population reproductrice en Europe dans la Garonne (France) (Gesner et al., 2010).

L'autre hôte connu, Salaria fluviatilis (Asso, 1801), a une distribution méditerranéenne et ne correspond pas à celui des populations françaises de M. auricularia. Par conséquent, depuis la disparition d’Acipenser sturio, il n’existe apparemment aucune espèce hôte disponible pour ces populations. Néanmoins, il a été observé un recrutement récent de M. auricularia dans plusieurs localités espagnoles et françaises (Nakamura et al., 2018; Prié et al., 2018). Alors que dans le bassin de l'Èbre, Salaria fluviatilis a pu empêcher l'extinction de cette moule, les populations françaises semblent utiliser une espèce hôte inconnue.

Toutes les espèces hôtes connues de M. auricularia partagent une caractéristique commune: elles ont une relation étroite avec le milieu marin. Les esturgeons sont des poissons anadromes. Salaria fluviatilis, bien que confinée à l'eau douce, peut tolérer des niveaux de salinité élevés et fait partie d'une grande famille d'espèces principalement marines. Par conséquent, nous avons émis l’hypothèse d’une relation similaire pour cette espèce de poisson hôte inconnue. L’identification de cette espèce hôte est d’une grande importance car 1) elle pourrait servir d’outil de conservation, 2) elle pourrait aider à comprendre les causes du déclin de M. auricularia et 3) elle pourrait améliorer les connaissances sur la relation hôte-moule parmi les espèces de Margaritiferidae.

Afin d'identifier ces espèces de hôtes de M. auricularia inconnues en France, une double approche a été suivie, comprenant l'évaluation de l'infestation dans la nature et sur les infestations artificielles en laboratoire.

L’infestation naturelle des communautés de poissons de la Creuse, de la Vienne et de la Charente ont été évaluées à partir de la pêche électrique directement en aval des populations de M. auricularia. Sur la base de la période de reproduction précédemment identifiée de ces populations (voir la section 3.2.1 de cette thèse), les dates de pêche électrique ont été choisies afin de coïncider avec la période d'enkystement des glochidies. Les branchies des poissons étourdis ont été inspectées sous microscope optique pour détecter la présence de glochidies enkystées. Bien que cette approche soit très utile car elle prend en compte les conditions d’infestation dans la nature, la faible abondance de moules ou d’hôtes peut rendre la collecte d’hôtes infectés peu pratique. De plus, l'observation de glochidies enkystées ne fournit que peu

26 de preuves d'un parasitisme réussi, car les glochidies peuvent se détacher avant la fin de la métamorphose.

Par conséquent, nous avons également mené des infestations expérimentales en laboratoire. Pour les infestations artificielles, des spécimens sauvages de plusieurs espèces de poissons hôtes potentiels ont été transportés au laboratoire. Ensuite, ils ont été plongés dans des bains d'infestation contenant des glochidies préalablement recueillies sur des moules gravides des populations de la Creuse, de la Vienne et de la Charente. Avant l'infestation, la viabilité des glochidies a été vérifiée en observant au microscope binoculaire leur réponse active après l'ajout de NaCl à une petite aliquote. Chaque infestation a été réalisée pendant 15 min dans des aquariums ou de petits récipients remplis d’eau de rivière avec différentes concentrations de glochidies et sous agitation constante. Afin de vérifier l'évolution des taux d'enkystement, plusieurs poissons ont été anesthésiés et sacrifiés pendant la période d'enkystement. Les branchies ont été excisées et analysées au microscope binoculaire pour compter des kystes.

Au cours des pêches électriques, nous avons analysé 966 individus appartenant à 29 espèces de poissons. Les glochidies de M. auricularia n'ont été trouvées que chez deux espèces: Gasterosteus aculeatus L, 1758 et Anguilla anguilla L, 1758. Les deux espèces ont un lien avec les environnements marins: les anguilles sont catadromes et Gasterosteus aculeatus habite les eaux saumâtres ainsi que les eaux marines et les populations anadromes sont relativement fréquentes.

Anguilla anguilla a déjà été testé pour l’infestation expérimentale avec glochidies de M. auricularia dans 2 études indépendantes, mais aucune glochidie enkystée n'a été trouvée au- delà d'une semaine après l'infestation (Araujo et al., 2001; López & Altaba, 2005).

En 2017, nous avons testé Gasterosteus acuelatus pour l’infestation artificielle en laboratoire et 27 jours après l'infestation, des juvéniles vivants complètement transformés ont été retrouvés, ce qui indique que l'espèce est un hôte approprié pour les glochidies de M. auricularia. Néanmoins, nous avons observé une perte élevée de glochidies pendant la période d'enkystement. Cela semble être fréquent chez les hôtes universels comme nous considérons G. aculeatus, car il a été signalé comme hôte de 14 espèces de moules unionidées en Europe, en Amérique du Nord et en Asie, et est le poisson-hôte le plus couramment utilisé par les moules européennes d'eau douce (Lopes-Lima et al., 2017). En plus de la forte perte de glochidies au cours de la période d'enkystement, G. aculeatus n'est pas présent dans les rivières de la Creuse

27 et de la Vienne. Par conséquent, nous avons suposé qu'une autre espèce de poisson pourrait également être responsable du maintien des populations françaises de M. auricularia.

Une des espèces anadromes les plus abondantes dans les rivières françaises qui coexiste avec des populations encore vivantes de M. auricularia est la lamproie marine Petromyzon marinus L, 1758. Nous avons testé cette espèce en 2018 et nous avons constaté que les glochidies de M. auricularia se métamorphosaient avec succès sur la lamproie marine. Un total de 13.827 moules juvéniles vivantes ont été recueillies dans un seul P. marinus de 1.200 g. À notre connaissance, il s’agit de la première fois que cette espèce est confirmée comme hôte de toute moule d’eau douce.

Bien que nous n’ayons capturé aucun spécimen de P. marinus lors des pêches électriques, les informations disponibles sur sa répartition géographique, sa phénologie et son habitat suggèrent qu’il pourrait constituer un bon hôte écologique pour M. auricularia. En ce qui concerne sa distribution, il est connu que les rivières étudiées abritent de grandes populations de lamproies marines et que des frayères à Petromyzon marinus ont été observées dans les habitats de M. auricularia. En effet, les lamproies marines sont également présentes dans le reste des rivières où des populations de M. auricularia recrutent encore en France. De plus, la répartition historique de M. auricularia correspond à l'aire de répartition européenne de Petromyzon marinus.

Les lamproies marines adultes ne passent que quelques mois de leur vie dans des rivières, ce qui signifie que le moment de la libération des glochidies est d’une importance cruciale. Dans les rivières françaises, la migration de P. marinus a lieu de décembre à fin juin avec un pic en mars et avril (Taverny & Élie, 2010), période qui correspond à la période de libération des glochidies des populations françaises de M. auricularia (Soler et al., 2018b). Cette période coïncide également avec celle de la migration des adultes de l’esturgeon européen qui a lieu d’Avril à Mai (CEMAGREF, 1994). Ainsi, la période de reproduction de M. auricularia, plus précoce par rapport à d'autres margaritiferidés, peut être liée à l'arrivée d'adultes anadromes sur leurs frayères. Néanmoins, les stades juvéniles des hôtes semblent être les plus fortement infestés avec des glochidies de moules d'eau douce (Karna & Millemann, 1978; Modesto et al., 2018; Young & Williams, 1984). Étant donné le grand nombre d'ammocètes présents pendant la période de reproduction de M. auricularia, on pourrait s'attendre à ce qu'ils soient également infestés par les glochidies.

28 Nos découvertes rapportées dans Prié et al. (2018) suggèrent que M. auricularia est récemment éteint dans les bassins du Rhône et de la Seine mais persiste dans les bassins de l'Èbre, de la Loire, de la Charente, de la Garonne et de l'Adour. La disparition de M. auricularia dans les bassins du Rhône et de la Seine pourrait être liée à la disparition de A. sturio et P. marinus aux dix-neuvième et vingtième siècles. Par ailleurs, les populations de M. auricularia des bassins de la Charente, de la Loire et de l’Adour auraient pu subsister jusqu’à aujourd’hui après l'extinction de A. sturio puisque dans ces bassins subsistent encore des lamproies marines.

La découverte de la lamproie marine en tant qu'espèce hôte de M. auricularia étend le nombre d'hôtes physiologiques à sept espèces appartenant à quatre familles différentes, ce qui suggère que M. auricularia pourrait afficher un comportement plus généraliste par rapport à l'utilisation d'hôtes que prévu auparavant. Des études antérieures suggèrent que les espèces de poissons envahissantes ont moins de possibilités de devenir hôte pour les moules d’eau douce que les espèces indigènes, probablement en raison d’un mécanisme co-évolutif de compatibilité entre les moules et les espèces de poissons hôtes (Taeubert et al., 2012; Douda et al., 2013; Salonen et al., 2016). Selon les données disponibles, la grande majorité des espèces de moules d'eau douce n'utilisent que des hôtes indigènes, tandis que les glochidies de quelques espèces généralistes se métamorphosent avec succès sur des poissons non indigènes (Huber & Geist, 2017; Modesto et al., 2018 ; Teixeira et al., 2018).

Compte tenu de l'ubiquité du poisson exotique Silurus glanis L, 1758 concordante avec celle des populations de M. auricularia en France et en Espagne, nous avons décidé de soumettre cette espèce à des infestations expérimentales en 2018. Vingt-quatre jours après l'infestation, des juvéniles ont été collectés, indiquant que S. glanis est un hôte physiologique des glochidies de M. auricularia. Néanmoins, nous n'avons pas trouvé glochidies de M. auricularia attachées à aucun des 27 spécimens inspectés lors de nos pêches électriques, ce qui suggère que Silurus glanis n'est pas un bon hôte écologique.

Les résultats obtenus avec Silurus glanis semblent corroborer l'idée que M. auricularia a un comportement généraliste par rapport à l'utilisation d'hôtes contrairement aux autres espèces de la famille Margaritiferidae pour lesquelles leurs hôtes sont connus, qui utilisent généralement très peu d'espèces d'une même famille. Etant donné qu'une range restreinte d'espèces hôtes implique généralement un risque d'extinction plus important, nos résultats semblent paradoxaux, étant donné que l'espèce est en danger critique d'extinction.

29 Tous les hôtes de M. auricularia actuellement connus ont une tolérance notable à la salinité, comme c'est le cas de G. aculeatus et P. marinus. Silurus glanis semble s'écarter légèrement de ce schéma d'halotolérance, car il vit principalement en eau douce. Néanmoins, il a été observé qu'il pénètre occasionnellement dans les eaux saumâtres de la mer Baltique, de la Mer Noire et de la Méditerranée et qu'il peut même frayer dans des eaux salées (Berg, 1964; Frimodt, 1995).

Il a été observé que les margaritiferidés utilisent fréquemment des poissons hôtes anadromes (Curole et al., 2004; Araujo et al., 2017): Margaritifera margaritifera, M. Laevis, M. falcata, M. dahurica et M. middendorffi n'utilisent que des salmonidés. (Murphy, 1942; Karna & Milleman, 1978; Kobayashi & Kondo, 2005; Kondo & Kobayashi, 2005; Klishko & Bogan, 2013). Cependant, la découverte récente de poissons hôtes exclusivement d’eau douce de la famille Hiodontidae pour M. monodonta (Sietman et al., 2017) et Esocidae pour M. marrianae et M. hembeli (Fobian et al., 2017; P. Johnson 2018, communication personnel) remet en question cette hypothèse, bien que les tolérances de salinité de ces poissons soient inconnues.

Cette préférence est peut-être apparue parce qu'elle était avantageuse pour la dispersion des moules. Si l'affinité pour les hôtes euryhalins a évolué au cours des premières étapes de leur évolution, il est possible que, puisque les esturgeons et les lamproies étaient présents lorsque les premiers margaritiferidés sont apparus, ils aient été utilisés depuis très longtemps. En outre, cette relation aurait pu être façonnée avant même l’apparition du premier Margaritiferidae. Les moules d'eau douce constituent un groupe ancien issu des bivalves marins du Triasique (Haas, 1969; Watters, 2001). Ainsi, si le parasitisme a évolué pendant la colonisation des milieux d'eau douce, l'unionidé ancestral aurait dû utiliser des hôtes euryhalins. Comme les Margaritiferidae pourraient être considérés comme la famille basale au sein de l'ordre Unionoida (Strayer, 2008), nous émettons l'hypothèse que les Margaritiferidae actuels auraient pu conserver ce trait d'utiliser des hôtes euryhalines d'ancêtres unionidés.

L’esturgeon européen est très difficile à réintroduire dans les rivières car il est pratiquement éteint et S. glanis est une espèce envahissante qui ne devrait pas être introduite dans les rivières où elle n'est pas indigène en raison des impacts écologiques qu'elle peut avoir sur la faune ichtyologique (Doadrio, 2001; Carol, 2007; Syväranta et al., 2009; Guillerault et al., 2017; Cucherousset et al., 2018). Par conséquence, seuls S. fluviatilis, G. aculeatus et P. marinus pourraient être utilisés pour une technique de renforcement des populations de M. auricularia simplifiée et rentable. Cette technique, utilisée avec succès pour d’autres espèces (Altmüller & Dettmer, 2006; Araujo et al., 2015; Carey et al., 2015) consiste à capturer des espèces hôtes à

30 proximité des populations de moules, à les infester avec des glochidies et à les relâcher immédiatement après dans la nature, afin d’augmenter les possibilités de recrutement de cette espèce patrimoniale européenne.

Sur la base de leur distribution géographique, S. fluviatilis peut être utilisé exclusivement dans le bassin de l'Èbre, tandis que G. aculeatus peut être utilisé dans les quelques rivières où il est naturellement présent. Par ailleurs, P. marinus pourrait potentiellement être utilisé sur l’ensemble de l’aire de répartition de M. auricularia, en particulier dans les bassins français où il est encore abondant. En France et en Espagne, P. marinus est considéré comme menacé et, dans certaines zones, il a disparu ou est devenu de plus en plus rare, principalement en raison de la perte de son habitat liée à la construction de barrages, à la perturbation de son milieu naturel et à la surpêche (Mateus et al., 2012; Hansen et al., 2016). Par conséquent, les mesures visant à améliorer la continuité longitudinale des fleuves favorisant l'accès de P. marinus aux zones habitées par M. auricularia devraient être abordées. Ce type de mesures favoriserait non seulement le renforcement des populations de M. auricularia, mais également celles de P. marinus en élargissant leurs habitats de frai disponibles. En outre, Limm & Power (2011) ont signalé que les larves de la lamproie du Pacifique (Petromyzon tridentatus Richardson, 1836) croissaient plus rapidement lorsqu'elles se trouvaient près des lits de Margaritifera falcata (Goul, 1850), où les moules capturent, concentrent et déposent de la nourriture près de leurs terriers. Une relation mutualiste potentiellement similaire entre M. auricularia et P. marinus devrait être étudiée dans de futures études.

6. Améliorer les connaissances sur les premiers stades de la vie de M. auricularia

Après la métamorphose dans les branchies des poissons (Araujo et al., 2002), les juvéniles excystés de M. auricularia doivent tomber dans des sites benthiques appropriés pour survivre. Les juvéniles pénètrent dans les espaces interstitiels des sédiments bien aérés à quelques centimètres de profondeur et y restent plusieurs années jusqu’à ce qu’ils atteignent une taille de 1 ou 2, leur permettant de contrer la traînée de l’écoulement. Cette période juvénile est le stade biologique le plus fragile chez les moules d'eau douce, tant chez les populations sauvages que chez les populations captives (Young & Williams, 1984; Jansen et al., 2001). Au cours de cette étape critique, une transition entre l'alimentation par le pied cilié et une alimentation suspensivore, facilitée par le développement de nouvelles structures anatomiques, doit avoir lieu.

31 Bien que les publications et les programmes de conservation concernant la reproduction en captivité d’autres espèces de naïades aient récemment considérablement augmenté, une mortalité élevée chez les juvéniles est toujours considérée comme un obstacle majeur à l’élevage de moules d'eau douce (Nichols et Garling, 2002; Jones et al., 2005). Il est important de dire que le régime alimentaire des moules juvéniles n'est pas encore connu avec précision. Pour améliorer l'élevage artificiel de ces mollusques, il est essentiel de mieux comprendre la diète optimale et le développement des organes d'alimentation des juvéniles nouvellement émergés.

6.1 Description des modifications anatomiques associées à l'alimentation des juvéniles de moules d'eau douce

L’objective de cette étude était de vérifier l’hypothèse selon laquelle les moules d'eau douce juvéniles subissent une deuxième métamorphose, qui leur permet de passer d'un mode d’alimentation dépositivore par le pied cilié à une alimentation suspensivore, et que, quelle que soit la disponibilité de nourriture, de nombreux juvéniles à ce stade sont incapables de réussir la transition de mode d'alimentation, ce qui entraîne une mortalité élevée.

Bien qu'il ait été initialement prévu d'inclure M. auricularia dans cette étude, cela n'a pas été possible car nous n'avons pas pu obtenir de juvéniles des différents stades de développement requis par l'expérience. Néanmoins, l'étude a été réalisée sur des juvéniles de la très proche M. margaritifera et Unio mancus Lamarck, 1819. Des juvéniles de Margaritifera margaritifera ont été recueillis à partir d'infections de Salmo salar L, 1758, dans une installation d'élevage de moules en Galice (nord-ouest de l'Espagne) avec des glochidies provenant du fleuve Arnego (bassin de Ulla). Les juvéniles de U. mancus ont été recueillis à partir d'infections de Barbus meridionalis Risso, 1827, avec des glochidies du lac Banyoles (Espagne). Les juvéniles des deux espèces ont été nourris avec des algues commerciales. Entre 5 et 10 juvéniles de chaque espèce ont été sacrifiés tous les 10 jours jusqu'au jour 90, puis une fois par mois jusqu'au jour 360. Des échantillons ont été préparés pour la microscopie électronique à balayage afin d'observer le développement des structures anatomiques impliquées dans les activités d'alimentation incluant les branchies, les palpes labiaux et les cils du pied et du manteau.

La longueur moyenne des juvéniles nouvellement émergés était respectivement de 350 et 260 mm pour M. margaritifera et U. mancus. La bouche ciliée de U. mancus était beaucoup plus grande que celle de M. margaritifera; cependant, le pied, la marge du manteau et les branchies

32 étaient très similaires chez les deux espèces à ce moment initial. Les caractères les plus remarquables étaient le pied cilié, avec une rainure ventrale marquée, la bordure ciliée du manteau et la présence d'une branchie préliminaire formée de trois filaments.

Il a été observé que le développement anatomique des structures d’alimentation est le résultat d’une augmentation globale de la taille de l’ plutôt que strictement en fonction de son âge. Pendant la croissance des juvéniles, on a observé le développement des structures anatomiques suivantes impliquées dans l’alimentation.

Le nombre de filaments branchiaux a augmenté pendant la croissance. La réflexion interne des branchies a commencé 130 jours (1 mm) et 60 jours après l'émergence (1,1 mm) chez M. margaritifera et U. mancus respectivement. La rainure ventrale de la branchie interne, utilisée pour transporter les aliments vers la bouche, est formée 210 jours (1,8 mm) et 60 jours après l'émergence (1,1 mm) chez M. margaritifera et U. mancus respectivement. L'apparition des premiers filaments de la branchie externe a été observée 210 jours après l'émergence chez M. margaritifera et 180 jours chez U. mancus, lorsque les juvéniles mesuraient environ 1,8 à 1,9 mm.

Chez U. mancus, des palpes labiaux primordiales avec cils étaient présents à 70 jours, tandis que chez M. margaritifera, ils apparaissaient à 150 jours, lorsque les juvéniles mesuraient environ 1,4 mm. Au cours de la croissance, les palpes labiaux se sont pliés et de nombreux cils ont recouvert les surfaces internes.

Chez M. margaritifera, le pied et la bordure du manteau étaient fortement ciliés au bout de 30 jours, mais ils commençaient à diminuer entre 40 et 80 jours (700 à 800 µm). À 210 jours, les cils et la rainure ventrale du pied étaient plus courts qu’à 100 jours. Chez U. mancus, la bordure du manteau était recouverte de cils à 70 jours, bien que ceux-ci soient déjà plus courts que ceux observés aux stades plus jeunes. À 345 jours, le pied était recouvert de courts cils et à 395 jours (3,8 mm), la rainure ventrale ne restait plus que dans la partie postérieure.

D'après l'observation des spécimens vivants, les juvéniles de M. margaritifera et de U. mancus nouvellement émergés se nourrissent d'abord à l'aide des cils du pied et de la marge du manteau, similaires à ceux observés chez des juvéniles post-métamorphosés de d'autres espèces de moules d'eau douce (Kovitvadhi et al., 2007; Lasee, 1991; Trump, 2010; Uthaiwan et al., 2001). Ces observations suggèrent que l'initiation à l'alimentation suspensivore commence lorsque les cils des parties postérieures des filaments branchiaux internes se rejoignent pour former un trou

33 virtuel pour le courant d'eau. En effet, les filaments branchiaux postérieurs forment une structure sous la forme de « panier » avec une chambre dorsale et une chambre ventrale et avec une ouverture postérieure ciliée. Cette ouverture permet à l'eau d'entrer, ce qui est facilité par les cils du manteau et des branchies, similaire à l'ouverture inhalante (ou siphon) ultérieure chez l'adulte. Dans ce nouveau mode d'alimentation suspensivore, un courant postéro-antérieur opposé à celui utilisé pour l'alimentation par pédale (c'est-à-dire antéro-postérieur) est généré. En outre, la fonction de ce « panier » s’améliore après la formation des palpes labiaux et des rainures ventrales des branchies internes, qui dirigent la nourriture dans la bouche et, à mesure que l’alimentation suspensivore s’améliore, les cils du pied deviennent plus courts et moins nombreux.

Compte tenu des taux élevés de mortalité juvénile signalés pour U. mancus et M. margaritifera dans les habitats naturels (Araujo et al., 2015; Eybe et al., 2015; Hastie & Young, 2003), la mortalité précoce des juvéniles peut être due à l’incapacité de réussir la transition des modes d’alimentation, facilitée par le changement des structures anatomiques. Le passage d’une alimentation dépositivore par le pied cilié à une alimentation suspensivore survient environ 150 à 200 jours après l'émergence chez M. margaritifera et environ 70 jours chez U. mancus, après que les juvéniles aient une longueur supérieure à 1 mm, moment qui coïncide avec celui d'une mortalité élevée. Une fois cette métamorphose alimentaire est terminée, la mortalité juvénile diminue.

6.2 Elevage artificiel des juvéniles de M. auricularia

Différentes stratégies de conservation ont été utilisées afin de protéger les populations de moules d'eau douce. Au moins dans certains cas, le repeuplement des populations avec des juvéniles élevés artificiellement semble être la seule option envisageable pour restaurer des populations gravement décimées. En Europe, la plupart des expériences en matière de reproduction artificielle de moules d'eau douce proviennent d'études sur M. margaritifera. Différentes approches ont été utilisées pour la reproduction artificielle de leurs juvéniles, dont l'intensité de soin des moules varie. Malgré le système d’élevage utilisé, le succès global de l’élevage des juvéniles semble être fortement déterminé par les taux de survie au cours des premiers mois de vie. Il a été observé qu’une fois que les juvéniles atteignent environ 1 mm de long, la mortalité diminue généralement. Bien que l'identification des diètes optimales permettant de maximiser la croissance et la survie au cours de cette étape critique constitue une

34 partie importante des efforts déployés pour élever les juvéniles, les besoins nutritionnels de ces juvéniles restent largement inconnus.

L’objective de ce travail était d'identifier les systèmes d'élevage qui maximisent les taux de survie et de croissance des juvéniles de M. auricularia au cours de leurs premiers mois de vie.

Entre 2015 et 2017, des spécimens gravides de M. auricularia ont été collectés au printemps dans les rivières Charente et Creuse et conservés au laboratoire pour la récolte de glochidies. Pour les infestations induites, des spécimens d'Acipenser baerii Brandt, 1869 achetés dans des fermes piscicoles ont été utilisés. Les poissons infestés ont été maintenus dans des bassins intérieurs et la température contrôlée régulièrement afin de calculer les degrés-jours nécessaires pour l'achèvement de la métamorphose, en suivant des études préalables (Araujo & Ramos, 2000; Araujo et al., 2001, 2002 et 2003). Les juvéniles récemment excystés obtenus chaque année ont été collectés et transférés vers différents systèmes d'élevage sur la base d'études antérieures (Eybe et al., 2013; Araujo et al., 2015): 1) un système semi-naturel extérieur de circulation alimentée constamment par de l’eau naturelle, 2) un système intérieur de recirculation d'eau et 3) des chambres de culture statiques. Plusieurs sources de nourriture ont été testées, notamment les eaux naturelles, le phytoplancton, les détritus, la protéine animale et les bactéries. De plus, l'effet du sédiment a été exploré.

Après infestation, le nombre total de juvéniles libérés par le poisson hôte pour être utilisés dans les différents systèmes au cours des expériences de 2015-2017 était de 38.033. La libération des juvéniles par les poissons hôtes s'est produit entre les jours 30 et 67 après l'infection (IP), en fonction de la température de l'eau (représentant un minimum de 531 et un maximum de 1.071 degrés-jours). En 2016, la période d'excystement était plus longue et retardée et nécessitait plus de degrés-jours, probablement en raison des températures de l'eau plus basses pendant l'enkystement. Ces données suggèrent que la vitesse à laquelle les glochidies se métamorphosent dépend de la température et peut être réduite à des températures plus basses. Araujo, Camara et Ramos (2002) ont déjà noté cet effet. En outre, Araujo, Quirós & Ramos (2003) ont signalé une métamorphose infructueuse lorsque les températures atteignaient 24 °C, ce qui suggère que ce processus est interrompu lorsque les températures sont élevées. Par conséquent, le calcul des degrés-jours en tant que somme de la température moyenne quotidienne peut ne pas refléter ces effets et doit inclure un facteur de correction pour la température. Cependant, des recherches supplémentaires sont nécessaires pour établir les valeurs de température seuil auxquelles la métamorphose est réduite ou interrompue. Ceci est

35 important car une meilleure compréhension des degrés-jours requis pour l'achèvement de la métamorphose est essentielle afin de minimiser le contrôle laborieux de la présence de juvéniles et de la période de jeûne des poissons infestés. En outre, ces informations peuvent aider à optimiser et à standardiser la reproduction artificielle de M. auricularia et à éviter la perte de juvéniles en raison d'une collecte retardée (Hastie & Young 2003; Taubert, Gum & Geist, 2013).

Parmi les systèmes d'élevage testés, à la fois dans le système à écoulement semi-naturel extérieur alimenté en continu par des eaux naturelles et dans le système d'eau à recirculation intérieure utilisés respectivement en 2015 et 2016, aucun juvénile n'a été maintenu en vie pendant plus de 1,5 mois. Les mauvais résultats obtenus en 2015 pourraient être en partie dus à l'utilisation d'eau et de sédiments qui n'étaient pas originaires des habitats de M. auricularia. Néanmoins, en 2016, ces éléments ont été obtenus à partir des mêmes habitats que ceux où se trouve M. auricularia. Par conséquent, d'autres facteurs doivent être identifiés pour améliorer ce système d'élevage. Compte tenu du statut de conservation critique de M. auricularia, ceci devrait être une priorité car cette approche permet de produire un grand nombre de juvéniles avec un investissement en ressources relativement faible (Araujo et al., 2015; Beaume et al., 2016).

Les taux de survie des juvéniles de M. auricularia élevés dans le système de chambres de culture statiques testé en 2017 étaient plus élevés et nous ont permis de formuler d'autres observations. Cette année, les juvéniles ont été transférés dans des boîtes en verre circulaires (14 cm de diamètre) remplies d'eau de la Vienne et stockées à température constante de 18 ° C pendant 110 jours. Les moules étaient nourries avec différents régimes alimentaires trois fois par semaine au moment du changement de l'eau. La nourriture était composée d'algues (A), de détritus (H), d'algues + détritus (S), de jaune d'œuf (E) et de bactéries (B). Chacun de ces traitements consistait en trois répliques et chaque boîte contenait 200 juvéniles. Afin de faciliter le contrôle de la croissance et de la survie des juvéniles, nous n'avons pas ajouté de substrat dans les boîtes. De plus, quatre boîtes de dimensions variables contenant un nombre variable de juvéniles ont été utilisées pour tester un régime alimentaire combiné d'algues et de détritus en présence de sédiment.

Après trois jours du début de l'expérience, la plupart des juvéniles élevés dans le traitement des bactéries sont morts probablement à cause des conditions physico-chimiques inadéquates du milieu de culture. Dans tous les traitements sans sédiment, une mortalité élevée a été observée

36 dans les premières semaines de l'expérience. Les juvéniles présentaient des adhérences organiques dans les coquilles, apparemment une croissance bactérienne jaune-brun, qui empêchait l'ouverture et la fermeture correctes des valves. Cependant, dans le traitement des algues + détritus, dans lequel contrairement aux autres traitements, les boîtes avaient une couche de substrat, ce phénomène n’a pas été observé et les valves des juvéniles étaient complètement propres. En conséquence, du substrat a été ajouté à toutes les boîtes et il a été observé que les juvéniles étaient progressivement plus propres et que la mortalité se stabilisait. Cela indique que la présence de substrat dans les cultures semble être d'une grande importance car elle facilite le nettoyage des valves.

La survie était supérieure dans le traitement algues + détritus suivi du traitement aux détritus, ce qui semble indiquer que les détritus sont d'une grande importance pour le maintien de bonnes conditions physico-chimiques dans le milieu de culture. Les taux de croissance les plus élevés concernaient les traitements dans lesquels les algues étaient utilisées comme source de nourriture (traitements aux algues et aux algues + détritus), ce qui suggère que les algues fournissent une combinaison de nutriments adaptée au développement des juvéniles. Le plus grand individu à la fin de l'expérience a été obtenu dans le traitement combiné (algue + détritus), mesurant 617 µm de long à 105 jours. Néanmoins, dans le traitement E (jaune d'œuf), des taux de croissance comparables à ceux obtenus avec les traitements contenant des algues ont été obtenus jusqu'à 75 jours après l'excystment. Cela suggère que le jaune d'œuf est potentiellement une ressource alimentaire avec laquelle les juvéniles peuvent se développer. Cependant, tous les spécimens sont morts vers le 75e jour; il n'a donc pas été possible de faire d'autres comparaisons à la fin de l'expérience. Enfin, l'association algues + détritus avec des sédiments semble constituer le meilleur des régimes testés, car elle optimise la croissance et la survie des juvéniles.

Dans le système de chambres de culture statiques testé en 2017, le taux de survie maximal après 110 jours était de 34%. Bien que les taux de survie restent faibles, ces résultats semblent indiquer que l'élevage de juvéniles de M. auricularia dans de petites boîtes peut donner de meilleurs résultats. Nakamura et al. (2018) ont signalé des résultats similaires, indiquant qu'après une décennie d'expériences infructueuses dans l'élevage de M. auricularia à Aragón (Espagne), ils ont été en mesure de maintenir en vie les juvéniles au-delà du premier mois lorsqu'ils ont commencé à utiliser des petites boîtes comme système d’élevage. Ils ont obtenu environ 50 juvéniles maintenus en vie parmi les cohortes de 2014, 2015, 2016 et 2017

37 (Nakamura, communication personnelle). Un facteur clé de ces résultats prometteurs semble être lié à la production d’un grand nombre de juvéniles. Entre 2014 et 2017, le nombre moyen de juvéniles de M. auricularia utilisés dans les installations d’élevage d'Aragon était d'environ 425.000 par an (Nakamura et al., 2018). Néanmoins, nos résultats et ceux obtenus à Aragón indiquent que les taux de survie utilisant cette méthodologie sont encore trop bas pour une réintroduction durable et à grande échelle de M. auricularia dans la nature. Par conséquent, des recherches supplémentaires sont nécessaires pour obtenir de meilleurs taux de survie.

7. Remarques finales

Comme cela est reflété dans les sections précédentes, les études incluses dans cette thèse offrent une amélioration de la connaissance de M. auricularia par rapport à divers aspects de sa biologie, de son écologie et de sa biogéographie, ce qui contribuera à aborder la conservation de l'espèce dans une perspective plus large.

D'un point de vue chronologique dans l'ontogenèse de l'espèce, le processus d'incubation des embryons jusqu'à la formation de glochidies et les modifications morphologiques qui se produisent dans les branchies ont été étudiés pour la première fois. D'autre part, la fécondité de deux populations françaises a été vérifiée et une estimation de la fécondité de l'espèce a été proposée pour la première fois. De même, la manière dont les gloquidies sont libérées a été décrite, en examinant leurs implications dans la stratégie d’infestation des poissons hôtes. La période de libération des glochidies, jusqu'ici inconnue dans les populations françaises, a été définie et comparée avec celle des populations espagnoles ce qui a permis de détecter un gradient latitudinal lié à la température de l'eau. De même, de nouvelles espèces de poissons hôtes ont été décrites et leur disponibilité en termes de géographie et de phénologie a été mise en relation avec celle des gloquidies dans la nature. La découverte de ces nouvelles espèces est un outil exceptionnel qui peut être utilisé pour la conservation de l'espèce et, en autre, elle ajoute de nouveaux éléments pour mieux comprendre les causes du déclin de l'espèce. En ce qui concerne le processus de métamorphose des gloquidies dans les branchies des poissons, nous fournissons données sur leur durée sous différents régimes de température, ce qui peut s'avérer très utile pour de futurs programmes de conservation. De même, les modifications morphologiques facilitant une transition du mode d'alimentation aux premiers stades de la croissance des juvéniles, qui semblent être en grande partie responsables de la mortalité élevée qu'elles subissent, ont été étudiées chez deux autres espèces de moules d'eau douce.

38 D'un point de vue écologique et de la conservation, la répartition historique et actuelle de l'espèce a été étudiée et les connaissances sur l'état de conservation des populations actuellement connues à l'échelle mondiale ont été mises à jour. Dans ce sens, une étude approfondie des deux populations françaises les plus importantes a été réalisée, ce qui a permis de décrire plus précisément leur état de conservation du point de vue de leur répartition, de leur abondance et de la structure de leur population, ainsi que de leur évolution temporelle à moyen et court terme. Cette étude a également permis d’améliorer les connaissances sur les exigences écologiques de l’espèce, en montrant que celle-ci tolère une large gamme de salinités et de concentrations de calcium. D'autre part, une nouvelle menace potentielle, celle d'un poisson capable de parasiter les moules d'eau douce. Enfin, les premières expériences d’élevage en captivité de l’espèce en France ont été réalisées; leurs résultats, même s’ils ne permettent pas la réintroduction à grande échelle de l’espèce actuellement, constituent un point de départ important pour la poursuite des travaux.

39

40 Content

Acknowledgments/Remerciements ...... 3 Abstract ...... 5 French Abstract/Résumé ...... 7 Spanish Abstract/Resumen ...... 9 French long abstract/Résumé significatif ...... 11 Content ...... 41 List of contributions ...... 45 1. INTRODUCTION ...... 51 1.1 General aspects of freshwater mussels ...... 52 1.1.1. Terminology and basic anatomy ...... 52 1.1.2. Evolution and global diversity ...... 55 1.1.3. Life-history ...... 57 1.1.4. Conservation ...... 64 1.2. Margaritifera auricularia ...... 67 1.2.1. Systematics ...... 67 1.2.2. Species description ...... 68 1.2.3. Brief history about the knowledge of the species ...... 69 1.2.4. Distribution ...... 70 1.2.5. Populations structure ...... 72 1.2.6. Reproductive biology ...... 74 1.2.7. Conservation ...... 81 2. OBJECTIVES AND METHODOLOGICAL APPROACH ...... 84 2.1. To update and improve the knowledge on the distribution and status of Margaritifera auricularia populations in France ...... 85 2.1.1. To clarify the historical distribution ...... 86 2.1.2. To update the available information on the status of living populations ...... 86 2.1.3. To describe the status of the Charente and Creuse-Vienne populations ...... 86 2.1.4. To increase the information about the threat factors for Margaritifera auricularia in France ...... 88 2.2. To improve the knowledge about the reproductive biology of Margaritifera auricularia ...... 88 2.2.1. Identification of the reproductive period of French populations ...... 89

41 2.2.2. Description of the brooding process in Margaritifera auricularia ...... 90 2.2.3. Identification of host fish species for French populations of Margaritifera auricularia ...... 91 2.3. To improve the knowledge about the early life stages of Margaritifera auricularia ... 92 2.3.1. Juvenile feeding morphology of two freshwater mussel species ...... 92 2.3.2. Artificial rearing of Margaritifera auricularia juveniles ...... 93 3. RESULTS ...... 97 3.1. Status of M. auricularia populations ...... 97 3.1.1. ARTICLE 1: Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe...... 97 3.1.2. SCIENTIFIC REPORT 1: Conservation status of two French Margaritifera auricularia populations...... 119 3.1.3. ARTICLE 2: Rhodeus amarus (Bloch, 1782): a new potential threat for Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae)...... 145 3.2. Reproductive biology ...... 159 3.2.1. ARTICLE 3: Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae)...... 159 3.2.2. ARTICLE 4: Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae)...... 169 3.2.3. ARTICLE 5: An unexpected host for the endangered Giant Freshwater Pearl Mussel Margaritifera auricularia (Spengler, 1793) as a tool against the “native species meltdown” effect...... 175 3.3. Research on the early stages of the European margaritiferids ...... 209 3.3.1. ARTICLE 6: Who wins in the “weaning” process? Juvenile feeding morphology of two freshwater mussel species...... 209 3.3.2. SCIENTIFIC REPORT 2: First experiences of captive breeding of juveniles of Margaritifera auricularia in France...... 225 4. GENERAL DISCUSSION ...... 265 4.1 Current status of the populations ...... 266 4.2. Reproductive biology ...... 268 4.3 Early stages of Margaritifera auricularia ...... 276 4.4. Implications for conservation ...... 281 5. CONCLUSIONS ...... 291

42 6. CONCLUSIONES ...... 295 7. BIBLIOGRAPHY ...... 299 Appendices ...... 351 Appendix 1 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 1 ...... 353 Appendix 2 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 2 ...... 363 Résumé ...... 373 Résumé en anglais ...... 373

43

List of contributions

The following six research articles and two scientific report present the basis of this dissertation and are included in the results section. At the time of the thesis submission, four of the articles have been already published, one has been accepted and the remaining one is in the process of review. The reports include other studies carried out by Joaquín Soler during the thesis and although they are still in preparation, they are presented here since they provide valuable information for the discussion of the topics covered in the thesis. Hereinafter, reference to the original articles is made by citation, i.e. Article 1, Article 2, Article 3, Article 4, Article 5, Article 6 and Scientific Report 1 and Scientific Report 2 as listed below. Although not included in the thesis, Joaquín Soler has actively contributed to the preparation of two publications dedicated to Margaritifera auricularia, which are based largely on the information included in this thesis. However, the index of these publications is included in Appendices 1 and 2.

Research papers:

1. PRIE, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., PATRY, N., ADAM, B., LEGRAND, N., JUGE, P., RICHARD, N., & WANTZEN, K. 2018. Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe. Hydrobiologia, 810, 157–175. 2. SOLER, J., WANTZEN, K. M., & ARAUJO, R. Rhodeus amarus (Bloch, 1782): a new potential threat for Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Freshwater Science. Accepted. 3. SOLER, J., WANTZEN, K. M., JUGE, P., & ARAUJO, R. 2018. Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84, 182–189. 4. SOLER, J., BOISNEAU, C., WANTZEN, K. M., & ARAUJO, R. 2018. Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 1–4. doi:10.1093/mollus/eyy038.

45 5. SOLER, J., BOISNEAU, C., JUGE, P., RICHARD, N., GUEREZ. Y., MORISSEAU, L., WANTZEN, K. M., & ARAUJO, R. An unexpected host for the endangered Giant Freshwater Pearl Mussel Margaritifera auricularia (Spengler, 1793) as a tool against the “native species meltdown” effect. Aquatic Conservation: Marine and Freshwater Ecosystems. Submitted. 6. ARAUJO, R., CAMPOS, M., FEO, C., VARELA, C., SOLER, J. & ONDINA, P. 2018. Who wins in the weaning process? Juvenile feeding morphology of two freshwater mussel species. Journal of Morphology, 279(1): 4–16.

Scientific reports: 1. SOLER, J. Conservation status of two French M. auricularia populations 2. SOLER, J. First experiences of captive breeding of juveniles of M. auricularia in France

Original contribution of Joaquin Soler in the collective works:  ARTICLE 1: Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe.

Joaquín Soler contributed with his work in this collaborative paper. He made the last M. auricularia surveys in the Charente and Vienne-Creuse rivers. He also participated in previous surveys in the Ebro River and updated all the information on the Spanish populations of M. auricularia including all this information in a geographic information system. Joaquín Soler conducted the experiments required to identify Gasterosteus aculeatus as a host fish for M. auricularia, which is detailed in an independent paper.  ARTICLE 2: Rhodeus amarus (Bloch, 1782): a new potential threat for Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

The first author, Joaquín Soler, found the first specimens of M. auricularia with the gills occupied by the eggs of the fish Rhodeus amarus. This finding occurred when he was studying the freshwater mussels in the aquaria for the studies of captive breeding. He wrote all the paper and conducted the study, and made the photograph. He studied all the cited literature. He contributed in the field surveys to collect the fish. All research steps were discussed with the research supervisors: R. Araujo, C. Boisneau and K. M. Wantzen, who also collaborated in the drafting of the manuscript.

46  ARTICLE 3: Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

The first author, Joaquín Soler, was responsible for writing the manuscript, developing the conceptual idea and conducting the study, including the histological and scanning work and the field and laboratory observations. He also made the photographs and identified the knowledge gaps and research questions. He conducted the surveys in the Charente and Vienne-Creuse rivers and revised specimens stored at the Museo Nacional de Ciencias Naturales collection. All research steps were discussed with the research supervisors: R. Araujo and K. M. Wantzen, who also collaborated in the drafting of the manuscript.  ARTICLE 4: Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

The first author, Joaquín Soler, was responsible for writing the manuscript and conducting the study, including all the laboratory work on the required for the natural and artificial infestation assessments. He participated in the electrofishing surveys and the surveys conducted for collecting the mussels. All research steps were discussed with the research supervisors: C. Boisneau, R. Araujo and K. M. Wantzen who also collaborated in the drafting of the manuscript.  ARTICLE 5: An unexpected host for the endangered Giant Freshwater Pearl Mussel Margaritifera auricularia (Spengler, 1793) as a tool against the “native species meltdown” effect.

The first author, Joaquín Soler, wrote all the paper and conducted the study, including all the work made at the laboratory conducted to the infectation of the suitable fish of the species Silurus glanis and Petromyzon marinus with the M. auricularia glochidia. He identified the knowledge gaps and research questions and developed the conceptual idea and analytical framework. He contributed in the field surveys to collect the fish. All research steps were discussed with the research supervisors: C. Boisneau, R. Araujo, and K. M. Wantzen who also collaborated in the drafting of the manuscript.  ARTICLE 6: Who wins in the “weaning” process? Juvenile feeding morphology of two freshwater mussel species.

This paper did not deal with M. auricularia, but this is because there were not enough juveniles of this species for the experiments made here. Joaquín Soler processed many of the samples for ulterior study: relaxation of the specimens with MgCl2, fixation in 2.5% glutaraldehyde, cleaning with PBS buffer and dehydratation through a graded ethanol series. He also made part

47 of the scanning electron microscopy analyses and contributed to the rearing of the juveniles of Unio mancus in the laboratory of Banyoles (Spain). He also contributed to the discussions and the interpretation of the results, especially regarding the morphological development of the gills.

Other contributions:  SOLER, J., ARAUJO R. & WANTZEN K. M. 2018. The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 1 – Synopsis on the current literature. University of Tours, France, CNRS UMR CITERES, LIFE+ project 13BIO/FR/001162 “Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe”. Tours (France) 68 pp.

Joaquín Soler wrote all the paper in collaboration with R. A. and K. M. W. A great part of this Manual is based in the results of the J. Soler Doctoral Thesis.  WANTZEN K. M. & ARAUJO R. (eds) 2018. The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 2 – Technical Manual: Monitoring, artificial reproduction, rearing techniques, and suggestions for habitat conservation. With contributions by Karl M. Wantzen, Rafael Araujo, Joaquin Soler, Catherine Boisneau, Nina Richard, Philippe Jugé, Yann Guerez, Laure Morisseau, Michèle De Monte, Keiko Nakamura and Vincent Prié. University of Tours, France, CNRS UMR CITERES, LIFE+ project 13BIO/FR/001162 “Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe”. Tours (France) 109 pp.

Joaquín Soler contributed with his work in this collaborative paper. The main part of the results obtained with the French populations of M. auricularia in this Manual is based in the results of the J. Soler Doctoral Thesis.

48

Introduction

49 1. INTRODUCTION

This thesis is devoted to the study of Margaritifera auricularia, a naiad or freshwater mussel species that is nowadays facing the extinction. When I tell untrained people the subject of my study, almost invariably a first reaction of surprise arises: "I did not know there were bivalves in the rivers". Most people today associate bivalves exclusively with marine systems and are not aware that bivalves live in freshwater at all. Once the first phase of surprise is over, the first question they ask is, "Can you eat 'em?" When I answer that this species, like the rest of freshwater mussels, does not have a special culinary value, the initial reaction of surprise usually becomes an expression of disappointment and condescension. Although the majority does not dare to say it, people ask themselves why someone should dedicate so much effort to trying to preserve a bivalve that is practically unknown and that, in addition it is not edible?

Most likely, this reaction would not be the same if the study subject had been a big cat or an elephant. At first glance, bivalves seem bored and without great interest. Freshwater mussels are fundamentally filter feeders, so they do not chase or catch their prey. They spend most of their lives buried in the bottom of rivers or lakes moving slowly, if at all, and do not perform striking migrations. The reproduction occurs by males releasing sperm into the water, so there are no spectacular battles between rival males. However, these are able to fascinate those who study them because they keep secrets that only reveal on close and careful inspection.

Freshwater mussels are a conspicuous and important element of aquatic ecosystems. They can form aggregations of more than 100 animals per m2 known as mussel beds where they can dominate the benthic biomass. As filter feeders, they are important in nutrient cycling, influencing the water chemistry and clarity (e.g., Vaughn & Hakenkamp, 2001). Their wastes can enhance local populations of algae (Vaughn et al., 2007) and macroinvertebrates (Vaughn & Spooner, 2006) and their shells provide habitat for other benthic organisms (Spooner & Vaughn, 2006; Vaughn & Spooner, 2006).

Freshwater mussels also provide important direct services to humans, such as water purification. In addition, they have been harvested as a source of pearls, mother-of-pearl, and human food since prehistoric times (e.g., Kunz, 1898, Morrison, 1942, Claassen, 1994; Anthony & Downing, 2001; Walker et al., 2001).

51 Despite having survived for hundreds of millions of years, they are now being decimated globally due to human activities and today they are one of the most threatened animal groups on the planet (Bogan, 1993; Young et al., 2001; Lydeard et al., 2004; Strayer et al., 2004; Strayer, 2008). This is the case of Margaritifera auricularia, which is considered one of the rarest and most endangered bivalve species in Europe (Araujo & Ramos, 2001a; Prié, 2010a).

Besides the practical interest, due to its ecological and economic importance, the study of freshwater mussels is fascinating because of its complex and unique life cycle. In addition, their better understanding is essential to successfully address their conservation. This thesis deals fundamentally with improving the knowledge of certain aspects of the biology of Margaritifera auricularia in order to provide tools for its conservation.

1.1 General aspects of freshwater mussels

1.1.1. Terminology and basic anatomy

Freshwater bivalves of the Order Unionoida Gray 1854 are also known as freshwater mussels, freshwater clams or naiads. In Greek mythology, Naiads were nymphs who inhabited and gave life to fresh waters. Naiads are bivalve mollusks and share the same basic anatomical plan with their marine counterparts. Here I will briefly describe their basic anatomy for a better understanding of the following sections.

Despite their great diversity, the general anatomical pattern of bivalves is very simple and homogeneous. The shell is formed by two valves that encompass the body of the animal and are joined by a proteinaceous ligament located near the umbo or apex. Under the umbo runs the hinge that usually has teeth on its inner face fitting between those of the opposite valve (Figure 1). In its internal part, the shells have muscle marks or scars. The insertion of the adductor muscles, whose contraction allows closing the valves, leaves characteristic rounded or oval marks. The pallial line is the mark formed by the bivalve mantle margin on the inner edge of the valves.

As a rule of basic orientation for Unionoids, it should be taken into account that the anterior region of those bivalves is the one closest to the umbo or apex of the shell, and therefore the posterior region is the opposite. On the other hand, the dorsal part is that in which the umbo, hinge and hinge ligament are located, and the ventral part is the opposite one (Figure 1).

52

Figure 1. External and internal view of the right (upper) and left (below) valves of Margaritifera auricularia (modified from Araujo et al., 2009).

Under the valves, and covered by these, are the mantle lobes, which form a kind of bag including the rest of the animal body (Figure 2). The cells of the outer face of the mantle release the calcium carbonate that makes the shell grow. The posterior edges where the two lobes of the mantle meet has two openings that form the siphonal apertures: one superior (excurrent) and another inferior (incurrent); there is a third opening through which the foot protrudes.

The foot is a very muscular and extensible organ, which allows the animals to bury themselves and move on the substrate. It is, together with the gills, the largest organ of the naiads. In young

53 specimens of some species the foot presents a gland called byssus that forms adherent filaments with which the specimens are attached to the substrate (this byssus is easily noticeable in marine mussels of the genus Mytilus).

Figure 2. Internal anatomy of a freshwater mussel. Arrows show the direction of water within the animal; arrows with circular stems show water flowing through and into the gills; arrows with dashed stems indicate interstitial sediment water entering the mantle/shell gape (modified from Haag, 2012).

The gills are placed between the mantle and the foot, two at each side of the foot (Figure 2). The gills of suspension feeding bivalves are extraordinary organs that are involved in feeding, gas exchange and ion transport although, in some marine and many freshwater species, they are also used to brood larvae during the reproductive period. Gills of freshwater mussels are elongated sheets extending into the mantle cavity and are composed of inner and outer demibranchs. Each demibranch is composed of ascending and descending lamellae, which are organized as filaments surrounding central water channels. The lamellae are joined by connective tissue and in most cases these connections are developed as continuous septa, running parallel to the gill filaments towards the edge of the gill that form the water tubes. However, in the Margaritiferidae there are no water tubes, and the lamellae of each demibranch

54 are intermittently united by irregularly scattered junctions (for further details on gills anatomy see Ortmann, 1911; Cox, 1969; Kays et al., 1990; McElwain & Bullard, 2014).

The gills present numerous cilia and their movement originates the water current that allows the filter-feeding mode of bivalves. The water enters the mantle cavity through the inhalant aperture, passes through the gills until the suprabranchial chamber and exits through the exhalant siphon (Figure 2). The food particles present in the water are retained by the gills and driven by cilia to the anterior area where they are selected by the labial palps that surround the mouth.

1.1.2. Evolution and global diversity

Among the approximately 20,000 living species of bivalves, only about 1,000 species live strictly in fresh waters (Haszprunar et al., 2008). Freshwater representatives occur in most major bivalve groups, indicating that there have been multiple independent bivalve invasions of fresh waters around the world. However, the greatest freshwater radiation has occurred in the order Unionoida, which comprises nearly 85% of the freshwater bivalve diversity (Haag, 2012).

Freshwater mussels (order Unionoida) belong to the bivalve subclass Palaeoheterodonta, together with their marine sister-group Neotrigonia (order ). The Trigoniida is an ancient lineage that was diverse and widespread in the Mesozoic (250-65 million years ago or mya) but is represented today by only six or seven surviving species restricted to marine waters of Tasmania and Australia (Giribet, 2008). Curole & Kocher (2002) based on DNA analyses estimated that the subclass Paleoheterodonta diverged from the rest of at approximately 500 mya (Middle Cambrian). Unionids first appeared in the Triassic (250-200 mya), and by the Cretaceous (145-65 mya) the group attained morphological and taxonomic diversity comparable to the recent fauna (Haas, 1969; Good, 1998; Watters, 2001).

The circumstances that led to the invasion of freshwaters by unionoids are unknown but their success is likely related to two aspects of their life histories: parental care and parasitism. Freshwater mussels retain their eggs in brooding chambers located in their gills, developing obligatory parasite larvae that attach to fish while metamorphosing into juvenile bivalves. This strategy prevents their offspring from being “flushed” downstream, and is a very effective mean of dispersal and colonization of upstream areas of river basins.

55

Figure 3. Distribution map of Margaritiferidae (modified from Lopes-Lima et al., 2018).

Freshwater mussels are represented by approximately 900 species distributed worldwide except in the Antarctica and the Pacific islands (Bogan, 2008). Six families make up the order Unionoida: Unionidae, Margaritiferidae, , , and . By far, the most diverse family is Unionidae, with 707 described species distributed throughout North America, Europe, and Asia (Graf & Cummings, 2007). The closely related family Margaritiferidae is considered an ancestral group of freshwater mussels and apparently diverged from Unionidae at a minimum of 230 mya (Curole & Kocher, 2002). Margaritiferidae present one of the lowest diversity within Unionoida, with only 16 currently recognized species which are found in Asia, Europe, North Africa and North America (Figure 3) (Lopes-Lima et al., 2018). However, their distribution is localized and is apparently a relic of a previous wider range (Smith, 2001). The remaining families of the Unionoida are restricted to only one or two biogeographical regions, and none of these families occur in Europe. Indeed, this continent possess one of the lowest freshwater mussel diversity with only 16 recognized species (Figure 4), 11 of which currently live in France (Lopes-Lima et al., 2017).

56

Figure 4. Shells of representative freshwater mussel taxa in Europe (Lopes-Lima et al., 2017)

1.1.3. Life-history

A generalized depiction of freshwater mussels’ life histories describe this group as long-lived animals, with slow growth, high fecundity and late maturity. This characterization is likely derived from the best-studied freshwater mussel in the world, Margaritifera margaritifera (Haag, 2012). However, naiads are able to adapt to the conditions of nearly all freshwater habitats and this requires adaptability and thus variability of their life-history traits.

1.1.3.1. Growth and longevity Freshwater mussels are famous for the high longevities of some of their representatives. Margaritifera margaritifera may live for about 200 years (Bauer, 1992; Ziuganov et al., 2001) and is considered to have one of the longest life spans reported for any noncolonial animal (Hurlbert et al., 2007). However, the longevity of freshwater mussels vary widely among and within species and even M. margaritifera may live less than 40 years in some places (San Miguel et al., 2004). Other naiad species live more than 50 years, although the majority of them commonly live between 15 and 40 years (Haag, 2012).

57 Life span is a life-history trait with potentially important ecological costs and benefits, and therefore should be under strong selective pressure (Bauer, 1992). It has been observed that long-lived freshwater mussel species such as Margaritifera margaritifera, have lower growth rates than short-lived species such as Anodonta and Unio species (Bauer, 1992; Aldridge, 1999; Hochwald, 2001). It is often assumed that growing faster is better, presenting a series of benefits that include earlier age at maturity and lower risk of predation at smaller sizes (Arendt, 1997; Metcalfe & Monaghan, 2003). However, most organisms do not grow as fast as they can, even when conditions are favorable (Metcalfe & Monaghan 2003). This is because growing faster implies costs, reducing the fitness of adults and leading to lower life spans (Bauer, 1992; Arendt, 1997; Metcalfe & Monaghan, 2003; Rose et al., 2009).

Freshwater mussels grow along their whole life span but the rate at which growth is accomplished varies along their lives. Like most organisms, their growth is usually fast during the first few years of life slowing down with age, as energetic resources are diverted to reproduction and maintenance. However, the amount of energy allocated to growth versus other functions and the timing of this allocation may vary among species (Haag, 2012).

In all bivalves, the shell is secreted by the mantle and seasonal variation in shell deposition produces rings which are observable in both the external surface and cross sections of the shell (Figures 1 and 5). These annual rings provide a detailed growth record similar to those found in trees or in fish otoliths and scales and have been utilized for age and growth estimation in bivalves (e.g., Schöne et al., 2004; Haag & Commens-Carson, 2008).

Figure 5. Growth rings observed in a cross section of Margaritifera auricularia shell treated with Mutvei’s solution to get a better contrast of the growth rings. One liter of Mutvei's solution consists of 500 ml 1% acetic acid, 500 ml 25% glutaraldehyde and ca. 5 to 10 g alcian blue powder (Schöne et al., 2005).

1.1.3.2. Reproductive strategies The reproductive cycle of naiads is truly unique and one of the keys to their success colonization of fresh waters but, in some cases, it may be responsible for the vulnerable situation that most

58 species are currently experiencing. All freshwater mussels share a complex life cycle, requiring a vertebrate host, usually a fish, during their parasitic larval stage. The cycle starts when females inhale the sperm that was released by males in the water, fertilizing the eggs that are located in brooding chambers formed by the gills. The eggs develop into parasitic larvae (in most families called glochidia) which are released in the water by the female. These microscopic thin-shelled larvae were presumed in the 19th century to be a parasitic species of fish (Glochidium parasiticum). The glochidium usually has hooks to attach itself to the fish’s body (fins or gills), where it becomes encapsulated for a variable period (between weeks and months), before metamorphosing into a free-living juvenile. Then, juveniles fall to the river substratum, grow to adult size and start another cycle (Figure 6).

Figure 6. Basic life cycle of a unionoid mussel (adapted from Reis, 2006)

However, numerous variations to this general cycle have been reported among freshwater mussel diversity. Although most species are dioecious, some are simultaneous hermaphrodites and occasional hermaphrodite individuals are found in several predominantly dioecious species (Schalie, 1970; Kat 1983). Bauer (1987) reported that Margaritifera margaritifera individuals are able to change from dioecious to hermaphrodite in response to decreasing population density.

The number of eggs that an individual mussel can produce is usually correlated with its length (Bauer, 1998; Hochwald, 2001; McIvor & Aldridge, 2007). Since fecundity increases with the animal size, this allometric constraint is probably a factor contributing to higher growth rates. In addition, for similarly sized mussels, more glochidia can be produced with the same energetic investment if the glochidia are smaller (Bauer, 1994). The brooding pattern is also variable

59 among naiads. The eggs can be brooded in all four demibranchs (tetrageny), only in the inner ones (endobranchy) or in the outer demibranchs (ectobranchy) (Heard, 1998). Furthermore, some species such as the European Anodonta anatina usually broods in the outer demibranchs although tetragenous populations have been observed (Lopes-Lima et al., 2016). Variations in brooding patterns seem to be intrinsic and species-specific, and have been used as an important character in the study of unionoid evolution and systematics (Graf & Cummings, 2006). Freshwater mussels have different strategies in relation to the length of the brooding period, being either short-term (tachytictic) or longterm (bradytictic) brooders. Tachytictic species spawn their gametes in spring and brood the embryos only until glochidia have fully developed, whereas bradytictic mussels spawn in late summer and brood their glochidia over winter, releasing them in early spring (Graf & Ó Foighil, 2000). Climate (Sterki, 1903; Ortmann, 1909; Graf, 1997) and synchronization with seasonal host activity (Zale & Neves, 1982; Young & Williams, 1984a) are considered important factors that determine brooding patterns.

Finally, the way in which larvae are released is probably one of the most extraordinary features of unionoid evolution. Many strategies have been developed by freshwater mussels in order to maximize the probability of their larvae to find an appropriate host. Margaritifera margaritifera females broadcast millions of larvae, and a typical healthy population may contain hundreds of thousands of mussels, so that the number of glochidia in the water during the reproductive period can be extremely elevated (Bauer, 1991). The behavior of Unio crassus constitutes an unusual modification of simple broadcast: gravid females move into shallow waters and spurt a stream of water with glochidia that spatters on the surface and, presumably, attracts host fish that are looking for falling insects (Vicentini, 2005). However, the most spectacular host attracting strategies are known from North American species of the subfamily Lampsilinae, which have developed mantle modifications that resemble the shape of a fish and are used as lures (Figure 7) (Haag et al., 1999; Barnhart et al., 2008; Haag, 2012). Other species “pack” their larvae in well-visible structures resembling invertebrates that are actively taken up by the fish.

60

Figure 7. Infestation strategy by Lampsilis reeveiana. The fish approaches (A) and bites the lure (B), then abruptly opens its mouth and expands its buccal cavity to inhale the ‘‘minnow’’ (C). Suction created by the fish ruptures the marsupium and extracts a cloud of glochidia (C, D) (Adapted after Barnhart et al., 2008).

1.1.3.3. Parasitic stage Once the parent mussel releases the glochidia, the availability of a suitable host becomes critical. Since glochidia lack storage tissues, they die within several days if unable to attach to an appropriate host (Murphy, 1942; Hochwald, 1988; Jansen et al., 2001). Although some species have developed highly specialized infestation strategies that maximizes infestation, the tremendous mortality rates of glochidia found in most species makes this the most critical stage in the freshwater mussels life cycle (Jansen et al., 2001).

After attachment, the glochidium becomes encapsulated by the host tissue, forming a cyst in which it metamorphoses into the free-living juvenile stage (Figure 8). The parasitic stage of unionoid mussels performs several functions, including protection, nutrition and dispersal. Most freshwater mussel species parasitize fish (Watters & O'Dee, 1998), which carry the glochidia upstream and potentially far from their progenitors. This is particularly important since adult mussels are almost sessile, and because of the unidirectional nature of the flow in rivers and streams.

61

Figure 8. Encysted glochidia of Margaritifera auricularia in Acipenser baerii gill filaments.

Once glochidia are released by the mussels, theoretically any kind of fish could be infested by them. However, fishes develop immunologic responses that glochidia must be able to overcome in order to be physiologically compatible with their host (Bauer, 1997; Bauer & Wächtler, 2001; Haag, 2012). In addition, even adequate fish species can develop an immunological memory, i.e. previously glochidia-infected fish cannot be successfully infected any more (Kat, 1984; Barnhart et al., 2008). Regardless of the exact cause of parasite rejection and death, the few available quantitative data show that even on known host fish with no or little previous exposure, glochidial mortality during the parasitic stage may differ between mussel species and can be very high (Jansen, et al., 2001). Moreover, not all physiological hosts may be infested under natural conditions since an ecological compatibility between the mussel and the host is also required. Ecological hosts need to be physically available in terms of glochidia-host geographic co-occurrence and phenological match (i.e., synchronous occurrence of the fish and the glochidia within the same habitat) (Levine et al., 2012).

While the glochidia of most freshwater mussels develop successfully until the juvenile stage only in a limited number of host species, some generalist mussel species are capable of metamorphosing on numerous hosts (Jansen et al., 2001; Barnhart et al., 2008; Strayer, 2008). This is a particularly important characteristic of species because it can determine their degree of susceptibility to changes in ecosystems: a narrow range of host species usually implies a larger risk of extinction (Strayer, 2008).

62 Although the underlying basis for host specificity is largely unknown, it has been proposed that the preference for hosts has a heritable component resulting from co-evolutionary processes (Cummings & Mayer, 1993; Roe et al., 1997; Graf & Cummings, 2016). Thus, a better knowledge of host-mussel relationships could be useful for improving our understanding of the evolutionary history of unionoids.

Nowadays this knowledge is urgent given the worldwide worrying conservation status of most freshwater mussels (Strayer et al., 2004; Lopes-Lima et al., 2017). The identification of host species is essential for conservation programs focused on the reinforcement of populations with low or absent recruitment by means of the reintroduction of juvenile mussels into the wild (Brady & Gatenby, 2018).

1.1.3.4. Post-parasitic stage Once the metamorphosis is completed, excysted juveniles must be dropped from their hosts in suitable benthos sites to survive. Juveniles penetrate the interstitial spaces of well-aerated sediment some centimeters deep and remain there for several years until they grow to a size of 1 or 2 cm that permit them to counter the drag of the flow. In most species, benthic life starts before the gills are fully developed, and thus filter-feeding is not possible at the first stages of life. The active post parasitic food uptake occurs by water flow mainly caused by the cilia of their highly developed foot. Pedal-feeding by juvenile mussels may extend to 140 days after the metamorphosis (Gatenby et al., 1997), although it depends on the growth rate of the concerned species. The adult mode of feeding is gradually adopted as differentiation of the gills increases (Wachtler et al., 2001).

Despite reaching this stage, juveniles are still fragile and suffer high mortality. Young & Williams (1984b) calculated mortality rates of 95% per year for juvenile Margaritifera margaritifera in Scotland. The high mortality during juvenile stage is believed to result from the mussels falling into unfavorable habitats, predation or caused by the inability to successfully adapt to the transition of feeding modes (Bauer, 2001; Wachtler et al., 2001).

As a result of the high mortalities experienced during their early life stages, less than 1% of the individuals reach the adult stage (Jansen et al., 2001). Despite the very low survival of their offspring, freshwater mussel used to be found in extremely dense populations, suggesting that reproductive rate did not limit size of local populations. It has been proposed that naiads have evolved as organisms that are generally long-lived and highly fecund which confers them a very

63 high life-time fecundity, in order to compensate the early life stages high mortalities (Jansen et al., 2001).

1.1.4. Conservation

Freshwater mussels are amongst the most threatened invertebrates in the world (Bogan, 1993; Araujo & Ramos, 2000a; Young et al., 2001; Lydeard et al., 2004; Strayer et al., 2004; Strayer, 2008). In North America, 74% of the known 300 species are threatened and 37 species are considered extinct (Williams et al., 1993; Lydeard et al., 2004). In Europe, of the 16 species recognized, 3 are critically endangered, 2 are vulnerable, and 5 are near threatened (Lopes-Lima et al., 2017). Indeed, both European Margaritifera species (M. margaritifera and M. auricularia) are critically endangered and listed as “fauna requiring special measures to be taken for their protection” under the provision of the Bern Convention on the Conservation of European Wildlife and Natural Habitats. Although the conservation of freshwater mussels in other parts of the world is less well known, the situation seems to be similar (see a review in Patterson et al., 2018).

Many countries or regions have created their own legal instruments to protect freshwater mussels species and habitats. In Europe, the Habitats Directive (Council Directive 93/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora) was created to this purpose, and freshwater mussel species such as Margaritifera auricularia and Unio crassus are included in Annex IV (Animal and plant species of community interest in need of strict protection), whereas M. margaritifera, U. mancus (= U. elongatulus) and Microcondylaea bonelli (= M. compressa) are listed in Annex V (Animal and plant species of community interest whose taking in the wild and exploitation may be subject to management measures).

Nevertheless, mussels are not the only imperiled group of animals in freshwater ecosystems. Nearly 12,000 species of freshwater invertebrates are either extinct or imperiled globally (Strayer, 2006). Of the approximately 600 species of crayfish worldwide, 32% are in danger of extinction (Richman et al., 2015). Although only about 5,800 of the 15,570 described species of freshwater fish (37%) had been assessed, 30% of them are listed as extinct, extinct in the wild or threatened with extinction (Carrizo et al., 2013). These statistics clearly show that freshwater ecosystems and the species that inhabit them are being globally decimated. In fact, extinction rates in freshwater ecosystems have been estimated to be five times higher than in terrestrial systems (Ricciardi & Rasmussen, 1999).

64 Multiple factors derived of human activities have been identified as responsible for the worldwide decline of freshwater mussels, including habitat loss, pollution, host species decline, invasive species and overexploitation (Bauer, 1988; Bogan, 1993; Neves, 1999; Araujo & Ramos, 2000a; Young et al., 2001; Lydeard et al., 2004; Strayer, 2006). These impacts may reduce the distribution range of species, leading, according to Strayer (2008), to 1) a diminished role of freshwater mussel species in local communities and ecosystems; 2) loss of genetic diversity; 3) increased distance and reduced dispersal among the remaining populations, which may lead to further losses of populations due to metapopulation dynamics; and 4) increased risk of extinction for species.

Habitat destruction and degradation due to dam construction has been identified as one of the most important causes for freshwater mussels decline (Bogan, 1993; Layzer et al., 1993; Neves 1999; Vaughn & Taylor, 1999; Lydeard et al., 2004; Strayer, 2006). There are currently more than 45,000 large dams and one million smaller dams on the world’s streams and rivers (Jackson et al., 2001, Malmqvist & Rundle, 2002). Dams create upstream lentic conditions that are inadequate for lotic species, significantly modify the hydrology of the river downstream of the dam, and constitute a barrier for species dispersal along the river (Layzer et al., 1993; Neves, 1999; Vaughn & Taylor, 1999; Strayer, 2006; Araujo & Álvarez-Cobelas, 2016). Other significant causes of habitat loss are sedimentation, channelization and dredging (Bogan, 1993; Lydeard et al., 2004; Strayer, 2006). Sedimentation often results from other habitat modifications such as dam construction, as well as of poor agricultural practices. It causes reduced oxygen concentrations in the interstitial substrate water, which is fundamental for the survival and development of juvenile mussels, and therefore, is an important factor leading to recruitment failure (Buddensiek et al., 1993; Gatenby et al., 1997). Channelization modifies drastically the habitat and dredging is known to have caused the elimination of whole mussel beds (Killeen et al., 1998; Aldridge, 2000).

Water quality and pollution have become major problems to freshwater mussels (Bauer, 1988; Bogan, 1993; Neves, 1999, Lydeard et al., 2004; Strayer, 2006). Pollution was considered the main cause of decline of Margaritifera margaritifera in Europe (Bauer, 1988). Strayer & Malcom (2012) found that concentrations of un-ionized ammonia greater than 0.2 mg N/L in interstitial water were correlated with recruitment failure in southeastern New York rivers.

The disappearance of glochidia natural hosts disrupts the reproductive cycle of naiads, so that recruitment of juveniles is no longer possible. Several studies have shown the importance of

65 the availability of appropriate host fishes during the reproductive period for the dispersal and population status of unionoid mussels (Bauer, et al., 1991; Watters, 1992; Bogan, 1993; Vaughn & Taylor, 1999; Österling, et al., 2008). When host fish disappear, mussel populations decline and finally become extinct (Kat & Davis, 1984; Watters, 1995), whereas reintroduction of host fish species may result in recovery of previously disappeared mussel populations (Smith, 1985). Since identification of suitable hosts is a key element for freshwater mussel conservation, many studies have examined to this issue (e.g., Lefevre & Curtis, 1912; Stern & Felder, 1978; Zale & Neves, 1982; Berrie & Boize, 1985; Haag & Warren Jr., 1997; O'Dee & Watters, 1998; Araujo et al., 2001).

Introduction of invasive species is a relatively recent threat to freshwater mussels, but has the potential to become one of the major concerns for their conservation (Bogan 1993; Lydeard et al. 2004). Invasive bivalve species such as the zebra mussel (Dreissena polymorpha) and the Asian clam (Corbicula fluminea) have been reported to cause the decline or extinction of native freshwater mussel populations (Bogan, 1993; Parker et al., 1998; Ricciardi et al., 1998; Neves, 1999; Yeager et al. 1999; Burlakova et al. 2000; Lydeard et al. 2004; Strayer, 2006). They compete with native mussels for resources and can drastically modify the water quality and substrate. Introduced fish species may also be an important threat to freshwater mussels by significantly contributing to the disappearance of native fish faunas (Ribeiro et al. 2009), which are essential as hosts for unionoids.

In some areas of the world, overexploitation has been a major cause of decline of some freshwater mussel species with commercial interest. However, in most cases the associated industries have decreased in importance or are now extinct. The freshwater pearl mussel Margaritifera margaritifera has been heavily harvested in Europe for freshwater pearls, but that activity is now extinct (Young et al., 2001). In the United States, mussels from the Mississippi drainage support a multi-million dollar commercial shell industry for the manufacture of buttons (Neves, 1999) although this activity is now regulated and managed to prevent overexploitation.

The high level of imperilment in global freshwater mussel populations combined with their important ecological function is causing great concern among scientists, prompting the creation of freshwater mussels conservation programs around the world. Artificial propagation has become a key conservation management strategy for restoration and recovery of freshwater mussels. Large numbers of juveniles can be produced in captivity using identified host fishes

66 and then reintroduced to the wild. The release of juveniles raised to larger size is desirable because of the high natural mortality of newly metamorphosed juveniles (Jones et al 2006; Araujo et al., 2018). Because of the difficulty of culturing juveniles, the release of native host fishes infected with glochidia has been shown to be an economic alternative that has been successfully applied in different conservation programs (Altmüller & Dettmer, 2006; Araujo et al., 2015; Carey et al., 2015). The action plans of many species already include artificial propagation as necessary for their recovery (e.g., Araujo & Ramos, 2001a). Nevertheless, these strategies should be regarded as temporal measures while the natural recruitment is recovered by habitat conservation measures (Gum et al., 2011; Patterson et al., 2018).

1.2. Margaritifera auricularia

1.2.1. Systematics

Margaritifera auricularia was first described by Spengler (1793) as Unio auricularius. Nevertheless, Spengler erroneously cited the locality of the type specimen from East India. Lamarck (1819) redescribed the same species with European specimens from French rivers as Unio sinuata. Haas (1910) introduced the subgenus Pseudunio to differentiate this species from Margaritifera margaritifera, but one year later, Pseudunio was synonymized as Margaritana (=Margaritifera) by Ortmann (1911). Pseudunio was later used by Haas (1969) in order to separate Margaritifera auricularia and Margaritifera marocana (Pallary, 1918) from the rest of species belonging to the genus Margaritifera.

Since the beginning of this century, several phylogenetic studies have presented the family Margaritiferidae as monophyletic divided into three to four major clades although keeping Margaritifera as a single genus (Huff et al., 2004; Graf and Cummings, 2007; Araujo et al., 2017). Nevertheless, other authors have pointed out the necessity of splitting the family in several subgenera, including Pseudunio (Bolotov et al., 2016; Lopes-Lima et al., 2018). Since a consensus regarding the nomenclature has not still been reached, I will use the genus name Margaritifera Schumacher, 1816, rather than Pseudunio Haas, 1910.

67 Phylum Cuvier, 1795

Class Bivalvia Linnaeus, 1758

Subclass Palaeoheterodonta Newell, 1965

Order Unionoida Stoliczka, 1870

Superfamily Unionoidea Rafinesque, 1820

Family Margaritiferidae Henderson, 1929

Genus Margaritifera Schumacher, 1816

Species: Margaritifera auricularia (Spengler, 1793)

Synonyms:

Unio margaritifera (Draparnaud, 1801)

Unio sinuatus (Rossmassler, 1844)

Unio crassissimus (Ferrussac, 1844)

Pseudunio auricularius (Spengler, 1913)

Unio sinuata Lamarck, 1819

Unio margaritanopsis Locard, 1889

Common names:

French: Grande mulette.

English: Giant freshwater pearl mussel, Spengler’s freshwater mussel.

Spanish: Almeja perlífera gigante de río, Perla de río, Margaritona, Náyade auriculada.

1.2.2. Species description

As indicated by its common name, the Giant freshwater pearl mussel is a very big naiad attaining a maximum length of 20 cm. The shell has a black periostracum and flattened umbones (Figure 1). The valves are elongated posteriorly and shortened and very thick in the anterior

68 part. Growth lines are visible in the external part of the shell, although in aged specimens they become densely packed near its edge, making it very difficult to count them for age estimations. In adults, the shell border is usually sinuated giving an ear shape to the shell, which has inspired its Latin specific name “auricularia”. The internal part of the valves are white due to the presence of a thick layer of nacre. Valves have well marked adductor muscle scars and pallial line. The hinge of left valve has two cardinal pyramidal teeth and two long lateral teeth behind it. The hinge of right valve present one small cardinal teeth and a lateral one that fits between the two lateral teeth of the opposite valve.

Although the anatomy of Margaritifera auricularia has not been well studied, it shares the common pattern of the naiads (Figure 2). Nevertheless, as in the rest of margaritiferids, it is characterized by some anatomical features different from the Unionidae (Araujo et al., 2017; Lopes-Lima et al., 2018): 1) the inhalant siphon papillae are small and branched; 2) a supra- anal aperture to the exhalant siphon is lacking; 3) the diaphragm dividing the infra and suprabranchial cavities is incomplete, and 4) in the gills, the lamellae of each demibranch are intermittently united by irregularly scattered junctions, rather than by continuous septa.

1.2.3. Brief history about the knowledge of the species

As evidenced by the finding of shell remains of Margaritifera auricularia in Chalcolithic archaeological sites, the existence of the species has been known by humans from prehistoric times (Araujo & Moreno, 1999; Llorente et al., 2015). However, despite the improvements achieved in recent decades, the scientific knowledge of the species is still relatively scarce specially if compared with the well-studied Margaritifera margaritifera.

From its description by Spengler in 1793 until the beginning of the 20th century, the literature devoted to Margaritifera auricularia was restricted to taxonomic descriptions and records on its distribution. However, some citations of the species in the early literature were misidentifications of M. margaritifera (Valledor & Araujo, 2006). A remarkable source for the early scientific knowledge on M. auricularia comes from the famous malacologist Fritz Haas. The beginning of the First World War surprised him during a field survey in the south of France where he was arrested accused of being a spy. Fortunately, the French authorities allowed him to leave to the neutral Spain. There, he published a couple of studies (Haas, 1916, 1917) describing an important population of the species and summarizing the knowledge that was held about the species until that moment.

69 "Until after the first half of the nineteenth century, it was known only in France. It was found there in many rivers, in some of which it was so frequent that nacre was industrially exploited for buttons and similar objects. Its anatomy and habits are not known and only some data on the color of the soft parts and its marked predilection for the deepest places of the rivers is available. Then, around the year 80, news of its appearance in some Italian rivers and in the Ebro and Tagus reached the ears of zoologists. Data on the current distribution of the species were completed in the last five years with its fossil discovery in West and Central Germany and in England” (Haas, 1916).

After Haas (1916, 1917), the Spanish naturalist Azpeitia (1933) cited the species in the Canal Imperial de Aragón, which was the last reference of living specimens in Spain. In France, the last documented finding of a living animal before dates from 1952, from a small tributary of the Loire River between Blois and Tours (Moolenbeek, 2000).

After a long period without new references, some dead specimens of the species were first rediscovered in Spain in 1985 (Altaba, 1990) and in France in 1989 (Nesemann & Nagel, 1989). The first living individuals were rediscovered in 1996 in Spain (Araujo & Ramos, 1996) and in 1998 in France (Nienhuis, 2003). Since 1998, the biology, distribution and life cycle of the Giant Freshwater Pearl Mussel in Spain have been described (Araujo & Ramos, 1998a, b; 2000a, b; Araujo et al., 2000, 2001, 2002, 2003; Grande et al., 2001; Gómez & Araujo, 2008). Since then, very few reports of the species have been released in national and international congresses (e.g., Nakamura et al., 2015), but except for papers by Araujo & Álvarez-Cobelas (2016) and Nakamura et al. (2018), there are no new scientific results published since 2008. In France, focused surveys have led to the rediscovery of many populations since 2007, but most of these results remains unpublished or are only available as grey literature.

1.2.4. Distribution

The Giant Freshwater Pearl Mussel is a species difficult to observe: it lives downstream in large rivers, a habitat that is difficult to survey due to deepness, turbidity, current and often navigation. Hence, not surprisingly, it has been overlooked by malacologists of the 20th century.

70 Margaritifera auricularia was once present on most major rivers in Western Europe. In the 19th century, the species was recorded on many French rivers from the Seine, Loire, Garonne, Adour and Rhone basins (Prié et al., 2011). On the other hand, the species was known on the Rhine in Germany, on the Tagus in Spain (Araujo & Ramos, 2000b) and in the Po basin in Italy until the 19th century (Bourguignat, 1883; Araujo & Ramos, 2001b). It is only known in the Thames in England from fossil specimens (Preece et al., 1983) (Figure 9).

Figure 9. Historical (black dots) and current (shells = grey dots; living populations = red dots) distribution of Margaritifera auricularia (from Prié et al., 2011)

The last surviving populations are currently found only in France and Spain. Until the beginning of this thesis, the known distribution of the species was synthetized by Prié et al. (2011). In Spain, the Giant Pearl Mussel is only present in the Ebro basin, from the Zaragoza region to the Ebro delta (Altaba, 1997, 2001; Araujo & Ramos, 1998b, 2000b; Araujo et al., 2000; Gómez & Araujo, 2008) (Figure 9). In France, the known living populations are located in the Vienne and Creuse (Cochet, 2001a, b, 2002), in the Charente (Nienhuis, 2003; Prié et al., 2008a), the Dronne, the Adour, the Luy and the Save (Prié et al., 2010). Empty but fresh shells have been found in the Oise (Prié et al., 2007, 2008b, which consider it as recently extinct in the Oise), the Louts (Cochet, 2004), in the Vesle, the Aisne, the Loire, the Dordogne, the Aveyron, the Isle (Prié et al., 2010). Fragments of shells were also harvested in the Indre by Nagel, Cochet and Dohogne (2008) and at several places in the Dordogne-Garonne basins (Prié & Bousquet, 2010) (Figure 9).

71 1.2.5. Populations structure

In Spain, the size of the living populations is estimated to around 5,000 specimens in the Canal Imperial, 80 in the Canal de Tauste and 70 individuals in the Lower Ebro (Araujo & Ramos, 2001a; Gómez & Araujo, 2008; Araujo & Álvarez-Cobelas, 2016). In France, Cochet (2001a) estimated that the population of the Vienne and Creuse rivers should be of about 1,000 individuals. Prié et al. (2010) reported an estimation of a few hundred individuals between the Dronne and Adour rivers and about 100,000 in the Charente (Prié, 2010b). Following these estimations, about 90% of the world population is found in France, containing most of the living populations and the largest one known in the Charente.

Araujo & Ramos (2000a), in a study comprising the measurement of 438 individuals, reported that the Canal Imperial population was dominated by individuals between 15 and 17 cm and classes less than 12 cm were non-existent, indicating a recruitment failure (Figure 10). Nevertheless, a few young specimens of about 5-10 cm have been recently found in this population (Nakamura et al., 2018). In the Vienne River, Nienhuis (2003) found that the 19 living individuals were between 11 and 16 cm. Cochet (2004) reported that 80% of 292 individuals measured were over 13 cm but found 9 individuals smaller than 11 cm, suggesting a recruitment in the Vienne in the relatively recent past. Furthermore, four juveniles (7-8 to 8- 10 cm) were discovered in 2009 in this same river (Prié et al., 2011). In the Oise River (Seine Basin), Prié et al. (2008b) harvested and measured 163 shells, showing that the former population in this river was ageing (over-representation of adult stages, particularly the size class between 13 and 14 cm). In the Charente River, Nienhuis (2003) reported that the majority of the 42 living animals measured were between 10 and 15 cm, except for 2 smaller individuals, measuring 7-8 cm. From the same river, Prié (2008a) reported that the dominant size classes of living individuals measured were between 11 and 12 cm.

72

Figure 10. Size frequencies structure of the Margaritifera auricularia population on the Ebro Basin (from Araujo & Ramos, 2000a)

Size classes reflect age classes. However, the age estimations of Margaritifera auricularia are almost inexistent until the beginning of this thesis. Altaba (pers. communication in Prié et al., 2011) proposed that age-specific growth of the Spanish individuals of this species follows a hyperbolic equation:

size = (169.0 × age) / (age + 4.86)

Altaba (in Prié et al., 2011) estimated that individuals are sexually mature at a length of 16 cm, which would correspond to about 50 years. Following this equation, Prié et al. (2011) estimated that the largest known individuals in the Ebro should be 159 years old, for a length of 20 cm.

On the other hand, Nienhuis (2003), counting the shell rings on the periostracum, estimated that the larger animals that he found in the Charente (14.4 cm) had a presumed age of 43 years. Moreover, he estimated that most animals in this river with a length of 11-12 cm had an age of about 29 years.

More recently, Nakamura et al. (2018) described the growth pattern of the species in the Ebro basin and found that the generalized von Bertalanffy and Richards models were the best fitted for the studied populations (Figure 11). They found that preadults with a median shell length of 9.3 cm corresponded to ages from 11 to 14 years. Adult’s shells with an estimated age of 38- 50 years measured in most cases between 14.4 and 15.3 cm.

73

Figure 11. Fitted curves of the generalized von Bertalanffy growth model for three populations of Margaritifera auricularia from three Ebro basin (from Nakamura et al., 2018)

According to these size classes, it seems that the currently known populations in Spain and France are senescent and the recruitment is very scarce or nonexistent. It has been proposed that this non-recruitment situation could be explained by a change in the environmental quality (especially water quality), which could be detrimental for juveniles and / or by the extinction of their host fish (Araujo & Ramos, 2000a; Prié et al., 2011; Araujo & Álvarez-Cobelas, 2016).

1.2.6. Reproductive biology

In the only population where the sex ratio has been studied (Canal Imperial de Aragón), Margaritifera auricularia is mainly hermaphrodite, with only some individuals being dioecious (Grande et al., 2001). Female and male gonads are mixed within the visceral mass without a specific location (Figure 12). In this population, the gametogenesis occurs from December to March (Grande et al., 2001), but the period when the eggs are transferred to the gills and the duration of brooding are still unknown. Furthermore, the number of larvae harbored in the gills of a gravid mussel and their distribution in the demibranchs have not been determined.

74

Figure 12. General aspect of the gonad of Margaritifera auricularia showing mixed female (left) and male (right) tissues (R. Araujo, unpublished picture)

Previous knowledge of the reproductive season of Margaritifera auricularia is based on Ebro basin populations. Haas (1917) reported that in the Ebro River, the species does not incubate between mid-July and early September. Results from a drift-net study by Araujo et al. (2000) indicated that in the Canal Imperial, M. auricularia has only one reproductive period per year. In this population, gravid individuals have been found in February and the release of glochidia has been reported to take place between February and March-April, with a peak in mid-March (Araujo et al., 2000; Araujo & Ramos, 2001b). Nevertheless, the release period of glochidia in French populations is still unknown. Furthermore, the way in which glochidia are released, which further determine the host infestation strategy, has not been described for M. auricularia.

The glochidium of Margaritifera auricularia was described by Araujo & Ramos (1998a) as the biggest of the family Margaritiferidae (length: 127-144 μm; height: 120-142 μm; width: 54-71 μm). It is of the hookless type, but presents many minute teeth at the ventral border, that allow successful attaching only to fish gill filaments (Figure 13).

75

Figure 13. Glochidium of Margaritifera auricularia observed under scanning electron microscopy (Araujo & Ramos, 2001b)

Like most of freshwater mussels, Margaritifera auricularia depend on fish for hosting the metamorphosis of their glochidia. Although tested in 22 fish species (Table 1), complete metamorphosis of glochidia into juveniles has only been proven to occur in four sturgeon species (Acipenser sturio, A. naccarii, A. ruthenus and A. baerii) and the river blenny (Salaria fluviatilis) (Figures 14, 15; Araujo & Ramos, 1998b, 2000b; Araujo et al., 2001, 2003; Altaba & López, 2001; López & Altaba, 2005; López et al., 2007; Nakamura et al., 2012). All these host species have been identified by experimental infestations conducted under laboratory conditions. Araujo et al. (2001) conducted electrofishing surveys in the Canal Imperial but they did not find any naturally infested fish.

Table 1. Previous results of artificial infestation in Margaritifera auricularia. ? = not reported value.

Duration of Number of Species encystment in Metamorphosis References specimens days Araujo & Ramos, 5 30 Yes Acipenser baerii 2000a 4 49 ? López & Altaba, 2005 Acipenser naccarii 9 30 Yes Araujo et al., 2003 Acipenser sturio 2 19-39 Yes López et al., 2007 Acipenser ruthenus ? ? Yes Nakamura et al., 2012 Salaria fluviatilis 2 42-45 Yes Araujo et al., 2001

76 Duration of Number of Species encystment in Metamorphosis References specimens days 6 31 Yes Araujo et al., 2003 11 26-40 Yes López & Altaba, 2005 Achondrostoma arcasii 1 3 No Araujo et al., 2001 Alburnus alburnus 10 ? No López & Altaba, 2005 3 4 No Araujo et al., 2001 Anguila anguila 8 9 No López & Altaba, 2005 Barbus haasi 2 ? No López & Altaba, 2005 Barbus haasi and 12 6 No Araujo et al., 2001 Luciobarbus graellsii Carassius auratus 10 6 No López & Altaba, 2005 Chelon auratus 9 ? No López & Altaba, 2005 Cobitis paludica 1 0 No López & Altaba, 2005 Cyprinus carpio 7 0 No López & Altaba, 2005 45 15 ? López & Altaba, 2005 Gambusia holbrooki 9 8 ? Araujo et al., 2003 Gobio gobio 4 0 No López & Altaba, 2005 Luciobarbus graellsii 35 2 No López & Altaba, 2005 Mugil cephalus 4 ? No López & Altaba, 2005 Parachondrostoma 20 6 No López & Altaba, 2005 toxostoma Scardinius erythrophthalmus 5 0 No López & Altaba, 2005 Tinca tinca 1 ? No López & Altaba, 2005

The European sturgeon, Acipenser sturio, is considered the primary natural host of Margaritifera auricularia. This assumption is based on the following facts: 1) the historic distribution of A. sturio matches with the original geographic range of M. auricularia (Araujo & Ramos, 2000b, 2001b), 2) both species occur together in Pleistocene deposits (Preece et al., 1983; Preece, 1988) and 3) both species have suffered dramatic population declines in the late 19th and early 20th centuries. Nowadays, A. sturio is virtually extinct and only one remaining reproductive population in Europe is known in the Garonne River (France) (Gesner et al., 2010). On the other hand, neither A. naccarii, nor A. ruthenus or A. baerii co-occur with current M. auricularia populations, and the distribution of S. fluviatilis only overlaps with that of the mussel in the Ebro basin.

77

Figure 14. Margaritifera auricularia releasing glochidia and developing embryos near an Acipenser baerii specimen (R. Araujo, unpublished picture)

Figure 15. Salaria fluviatilis on a specimen of Margaritifera auricularia (R. Araujo, unpublished picture)

Thus, the scarce but relatively recent recruitment observed in several French populations could not be explained by the presence of the previously identified host species. Therefore, French populations of Margaritifera auricularia appear to use an unknown host fish species.

The duration of the encystment period of Margaritifera auricularia glochidia in the fish gills varies depending on temperature. It usually takes 4 weeks (30 days at 23-24°C), although longer periods have been reported with lower temperatures (50 days at 18-22°C or even 65 days at 16- 17°C) (Araujo & Ramos, 2000a; Araujo et al., 2002).

78 The released juveniles have spheroidal shells (Figure 16F) with a finely ciliated foot, and their mean measurements are: length = 190 μm (n = 1), width (from the umbo to the ventral border) = 193 μm (n = 4), and height = 210 μm (n = 4). The growth rate of M. auricularia glochidia during metamorphosis (length, 41%; width, 53%, and height, 238%) is the smallest reported for the genus (Araujo et al., 2002).

Figure 16. Development of Margaritifera auricularia glochidia observed by scanning electron microscopy. A and B: Glochidium; C: Encysted glochidia on the gills of Acipenser baeri 4 hours after infestation; D and E: Completed encystment; F: Juvenile (adapted from Araujo & Ramos, 1998a; Araujo et al., 2002).

The only peer-reviewed published study on the growth of Margaritifera auricularia juveniles in captivity is that of Araujo et al. (2003). The authors tested the survival capacity of about 2,500 recently released juveniles in a nutrient laden media recreated in tanks filled with well and green water, sediment from the Canal Imperial, soil and vegetation from the Ebro River. Juveniles were kept alive for at least 4 weeks, although the highest mortalities were observed during the first two weeks. In the most successful tank, 13 juveniles were found after 39 days of culture, with maximum dimensions of 325 µm in length and 350 µm in height (Figures 17 and 18).

79

Figure 17. Mean Margaritifera auricularia juvenile shell size during a 6-week culture (from Araujo et al., 2003)

Figure 18. Juvenile shells of Margaritifera auricularia after release (left) and at 6 weeks old (from Araujo et al., 2003). Scale bar: 40 μm.

More recently, Nakamura et al. (2015) presented a poster at an international meeting with new results on the artificial rearing of juveniles of Margaritifera auricularia. These authors tested different feeding treatments culturing the juveniles in plastic boxes. A survival rate of 20-70% was achieved at 120 days by feeding the juveniles with phytoplankton and detritus. When the experiment time was doubled to 30 weeks, the survival rate declined dramatically but some individuals were kept alive for one year reaching a maximal length of 3.5 mm. In 2017, in a study focused on modelling the growth of the species, these authors reported that the mean length of 5 juveniles 60 weeks old was 3.23 mm (Nakamura et al. 2018).

80 1.2.7. Conservation

Margaritifera auricularia is considered one of the rarest and the most threatened bivalve species in Europe (Araujo & Ramos, 2001; Prié, 2010a). The population decline in the recent past has been estimated to be over 90% (Prié, 2010a) and today it is nearly extinct, with only a few remaining and ageing populations in Spain and France.

 Although there are important knowledge gaps that prevent accurately identifying the causes of its decline, it has been proposed that these are manifold and similar to those of other endangered freshwater mussel species. The following are the main threats that have been related to the current critical situation of Margaritifera auricularia: The near disappearance of its primary host, Acipenser sturio (Altaba, 1990; Araujo & Ramos, 2000a; López et al., 2007) in Europe and the rarefaction of Salaria fluviatilis in the Ebro basin (Araujo & Álvarez-Cobelas, 2016).  The physical degradation of watercourses. In Spain, direct impacts on the populations of M. auricularia have been reported due to alterations in the channels where the species is present, mainly due to the dredging and paving of their beds (Gomez & Araujo, 2005). In addition, in the lower part of the Ebro River, local extinctions have been related to flow water diversion for hydroelectric power plants (Araujo & Álvarez-Cobelas, 2016). Siltation due to human activities such as intensive agriculture and dams has been also identified as an important threat (Araujo & Ramos, 2001a; Prié et al., 2011). Furthermore, dams constitute obstacles preventing the free movement of migratory fish and altogether with overfishing and water pollution, are related to the decline of diadromous fish species such as Acipenser sturio in Europe (Limburg & Waldman, 2009; Mateus et al., 2012).  Overharvesting from the 18th to the beginning of the 20th century. Despite its common name, M. auricularia does not produce pearls. However, it is the thickness of its mother- of-pearl which is very striking and for which it has been exploited commercially. In the low sections of the Charente and Ebro rivers, extraordinarily high quantities of specimens were extracted for the manufacture of buttons and knife handles (Bonnemère, 1901; Álvarez-Halcón, 1998; Araujo & Álvarez-Cobelas, 2016).  Introduction of invasive species. The effects of the Asian Clam (Corbicula fluminea) on M. auricularia populations have not been well studied but it could be of major concern given the filtering capacities of those small bivalves. On the other hand, the Zebra

81 Mussel (Dreissena polymorpha) seem to be harmful by fixing on the valves of the Giant Pearl Mussel, hindering the movement of their valves (Araujo, 2006).  Water pollution has also been identified as an important threat factor, although tolerances to pollution at the adult stage and especially at the larval and juvenile stages have not been determined by ecotoxicological studies.

Globally, Margaritifera auricularia is assessed as Critically Endangered by the UICN (Prié, 2010a). The species is also listed on Annex IV of the EEC Habitats Directive, which includes animal and plant species of European interest requiring strict protection, and on Annex II of the Bern Convention, which includes animal species requiring special protection. At a national level, the species is protected by Spanish and French laws and both countries have elaborated National Strategies and Action Plans for its conservation (MARM, 2009; Prié et al., 2011) following the recommendations proposed in the European Action Plan for Margaritifera auricularia (Araujo & Ramos, 2001a).

As a consequence of this status of protection, it is strictly forbidden to disturb or sample juveniles or adults of the species, or to modify their habitats, without previous permission. The manipulation involving the death of animals is of course forbidden and the number of individuals that can be manipulated is usually very small. Permits need to be demanded at the regional authorities for environmental conservation, and may be very time-consuming, as the responsible evaluation teams often do not decide ad hoc, but wait for regular meetings. This has important consequences for the study of a species with which it is already difficult to work due to the deep, fast-flowing and turbid nature of the habitats in which it is found, which in many cases requires the use of scuba diving techniques for its location.

Nevertheless, despite the difficulties and given its critical situation, several conservation programs have been devoted to the species. Among these, the LIFE projects stand out for their importance. LIFE is the European Union’s financial instrument supporting environmental, nature conservation and climate actions projects throughout the EU. In Spain, two LIFE projects have dealt with the conservation of Margaritifera auricularia: the LIFE04 NAT/ES/000033 project leaded by the Regional Government of Catalonia between 2001 and 2004 and the LIFE04 NAT/ES/000033 project leaded by the Regional Government of Aragón between 2004 and 2007. After the end of the last project, the Regional Government of Aragón have maintained the financial support and nowadays this is the longest conservation program for M. auricularia with more than 13 years of experience. This conservation program have devoted great efforts

82 to captive breeding of the species, but more than a decade of attempts have been unsuccessful because juveniles died after the first weeks of life. However, as indicated in the previous section, Nakamura et al (2015) reported that some live juveniles remained alive after a year.

In France, the LIFE+ project “LIFE13 BIO/FR/001162: Conservation of the Giant Pearl Mussel in Europe” is the most comprehensive conservational effort deployed on the species up to date. Leaded by the Université Francois Rabelais and the Conseil Départemental Charente-Maritime, it was started in 2014. The overall goal of this project is the conservation and the reinforcement of populations of this critically endangered species in France and more widely in Europe.

83 2. OBJECTIVES AND METHODOLOGICAL APPROACH

As described in the previous section, Margaritifera auricularia is one of the most endangered freshwater mussels species in Europe. Although likely widespread in most western European rivers at the beginning of the 20th century, nowadays only a few remaining populations of this species persist in Spain and France. The almost lack of young mussels indicates that reproduction is strongly reduced in these populations, representing the most critical problem for the survival of the species.

While the smaller pearl mussel species known from European headwater streams, Margaritifera margaritifera, has received considerable attention in the past years, M. auricularia is facing the threat of extinction without receiving adequate scientific support. Although knowledge of M. auricularia has improved considerably in the last two decades, there are still important gaps that hinder its effective conservation. Furthermore, although about 90% of the living populations occur in French rivers, most of the biological knowledge of this species comes only from studies of Spanish populations.

The main objective of the research work presented in this thesis is to provide the necessary knowledge for the implementation of Margaritifera auricularia conservation programs in France. Therefore, the work presented here focuses mainly on the study of French populations, most of them reduced in number of specimens, which makes them more vulnerable. Therefore, most of the works described in this thesis have been carried out in the largest populations in France, which are located in two different river basins, the Loire (Vienne and Creuse rivers) and the Charente (Figure 19).

Given the limited knowledge about French populations and the variety of factors that threatens the survival of Margaritifera auricularia, this thesis includes the study of very diverse issues. The objectives pursued in this thesis, as well as the methodological approach used to achieve them, are described below.

84

Figure 19. Populations of M. auricularia studied in two different river basins, the Charente and the Loire.

2.1. To update and improve the knowledge on the distribution and status of Margaritifera auricularia populations in France

The conservation status of a species is an indicator of how likely is to become extinct in the near future. Different factors are used in order to assess the conservation status of a species, including the number of remaining individuals, the overall trends in the populations over time, the breeding success rates and the known threats. This information is essential in order to stablish adequate preservation measures.

The aim of this section is to improve the current knowledge on the French distribution and status of the species in order to serve as the basis for the execution of further conservation measures. Furthermore, the approaches will make it possible to better target the ecological requirements of the species and to evaluate more finely its geographic reduction during the last

85 century, which further may facilitate the investigations on the causes of its decline. The specific objectives pursued are described in the next sections.

2.1.1. To clarify the historical distribution

Although the ancient texts showed that Margaritifera auricularia was widespread in Europe, the precise distribution of this species before the current collapse of the populations remains poorly known, especially in France. To overcome this lack of knowledge about the biogeography of the species, an exhaustive bibliographic and museum collections review was performed on French populations.

The results of this research are included in ARTICLE 1 (Section 3.1.1).

2.1.2. To update the available information on the status of living populations

Since the rediscovery of the current French populations, numerous focused field surveys have been performed in France and Spain, but most of these results are unpublished or available only as grey literature. A review of all the grey literature related to Margaritifera auricularia in France and Spain was accomplished, summarizing the information on the current distribution of the species, the population sizes and their reproductive status.

Margaritifera auricularia generally lives in downstream ecosystems. Surveying this habitat is difficult because it is often deep, turbid, strongly flowing and navigable. In these cases, surveys are performed by scuba diving. In more accessible populations, snorkelling or wading with viewing glasses allow efficient surveys. Margaritifera auricularia populations have been estimated using different methods, including exhaustive counts of observed living individuals, statistical analyses and subjective appreciation based on the density of specimens observed.

The results of this research are included in ARTICLE 1 (Section 3.1.1).

2.1.3. To describe the status of the Charente and Creuse-Vienne populations

Detailed data on the size and structure of populations are required to design conservation measures and to provide base-line data for further investigations to reveal changes in the studied population. This information is limited for most Margaritifera auricularia French populations. The population structure in the Charente and Vienne-Creuse rivers has been analyzed previously, but it was based on reduced sample sizes or was only estimated from shells of dead

86 individuals, which may offer an inaccurate interpretation of the actual situation. Furthermore, the evolution of these populations has been little studied since its rediscovery and there is scarce information to support the selection of an appropriate methodology.

The aim of this study was to 1) to assess basic ecological characteristics of both populations such as distribution, abundance and population structure, 2) to compare our results with previous studies when possible in order to detect temporal changes, 3) to describe the displacement capability of the species and, 4) to discuss the suitability of different methodologic approaches for future studies focused on detecting temporal changes in Margaritifera auricularia populations.

The Charente and Vienne-Creuse populations were chosen for this study since they are the largest ones in France and because there are base-line data for comparison with our results. In order to establish these comparisons, we followed the sampling methods established in the former studies. In the Vienne-Creuse, 14 stations distributed along a 54 km long stretch were sampled using bathiscopes and by snorkeling. In the Charente River, 12 stations distributed along a 25 km long stretch of the river were studied. At each station, 20 m long transects were studied by scuba diving.

The characterization of the structure of the populations was based on biometric measurements. Shell length of dead shells and living M. auricularia individuals was measured to the nearest 0.1 mm with Vernier calipers. To infer the population structure and detect possible recent recruitment, a length–frequency distribution of 10 mm intervals was used.

In order to identify the biometrically measured specimens, and to permit future tracking actions, a small numbered tag was glued on one of the valves. Additionally, in the Vienne-Creuse population, the precise location of marked specimens was recorded using a centimeter-accurate GPS. One year later, georeferenced specimens were searched again in order to assess the mortality rate and the horizontal displacement.

The results of this research are included in SCIENTIFIC REPORT 1 (Section 3.1.2).

87 2.1.4. To increase the information about the threat factors for Margaritifera auricularia in France

Multiple factors have been identified as responsible for the worldwide decline of M. auricularia, including habitat destruction, commercial overharvesting, water and sediment pollution and host fish loss due to exotic fishes and changes of the natural flow and flood regime by dam construction. Introduction of invasive species is a relatively recent threat to freshwater mussels, but has the potential to become one of the major concerns for their conservation. The effects of invasive bivalve species such as the zebra mussel, Dreissena polymorpha (Pallas, 1771), and the Asian clam, Corbicula fluminea (O. F. Múller, 1774), on Margaritifera auricularia populations have not been well studied but it could be of major concern given the filtering capacities of those small bivalves. Introduced fish species may also be an important threat to freshwater mussels by significantly contributing to the disappearance of native fish faunas, which are essential as hosts for unionoids.

The objective of this study was to alert the scientific community and biodiversity managers of a new potential threat to Margaritifera auricularia, derived from the recent colonization of their habitats by an exotic fish species. The presence of the European bitterling, Rhodeus amarus (Bloch, 1782), was detected during electrofishing surveys that were carried out to identify host fish of M. auricularia in the Vienne, Creuse and Charente rivers. This cyprinid is known for its unusual life cycle characterized by its obligatory symbiosis of spawning in freshwater mussels, which may be detrimental for their hosts. Field and laboratory observations permitted to verify for the first time that R. amarus uses M. auricularia as a host of their eggs and embryos. This finding was contextualized by means of an extensive literature review and its implications for the conservation of endangered freshwater mussel species is discussed.

The results of this research are included in ARTICLE 2 (Section 3.1.3).

2.2. To improve the knowledge about the reproductive biology of Margaritifera auricularia

Given that one of the most important problems affecting the conservation of the species is the lack of reproduction in natural populations, expanding the knowledge on this topic was one of the main objectives of this thesis. The reproductive biology of French populations has never been studied and basic questions have not yet been addressed. Is the current lack of recruitment

88 due to infertility of the mussel specimens? If they are still fertile, which is the reproductive season and what abiotic variables may determine it? What host fish species are used? Is the host availability limiting the reproduction? To what extent is the species overall decline related to the host availability? Answering to these questions is the main goal of this section as detailed in the following specific objectives.

2.2.1. Identification of the reproductive period of French populations

Although the reproductive period of the Spanish populations has been previously identified, this information is still unknown for French populations. This basic knowledge is essential in order to establish future conservation measures. Collecting gravid females at the right time of the year is essential to carry out the corresponding captive breeding programs. Collection should occur when the brooded larvae are mature in order to avoid large maintenance periods of adult mussels in the laboratory. In addition, the larvae of short-term brooders (tachytictic) species (as supposed for all margaritiferids) are generally mature for a very short period of time, so missing the adequate moment can cause the failure of one year of the breeding program. Furthermore, the timing of glochidial release is of crucial importance in relation to host-fish availability and may give cues to identify the host species that are being used in the wild.

Based on the available information on the reproductive season of the species in Spain, where gravid individuals were found in February and the glochidia release occurred between February and March-April, the Margaritifera auricularia populations from the Creuse and Charente rivers were studied between March and May of the years 2015-2017. Specimens were collected by scuba diving and regularly inspected for gravidity and the stage of embryos development, both in the field and in the laboratory. Field assessments of gravidity were done by observing samples of the gills content obtained with the aid of a disposable syringe under an optical microscope. Gravid mussels were transported to the laboratory and maintained in aquaria filled with aerated river water. Larval material released by each mussel was inspected daily under a binocular microscope and the dates of glochidia emission recorded. The identified period of glochidia release of each population was compared with that of the Spanish populations in order to check for differences due to latitudinal gradients related to water temperature.

The results of this research are included in ARTICLE 3 (Section 3.2.1).

89 2.2.2. Description of the brooding process in Margaritifera auricularia

Although previous studies have greatly improved the knowledge on the reproductive biology of Margaritifera auricularia, except for the brief description provided by Haas in the beginning of the 20th century, no studies have characterized the anatomy of the gills and changes in the marsupium during brooding, or embryo development times and brood size in this species. This information is basic to better understand other biological traits and is valuable for taxonomic studies.

Observations of the gills of Margaritifera auricularia were made using specimens collected at the Canal Imperial (Ebro Basin, Spain) and preserved in the malacological collection of the Museo Nacional de Ciencias Naturales (MNCN) in Madrid. Due to the rarity of this species, we did not sacrifice any additional specimens for this study. Three specimens (two gravid and one non-gravid) were processed for histological analyses. Portions of the central part of the outer and inner demibranchs were excised, gradually dehydrated through an ethanol series, embedded in Paraplast® and serially sectioned (5–10 μm) with a microtome. The slides were then stained with haematoxylin–eosin and Heidenhain’s AZAN and observed under an optical microscope mounted with a digital camera.

In order to estimate the duration of the brooding process and the number of larvae harbored by a single mussel, individuals from the Creuse and Charente populations were collected by scuba diving during the identified reproduction period. The duration of egg maturation of embryonic and larval development was estimated by counting the number of days between the first observation of undivided eggs and the presence of mature glochidia in two specimens from the River Creuse. To estimate the total number of glochidia harbored by a single mussel, we collected all larval material individually released by three specimens from both populations. We diluted and homogenized this material in a known volume of river water and then counted the number of glochidia in three 1 ml samples in a Sedgewick Rafter counting chamber for each specimen.

The results of this research are included in ARTICLE 3 (Section 3.2.1).

90 2.2.3. Identification of host fish species for French populations of Margaritifera auricularia

The current knowledge on the Margaritifera auricularia host fish cannot explain the recruitment observed recently in the French populations, suggesting that these populations are using an unknown host species. The identification of this host species is of great importance since 1) it could be used as a conservation tool, 2) it may help in understanding the causes of the decline of M. auricularia, and 3) it may improve the knowledge on the host-mussel relationship within the Margaritiferidae species.

A double approach was followed, including the assessment of natural infestation in the wild and laboratory studies on artificial infestations.

The fish communities of the Creuse, Vienne and Charente rivers were assessed for natural infestation by electrofishing directly downstream of Margaritifera auricularia populations. Based on the previously identified reproductive period of these populations, the electrofishing dates were chosen in order to coincide with the glochidial encystment period. The gills of the stunned fish were inspected for encysted glochidia under optical microscopes. A total of 966 specimens belonging to 29 fish species were captured.

For artificial infestations, wild specimens of several potential host fish species were transported to the laboratory. Then, they were immersed in infestation baths containing glochidia previously collected from gravid mussels from the Creuse, Vienne and Charente populations. Prior to infestation, glochidia were checked for viability by observing their active response after the addition of NaCl to a small aliquot under a binocular microscope. Each infestation was carried out for 15 min in aquaria or small containers filled with river water and different glochidial concentrations under constant agitation.

In order to check the evolution of the encystment rates, several fish individuals were anesthetized and sacrificed during the encystment period. The gills were excised and analyzed under a binocular microscope for cyst counting. In order to analyze the role of identified host species on the overall decline of Margaritifera auricularia, a literature review on the historic distributions of the fish was performed and compared with the geographic evolution of M. auricularia distribution.

The results of this research are included in ARTICLES 4 and 5 (Section 3.2.2 and 3.2.3).

91 2.3. To improve the knowledge about the early life stages of Margaritifera auricularia

After the metamorphosis in the fish gills, the excysted juveniles of Margaritifera auricularia must be dropped from their hosts in suitable benthos sites in order to survive. Juveniles penetrate the interstitial spaces of well-aerated sediment some centimeters deep and remain there for several years until they grow to a size of 1 or 2 cm that permit them to counter the drag of the flow. This juvenile period is the most fragile life stage in freshwater mussels, both in wild and captive populations. In the course of this critical stage, a transition between pedal to filter feeding must be achieved facilitated by the development of new anatomical structures. Although recently there has been a considerable increase in publications and conservation programs concerning the captive breeding for propagation of other naiad species, high juvenile mortality is still considered the bottleneck in freshwater mussel rearing. It is important to say that the diet of the juvenile mussels is still not precisely known. To improve artificial rearing of these mollusks, a better understanding of the optimal diet and the development of the feeding organs of newly emerged juveniles is crucial.

2.3.1. Juvenile feeding morphology of two freshwater mussel species

The aim of this study was to test the hypothesis that freshwater mussels juveniles experience a second metamorphosis enabling them to change from a deposit feeding by the ciliated foot to a suspension feeding, and that regardless of the availability of food, many juveniles at this stage are unable to successfully transition feeding modes, thus resulting in high mortality.

Although it was initially planned to include Margaritifera auricularia in this study, it was not possible since we could not obtain juveniles of the different development stages required by the experiment. Nevertheless, the study was performed with juveniles of the closely related Margaritifera margaritifera and Unio mancus. Margaritifera margaritifera juveniles were collected from infections of 0+ Salmo salar Linnaeus, 1758, in a mussel rearing facility in Galicia (NW Spain) with glochidia specimens from the Arnego River (Ulla basin). The juveniles of Unio mancus were collected from infections of Barbus meridionalis Risso, 1827, with glochidia specimens from Banyoles Lake (Spain). Juveniles of both species were fed with commercial algae. Between 5 and 10 juveniles of each species were sacrificed every 10 days until day 90, and then once a month until day 360. Specimens were prepared for scanning

92 electron microscopy in order to observe the development of the anatomical features implicated in feeding activities including the gills, labial palps and the cilia of the foot and mantle.

The results of this research are included in ARTICLE 6 (Section 3.3.1).

2.3.2. Artificial rearing of Margaritifera auricularia juveniles

Different conservation strategies have been employed in order to protect freshwater mussel populations. At least in some cases, restocking populations with artificially reared juveniles seems to be the only feasible option for restoring severely depleted populations. In Europe, most of the experience on freshwater mussel propagation comes from studies on Margaritifera margaritifera, and different approaches have been used for the artificial breeding of their juveniles differing in the intensity of mussel care required. Despite the rearing system employed, the overall success in the juvenile culturing seems to be strongly determined by the survival rates during the first months of life. It has been observed that once the juveniles reach approximately 1 mm in length, mortality usually decreases. Although an important part of the efforts to rear young juveniles is the identification of optimal diets that maximizes growth and survival during this critical stage, the nutritional requirements remain largely unknown.

The aim of this work was to identify rearing systems that maximize the survival and growth rates of Margaritifera auricularia juveniles during their first months of life.

Between 2015 and 2017, gravid Margaritifera auricularia specimens were collected in spring from the Charente and Creuse rivers and maintained in the laboratory for glochidia harvesting as described in Objectives 2.2.1 and 2.2.3. For induced infestations, specimens of Acipenser baerii Brandt, 1869 purchased from fish farms were used. Infested fish were maintained in indoor tanks and the temperature was regularly monitored in order to calculate the degrees/days required to complete the metamorphosis, which was based on previous studies. Recently excysted juveniles obtained each year were collected and transferred to different rearing systems: 1) an outdoor semi-natural flow-through system supplied by wild water, 2) an indoor recirculating water system and 3) static culture chambers. Several food sources were tested, including wild water, phytoplankton, detritus, animal protein and bacteria. In addition, the effect of the sediment granulometry was explored.

The results of this research are included in SCIENTIFIC REPORT 2 (Section 3.3.2).

93

94

Results

95

96 3. RESULTS

3.1. Status of M. auricularia populations

3.1.1. ARTICLE 1: Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe.

Résumé La grande mulette (Margaritifera auricularia), espèce en danger critique d'extinction, a été présumé éteint avant sa redécouverte en Espagne en 1985 et en France en 2000. Depuis lors, de nombreuses programmes d’inventaires ont été mises en place pour rechercher des populations vivantes en France et en Espagne. Cet article présente un compte rendu à jour de la répartition de l’espèce sur la base des données disponibles, à savoir la littérature, les collections de musées et les enquêtes de terrain récentes, et fournit des données moléculaires non publiées pour la France. Il existe encore trois populations de grande mulette dans l'Èbre en Espagne et huit populations en France (deux dans le bassin versant de la Loire, une dans le bassin versant de la Charente, deux dans le bassin de la Garonne et trois dans le bassin de l'Adour). La plus grande population vit dans la Charente avec environ 100 000 individus. Le recrutement est très rare dans toutes les populations, mais des spécimens vivants estimés avoir moins de 10 ans ont été trouvés dans l'Èbre en Espagne et dans les fleuves Vienne, Charente, Dronne et Adour en France. La redécouverte récente des populations en France est principalement résultat d'enquêtes spécialisées intensives, y compris la plongée. Les progrès ultérieurs des connaissances montrent que les grands fleuves restent une terra incognita pour les hydrobiologistes.

97

98 Hydrobiologia (2018) 810:157–175 https://doi.org/10.1007/s10750-017-3456-0

FRESHWATER BIVALVES

Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe

Vincent Prie´ . Joaquin Soler . Rafael Araujo . Xavier Cucherat . Laurent Philippe . Nicolas Patry . Benjamin Adam . Nicolas Legrand . Philippe Juge´ . Nina Richard . Karl M. Wantzen

Received: 1 October 2016 / Revised: 5 November 2017 / Accepted: 28 November 2017 / Published online: 22 December 2017 Ó Springer International Publishing AG, part of Springer Nature 2017

Abstract The critically endangered Giant Freshwa- France. There are still three populations of the Giant ter Pearl Mussel Margaritifera auricularia was pre- Freshwater Pearl Mussel in the Ebro River in Spain, sumed extinct before its rediscovery in Spain in 1985 and eight populations in France (two in the Loire and France in 2000. Since then, numerous surveys watershed, one in the Charente watershed, two in the have been set up to search for living populations in Garonne watershed and three in the Adour watershed). France and Spain. This article presents an up-to-date The biggest population lives in the Charente River account of species distribution based on available data, with an estimated 100,000 individuals. Recruitment is i.e. the literature, museum collections and recent field very scarce in all populations but living specimens surveys, and provides unpublished molecular data for estimated to be less than 10 years old have been found in the Ebro in Spain and in the Vienne, Charente, Dronne and Adour rivers in France. The recent Electronic supplementary material The online version of this article (https://doi.org/10.1007/s10750-017-3456-0) con- rediscovery of populations in France was mainly a tains supplementary material, which is available to authorized result of intensive dedicated surveys including scuba users diving. Subsequent advances in knowledge show how large rivers and downstream ecosystems remain a Guest editors: Manuel P. M. Lopes-Lima, Ronaldo G. Sousa, terra incognita for the hydrobiologist. Lyuba E. Burlakova, Alexander Y. Karatayev & Knut Mehler / Ecology and Conservation of Freshwater Bivalves

V. Prie´ (&) Á L. Philippe Á N. Patry Á R. Araujo B. Adam Á N. Legrand Museo Nacional de Ciencias Naturales - C.S.I.C, c/Jose´ Biotope, Service International, Diversification, Gutie´rrez Abascal 2, 28006 Madrid, Spain Innovation, 22 Bd Mare´chal Foch, 34140 Me`ze, France e-mail: [email protected] X. Cucherat 10 rue Louis Aragon, 59147 Gondecourt, France V. Prie´ ´ Institut de Syste´matique, Evolution, Biodiversite´ ISYEB – P. Juge´ Á N. Richard UMR 7205 – CNRS, MNHN, UPMC, EPHE, Muse´um Universite´ Franc¸ois-Rabelais, CETU Elmis Inge´nieries, national d’Histoire naturelle, Sorbonne Universite´s, 57, 11 Quai Danton, 37500 Chinon, France rue Cuvier, CP26 F-75005 , France K. M. Wantzen J. Soler Á K. M. Wantzen UNESCO River Culture – Fleuves et Patrimoines Chair, Universite´ Franc¸ois Rabelais, UMR 7324 – CITERES, 33 Universite´ Franc¸ois Rabelais, UMR 7324 – CITERES, 33 Alle´e Ferdinand de Lesseps, 37204 Tours Cedex 03, Alle´e Ferdinand de Lesseps, 37204 Tours Cedex 03, France France 123 158 Hydrobiologia (2018) 810:157–175

Keywords Distribution Á Museum collections Á An extensive review of all available data on Historical data Á Scuba-diving surveys Á Large rivers Á Margaritifera auricularia’s distribution is provided Conservation here for the first time, together with new data from museum collections and recent field surveys. This article clarifies the past and present distributions of the species, presents the results of the last ten years’ Introduction surveys in France and Spain and discusses conserva- tion perspectives. Freshwater ecosystems are the most threatened ecosystems worldwide (Dudgeon et al., 2006), and freshwater bivalves rank amongst the most threatened Materials and methods animals in the world (Lydeard et al., 2004; Lopes- Lima et al., 2016). One of them, the Giant Freshwater Bibliography review Pearl Mussel Margaritifera auricularia (Spengler, 1793), figures amongst the most imperilled bivalve The bibliography since 1793 (species description date) species. Although it was considered widespread in has been extensively reviewed. Local publication and most of the western Europe rivers at the beginning of the grey literature were also consulted when available. the twentieth century, it is now considered as critically Bibliographic data were generally imprecise, but endangered by the IUCN (Araujo & Ramos, 2001; allowed figuring a broad image of the original Prie´, 2010). The Giant Freshwater Pearl Mussel had distribution and ecology of M. auricularia (Fig. 1). become so rare during the twentieth century that it was Bibliography review thus provided the first indications not even considered when the European Habitat for where to look for this species. Directive species lists have been established. Indeed, the Giant Freshwater Pearl Mussel is difficult to Museum collections observe: it lives downstream in large rivers, a habitat that is difficult to survey due to deepness, turbidity, A first review of museum collections had been current and often navigation. Hence, not surprisingly, performed by Araujo & Ramos (2000a) at a global it has been overlooked by malacologists of the scale. This review mostly aimed at large national twentieth century. However, it nowadays still survives museum collections and included also Margaritifera as a few populations in south-west France and eastern marocana (Pallary, 1918), a species living only in Spain. Morocco (Araujo et al., 2009a). We then inventoried The species was first rediscovered in Spain in 1985 all the regional museum and universities’ collections (Altaba, 1990) and in France in 2000 (Cochet, 2001). in France. Fifty-eight local natural history collections Since 1998, the biology, distribution and lifecycle of were identified. Each of them was contacted and the Giant Freshwater Pearl Mussel in Spain have been questioned about the presence of malacological col- described (Araujo & Ramos, 1998a, b; Araujo & lections, freshwater bivalves and eventually M. auric- Ramos, 2000a, b; Araujo et al., 2000; Grande et al, ularia specimens. When M. auricularia specimens 2001; Araujo et al, 2001, 2002, 2003;Go´mez & were recorded in the inventories or discovered in the Araujo, 2008). Since then, very few reports of the collection by the curator, pictures were sent to us to species in Spain have been released in national and confirm identification. Eventually, some of the most international congresses (e.g. Nakamura et al., 2015; important collections (Muse´e des Confluences in Online Resource 1), but, apart from Araujo & A´ lvarez- , Museum d’Histoire Naturelle in Bordeaux, Cobelas (2016), there are no new scientific results Museum d’Histoire Naturelle in Toulouse, Museum published since 2008. In France, focused surveys have national d’Histoire naturelle in Paris, Museum d’His- led to the rediscovery of many populations since 2007, toire Naturelle in Lille, Museum d’Histoire Naturelle but most of these results are unpublished (but see Prie´ in , Museum d’Histoire naturelle in Orleans, et al., 2007, 2008, 2010) or available only as grey University of Rennes, University of ) were literature (Online Resource 1). visited by one of us.

123 Hydrobiologia (2018) 810:157–175 159

Fig. 1 Fossil (black crosses) and historical data (white dots for precise locations, blue lines for rivers’ names only) collected from the literature and Museum collections; and subsequent intensive field surveys locations (polygons)

Specimens collected since 2000, year of the redis- studies. These impact studies were triggered when M. covery of the species in France, were not included in auricularia was living—or when available data sug- the results presented here. gested that it could still be living—in an area impacted by a development project. The results of these impact Field surveys and population sizes studies are generally not published, consisting only in various cryptic reports (but see Prie´ et al., 2007, 2008; Numerous field surveys aiming at freshwater mussels Araujo & A´ lvarez-Cobelas, 2016). We here summa- have been performed in France and Spain (Fig. 1, rize for the first time all the grey literature related to M. Table 1). auricularia in France and Spain (Online Resource 1). These dedicated surveys aimed at places most Margaritifera auricularia mainly lives in down- likely to host the species, i.e. places identified by the stream ecosystems. Surveying this habitat is challeng- literature data, museum collection data or, for France, ing because it is often deep, turbid, strongly flowing species habitats modelling (Prie´ et al., 2014). More- and navigable. In the Ebro historical channels, sam- over, some surveys took place into the frame of impact pling depends on the hydraulic works made by the

123 160 Hydrobiologia (2018) 810:157–175

Table 1 Summary of the methods used in France for surveying M. auricularia Coastal River Year Use Scuba diving Other methods Authors drainage of a involved boat

Somme Somme River 2011 X 4 divers, 5 days XC, VP Seine Seine River 2011 X Dredging (2 persons for 4 days) XC (downstream) Seine River 2015 Snorkelling (2 persons for one days) XC (upstream) Oise River 2007–2008 X 3 divers, 20 days VP, LP, XC Aisne River 2011 2–4 divers, 5 days XC, VP Rhoˆne Saoˆne and 2016 4 divers, 5 days Wading with viewing glasses (2 VP, LP, NL, Doubs Rivers persons, 5 days) NP, BA Loire Loire, Indre 2010–2011 2–3 divers, about 5 days Wading with viewing glasses (2-4 LP, VP and Cher persons, estimated to about Rivers 10 days altogether) Vienne and 2009–2016 2 to 6 divers, over 20 days Wading with viewing glasses and VP, XC, LP Creuse altogether snorkelling (2 to 6 persons, Rivers estimated to over 30 days altogether) Charente Charente River 2007–2016 X 2–3 divers, about 20 days VP altogether Garonne Dronne and 2012–2014 2 divers, about 10 days Wading with viewing glasses (1–2 VP Isle Rivers altogether persons, about 5–10 days altogether) Dordogne 2016 X 2 divers, 3 days altogether Wading with viewing glasses (1 VP River person, about 5 days) Ve´ze`re River 2016 3 divers, 3 days Wading with viewing glasses and VP, LP, NP, snorkelling (2 persons 3 days) BA, NL Garonne River 2016 3 divers, 2 days Wading with viewing glass (1 VP, NL person, 30 days) Save River 2009 2 divers, 1 day Wading with viewing glass and VP, BA snorkelling (1 person, 5 days) Adour Adour River 2012–2014 2 divers approx. 9 days Wading with viewing glasses and VP, BA, NL snorkelling (2 to 6 persons, 10 days) Arros River 2016 3 divers, 3 days Wading with viewing glasses (6 VP, BA, NL persons, 4 days) For Spain, survey methods were detailed and reported in Go´mez & Araujo (2008) and Araujo & A´ lvarez-Cobelas (2016)

Confederacio´n Hidrogra´fica del Ebro; it is necessary used to shuttle the divers from a place to another. For to decrease the water level in order to wade the others, divers dove from the river banks, and sampling channel bottom to find the specimens (Go´mez & plans were then constrained by river accessibility. Araujo, 2008). In France, some populations are readily Population sizes given here were estimated based accessible, living in the banks (Vienne River) or in on exhaustive counts of observed living individuals shallow waters (Creuse, Luy or Arros River). For those (Luy, Creuse and Vienne Rivers); statistical analyses populations, snorkelling or wading with viewing (Ebro, Arros and Charente Rivers), or in the worst glasses allowed efficient surveys. However, cumber- case, by a subjective appreciation based on the density some methods based on a team of scuba divers were of specimens observed (Dronne, Adour and Save needed in most cases. For some surveys, a boat was Rivers).

123 Hydrobiologia (2018) 810:157–175 161

The Seine (downstream) and Eure Rivers could GeoDa software (Anselin et al., 2006). Suitable habitat only be surveyed by dredging. The dredger used had length in the whole river was delimited downstream by an aperture of 50 cm, a 25-mm mesh, weighed 11 kg the limit of the mud cover due to the influence of the and was propelled by a 30-horsepower engine Zodiac Saint-Savinien’s impoundment, upstream by the limit by means of a 30-m-long rope. In the Eure River, of the living population. Between these limits, the different biotopes and flow facies were aimed at (mud, substrate and general ecological quality of the river sand, stones, riffles, vegetation). In the Seine River, was very homogenous. In this stretch of favourable water was up to 6 m deep and too troubled for habitat, live specimens have been observed wherever operators to see the river bed. Catches were then we have dived between 2010 and 2016, thus confirm- randomly positioned. Catches were 8–10 m long in the ing that the population is uniformly distributed. Eure River, and up to 40-50 m long in the Seine River. Fourteen sampling surveys were undertaken Sediment collected by the dredger was pulled up and between September 2000 and June 2006 in the Ebro sorted out on the boat. Wading surveys were adopted River, covering a total length of 25 km, wading in upstream the Seine River. shallow waters and with a team of divers in the deeper In the Somme River, a boat was used to shuttle parts of the river. Divers used submerged ropes to divers and 82 bank-to-bank transects were sampled on perform bank-to-bank or longitudinal transects (sur- a 26-km-long river stretch. In the Oise River, the vey methods are detailed and reported in Go´mez & divers were also transported by boat from a spot to Araujo, 2008; Araujo & A´ lvarez-Cobelas, 2016). another, but diving plans were constrained by river In the Dronne and Isle Rivers, about 100 km stretch condition (from very strong current to muddy bot- of each river upstream their confluence was surveyed, toms). Areas with very strong current were sampled both by wading and scuba diving from the banks. The combining scuba-diving and climbing techniques, estimation of the population size was based on with a 100-m-long static rope secured on a tree on author’s appreciation only, and is likely underesti- the bank. The diver used a climbing harness and mated: over 50 specimens have been observed during caving equipment in addition to scuba- diving gear to the surveys, with a subpopulation of 30 specimens in progress on the rope. Fins were used to go from side to the lower location (exhaustive count). We estimate side in the current, allowing to cover ca. 90-m-long that about half of the living individuals have been cone-shaped surface on the river bottom. Altogether, observed during surveys, which is unlikely, given the 115 dives have been carried out on a 35-km-long detection probability estimate in this large river. stretch of the river, from the confluence with the Aisne About 60 km of the Save River was surveyed by River downstream to a few kilometres upstream the wading and scuba diving, aiming at an exhaustive town of Sempigny. Upstream this stretch of river, count of the few remaining specimens which were surveys were carried out wading randomly in found only in the lower section of the river. Most of the suitable habitats. sampling in the Adour River was undertaken by In the Charente River, the population was estimated wading and snorkelling, with scuba divers requested based on scuba-diving transects’ surveys. A boat was only for a few deeper places. As for the Dronne River, used to shuttle the divers from a transect to another. A a few specimens were found in isolated places, with 20-m-long line was settled down on the bottom of the the biggest subpopulation numbering about ten spec- river, and scuba divers counted every living specimen imens. Population size is estimated based on experts’ left and right of the line at a distance of 2 m. Each appreciation only. The Arros River is highly impacted sample then covered 80 m2. Transects were repeated by agriculture practices. The remaining favourable every kilometre in the river stretch where mussels habitats were found isolated between the numerous were present, and then every 3 km downstream and impoundments’ influences. A first survey was con- upstream the population’s distribution limits. A total ducted by scuba divers, but the deepest places did not of 43 transects were repeated on a stretch of 60 km. have suitable habitats. A more intensive survey was Detection probability has been estimated at 75% using then organized by a team wading with viewing glasses. iterated observations analysed with the software The total length of river stretches having suitable habi- MARK (White & Burnham, 1999). Geographical tats was 54 km. Within this 54-km stretch, sixteen statistics (Anselin, 1996) were performed using sites were sampled. On each sampled site, stretches of 123 162 Hydrobiologia (2018) 810:157–175

100 m–1 km were exhaustively surveyed. Population or frozen, were ground to a powder in liquid nitrogen size was estimated based on average densities before adding 600 mL of CTAB lysis buffer [2% observed during surveys, multiplied by favourable CTAB, 1.4 M NaCl, 0.2% b-mercaptoethanol, 20 mM habitat’s surface. In the Luy River, divers explored the EDTA, 0.1 M TRIS (pH 8)] and subsequently digested deepest pools, while most of the river can be explored with proteinase K (100 mg ml-1) for 2–5 h at 60°C. by wading. The main population is found in a very Total DNA was extracted according to standard shallow place, and exhaustive counts were performed phenol/chloroform procedures (Sambrook et al. three times (during the years 2010, 2011, and 2012) by 1989). For French specimen, DNA was extracted five persons wading in a line, about 1 m apart, using the Nucleospin Tissue Kit (marketed by ensuring efficient scanning of every single place of Macherey–Nagel), according to the manufacturer’s the river bed. However, detection probability is never protocol. Extractions, amplifications and sequencing 100%. Some specimens may spend some time com- were performed by Genoscreen (France). pletely buried in the sediment and are overlooked (see To test genetic variability between populations, we below the results for the Luy River). The results of examined fragments of two mitochondrial genes, COI these assumed exhaustive counts are therefore likely and 16S, used previously by Huff et al. (2004); these underestimated. showed the greatest phylogenetic resolution power for The most intensive surveys took place in the Vienne relationships amongst margaritiferids. 28S nuclear and Creuse Rivers. The surveys aimed at providing gene fragments were also amplified, but different exhaustive counts of all living specimens. Observers fragments were targeted for French and Spanish with viewing glasses and divers (depending on the specimens. The COI, 16S and 28S genes were depth) were lined 1 m apart and moved forward amplified by polymerase chain reaction (PCR) using upstream, ensuring efficient scanning of every single the protocol described by Prie´ & Puillandre (2014) for place of the river bed. Sampling was reiterated several French specimens, and described by Machordom et al. times between 2009 and 2016 using the same methods. (2013) and Araujo et al. (2016a, b) for Spanish In this study, when shells only have been collected, specimens. The amplified fragments were purified by we considered ‘‘ancient shells’’ those that were worn ethanol precipitation prior to sequencing both strands and uncomplete, with neither periostracum nor liga- using BigDye Terminator kits (Applied Biosystems, ment remains. ‘‘Recent shells’’ include shells with at ABI). Products were electrophoresed on an ABI 3730 least periostracum and ligament remains. We consider genetic Analyser (Applied Biosystems). The forward as ‘‘juveniles’’ specimens with shell length lower than and reverse DNA sequences obtained for each spec- 11 cm, and ‘‘subadults’’ those specimens with lengths imen were aligned and checked using the Sequencer ranging from 11 to 14 cm. Occasionally, some adult program (Gene Code Corporation) after removing specimens had very short shells, especially in the primer regions. Sequences were automatically aligned Charente River, but these were obviously very old using ClustalW multiple alignments implemented in given the growth lines density and shell wear. BioEdit 7.0.5.3 (Hall, 1999). The accuracy of auto- matic alignments was confirmed by eye. Genebank Genetic analyses accession numbers are provided in Table 2.

Tissue samples have been collected from ten speci- mens from the Ebro River in Spain, and ten specimens Results from the Vienne River (Loire watershed), two spec- imens from the Luy River (Adour River watershed), Bibliography two specimens from the Charente River and one specimen from the Save River (Garonne River water- Available literature provided valuable data, albeit shed) in France. Foot tissue samples were snipped in generally with neither precise location nor date. the field and preserved in 90° ethanol for molecular Nevertheless, a first historical distribution map could analysis. be drawn from the ancient literature data. Margari- For Spanish specimens, DNA was extracted using tifera auricularia is known from the Netherland, CTAB protocol: tissue samples, preserved in ethanol England and Germany from fossil records only. 123 Hydrobiologia (2018) 810:157–175 163

Table 2 Genes and Genbank accession numbers of French specimens used for DNA analyses Coastal drainage River Specimen voucher number Genbank accession number COI 16S 28S

Charente Charente MNHN-IM-2009-12596 MF494673 MF494681 MF494677 MNHN-IM-2009-12597 MF494674 MF494682 MF494678 Garonne Save MNHN-IM-2009-12601 MF494675 MF494683 MF494679 Adour Luy MNHN-IM-2009-12662 MF494671 MF494696 MF494676 MNHN-IM-2009-12663 MF494672 MF494697 Loire Vienne Maur91 MF494670 MF494695 MF494680 MNHN-IM-2009-12611 MF494661 MF494684 MNHN-IM-2009-12615 MF494662 MF494685 Maur70 MF494663 MF494686 Maur72 MF494664 MF494687 Maur74 MF494665 MF494688 Maur76 MF494666 MF494689 Maur77 MF494667 MF494690 Maur78 MF494668 MF494691 Maur85 MF494669 MF494693 Maur79 MF494692 Maur88 MF494694

However, some shells collected in the Unstrut River in the Canal de Tauste, where there were about 5,000 live Germany are very well preserved and perhaps might specimens. The more recent data published about date back to historical times, at least until the early these Spanish populations were recorded in Araujo & Middle Ages (Bo¨ssneck et al., 2006). Fossil data in Ramos (2000b), Go´mez & Araujo (2008) and Araujo Spain includes a Mediterranean Quaternary river in &A´ lvarez-Cobelas (2016). Yecla (Murcia) with 129,000–140,000-year-old spec- imens (Andre´s & Ortun˜o, 2014) and many other Museum collections Atlantic rivers with 5,000-year-old specimens (Araujo & Moreno, 1999). In France, fossil data near Marseille The Museum collections have been examined first by (coming from archaeological excavation) and in Araujo & Ramos (2001) at a wide scale, focusing Massif Central (found amongst fossils collected in a mainly on national museums worldwide. Prie´ et al. cave) were presumably a result of human (unpublished data, Online Resource 1) have focused transportation. on French regional collections only. Out of the 58 According to historical data collected, Margari- collections identified, 25 had at least one specimen of tifera auricularia was only found in large rivers, in a M. auricularia (Fig. 2A): Muse´e du Chaˆteau in calcareous substrate, in France, Spain and Italy. In Annecy, Muse´e des Confluences in Lyon, Museum France, historical data mainly come from the Atlantic of Perpignan, Muse´e zoologique of , and Channel sea watersheds, with only one occurrence Muse´um—Aquarium of Nancy, Museum of Auxerre, in the Mediterranean drainages, in the Saoˆne River Muse´um d’histoire naturelle in Bordeaux, Muse´um (Rhoˆne tributary). In Italy and Spain, the species is d’histoire naturelle in Bourges, Muse´um d’histoire historically known from two Mediterranean water- naturelle in Grenoble, Museum d’Histoire Naturelle in sheds, the Po and Ebro Rivers (Araujo & Ramos, Nantes, Muse´um d’Histoire Naturelle in Toulouse, 2000a). In Spain, M. auricularia lived in two historical Museum d’Histoire Naturelle Victor Brun in Mon- channels from the Ebro River, the Canal Imperial and tauban, Museum d’Histoires Naturelles in Colmar,

123 164 Hydrobiologia (2018) 810:157–175

Muse´um of Orle´ans, Muse´um of , Muse´um Fig. 2 A Location of main museum collection investigated (dotc Lecoq in Clermont-Ferrand, Muse´um national d’His- size according to number of M. auricularia specimens). B number of specimens held in Museum collections per main toire naturelle in Paris, Muse´um national d’histoire watersheds (the Saoˆne River is actually a tributary of the Rhoˆne, naturelle in Lille, Paraclet centre of ONEMA in but all the specimens are located in the Saoˆne and none Boves, Poˆle muse´al of , Universite´ of Bour- elsewhere in the Rhoˆne) gogne in Dijon, Universite´ of Montpellier I, Universite´ of Rennes I, Museum d’histoire naturelle in la Rochelle, Museum of Cherbourg-Octeville. Part of In France, field surveys allowed finding ancient the data from Museum collections were fossil spec- shells in the Seine, in the Vesle and in the Aisne imens. A total of 400 non-fossil specimens were found Rivers; in the Saoˆne River (Rhoˆne watershed) near in Museum collections, including the 37 specimens Pontailler-sur-Saoˆne and in the Garonne River near already found by Araujo & Ramos (2001). Amongst Agen, findings which corroborate historical data. We them, 332 were localized at a river drainage scale. believe the species was extirpated long time ago in One-third of the specimens came from the Garonne those rivers. In the Oise River (Seine watershed), very watershed, 19% from the Saoˆne River (half of them recent shells have been found in 2007 and 2008, some coming from a single batch collected by Coutagne in of them still embedded in their natural position, 1879) and 17% from the Ebro River (Fig. 2B). Other suggesting that the species became extirpated very watersheds represented less than 30% of the Museum little time before the surveys took place. collections specimens. About 80% of the specimens The populations of the Creuse and Vienne Rivers dated were collected before the beginning of the (Loire watershed) are the most studied in France. They twentieth century. live in shallow and clear water, allowing regular surveys using viewing glasses or snorkelling. Field surveys and populations sizes Although these populations are rather small (about 250 specimens altogether), over 40 juveniles were A total of 2,500 km of rivers has been surveyed for M. found in the Vienne and Creuse Rivers, which auricularia in France and Spain for the last 10 years represent about 15% of the population. (see bibliography and Online Resource 1 for details). Three sites with a few tens of live specimens were These surveys covered most of the river stretches for discovered in the Dronne River, including one juvenile which the literature or museum collections data were of about ten cm. Additionally, some isolated individ- available. Eleven populations could be identified, uals were also observed, suggesting the population is eight in France and three in Spain, plus a single scarce but relatively widespread. In the Save River, individual found recently in the Ebro River (pers. only five live specimens were observed. Sampling ´ comm. from R. Alvarez-Halco´n to R. Araujo) conditions are difficult, with variable depth and upstream Zaragoza (Fig. 3, Table 3). In Spain, the current strength, and very low visibility. We can main population, with 5,000 live specimens, lives in therefore suppose that our detection probability is low. the Canal Imperial in Arago´n. Although there have But based on survey results, we estimate that the been some recent mortalities, some young specimens population should not exceed a few tens of living probably less than 10 years old have been observed individuals. It is likely rapidly declining given the bad during the last years (pers. comm. from J. Guerrero to condition of the river and the large number of recent R. Araujo). A new population was recently discovered shells collected compared to the very few living in the Quinto Ditch, with 25 live specimens, including specimens observed. The Adour drainage rivers were some subadults (Nakamura, com. pers.). The popula- known to host M. auricularia from both the literature tions on the Canal de Tauste still host several live and Museum collections data. In the Adour main- specimens and juveniles. The population of the lower stream, the population is now highly fragmented, with Ebro River is today practically testimonial (pers. com. only three sites identified, where live specimens could of the Generalitat of Catalonia to R. Araujo). See be found. One of them, the most upstream, is now Go´mez & Araujo (2008), Araujo (2012) and Araujo & extirpated (Prie´ et al., 2010). The total population is ´ Alvarez-Cobelas (2016) for more information. estimated to be about 300 specimens in the total length

123 Hydrobiologia (2018) 810:157–175 165

123 166 Hydrobiologia (2018) 810:157–175

Fig. 3 Results of last ten years’ field surveys and known past and actual distribution of M. auricularia. Fossil data (black crosses), historical data (white dots), shells collected in the last 10 years (orange dots) and still living populations (red dots) of the Adour mainstream, but we still need a better literature to host an important population of M. estimation based on an appropriate sampling protocol. auricularia (Bonneme`re, 1901). Shell fragments and On the Luy tributary, a population of about 150 very few live specimens had been found by naturalists specimens is found in a very small stretch of river. since 2003 (Nienhuis, 2003; P. Jourde pers. com.). Interestingly, although this River is very shallow Intensive field surveys performed in 2007, 2010 and (from 30 cm to 1.5 m), clear and easy to survey (hence 2016 led to the discovery of the largest population detection probability is optimal), successive counts of worldwide. Geographical statistics based on scuba- 2010, 2011 and 2012 lead to, respectively, 110, 96 and diving transects showed that the population was not 145 specimens. We suppose that a significant part of aggregated. Hence, the total population size could be the population lives buried in the sediment, which estimated by multiplying the average density by the biases the results of the counts. The Arros River had total surface of suitable habitat in the stretch of river been overlooked by the literature review and field inhabited by M. auricularia. The average population surveys up to 2016. Following the findings in Museum size in the Charente River was estimated to be about collections, dedicated field surveys were conducted in 100,000 (range 80,000–120,000) individuals, between 2016, allowing the rediscovery of a living population. the towns of Cognac upstream to Port-d’Envaux This population’s size was estimated to be about 200 downstream. individuals on the 54 km stretch of favourable habitat. The Charente River was known from the ancient

123 yrbooi 21)810:157–175 (2018) Hydrobiologia Table 3 Summary of literature data, museum collections (fossil data are not considered) and field surveys Country Coastal Literature Museum collections Dedicated field surveys (references with an * refer to Recent surveys results Estimated drainage (number of grey literature, summarized in Online Resource 1) population specimens) size

France Charente Charente X 9 Prie´ et al. (2007)*, Prie (2010), Prie´ & Mouton (2016)* Live specimens and 100. 000 juveniles Garonne Garonne X 125 Prie´ et al. (2016)* Ancien shells (mainstream) Isle X 0 Prie´ (2012)* Recent shells Dronne X 0 Prie´ (2012)*; Prie´ (2013)* Live specimens and [ 100 juveniles Save 0 Prie´ (2012)* Few live specimens, \ 30 declining population Adour Adour X 9 Prie´ (2012)* Live specimens and [ 300 (mainstream) juveniles Arros 6 Prie´ &Ne´ri (2016)* Live specimens and 200 subadults Luy 0 Prie´ (2012)* Live specimens and 150 subadults Loire Loire X 1 Recent shells (mainstream) Vienne 0 Cochet (2006)*, Philippe et al. (2009)*, Philippe et al. Live specimens and [ 100 (2010*, 2011*, 2012*) juveniles Indre 0 Dohogne (2008)*; Philippe et al. (2009)* Recent shells Creuse 0 Philippe et al. (2012)*; Philippe et al. (2013*, 2014*, Live specimens and [ 150 2015*, 2016*) juveniles Cher 1 Prie´ et al. (2011)*, Prie´ et al. (2016)* Nothing Seine Seine X 12 Cucherat et al. (2011)* Ancien shells (mainstream) Oise 0 Prie´ et al. (2007)* Recent shells Aube X 1 Cucherat et al. (2011)* Ancien shells Aisne X 6 Philippe et al. (2009)*, Cucherat et al. (2011)* Ancien shells Escaut Escaut X 2 No dedicated survey Somme Somme X 1 Cucherat & Prie´ (2011)* Nothing

123 Rhoˆne Saoˆne X 65 Prie´ et al. (2016)* Ancien shells Italy Po Po X 15 No dedicated survey 167 168 123 Table 3 continued Country Coastal Literature Museum collections Dedicated field surveys (references with an * refer to Recent surveys results Estimated drainage (number of grey literature, summarized in Online Resource 1) population specimens) size

Spain Ebro Upper Ebro X Araujo et al. (2009b)*, Araujo & A´ lvarez-Cobelas Live specimen (at 1 (2016); pers. comm. from R. A´ lvarez-Halco´ntoR. Gallur) Araujo Ribera alta Nakamura & Guerrero (2008), Araujo & A´ lvarez- Used to be 38–40 live Cobelas (2016) specimens, today likely extirpated Canal Imperial X55Go´mez & Araujo (2008), Araujo et al. (2009a, b), Araujo Live specimens and 4,000 de Aragon &A´ lvarez-Cobelas (2016); pers. comm. from J. juveniles Guerrero to R. Araujo Canal de X Araujo et al. (2009b)*; pers. comm. from J. Guerrero to Live specimens and 200 Tauste R. Araujo juveniles Quinto ditch Go´mez & Araujo (2008), Nakamura et al. (2017) Live specimens and 25 subadults Lower Ebro X Araujo et al. (2009b)*; Araujo & A´ lvarez-Cobelas (2016) Used to be 70 live specimens, today extirpated Tajo Tajo X 1 Villasante et al. (2016) Nothing Fossil data U.K. Thames Thames 17 No dedicated survey Germany Rhine Rhine X 3 No dedicated survey Netherland Rhine Rhine X No dedicated survey yrbooi 21)810:157–175 (2018) Hydrobiologia Hydrobiologia (2018) 810:157–175 169

Genetic diversity considered restricted to five watersheds: from north to south the Loire watershed (two close populations in Margaritifera auricularia is genetically remarkably the Vienne and Creuse Rivers), the Charente water- homogenous. The specimens from France and Spain shed, the Garonne watershed (two very isolated all shared the same 16S and COI haplotypes, but two populations, in the Dronne and Save Rivers), the specimens from Spain: specimens vouchered with Adour watershed (at least three isolated populations, FW1238-14 and FW1238-12; for COI T ? Ain one in the Adour itself, one in the Luy and one in the position 37, T ? A in position 50 and G ? Cin Arros) and the Ebro River (two populations, three in position 73; and for 16S T ? C in position 176. The channels and a small one remaining in the Ebro itself). French and Spanish specimens could not be compared As has been previously estimated (Prie´ et al., 2014), for 28S as different gene fragments were amplified. Margaritifera auricularia’s range contraction has However, within France, all specimens shared the probably reached about 90% in the last two centuries. same haplotype and within Spain, all specimens shared the same haplotype. Surveying downstream ecosystems

Large rivers are amongst the most difficult ecosystems Discussion to sample. Deepness, turbidity and water current are challenging conditions. In addition, large rivers are Historical and actual data subject to navigation, which makes scuba diving potentially hazardous. Nevertheless, scuba diving The number of specimens found in the various appears to be the most efficient way to produce data regional museum collections was unexpected. Mar- for species such as M. auricularia: despite malaco- garitifera auricularia is a large species that retained logical surveys undertaken with canoes and dredging, collector’s attention. Most data from museum collec- only a few shell fragments had been collected in the tions corresponded to the literature data, excepted Charente River before scuba- diving sampling had those from the Arros and Vezere Rivers in France. been set up. Scuba divers met hundreds of shells and Surprisingly, most French specimens came from the living specimens there. Similarly, scuba divers col- Garonne and Saoˆne Rivers, where the species is now lected the few living specimens that are probably dead believed to be extirpated or very rare. In contrast, very by now, in the main Ebro River in Spain (Araujo & few specimens came from the Charente River, where A´ lvarez-Cobelas, 2016). In the Oise River, a few the largest population is found nowadays, and where ancient shell fragments had been collected on the industrial fisheries were established to make nacre banks by amateur malacologists, but scuba diving shirt buttons (Bonneme`re, 1901). Similarly, museum allowed finding numerous shells in most of the river collections host no specimen from the Vienne or stretches investigated. In the Garonne River main- Creuse Rivers, where healthy populations live in stream, a malacologist spent about 20 days wading shallow and clear waters. In the Seine watershed, most and searching for shells on the gravelled banks. In shells came from upstream and the Aisne tributary, 2 days, a team of three divers found four shell while the Oise tributary seems to have host the last fragments. population. While bivalve surveys have been conducted in the The historical review confirmed that M. auricularia Saoˆne River (ex. Mouthon & Daufresne, 2006), no was once present as far as the Thames in England and shell fragments had ever been collected before 2016’s Netherlands and Germany where fossil specimens scuba-diving prospections. The advances in the dis- have been found and studied (Araujo & Ramos, 2001). tribution knowledge of M. auricularia in France and On historical times, we found museum records (recent Spain are directly linked to new investigation methods shells) from the Rhine in France or Germany (precise and scuba diving is so far the most efficient mean of location being unknown), the Seine and the Rhoˆne in survey for this species. France, the Poˆ in Italy and the Tajo in Spain, where the species is now believed to be extirpated (Araujo & Ramos, 2001). Today, Margaritifera auricularia is 123 170 Hydrobiologia (2018) 810:157–175

Conservation and further perspectives don’t know how these eutrophic and polluted waters may impact juvenile survival (Augspurger et al., 2007; Main threats Strayer & Malcom, 2012; Archambault et al., 2014). Invasive species probably add to the threats M. While overfishing may have contributed to the species auricularia is facing. Widespread invasive species decline in the past (Bonneme`re, 1901; Prie´ et al., 2011; such as Corbicula fluminea probably affect the Araujo & A´ lvarez-Cobelas, 2016), it is obviously river freshwater mussels of Europe like it has been demon- management and agriculture impacts that nowadays strated for other species in North America (e.g. Soussa cause the most important threats to the Giant Fresh- et al., 2014). However, no clear impacts have been water Pearl Mussel. Both causes are linked together, at described for M. auricularia, and the healthiest least in the southern part of the species distribution populations survive in rivers largely colonized by area: river management aims at providing freshwater Corbicula. The zebra mussel Dreissena polymorpha for corn culture, especially in summer. Hence, attaches to the valves of M. auricularia in the Ebro, numerous dams are built, even in small rivers, to probably affecting filtration efficiency. This phe- maintain pools for pumping water in the dry season. nomenon has not been observed in France, where the These dams produce lotic and silty conditions unsuit- zebra mussel remains at low densities in the rivers of able for the Giant Freshwater Pearl Mussel. Alto- the Atlantic coast. gether, these small dams can affect about than 70% of a given rivers stretch. In the Dronne, Arros and Save Habitat management Rivers in France for example, the Giant Pearl Fresh- water Mussel populations survive in the form of Contrarily to the Freshwater Pearl Mussel M. mar- dashed lines, only in riffles (shallow parts of streams garitifera, for which experiments of habitat manage- where water flows brokenly) with gravel or stony ments have proved to be successful (Altmu¨ller & bottoms, between long portions of lotic conditions. Dettmer, 2006), the Giant Freshwater Pearl Mussel Moreover, these dams constitute obstacles for poten- lives in downstream ecosystems. Attempts to imple- tial fish hosts. The presumed natural host fish of the ment broad scale watersheds management are there- Giant Freshwater Pearl Mussel, the European Stur- fore unrealistic. However, some realistic management geon Acipenser sturio, has been extirpated from objectives can be achieved to improve the habitat almost all European rivers mainly because of dams quality locally, in a short or middle term. The (Lepage & Rochard, 1995; Gesner et al., 2010). River deconstruction of the numerous impoundments (many management has been an important threat in Spain too. of them being disused) seems the most efficient way to Water regulation and the replacement of natural restore suitable riverbed conditions for the Giant bottoms with concrete have been responsible for a Freshwater Pearl Mussel. Although the negative massive death of Naı¨ads. Recently, there has been an impacts of pollution and eutrophication are not clearly unusual high mortality of adults in the Imperial known, they are for sure not needed for the species Channel (pers. com. from the Diputacio´n General de survival. Improving water quality through reasonable Arago´n to R. Araujo), but the causes are unknown. agricultural practices, with buffer strips or grass strips Despite being a probable cause of recruitment along waterways, should be a medium-term objective. failure, moderate levels of pollution and eutrophica- tion have not demonstrated to be a significant threat to Farming projects adult specimens. Some populations survive in highly human-impacted waters. For example, one of the Breeding farms have been established for many highest Giant Freshwater Pearl Mussel densities spot endangered mussel species. In Europe, there is an lies just downstream the Saintes sewage system in the abundant literature dealing with M. margaritifera Charente River. The same kind of conditions occurs at breeding farms. Some trials are also ongoing for U. the Canal Imperial in Arago´n with the water coming crassus and for various Unio species in Spain (Araujo from the Ebro River, which is highly polluted. Overall, et al., 2015). Regarding the Giant Freshwater Pearl the species survives in rivers that are highly impacted Mussel, attempts of ex situ breeding have been by agriculture and domestic effluents. But we still performed in Spain (Nakamura et al., 2015), and a 123 Hydrobiologia (2018) 810:157–175 171

LIFE project is ongoing in France to artificially breed recently considered as distinct species based on the species in controlled conditions. Juveniles have molecular divergences (Reis & Araujo, 2009; Araujo been successfully produced (Nakamura et al., 2015), et al., 2009b). But on the other hand, some species do but we still face obstacles in the rearing of these not show significant genetic divergences (ex. U. juveniles (although some juveniles are still alive, mancus, Prie´ et al., 2012; Potomida littoralis, Araujo Nakamura com. pers.). et al., 2016a, b; Froufe et al., 2016). (ii) The different populations known today have obvious morphological Genetic diversity differences in shell size and shape (Fig. 4). The specimens from the Charente River population have a The very low genetic diversity for the mitochondrial peculiarly small and conspicuously ear-like shell genes studied was unexpected as (i) the Giant Fresh- shape, contrasting to the Vienne and Dordogne Rivers’ water Pearl Mussel populations are geographically populations, which are larger and more elongated; and isolated for a long time; and (ii) strong morphological to the Arros and Save Rivers populations, which are differences are found between populations (Fig. 4). remarkable with their huge sizes. Some populations (i) The populations from France belong to the Atlantic live in deep coastal rivers (ex. Ebro, Vienne and drainage and the population from Spain to the Charente populations) while others seem to be Mediterranean drainage, two geographically isolated confined to shallow riffle sections of the upstream bioregions. Strong genetic divergences are observed rivers (ex Save and Adour populations), but these for other freshwater mussel species from the Iberian ecological traits are not linked to shell morphological Peninsula: U. delphinus from the ‘‘pictorum’’ lineage differences. and U. tumidiformis from the ‘‘crassus’’ lineage were Margaritiferidae are known to have very low mitochondrial DNA evolution rates (Araujo et al., 2016a, b; Bolotov et al., 2016). Population genetics based on microsatellites allowed to differentiate evolutionary units within the related species Margar- itifera margaritifera (Geist et al., 2010; Stoeckle et al., 2016) and M. marocana (Sousa et al., 2016). But first studies using microsatellites based on M. margari- tifera primers have failed to reveal any population structure in France (Prie´, unpublished data). If the ex situ breeding projects are successful, the population genetics question will become unavoidable.

The fish host issue

The known host fish of Margaritifera auricularia are sturgeon species Acipenser sturio, A. nacari and A. baeri, the River Blenny Salaria fluviatilis and the Eastern Mosquitofish Gambusia holbrooki (Araujo & Ramos, 1998b; Araujo et al., 2000, 2001; Altaba & Lopez, 2001; Lopez & Altaba, 2005; Lopez et al., 2007). The only native Acipenser species in the area of occurrence of Margaritifera auricularia is the Euro- pean sturgeon A. sturio. This species became extir- pated from most European Rivers during the twentieth century. Nowadays, it is almost extinct, with the last Fig. 4 Morphological variability of Margaritifera auricularia. A Vienne River; B–C Charente River; D Dronne River; E Save documented case of natural reproduction dating back River; F Luy River; G Arros River, H Ebro River to 1994 in the Garonne River. The River Blenny is a 123 172 Hydrobiologia (2018) 810:157–175

Mediterranean species whose range does not overlap magnitude of the efforts allocated to surveying the with the French populations of M. auricularia. The species in its historical range, we now believe that Eastern Mosquitofish, an introduced species, lives in there are very few chances to rediscover unnoticed shallow and standing to slow-flowing waters. It is not populations (except maybe in north-east France). usually found in places favoured by Margaritifera Although some juveniles were found recently, they auricularia. Reported success as host fish for M. remain very scarce, and most extant populations seem auricularia glochidia was questionable. Experiments to live on borrowed time. Within the time lapse of this with other common fish species that occur within the study, some populations already became extirpated in distribution range of M. auricularia (Anguilla Angu- the Ebro and Adour Rivers. The status of the species illa, Barbus graellsii, Barbus haasi, Parachondros- therefore remains worrying. Priority populations for toma toxostoma, Cobitis paludicola, Liza aurata, conservation are the Charente River’s population, Mugil cephalus, Alburnus alburnus, Carassius aura- because it is by far the largest worldwide; the Vienne tus, Cyprinus carpio, Gobio gobio, Scardinus ery- and Creuse population, because it has the highest level throphthalmus and Tinca tinca) failed to produce of natural recruitment; the Adour watershed popula- juveniles (Araujo et al., 2001; Lopez & Altaba, 2005). tions, because they form an important and unique The actual knowledge on the M. auricularia host metapopulation; and the Ebro population because it is fish cannot explain the recruitment observed recently now the only remaining one in the Mediterranean in the Atlantic watersheds. We therefore suspect an drainage system. Conservation challenges for the next overlooked host fish species. For example, the Alosa years are (i) appropriate management of the rivers species, which are migratory fish and still breed in the which host the priority populations; (ii) the develop- watersheds where M. auricularia produces juveniles, ment of farming projects, in order to reinforce existing are good candidates (Llorente et al., 2015). But there populations; (iii) research on fish hosts for better must be another fish host in order to explain recruit- comprehension of threats to the species, ecological ment in the Dronne and Charente Rivers, which are requirements, to understand what those habitat factors isolated from the sea by impoundments, or in very are which drive the species’ success with recruitment, upstream populations such as those of the Arros or population genetics to plan conservation efforts Aisne Rivers, where migratory fishes do not breed. according to the genetic diversity of the remaining The other hypothesis could be that reproduction populations; and (iv) wide scale development of occurs in France periodically, taking advantage of modern survey methods such as scuba diving and accidental releases of A. baeri, a common species on environmental DNA in order to discover the unnoticed French fish farms (but not in Spain). We have recently populations which potentially remain. succeeded in completing the full cycle on the three- Despite these efforts, we may fail to save the Giant spined stickleback Gasterosteus aculeatus in con- Freshwater Pearl Mussel from extinction. However, trolled conditions (Soler et al., in prep.). As this current research helps to shed light on the obscure species is widespread within the range of M. auricu- downstream river ecosystems’ ecological functions laria and tolerant to brackish waters, it could also be a and the threats to it, as well as to develop exploration good candidate as a natural fish host. Finding the methods for this challenging environment. natural host fish species of M. auricularia in France is now vital for the survival and conservation of this Acknowledgements This work was conducted within the freshwater bivalve. scope of the LIFE project ‘‘Life13BIOFR001162 Conservation of the Giant Pearl Mussel in Europe’’. We thank Dominique Tesseyre from the Adour-Garonne Water Agency; Julie Marcinkowsky and Ge´rard Tardivo from the DREAL Centre Conclusion as well as the DREAL Picardie for providing financial support for large-scale surveys of M. auricularia in France; Elodie Hugues, Guillaume Me´tayer (Conseil Ge´ne´ral de Charente Margaritifera auricularia has become very rare in the Maritime), David Be´cart (Voies Navigables de France) and twentieth century, with an estimated range contraction Amandine Szurpicki (COSEA), Fre´de´rique Moinot and Olivier of 90%. Only three populations were known world- Guerri (EPIDOR) for financing focus surveys in the Charente, wide before 2007. Intensive surveys in the last decade Seine, Vienne and Garonne Rivers; and for Spain, the Government of Arag o´n, FMC Foret S.A., Enagas, Gas allowed for the rediscovery of nine more. Given the 123 Hydrobiologia (2018) 810:157–175 173

Natural, INYPSA, Hidroele´ctrica La Zaida, Edison Mission Araujo, R. & M. A. Ramos, 2001. Action Plan for Margaritifera Energy and EID Consultores. auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Council of Europe Publishing. Nature and environment, No. 117. Strasbourg. References Araujo, R., D. Bragado & M. A. Ramos, 2000. Occurrence of glochidia of the endangered Margaritifera auricularia Andre´s, D. & A. Ortun˜o, 2014. Margaritifera auricularia (Spengler, 1793) and other mussel species (Bivalvia: (Spengler, 1793): primeras evidencias sobre el antiguo Unionoida) in drift and on fishes in an ancient channel of cuaternario de Yecla (Murcia). Athene 24: 21–37. the Ebro River, Spain. Archiv fu¨r Hydrobiologie 148: Anselin, L., 1996. The Moran scatterplot as an ESDA tool to 147–160. assess local instability in spatial association. In Fischer, M., Araujo, R., D. Bragado & M. A. Ramos, 2001. Identification of H. Scholten & D. Unwin (eds), Spatial Analytical Per- the river blenny, Salaria fluviatilis, as a host to the glo- spectives on GIS in Environmental and Socio-Economic chidia of Margaritifera auricularia. Journal of Molluscan Sciences. Taylor and Francis, London. Studies 67: 128–129. Anselin, L., I. Syabri & Y. Kho, 2006. GeoDa: an introduction to Araujo, R., N. Ca´mara & M. A. Ramos, 2002. Glochidium spatial data analysis. Geographical Analysis 38: 5–22. metamorphosis in the endangered freshwater mussel Altaba, C. R., 1990. The last known population of the freshwater Margaritifera auricularia (Spengler, 1793): a histological mussel Margaritifera auricularia (Bivalvia, Unionoida): a and scanning electron microscopy study. Journal of Mor- conservation priority. Biological Conservation 52: phology 254: 259–265. 271–286. Araujo, R., M. Quiro´s & M. A. Ramos, 2003. Laboratory Altaba, C. R. & M. A. Lopez, 2001. Experimental demonstration propagation and culturing of juveniles of the endangered of viability for the endangered giant pearlmussel Margar- freshwater mussel Margaritifera auricularia (Spengler, itifera auricularia (Bivalvia: Unionoida) in its natural 1793). Journal of Conchology 38: 53–60. habitat. Bolleti de la Societat d’Historia Natural de les Araujo, R., C. Toledo, D. Van Damme, M. Ghamizi & A. Baleares 44: 15–21. Machordom, 2009a. Margaritifera marocana (Pallary, Altmu¨ller, R. & R. Dettmer, 2006. Erfolgreiche Arten- 1918) a valid current species inhabiting the Moroccan schutzmaßnahmen fu¨r die Flussperlmuschel Margaritifera rivers. Journal of Molluscan Studies 75: 95–101. margaritifera L. durch Reduzierung von unnatu¨rlichen Araujo, R., J. Reis, A. Machordom, C. Toledo, M. J. Madeira, I. Feinsedimentfrachten - Erfahrungen im Rahmen des Lut- Gomez, J. C. Velasco, J. Morales, J. M. Barea, P. Ondina & terprojekts. Informationsdienst Naturschutz Niedersachsen I. Ayala, 2009b. Las Nayades de la peninsula Iberica. 26: 192–204. Iberus 27: 7–72. Araujo, R. 2012. Margaritifera auricularia. In VV.AA., Bases Araujo, R., C. Feo, Q. Pou & M. Campos, 2015. Conservation of ecolo´gicas preliminares para la conservacio´n de las espe- two endangered European freshwater mussels. A three- cies de intere´s comunitario en Espan˜a: Invertebrados. year semi-natural breeding experiment. The Nautilus 129: Ministerio de Agricultura, Alimentacio´n y Medio Ambi- 126–135. ente. Madrid: 54 pp. Araujo, R., S. Schneider, K. J. Roe, D. Erpenbeck & A. Araujo, R. & M. A´ lvarez-Cobelas, 2016. Influence of Flow Machordom, 2016a. The origin and phylogeny of Mar- Diversions on Giant Freshwater Pearl Mussel Populations garitiferidae (Bivalvia, Unionoida): a synthesis of molec- in the Ebro River, Spain. Marine and Freshwater Ecosys- ular and fossil data. Zoologica Scripta. https://doi.org/10. tems, Aquatic Conservation. https://doi.org/10.1002/aqc. 1111/zsc.12217. 2622. Araujo, R., D. Buckley, K.-O. Nagel & A. Machordom, 2016b. Araujo, R. & R. Moreno, 1999. Former Iberian distribution of Potomida littoralis (Bivalvia, Unionidae) evolutionary Margaritifera auricularia (Spengler) (Bivalvia: Margari- history: slow evolution or recent speciation? Zoological tiferidae). Iberus 17: 127–136. Journal of the Linnean Society. https://doi.org/10.1111/zoj. Araujo, R. & M. A. Ramos, 1998a. Margaritifera auricularia 12470. (Unionoidea, Margaritiferidae), the giant freshwater pearl Archambault, J.-M., W. G. Cope & T. J. Kwak, 2014. Survival mussel rediscovered in Spain. Graellsia 54: 129–130. and behaviour of juvenile unionid mussels exposed to Araujo, R. & M. A. Ramos, 1998b. Description of the thermal stress and dewatering in the presence of a sediment glochidium of Margaritifera auricularia (Spengler, 1793) temperature gradient. Freshwater Biology 59: 601–613. (Bivalvia, Unionidae). Philosophical Transactions of the Augspurger, T., F. J. Dwyer, C. G. Ingersoll & C. M. Kane, Royal Society of London B 353: 1553–1559. 2007. Advances and opportunities in assessing contami- Araujo, R. & M. A. Ramos, 2000a. A critical revision of the nant sensitivity of freshwater mussel (Unionidae) early life historical distribution of the endangered Margaritifera stages. Environmental Toxicology and Chemistry 26: auricularia (Spengler, 1782) (Mollusca: Margaritiferidae) 2025–2028. based on Museum specimens. Journal of Conchology 37: Bolotov, I. N., I. V. Vikhrev, Y. V. Bespalaya, M. Y. Gofarov, 49–59. A. V. Kondakov, E. S. Konopleva, N. N. Bolotov & A. Araujo, R. & M. A. Ramos, 2000b. Status and conservation of A. Lyubas, 2016. Multi-locus fossil-calibrated phylogeny, the relict giant European freshwater pearl mussel Mar- biogeography and a subgeneric revision of the Margari- garitifera auricularia (Spengler, 1793). Biological Con- tiferidae (Mollusca: Bivalvia: Unionoida). Molecular servation 96: 233–239. Phylogenetics and Evolution 103: 104–121. 123 174 Hydrobiologia (2018) 810:157–175

Bonneme`re, L., 1901. Les mollusques des eaux douces de Reunio´n de Arqueomalacologı´a de la Penı´nsula Ibe´rica. France et leurs perles. Institut international de bibliogra- Santander: 125–134. phie scientifique, Paris. Lopes-Lima, M., R. Sousa, J. Geist, D. C. Aldridge, R. Araujo, J. Bo¨ssneck, U., R. Kleemann & L. Buttstedt, 2006. Die Große Bergengren, Y. Bespalaya, E. Bo´dis, L. Burlakova, D. Van Flussperlmuschel (Pseudunio auricularius Spengler, 1793) Damme, K. Douda, E. Froufe, D. Georgiev, C. Gumpinger, in Mitteldeutschland: Neue Befunde zur historischen und A. Karatayev, U¨ . Kebapc¸i, I. Killeen, J. Lajtner, B. fossilen Verbreitung (Bivalvia: Margaritiferidae). M. Larsen, R. Lauceri, A. Legakis, S. Lois, S. Lundberg, E. Malakologische Abhandlungen 24: 141–156. Moorkens, G. Motte, K.-O. Nagel, P. Ondina, A. Outeiro, Cochet, G., 2001. Rede´couverte d’une population vivante de la M. Paunovic, V. Prie´, T. von Proschwitz, N. Riccardi, M. Grande Mulette, Margaritifera auricularia, sur la Vienne Rudzı¯te, M. Rudzı¯tis, C. Scheder, M. Seddon, H. S¸ ere- et la Creuse. Recherches Naturalistes en Re´gion Centre 10: flis¸an, V. Simic´, S. Sokolova, K. Stoeckl, J. Taskinen, A. 3–16. Teixeira, F. Thielen, T. Trichkova, S. Varandas, H. Dudgeon, D., A. H. Arthington, M. O. Gessner, Z.-I. Kawabata, Vicentini, K. Zajac, T. Zajac & S. Zogaris, 2016. Conser- D. J. Knowler, C. Le´veˆque, R. J. N. Aiman, A.-H. Prieur- vation status of freshwater mussels in Europe: state of the Richard, D. Soto, M. L. J. Stiassny & C. A. Sullivan, 2006. art and future challenges. Biological Reviews. https://doi. Freshwater biodiversity: importance, threats, status and org/10.1111/brv.12244. conservation challenges. Biological Reviews 81: 163–182. Lo´pez, M. A. & C. R. Altaba, 2005. Fish host determination for Froufe, E., V. Prie´, J. Faria, M. Ghamizi, D. V. Gonc¸alves, M. Margaritifera auricularia (Bivalvia: Unionoida): results Emre Gu¨rlek, I. Karaouzas, U¨ . Kebapc¸i, H. Sereflisan, C. and implications. Bolletino Malacologico 41: 89–98. Sobral, R. Sousa, A. Teixeira, S. Varandas, S. Zogaris & M. Lopez, M. A., C. R. Altaba, T. Rouault & E. Gisbert, 2007. The Lopes-Lima, 2016. Phylogeny, phylogeography, and evo- European sturgeon Acipenser sturio is a suitable host for lution in the Mediterranean region: news from a freshwater the glochidia of the freshwater pearl mussel Margaritifera mussel (Potomida, ). Molecular Phylogenetics auricularia. Journal of Molluscan Studies 73: 207–209. and Evolution 100: 322–332. Lydeard, C., R. H. Cowie, W. F. Ponder, A. E. Bogan, P. Geist, J., H. So¨derberg, A. Karlberg & R. Kuehn, 2010. Drai- Bouchet, S. A. Clark, K. S. Cumming, T. J. Frest, O. nage-independent genetic structure and high genetic Gargominy, D. G. Herbert, R. Hershler, K. E. Perez, B. diversity of endangered freshwater pearl mussels (Mar- Roth, M. Seddon, E. E. Strong & F. G. Thompson, 2004. garitifera margaritifera) in northern Europe. Conservation The global decline of nonmarine mollusks. BioScience 54: Genetics 11: 1339–1350. 321–330. Gesner, J., P. Williot, E. Rochard, J. Freyhof & M. Kottelat, Machordom, A., R. Araujo, D. Erpenbeck & M.-A. Ramos, 2010. Acipenser sturio. The IUCN Red List of Threatened 2013. Phylogeography and conservation genetics of Species 2010: e.T230A13040963. https://doi.org/10.2305/ endangered European Margaritiferidae (Bivalvia: Union- IUCN.UK.2010-1.RLTS.T230A13040963.en. Accessed oidea). Biological Journal of the Linnean Society 78: 19 Dec 2016. 235–252. Go´mez, I. & R. Araujo, 2008. Channels and ditches as the last Mouthon, J. & M. Daufresne, 2006. Effects of the 2003 heat- shelter for freshwater mussels. The case of M. auricularia wave and climatic warming on mollusc communities of the and other naiads at the mid Ebro River basin, Spain. Saoˆne: a large lowland river and its two main tributaries Aquatic Conservation: Marine and Freshwater Ecosystems (France). Global Change Biology 12: 441–449. 18: 658–670. Nakamura, K. & J. Guerrero, 2008. Margaritifera auricularia, Grande, C., R. Araujo & M. A. Ramos, 2001. The gonads of un difıcil reto de conservacion. Quercus 265: 20–25. Margaritifera auricularia (Spengler, 1793) and Margari- Nakamura, K., M. A. Yanguas, I. Ruiz, I. de Blas, E. Elbaile & tifera margaritifera (L. 1758) (Bivalvia: Unionoidea). C. Salinas, 2015. The first year of Margaritifera auricu- Journal of Molluscan Studies 67: 27–35. laria (Spengler, 1793). Breeding in captivity in Aragon, Hall, T. A., 1999. BioEdit: a user-friendly biological sequence Spain. 2nd International seminar on the rearing of unionoid alignment editor and analysis program for Windows 95/98/ mussels. Tuesday 24th November – Thursday 26th NT. Nucleic Acids Symposium Series 41: 95–98. November 2015, Clervaux (Luxembourg). Huff, S. W., D. Campbell, D. L. Gustafson, C. Lydeard, C. Nakamura, K., L. Cucala, A. Mestre, F. Mesquita-Joanes, E. R. Altaba & G. Giribet, 2004. Investigations into the Elbaile, C. Salinas & M. A´ ngel Mun˜oz-Yanguas, 2017. phylogenetic relationships of freshwater pearl mussels Modelling growth in the critically endangered freshwater (Bivalvia: Margaritiferidae) based on molecular data: mussel Margaritifera auricularia (Spengler, 1793) in the implications for their and biogeography. Journal Ebro basin. Hydrobiologia. https://doi.org/10.1007/ of Molluscan Studies 70: 379–388. s10750-017-3103-9. Lepage, M. & E. Rochard, 1995. Threatened fishes of the world: Nienhuis, J. A. J. H., 2003. The rediscovery of Spengler’s Acipenser sturio Linnaeus, 1758 (Acipenseridae). Envi- freshwater pearlmussel Pseudunio auricularius (Spengler, ronmental Biology of Fishes 43: 28. 1793) (Bivalvia, Unionoidea, Margaritiferidae) in two river Llorente, L., E. Rosello´, A. Morales, C. Liesau, A. Daza & R. systems in France, with an analysis of some factors causing Araujo, 2015. Las na´yades (Mollusca, Unionoida) del its decline. Basteria 67: 67–86. Calcolı´tico de Camino de las Yeseras (San Fernando de Prie´, V., 2010. Margaritifera auricularia. The IUCN Red List of Henares, Madrid). In Gutie´rrez, I., D. Cuenca & M. Threatened Species 2010: e.T12798A3381899. https://doi. R Gonza´lez (eds) La Investigacio´n Arqueomalacolo´gica en org/10.2305/IUCN.UK.2010-4.RLTS.T12798A3381899. la Penı´nsula Ibe´rica: nuevas aportaciones. Actas de la IV en. Accessed 30 Sept 2016. 123 Hydrobiologia (2018) 810:157–175 175

Prie´, V. & N. Puillandre, 2014. Molecular phylogeny, taxonomy south-western Iberian Peninsula. Journal of Natural His- and distribution of French Unio species (Bivalvia, Union- tory 43: 1929–1945. idae). Hydrobiologia 735: 95–110. Sambrook, J., E.F. Fritsch & T. Maniatis, 1989. Molecular Prie´, V., L. Philippe & G. Cochet, 2007. Evaluation de l’impact Cloning. A Laboratory Manual. Second Edition. Cold d’un projet de canal sur les naı¨ades de l’Oise (France) et Spring Harbor, New York, 1626 p. de´couverte de valves re´centes de Margaritifera auricularia Soussa, R., A. Novais, R. Costa & D. L. Strayer, 2014. Invasive (Spengler, 1793) (Bivalvia: Margaritiferidae). MalaCo 4: bivalves in fresh waters: impacts from individuals to 178–182. ecosystems and possible control strategies. Hydrobiologia Prie´, V., G. Cochet & L. Philippe, 2008. Une population majeure 735: 233–255. de la tre`s rare Grande Mulette Margaritifera auricularia Sousa, R., S. Varandas, A. Teixeira, M. Ghamizi, E. Froufe & (Bivalvia; Margaritiferidae) dans la Charente (France). M. Lopes-Lima, 2016. Pearl mussels (Margaritifera MalaCo 5: 230–239. marocana) in Morocco: Conservation status of the rarest Prie´, V., P. Bousquet, A. Serena, E. Tabacchi, P. Jourde, B. bivalve in African fresh waters. Science of The Total Adam, T. Deschamps, M. Charneau, T. Tico, M. Bramard Environment 547:405–412. https://doi.org/10.1016/j. & G. Cochet, 2010. Nouvelles populations de Grande scitotenv.2016.01.003. Mulette Margaritifera auricularia (Spengler, 1793) (Bi- Stoeckle, B. C., R. Araujo, J. Geist, R. Kuehn, C. Toledo & A. valvia, Margaritiferidae) de´couvertes dans le sud-ouest de Machordom, 2016. Strong genetic differentiation and low la France. MalaCo 6: 294–297. genetic diversity of the freshwater pearl mussel (Margar- Prie´, V., L. Philippe & G. Cochet, 2011. Plan National d’Actions itifera margaritifera L.) in the southwestern European en faveur de la Grande Mulette Margaritifera auricularia – distribution range. Conservation Genetics. https://doi.org/ 2012-2017. Ministe`re de l’e´cologie, du de´veloppement 10.1007/s10592-016-0889-3. durable, des transports et du logement. Strayer, D. L. & H. M. Malcom, 2012. Causes of recruitment Prie, V., N. Puillandre & P. Bouchet, 2012. Bad taxonomy can failure in freshwater mussel populations in southeastern kill: molecular reevaluation of Unio mancus Lamarck, New York. Ecological Applications 22: 1780–1790. 1819 (Bivalvia: Unionidae) and its accepted subspecies. Villasante, F., J. Abad, R. Araujo & J. Balset, 2016. Aportacio´n Knowledge and Management of Aquatic Ecosystems. al conocimiento de la presencia de Margaritifera auricu- https://doi.org/10.1051/kmae/2012014. laria (Spengler, 1793) en el cauce del rı´o Tajo (Espan˜a). Prie´, V., Q. Molina & B. Gamboa, 2014. French Naiad (Bi- Iberus 34: 79–82. valvia: Margaritiferidae, Unionidae) species distribution White, G. C. & K. P. Burnham, 1999. Program MARK: survival models: prediction maps as tools for conservation. estimation from populations of marked animals. Bird Study Hydrobiologia 735: 81–94. 46: 120–138. Reis, J. & R. Araujo, 2009. Redescription of Unio tumidiformis Castro, 1885 (Bivalvia, Unionidae), an endemism from the

123

118 3.1.2. SCIENTIFIC REPORT 1: Conservation status of two French Margaritifera auricularia populations.

Résumé

Margaritifera auricularia (Spengler, 1793) est l'une des espèces de moules d'eau douce les plus menacées d'Europe. De nos jours, il ne reste que quelques populations en Espagne et en France. Afin de concevoir des mesures de conservation et de fournir des données de base pour des investigations ultérieures, nous avons étudié la répartition, l’abondance et la structure des populations de la Charente et de la Vienne et Creuse. Les densités les plus élevées dans les rivières étudiées ont été observées dans les zones urbaines ou près des rejets d’eaux usées, ce qui suggère que M. auricularia (même les jeunes individus) semble capable de tolérer un certain degré de pollution de l’eau. La distribution de fréquence des longueurs des coquilles indique que la population de la Charente est vieillissante et que le recrutement est presque inexistant. Dans cette rivière, bien que la comparaison avec les études précédentes doive être interprétée avec prudence, les résultats suggèrent que la population de M. auricularia semble être restée relativement stable au cours des 6 à 9 dernières années. Inversement, dans les rivières de la Vienne et de la Creuse, des preuves d'un important recrutement récent ont été trouvées, ce qui constitue l'événement le plus notable de la reproduction récente de M. auricularia signalé jusqu'à présent. Néanmoins, une étude comparative sur les abondances de cette population suggère un déclin de 45% au cours des 10 dernières années. Une évaluation du déplacement horizontal affiché pendant un an par des spécimens marqués dans les rivières Vienne et Creuse a confirmé que M. auricularia est essentiellement sédentaire. Cela suggère que le fait d’étudier le même endroit au cours de campagnes de suivi successives peut constituer une approche efficace pour évaluer les variations temporelles des populations. Compte tenu de la profondeur et des fortes densités de M. auricularia en la Charente, un relevé quantitatif des parcelles permanentes est proposé pour ce fleuve. Selon les caractéristiques de la population des rivières Vienne et Creuse, nous proposons que les futures actions de surveillance de ces populations soient basées sur le suivi d'individus géoréférencés et marqués avec pit-tags.

119

120 INTRODUCTION

The giant freshwater pearl mussel Margaritifera auricularia (Spengler, 1793) is considered the most endangered bivalve species in Europe (Araujo & Ramos, 2001; Prié, 2010a). Although likely widespread in most western European rivers at the beginning of the twentieth century, its decline in both abundance and distribution range has been estimated to be over 90% (Prié, 2010a) and it is now nearly extinct. Only a few populations remain in Spain (Ebro basin) and France (Loire, Charente, Garonne, and Adour basins) (Araujo & Ramos, 2000a; Cochet, 2001; Prié et al., 2018). It has been estimated that about 90% of the world population is found in France (Prié et al., 2011; Prié et al., 2018).

Since the end of 2014, the University of Tours, France, is leading the European LIFE13 project BIO/FR/001162 "Conservation of the Giant Pearl Mussel in Europe". To carry out a conservation program of this species in France, it is necessary to know the actual distribution and the conservation status of the populations. To assess the conservation status of freshwater mussel populations, it is desirable to use complete demographic studies (Hastie et al., 2000), but these require a larger investment in time and resources. Therefore, some authors propose to use the analysis of their density and size structure as indicators (Araujo & Ramos, 2000b; Prié et al., 2008). In this way, estimations on the size and structure of populations can be used to design conservation measures and to provide baseline data for further investigations to reveal changes in the studied population. Despite its importance, this information is limited for most French populations.

The two largest populations of Margaritifera auricularia in France are located in the Charente and Vienne-Creuse rivers (Cochet, 2002; Prié et al., 2008; Prié et al., 2018). The population from the river Vienne is one of the best studied in France. Its size was estimated by Cochet (2001) to be around 1,000 individuals. This author produced a map indicating the number of specimens found in different stations located along more than 54 km of river section (Cochet, 2002). The population of the Charente River has been much less studied because of the sampling difficulties imposed by the high depth, turbidity and flow velocities found in the habitats occupied by M. auricularia. However, Prié (2010b) estimated the population size to be about 100,000 individuals (the largest worldwide), based on the assessment of the density of specimens in 43 prospective stations located along more than 50 km of river section.

121 Concerning the population structure, it has been analyzed previously in the Charente and Vienne-Creuse rivers but only based on reduced sample sizes or estimated from shells of dead individuals (Cochet, 2001, 2002; Nienhuis, 2003; Prié, 2008, 2010b), which may result in an inaccurate interpretation of the actual situation.

Furthermore, the evolution of these populations has been little studied since their rediscovery. Therefore, there is insufficient information to support the selection of an appropriate methodology for detecting temporal changes in these challenging Margaritifera auricularia habitats. Moreover, results of individual samplings may be confounding, as the horizontal and vertical displacement capability of the species has never been described quantitatively, although Araujo & Ramos (2000b) described it as very sedentary.

The aim of this study was to 1) to assess basic ecological characteristics of these two populations such as distribution, abundance, and population structure, 2) to compare the results with the mentioned previous studies in order to detect temporal changes, 3) to describe the displacement capability of the species, and 4) to discuss the suitability of different methodologic approaches for future studies focused on detecting temporal changes in M. auricularia populations.

METHODS

Distribution, abundances and temporal changes

In summer 2016, extensive field surveys were performed in the Charente and Vienne-Creuse rivers in order to evaluate midterm temporal changes in abundances of Margaritifera auricularia populations. In both populations we studied almost all the sites where the higher densities of specimens were detected in previous studies (Cochet, 2002; Prié et al., 2008; Prié, 2010b). The stations surveyed in the Vienne-Creuse and the Charente are indicated in Figures 1 and 2, respectively. To facilitate comparison with previous studies, both the sampling methods (Cochet, 2002; Biotope, 2010) and the stations codes correspond to that of previous studies.

In the Vienne-Creuse rivers, sampling was done using bathiscopes and by snorkeling on the geographically delimited areas by Cochet (2002). We studied 14 stations distributed along a 54 km stretch located in the lower section of both rivers. In the lower section of the Charente River, 12 stations distributed along a 25 km stretch of the river were surveyed using scuba diving

122 methods. As described in the methodology by Prié (2010b), 20 m long transects located in the middle of the river were studied in each station. The transect was materialized by a cord fixed to the bottom of the river perpendicular to the flow of the current. Each transect was sampled by two divers. Each diver made a transect upstream and downstream of the rope with the arms extended in order to cover an area of 2 m. Thus, the density in each transect was calculated by counting the total number of individuals on the 20 m transect, that is, 80 m² of the bottom surface area.

In order to identify the specimens in future tracking actions, a small plastic numbered tag was glued on the left valve of all the specimens collected. Additionally, in the populations of the Vienne-Creuse rivers, the precise location of the marked specimens was recorded in 2016 using a centimeter-accurate GPS.

In order to evaluate the short-term changes of the population of these rivers, 4 stations, 2 at the Creuse and 2 at the Vienne, were revisited in summer 2017. Using a centimeter-accurate GPS, mussels were searched by snorkeling or with bathiscopes near the coordinates recorded in 2016. Individuals were identified by its numbered tags, described as live or death and their position was recorded. In order to determine the horizontal movements displayed by the mussels during one year, the distance between the 2016 and 2017 coordinates was calculated using the following equation:

Where Xb and Yb are the corresponding X and Y coordinates of 2017 and Xa and Ya are the corresponding X and Y coordinates of 2016.

Basic water quality parameters were obtained for the studied rivers, and the mean values corresponding to the years 2006-2014 were calculated. For the Charente River, the data correspond to the Taillebourg station and were downloaded from the “Système d'information sur l'eau du Bassin Adour-Garonne” (http://adour-adourgaronne.eaufrance.fr/). For the Vienne and Creuse rivers, the data correspond to the stations of Sauvegrain and Port de Piles respectively and were downloaded from the “Agence de l'eau Loire-Bretagne” (http://osur.eau- loire-bretagne.fr/exportosur/Accueil).

123

Figure 1. Location of studied stations on the Vienne and Creuse rivers (Map Source: Geoportail).

Figure 2. Location of surveyed stations on the Charente River (Map Source: Geoportail).

124 Population structure The characterization of the structure of the Charente and Vienne-Creuse populations was based on the biometric measurements made during the summer survey of 2016 and spring surveys performed between 2015 and 2018 in order to collect specimens for artificial reproduction (Soler et al., 2018 in section 3.2.1. of this thesis). Although in the Vienne and Creuse rivers these measurements were made in all the mentioned stations studied, in the Charente River this work was only done in the stations 30 and 35. Shell length of dead shells and living Margaritifera auricularia individuals was measured to the nearest 0.1 mm with Vernier calipers. All living specimens were carefully returned to the river in their original position after sampling aided by position markers (Figure 3) that were previously installed next to each captured individual. To infer the population structure and detect possible recent recruitment, a length- frequency distribution of 10 mm intervals was used.

Figure 3. Position markers installed next to each captured individual of Margaritifera auricularia. Length: 25 cm.

RESULTS AND DISCUSSION

Presence of the species

Twelve stations were studied in the Charente River between the 23th and 25th of June 2016. The visibility conditions were approximately of 1 m except in the stations 11 and 35 that presented low visibilities of about 50 cm. Specimens of Margaritifera auricularia observed in the Charente River were found between 4.2 and 7.5 m deep. The substrate of the studied transects was fairly homogeneous, usually formed by gravel and pebbles. However, in the stations 8 and 31 some parts of the transect were covered by sand. In these two stations, we

125 observed that in these sandy areas and therefore, less stable substrates, specimens of M. auricularia were poorly present. In areas with substrates formed by stones and gravel unclogged by fine sediments, the mussel densities were higher.

Table 1. Location of surveyed stations in the Charente River and density results of M. auricularia in 2016 and previous studies. (a quantitative densities from Prié, 2010b; b estimated data from a density distribution map from Prié 2010b; c quantitative data from Prié et al., 2008).

Density Geographic coordinates Density previous Station 2016 studies Latitude Longitude (ind/m2) (ind/m2) 8 45°42'49.13"N 0°33'33.02"O 0.275 0.400a 9 45°43'27.56"N 0°37'10.28"O 0.075 0.087a 10 45°43'50.69"N 0°38'20.89"O 0.100 0.100b 11 45°44'22.66"N 0°38'16.59"O 0.475 0.212a 19 45°47'29.71"N 0°38'52.63"O 0.025 0.020b 20 45°47'55.72"N 0°38'28.93"O 0.000 0.000b 22 45°48'41.31"N 0°38'34.18"O 0.013 0.020b 23 45°48'53.24"N 0°38'38.01"O 0.075 0.050b 24 45°49'21.32"N 0°38'23.31"O 0.050 0.050b 30 45°50'8.28"N 0°39'2.83"O 0.725 0.300c 31 45°50'11.64"N 0°39'11.16"O 0.087 0.230c 35 45°50'11.34"N 0°40'42.02"O 0.837 0.900c

The densities for each surveyed station in 2016 are provided in Table 1. The highest densities were observed at the stations located on urban areas: station 35 at Port d'Envaux, station 30 at Taillebourg, station 11 at Saintes, and station 8 at Chaniers. In the other hand, density values were much lower in the stations 19-24, at the sector between Saintes and Taillebourg (Figure 4). These results confirm the previous observations by Prié (2010b), who also reported the higher mussel densities in the same stations. In addition, Prié (2010b) surveyed a station located just in front of the Saintes treatment plant where he found one of the highest densities of Margaritifera auricularia in the Charente River. Although this may suggest a certain tolerance to moderate or poor water qualities (Prié, 2010b), further ecotoxicological studies are required in order to stablish the tolerance of M. auricularia to pollution at the adult stage and especially at the larval and juvenile stages.

126

Figure 4. Densities of Margaritifera auricularia in the Charente River observed during 2016 surveys (Map Source: Geoportail).

The water quality data of the studied rivers (Table 2) indicate that Margaritifera auricularia can tolerate moderate levels of nutrients as reflected by the mean values of sulphate and nitrate concentrations in the Charente River. This is corroborated by the data that Araujo & Ramos (2000b) reported for the Canal Imperial (Ebro basin, Spain), where concentrations of up to 373 mg/l of SO4 were recorded during low waters. On the other hand, it is important to highlight the wide tolerance of M. auricularia to different calcium concentrations and water conductivity values. Based on data from the only known population so far, Araujo & Ramos (2000b) defined the species as a hardwater species living in subsaline waters, since the calcium levels in the Canal Imperial ranged between 114 and 163 mg/l and the conductivity varied between 1077 and 1547 µS/cm. However, the Vienne and Creuse data, where the average calcium concentrations are 37.2 and 41.4 mg/l with conductivity values of 304.2 and 287.3 µS/cm, respectively (Table 2), indicate that M. auricularia can inhabit at a wide range of calcium concentrations and salinities.

127 Table 2. Mean values of basic water quality parameters for 2006-2014. For the Charente River, the data correspond to the Taillebourg station (30). For the Vienne and Creuse rivers, the data correspond to the stations of Sauvegrain (29) and Port de Piles (9) respectively.

Physico-Chemical Charente Creuse Vienne Parameter Dissolved oxygen (mg/l) 8.7 10.1 8.9 pH 8 7.8 7.8 Conductivity 25°C (µS/cm) 583.9 287.3 304.2 Ca2+ (mg/l) 94.6 41.4 37.2 - NO3 (mg/l) 25.8 9.1 12.2 2- SO4 (mg/l) 25.4 14.8 18.0

To detect temporal changes in these populations, we compared our data with previous studies. Unfortunately, quantitative data on the densities of Margaritifera auricularia observed by Prié (2010b) were only available for stations 8, 9 and 11. In the rest of stations, the densities were estimated based on a map produced by Prié (2010b) showing the distribution of M. auricularia densities (Figure 5), except for stations 30, 31 and 35 for which quantitative data were available from the study by Prié et al. (2008). These density values are provided in Table 1. In most stations, the densities observed in 2016 were very similar to those estimated by previous studies. Nevertheless, higher densities were observed in 2016 for stations 11 and 30 compared to those of previous studies. On the other hand, our results indicated lower densities in stations 8 and 31 when compared with former estimates.

Figure 5. Distribution map of Margaritifera auricularia densities in the Charente River estimated by Prié (2010b).

128 Although this comparison should be interpreted with caution, the results suggest that the population of Margaritifera auricularia in the Charente River seems to have remained relatively stable in the last 6-9 years. Furthermore, the data presented here constitute a baseline for future follow-up activities and account for the high densities of M. auricularia observed at the Charente River (Figure 6).

Nevertheless, locating the exact position of transects by scuba diving is very complicated at the Charente River because, even if the position is located on the surface by GPS, the margin of error when transferring the position of the transect to the bottom of the river can be of several meters. Therefore, sampling exactly the same location in successive follow-up campaigns is very difficult, and given the variability in the spatial distribution of the specimens, the use of these techniques can make the comparisons not very accurate. An approach that may be more effective in assessing temporal variations of the population of Margaritifera auricularia in the Charente River is the quantitative survey of permanent plots. This technique requires the installation of permanent grids that can be easily located, which are fixed to the substrate and allow studying the same area in successive sampling campaigns.

Figure 6. Margaritifera auricularia in the Charente River (Pictures from Gemosclera).

The Creuse and Vienne rivers were surveyed in August 2016. In these rivers, macrophytes developments are common in summer, which makes it difficult to find mussel by visual methods (Richard et al., 2015). The visibility conditions during the survey were exceptionally good since there were no macrophytes due to the strong floods occurred the previous spring. The prospected area in each station of the Creuse and Vienne rivers is presented in Figures 7- 17.

129

Figure 7. Stations 1 and 3 (Creuse). Figure 8. Station 9 (Creuse).

Figure 9. Station 10 (Vienne). Figure 10. Stations 13 and 14 (Vienne).

Figure 11. Station 18 (Vienne). Figure 12. Station 24 (Vienne).

130

Figure 13. Stations 29 and 30 (Vienne). Figure 14. Station 34 (Vienne).

Figure 15. Station 35 (Vienne). Figure 16. Station 38 (Vienne).

Figure 17. Station 8 (Vienne).

In 2016, 264 specimens of Margaritifera auricularia were found in the rivers Vienne and Creuse (Table 3). Interestingly, in most of the stations, they were distributed in an aligned way close to the shoreline (Figure 18). As occurred in the Charente, most of the individuals were found in gravel substrates. In this population, we observed a relative high proportion of juvenile specimens (see below).

131 Table 3. Number of specimens counted in the studied sampling stations of the Creuse and Vienne rivers in 2016 (this study) and 2002 (Cochet, 2002). (* In 2012, the A10 station was subject of a displacement of all the individuals within the framework of the construction of a bridge. **Autochthonous mussels + mussels transferred from station 10).

Difference Station River 2002 2016 2002-2016 (%) 3-Rhonne Creuse 22 43** - 1- Rhonne Creuse 10 9** - 9- Port de Piles Creuse Not studied 120** - 8-Anthony le tillac Vienne Not studied 18 - 10-A10 Vienne 42* 3 - 13-Mougon Vienne 16 6 -63 14-Mougon Vienne 24 5 -79 18-Tavant Vienne 10 3 -70 24-Riviere Vienne 8 0 -100 29-Sauvegrain Vienne 47 34 -28 30-Pont deviation Vienne 4 0 -100 Chinon 34-Thizay Vienne 11 13 +18 35-l'ile a Seguin Vienne 1 0 -100 38-Sain Germain Vienne 7 10 +43

Figure 18. Location of Margaritifera auricularia specimens (dots) at the station 29 (Sauvegrain) in 2016 recorded by centimeter-accurate GPS.

132 The most important populations were found on the Creuse River (stations 3 and 9) (Table 3). In station 9 (Port de Piles) 120 individuals, with a relative high proportion of specimens under 11 cm, were located, representing the 45% of the entire population of the Vienne-Creuse rivers (Table 3). In the middle of this station, there was a wastewater discharge which effect was evident by the development of algae downstream.

In the Vienne River, the greatest number of specimens was found in station 29 (Sauvegrain), located a few kilometers downstream of Chinon. The 34 individuals located in this station represent the 46% of the total Margaritifera auricularia specimens located in the Vienne River. This station is also in front of the discharge area of the treatment plant of Chinon. As noted before when describing the Charente and Creuse mussel populations, M. auricularia (even young individuals) seems capable of tolerating a certain degree of water pollution. Nevertheless, as we will see below, in the station 29, 13 specimens were lost between 2002 and 2016. Further ecotoxicological studies are required in order to investigate tolerance limits of M. auricularia to different pollutant concentrations, especially, in their early stages of life.

The stations 1, 3, and 9 on the Creuse River were strengthened in 2013 by transplanting specimens from the station 10 at the Vienne near the highway A10. Consequently, these three stations at the Creuse, as well as station 10 at the Vienne, were excluded from the following assessment of temporal changes in mussel abundances (Table 3). For the stations located on the Vienne (with the exception of station 10), there was a considerable decrease in the number of specimens found between 2002 and 2016 with the exception of stations 34 and 38, where we found more specimens in 2016. At the mentioned stations there was a decrease of 57 M. auricularia specimens, representing a 44.5% decline in 14 years.

It is important to note, however, that these results should be interpreted with caution as the probability of detection varies depending on factors such as visibility and sampling effort. This is clearly exemplified by the results obtained by Richard et al. (2015) in two stations of the Vienne River studied in 2013-2014. As shown in Table 4, these authors found considerably less mussels in both stations compared to our results of 2016, likely due to low visibilities caused by high macrophyte developments.

133 Table 4. Number of specimens registered in two sampling stations of the Vienne River in 2002 (Richard et al., 2015) and 2016 (this study).

Stations 2013/2014 2016 13 and 14 (Mougon) 2 11 29-Sauvegrain 4 34 Total 6 45

The low densities of Margaritifera auricularia found in the Creuse and Vienne rivers do not allow the above proposed method for monitoring the Charente population based on quantitative survey of permanent plots. Therefore, here we propose that future monitoring actions should be based on tracking precisely georeferenced and/or pit-marked individuals.

Population structures

The length of 352 living individuals (188 in the Charente and 164 in the Creuse and Vienne rivers) and 299 empty shells (115 from the Charente and 184 from the Vienne-Creuse rivers) were measured to the nearest 0.1 mm with Vernier calipers. Size of freshwater mussels are related to their age, so size-frequency histograms may give a preliminary idea of the age distribution of a mussel population. Nevertheless, studies using additional age structure information are more satisfactory, particularly in terms of providing detail on recent recruitment levels (Hastie et al., 2000). Although the age determination of the Charente and Vienne-Creuse populations is currently in progress, this research is not finished yet, thus the results are not given here except for some details that may serve to interpret the most significant results derived from size structure data.

The specimens from the Charente River population have a peculiarly small and conspicuously ear-like shell shape, contrasting to the specimens from the Vienne River population, which are larger and more elongated (Figures 19, 20). Haas (1917) reported similar differences in shell morphology of Margaritifera auricularia in the Ebro basin. He observed that individuals from sites with high current speeds (Ebro River at Mequinenza), had a more ear-like shaped shells compared to specimens from other sites with slower current velocities (Canal Imperial and Ebro River at Sástago). Variation of shell morphology is a common feature in freshwater mussels (Haag, 2012) and has been explained on the basis of genetic factors and/or environmental influences (Zieritz et al., 2010; Zajac et al., 2017). Nevertheless, many studies have reported a lack of match between intraspecific morphological and genetic patterns (Buhay et al., 2002; Machordom et al., 2003; Geist & Kuehn, 2005; Zieritz et al., 2010; Guarneri et al., 2014).

134 _ Figure 19. Margaritifera auricularia shell morphology. A) Creuse River; B) Charente River.

Figure 20. Lengths of dead and living individuals of Margaritifera auricularia from the Charente and Vienne-Creuse rivers.

Length–frequency distribution histograms of living animals and shells of dead specimens from the Charente population are presented in Figures 21 and 22 respectively.

135 Charente (n= 188) 60%

50%

40%

30%

20%

10% Percentage Percentage of individuals 0% 8_9 9_10 10_11 11_12 12_13 13_14 14_15 15_16 16_17 17_18 Length (cm)

Figure 21. Length–frequency distribution histogram of Margaritifera auricularia living specimens of in the Charente River (n= 188).

Charente (n= 115) 40% 35% 30% 25% 20% 15% 10%

Percentage Percentage of individuals 5% 0% 8_9 9_10 10_11 11_12 12_13 13_14 14_15 15_16 16_17 17_18 Length (cm)

Figure 22. Length-frequency distribution histogram of Margaritifera auricularia dead specimens of in the Charente River (n= 115).

In the Charente River, the length-frequency distribution histogram of living Margaritifera auricularia specimens shows a predominance of larger individuals between 11 and 13 cm and the absence of specimens smaller than 10 cm. The size distribution of dead individuals follow a similar pattern, although some smaller specimens (8-10 cm) were found, which could imply the presence of younger living individuals. Given the general smaller sizes of the specimens in this population, and the low proportion of individuals on the 10-11 cm size class compared to the 11-12 cm size class, it seems that recruitment of juveniles in this river is not occurring since several years ago. Nienhuis (2003) reported a similar population structure based on the

136 measurements of 42 living specimens although he found two individuals between 7 and 8 cm. He estimated that individuals between 11 and 12 cm (the most abundant size class he observed), were 29 years old. Nevertheless, this estimation was based on counting only the external annual rings of living specimens, a method that is considered to underestimate ages in bivalves (Neves & Moyer, 1998; Liu et al., 2017).

Prié et al. (2008) made a length-frequency distribution histogram based on the measurements of several tens of living specimens (red bars in Figure 23) very similar to the one presented in the Figure 21. Nevertheless, Prié (2010b), found a few live specimens between 4.2 cm and about 10 cm in stations 8, 9 and 11, although the exact number of these individuals was not reported. This finding implies that although scarce, there has been some recruitment recently at least in the most upstream area of the studied river section.

Figure 23. Charente population structure based on the measurement of living (red bars) and dead (blue bars) specimens (Prié et al., 2008).

Length-frequency distribution histograms of living animals and shells of dead specimens from the Vienne-Creuse population are presented in Figures 24 and 25 respectively.

137 Vienne & Creuse (n= 164) 30%

25%

20%

15%

10%

5% Percentage Percentage of infividuals

0% 8_9 9_10 10_11 11_12 12_13 13_14 14_15 15_16 16_17 17_18 Length (cm)

Figure 24. Length-frequency distribution histogram of Margaritifera auricularia living specimens of in the Vienne-Creuse rivers (n= 164).

Vienne & Creuse (n= 184) 50% 45% 40% 35% 30% 25% 20% 15% 10% Percentage Percentage of individuals 5% 0% 8_9 9_10 10_11 11_12 12_13 13_14 14_15 15_16 16_17 17_18 Length (cm)

Figure 25. Length-frequency distribution histogram of Margaritifera auricularia dead specimens of in the Vienne and Creuse rivers (n= 184).

Cochet (2001) assessed the Vienne-Creuse population structure based on the measurement of 292 dead specimens (Figure 26) concluding that there were a lack of recent recruitment in these rivers. The length-frequency distribution histogram of dead animals reported by Cochet (2001) is very similar to that obtained in our study with empty shells, and reveals a mortality in almost all size classes, although with greater representation of the older specimens. Nevertheless, since we ignore the time elapsed between the finding of these valves and the death of the specimens, it is not possible to evaluate mortality rates. Nienhuis (2003) produced a size-frequency distribution histogram based on only 19 living specimens in which more than 60% of the

138 individuals were larger than 14 cm. The smallest size class found in that study was 11-12 cm with only one specimen.

Figure 26. Vienne-Creuse population structure based on the measurement of 292 dead specimens of Margaritifera auricularia (from Cochet, 2001).

The results obtained in our study with living individuals differ markedly from previous results. As shown in Figure 24, the sizes of living specimens from the Vienne and Creuse rivers range between 8 and 18 centimeters, with 34% of the individuals belonging to size classes smaller than 12 cm. On the other hand, the low proportion of individuals in the size classes between 11 and 14 cm seems to indicate that this population experienced a recruitment failure during several years or decades, followed by a relatively recent reproduction event. In all the studied stations, with the exception of station 10 (where the individuals were translocated to the stations 1, 3 and 9), specimens smaller than 11 cm were found, indicating that this latest recruitment has occurred along the entire river section. A preliminary study of the age of the individuals based on counting external and internal growth rings of the shells of this population, suggests that the 10-11 cm size class corresponds to an age of 15-20 years (Nakamura et al., 2018; Soler, unpublished data). This implies that, after a time period without effective reproduction, an important recruitment event occurred between the late 1990s and the beginning of 2000. The small size of the mussels of this cohort, when the previous studies were conducted, would explain why the presence of small specimens was not detected. Nevertheless, this age study is still ongoing and the preliminary data offered here should be taken with caution.

Because juveniles are very small and inconspicuous in the river bed sediments, it is very difficult to detect them by visual methods, and therefore conventional surveys are strongly

139 biased against small size classes (Haag, 2012). In order to obtain unbiased estimates of size structure, excavation of substrate and its examination across a series of sieves is required to allow consistent detection of small individuals (Miller & Payne, 1988; Haag & Warren, 2007). In order to avoid a supplementary stress for these endangered populations, sediment samples were not taken in search of younger individuals in this study. Thus, the lack of smaller size classes in both studied populations could be due to a recruitment failure in the last years or to an artifact of the sampling strategy.

Horizontal movement and annual mortality rate assessment

Between the days 14 and 19 August of 2017, four different stations (2 in the Creuse and 2 in the Vienne) were revisited in order to make a real estimation of both the annual mortality rates of this population and the horizontal displacement of the specimens that were georeferenced in August 2016. During this survey, dense developments of macrophytes were observed covering large areas of Margaritifera auricularia habitats, which seems to be common during summer in the Creuse and Vienne rivers (Richard et al., 2015).

Of the 140 specimens studied that were already marked with a plastic label in 2016, one year later we collected 122 living and 2 dead specimens. The remaining 16 specimens could not be located. Although a more thorough analysis is still in progress, these preliminary data suggest that the annual mortality rate at these stations investigated could be between 1 and 13%, which is in line with the 44.5% decline in 14 years.

Despite the possible errors that may be associated to the georeferencing system, this study demonstrates that 71% (87 specimens) of the studied mussels moved less than 20 cm in one year (Table 5). Only 15 mussels (12%) were found at a distance larger than 35 cm, and 8 specimens (7%) at a distance greater than 50 cm. Three individuals were located more than 1 m away its 2016 position. The largest displacement was 1.8 m and was recorded in the station 29 (Sauvegrain). As shown in Table 5, the higher proportion of mussels that changed its position was observed in the station 29 (Sauvegrain). This may be related to the different macrophyte species composition observed in this station. The macrophyte community in this station was dominated by Cladophora sp. in contrast to the rest of the studied stations, where Ranunculus sp. and Myriophyllum sp. were the dominant taxa. Although no specific sampling was performed, it was observed that dense Cladophora masses covering the mussels were limiting

140 the water flow reaching the naiads. Thus, the possible anoxic conditions created by the proliferation of these algae could be responsible for the more frequent displacements observed in this station.

Table 5. Displacements observed in M. auricularia specimens from the Vienne and Creuse rivers between August 2016 and August 2017.

Mean distance Distance Distance Distance River Station Mussels (cm) (SD) > 20 cm > 35 cm > 50 cm 9-Port de Piles 36 15 (15,80) 6 (17%) 2 (6%) 2 (6%) Creuse 1 and 3- Rhonne 49 14,9 (15,03) 10 (20%) 4 (8%) 2 (4%) 8-Anthony le tillac 14 12,5 (9,31) 3 (21%) 0 (0%) 0 (0%) Vienne 29-Sauvegrain 23 40,3 (43,38) 16 (70%) 9 (39%) 4 (17%) Total 122 19,51 (24,81) 35 (29%) 15 (12 %) 8 (7%)

These data suggest that Margaritifera auricularia is mostly sedentary, as previously observed by Araujo & Ramos (2000b). Nevertheless, it can perform reduced displacements that seems to be more frequent under stressful conditions. These movements should be more important during low flow conditions when water velocities are not too high to be dragged the specimens downstream. These data suggest that given the reduced mobility of M. auricularia, georeferencing individuals could be a valuable method for tracking actions in these rivers. Nevertheless, the use of pit tags can improve the effectiveness of this method, so a combined use of both methods is recommended.

ACKNOWLEDGEMENTS

This work was conducted within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’. The paper is also issued under the auspices of the UNESCO Chair ‘Fleuves et Patrimoine-River Culture’. Thanks to the Charente-Maritime and Indre-Loire Prefectures for collection permits. Diving surveys were performed by Gemosclera and the Club Chinon Plongé. This work would not have been possible without the help of Beatriz Ontín, who collaborated during the 2016 and 2017 surveys of the Vienne and Creuse rivers. In these rivers, I also had the help of Philippe Jugé, Miguel Gailledrat, Margot Sicot, Valentin Viennot and Laurent Philippe and other members of the Biotope staff.

141 REFERENCES

ARAUJO, R. & RAMOS, M. A. 2000a. A critical revision of the historical distribution of the endangered Margaritifera auricularia (Spengler, 1793) (Mollusca: Margaritiferidae) based on museum specimens. Journal of Conchology, 37: 49–59.

ARAUJO, R. & RAMOS, M. A. 2000b. Status and conservation of the giant European freshwater pearl mussel (Margaritifera auricularia) (Spengler, 1793) (Bivalvia: Unionoidea). Biological Conservation, 96: 233–239.

ARAUJO, R. & RAMOS, M. A. 2001. Action plan for Margaritifera auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Nature and environment, No. 117. Council of Europe Publishing, Strasbourg.

BUHAY, J. E., SERB, J. M., DEAN, C. R., PARHAM, Q. & LYDEARD, C. 2002. Conservation genetics of two endangered unionid bivalve species, Epioblasma florentina walkeri and E. capsaeformis (Unionidae: Lampsilini). Journal of Molluscan Studies, 68: 385– 391.

COCHET, G. 2001. Redécouverte d’une population vivante de la Grande Mulette, Margaritifera auricularia, sur la Vienne et la Creuse. Recherches Naturalistes en Région Centre, 10: 3–16.

COCHET, G. 2002. La Grande Mulette (Margaritifera auricularia) dans la Vienne et la Creuse. DIREN Centre & DIREN Poitou-Charente, rapport et atlas: 40 + 26 pp.

GEIST, J. & KUEHN, R. 2005. Genetic diversity and differentiation of central European freshwater pearl mussel (Margaritifera margaritifera L.) populations: implications for conservation and management. Molecular Ecology, 14: 425–439.

GUARNERI, I., POPA, O. P., GOLA, L., KAMBURSKA, L., LAUCERI, R., LOPES-LIMA, M., POPA, L. O. & RICCARDI, N. 2014. A morphometric and genetic comparison of Sinanodonta woodiana (Lea, 1834) populations: does shape really matter? Aquatic Invasions, 9: 183–194.

142 HASTIE, L. C., YOUNG, M. R., BOON, P. J., COSGROVE, P. J. & HENNINGER, B. 2000. Sizes, densities and age structures of Scottish Margaritifera margaritifera (L.) populations. Aquatic Conservation: Marine and freshwater ecosystems, 10: 229–247.

LIU, S., WROBLEWSKI, J. S. & SCHNEIDER, D. C. 2017. Comparison of three methods for aging the Iceland Scallop Chlamys islandica. Journal of Shellfish Research, 36: 607–614.

MACHORDOM, A., ARAUJO, R., ERPENBECK, D. & RAMOS, M. A. 2003. Phylogeography and conservation genetics of endangered European Margaritiferidae (Bivalvia: Unionoidea). Biological Journal of the Linnean Society, 78: 235–252.

NAKAMURA, K., CUCALA, L., MESTRE, A., MESQUITA-JOANES, F., ELBAILE, E., SALINAS, C. & MUÑOZ-YANGUAS, M. A. 2018. Modelling growth in the critically endangered freshwater mussel Margaritifera auricularia (Spengler, 1793) in the Ebro basin. Hydrobiologia, 810: 375–391. Doi: 1010071075001731039

NEVES, R. J. & MOYER, S. N. 1988. Evaluation of techniques for age determination of freshwater mussels (Unionidae). American Malacological Bulletin, 6: 179–188.

NIENHUIS, J. A. J. H. 2003. The rediscovery of Spengler’s freshwater pearlmussel Pseudunio auricularius (Spengler, 1793) (Bivalvia, Unionoidea, Margaritiferidae) in two river systems in France, with an analysis of some factors causing its decline. Basteria, 67: 67–86.

PRIÉ, V., 2010a. Margaritifera auricularia. In: The IUCN Red List of Threatened Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN.UK.2010- 4.RLTS.T12798A3381899.en

PRIÉ, V. 2010b. La Grande Mulette Margaritifera auricularia sur la Charente, complément d’inventaire sur le site Natura 2000 FR 5400-472. Ligue pour la Protection des Oiseaux – Biotope, France.

PRIÉ, V., PHILIPPE, L. & COCHET, G. 2011. Plan National d’Actions en faveur de la Grande Mulette Margaritifera auricularia – 2012-2017. Ministère de l’écologie, du développement durable, des transports et du logement, Paris.

143 PRIÉ, V., COCHET, G., PHILIPPE, L., RETHORET, H. & FILALI, R. 2008. Une population majeure de la très rare Grande Mulette Margaritifera auricularia (Spengler, 1793) (Bivalvia : Margaritiferidae) dans le fleuve Charente (France). MalaCo, 5: 230–239.

PRIÉ, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., PATRY, N., ADAM, B., LEGRAND, N., JUGÉ, P., RICHARD, N. & WANTZEN, K. 2018. Margaritifera auricularia (Bivalvia: Unionida), a critically endangered flagship species for challenging downstream ecosystems exploration: results of last 10 years’ dedicated surveys. Hydrobiologia, 810: 157– 175.

RICHARD, N., JUGE, PH., RODRIGUES, S. 2015. Impacts de l'arasement du barrage de Maisons-Rouges sur la dynamique sédimentaire ainsi que sur les habitats et espèces faunistiques et floristiques de la Vienne et de la Creuse. Campagne 2012-15. CETU Elmis Ingénieries de l’Université François-Rabelais, Tours.

SOLER J., WANTZEN, K.M., JUGÉ, P. & ARAUJO, R. 2018. Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84: 182–189.

ZAJAC, K., ZAJAC, T. & CMIEL, A. 2017. What can we infer from the shell dimensions of the thick-shelled river mussel Unio crassus? Hydrobiologia,DOI 10.1007/s10750-017-3098-2

ZIERITZ, A., HOFFMAN, J. L., AMOS, W. & ALDRIDGE, D. C. 2010. Phenotypic plasticity and genetic isolation-by-distance in the freshwater mussel Unio pictorum (Mollusca: Unionoida). Evolutionary Ecology, 24: 923–938.

144 3.1.3. ARTICLE 2: Rhodeus amarus (Bloch, 1782): a new potential threat for Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

Résumé

Rhodeus amarus est un cyprinidé qui fraie dans la cavité du manteau des moules d'eau douce, ce qui entraîne un coût du fitness pour les moules. Étant donné que R. amarus étend son aire de répartition en Europe, son contact avec différentes espèces de moules d'eau douce augmente également. L’expansion considérable de cette espèce et la réduction du fitness qu’elle pourrait causer sur les moules peuvent être problématiques pour les espèces menacées d’extinction. Nous examinons ici les espèces de moules précédemment signalées comme hôtes de R. amarus et leur état de conservation. Nous rapportons également l'utilisation de M. auricularia, espèce en danger critique d'extinction, par R. amarus pour la première fois et nous signalons le danger potentiel de son expansion dans de nouvelles zones habitées par des espèces de moules avec des distributions spatiales réduites.

145 Running head: R.amarus: a threat for M. auricularia

Rhodeus amarus (Bloch, 1782): a new potential threat for Margaritifera auricularia

(Spengler, 1793) (Unionoida, Margaritiferidae)

Joaquín Soler, 1, 3 Karl M. Wantzen1, 2 and Rafael Araujo 3

1 Université François Rabelais, UMR 7324 – CITERES, 33 Allée Ferdinand de Lesseps, 37204

Tours cedex 03 France

2 UNESCO River Culture – Fleuves et Patrimoines Chair, Université François Rabelais, UMR

7324 – CITERES, 33 Allée Ferdinand de Lesseps, 37204 Tours cedex 03 France

3 Museo Nacional de Ciencias Naturales – C.S.I.C. c/ José Gutiérrez Abascal 2, 28006 Madrid,

Spain

Corresponding author:

Joaquín Soler e-mails: [email protected]; [email protected]; [email protected]

146 ABSTRACT

Rhodeus amarus is a cyprinid fish that spawns in the mantle cavity of freshwater mussels, resulting in a fitness cost for the mussels. Because R. amarus is expanding its geographic range in Europe, its contact with different freshwater mussel species is also increasing. The extensive expansion of this bitterling species, and the reduction in fitness that it could cause on mussels, may be problematic for species facing extinction risk. Here, we review the mussel species previously reported as hosts for R. amarus and their conservation status. We also report the utilization of the critically endangered M. auricularia by R. amarus for the first time and point out the potential danger of its expansion into new areas inhabited by spatially restricted mussel species.

Keywords: Rhodeus amarus, bitterling, Margaritifera auricularia, threat, endangered, freshwater mussels, conservation, invasive fish

MAIN TEXT

The European bitterling Rhodeus amarus (Bloch 1782) is a small cyprinid fish belonging to the subfamily Acheilognathinae that has an unusual life cycle characterized by its obligatory symbiosis of spawning in freshwater mussels. During the fish spawning season (April–August), female bitterlings develop long ovipositors that are used to introduce eggs (2.4–3.1 mm in length) into the gills of freshwater mussels through the mussel’s exhalant siphon (Smith et al. 2004). Male bitterlings then fertilize the eggs by releasing sperm into the mussel’s inhalant aperture. Female bitterlings release between one and six eggs per spawning, and, at the peak of the breeding season, multiple spawns can result in a single mussel hosting up to 40 to 50 eggs, although densities over 250 eggs have been recorded (Reichard et al. 2004; Smith et al. 2004). Bitterling embryos hatch approximately 36 hours after fertilization and, at this stage, are approximately 3.3 mm long. Four days after hatching, embryos start to move inside of the mussel’s gills. They continue to develop within the gills for approximately one month until their yolk sac has been absorbed. At this stage, bitterling larvae are approximately 10 mm long and exit the mussel via the exhalant siphon (Smith et al. 2004).

This symbiotic relationship was originally proposed as potentially mutualistic (Smith et al. 2004, Reichard et al. 2010), in which bitterlings use mussels as spawning sites where their

147 embryos develop to larval stages, and mussels use bitterlings as hosts for glochidia, their parasitic larvae. However, recent research suggests that European bitterlings should be regarded as parasites of mussels. Experimental evidence has shown that R. amarus rarely host glochidia of freshwater mussels (Kadlec et al. 2003, Reichard et al. 2006, Douda et al. 2013). Nevertheless, the presence of bitterling embryos in mussels causes a fitness cost for the hosts: they compete with the host for oxygen (Smith et al. 2001) and reduce water circulation over the mussel’s gills, potentially affecting their ability to filter feed (Mills et al. 2005). These effects may account for the significantly reduced growth rates of mussels infected with bitterling embryos (Reichard et al. 2006). In freshwater mussels, size is positively correlated with fecundity (Bauer 1994); therefore, a reduction in growth represents a fitness cost for mussels. In addition, mussels parasitized with multiple clutches of bitterling eggs under laboratory conditions experience high mortality; however, this observation has not been reported in wild populations (Spence and Smith 2013).

In this parasitic relationship, an arms race has developed between bitterlings and their hosts: mussels have evolved adaptations to eject bitterling eggs and embryos, while bitterling embryos have developed counter adaptations to avoid ejection (Mills and Reynolds 2002, Smith et al. 2004, Reichard et al. 2015). Evolution of these adaptations requires that host and parasite are in sympatry over a long period of time. Although R. amarus is considered a native species over much of its present-day range in Europe, recent studies have demonstrated that it expanded relatively recently (centuries to millennia before the present day) from the Black Sea region into central and western Europe (Bohlen et al. 2006, Van Damme et al. 2007, Bryja et al. 2010). During its expansion, R. amarus successfully used mussel species present in its historic range as hosts, such as Anodonta anatina (Linnaeus 1758), A. cygnea (Linnaeus 1758), Pseudanodonta complanata (Rossmässler 1835), Unio pictorum (Linnaeus 1758), U. tumidus Retzius, 1788 (Wiepkema 1961, Balon 1962, Reynolds et al. 1997; Smith et al. 2000; Smith et al. 2004) and U. crassus Philipsson, 1788 (Reichard et al. 2010; Tatoj, et al. 2017). Compared with unionid populations in the Pontic region, native mussel populations across continental Europe have limited adaptations to avoid or eject R. amarus eggs, further favoring the expansion of this bitterling species. Reichard et al. (2010) suggested that the shorter duration of sympatry of this species with mussels in continental Europe than in the Pontic region likely accounts for the lack of highly specialized adaptations.

148 Following its dramatic decline in abundance from 1960 to 1980, R. amarus was declared an endangered species in central and western Europe and included in Appendix II of the Habitat Directive (92/43/EEC). Nevertheless, since 1980, the distribution of R. amarus has expanded in many parts of Europe, particularly in eastern Europe where a considerable and rapid increase in abundance has also been observed (Kozhara et al. 2007). In France, this species was restricted to only the northeastern basins during the 19th century (Valenciennes 1848, Gehin 1868, Gensoul 1908); however, nowadays it has spread nearly throughout the entire country, especially in the southwest, where it is considered invasive (Kottelat and Freyhof 2007). The distribution of R. amarus has also spread in countries where it was recently introduced, such as England (Wheeler and Maintland 1973, Lever 1977), Denmark (Møller and Menne 1998, Møller 2012) and Italy (Confortini 1992; Carosi et al. 2016).

Two explanations have been proposed for the recent expansion of R. amarus: anthropogenic activities, including its introduction by anglers, aquarists and artificial connections of waterway systems, and climate change (Kozhara et al. 2007, Van Damme et al. 2007). Increases in temperature may generally favor the expansion of R. amarus as it is a relatively thermophilic species with an optimum temperature for reproduction (15–21 ºC) that is considerably higher than that of most other European freshwater fishes (reviewed in Souchon and Tissot 2012). In fact, predictions of the future prevalence of the species under different climate change scenarios for 2080 also suppose a further increase in its distribution range, mainly through the colonization of colder upstream habitats (Buisson et al. 2008).

The recent expansion of R. amarus into new habitats and new geographical areas has led to its contact with different freshwater mussel species. This bitterling species appears able to spawn in all unionid mussels with which it shares a habitat, as clearly exemplified by its successful colonization of the Bronx River (New York, USA) where it spawns in the Nearctic unionids

Pyganodon cataracta (Say 1817) and Elliptio complanata (Lightfoot 1786) (Smith et al. 2004). Following its recent expansion through western Europe, R. amarus was observed to ovoposit in Unio mancus Lamarck, 1819 and Potomida littoralis (Cuvier 1798) (Prié 2017).

The extensive expansion of this bitterling species, and the reduction in fitness that it could cause, may be problematic for mussel species at risk of extinction as in the case of the European giant freshwater pearl mussel Margaritifera auricularia (Spengler 1793), which is considered to be the most imperiled bivalve species in Europe (Araujo and Ramos 2001, Prié 2010). Although likely widespread in most western European rivers at the beginning of the 20th

149 century, its decline in both abundance and distribution has been estimated to be over 90% (Prié 2010). This species is now thought to be nearly extinct with only a few remaining populations known in Spain (Ebro Basin) and France (Loire, Charente, Garonne and Adour basins) (Araujo and Ramos 2000; Prié et al. 2018).

During the spring of 2016 and 2017, 39 R. amarus individuals were captured by electrofishing immediately downstream of M. auricularia populations in the Charente and Creuse rivers (France). Their gills were inspected for infestation by mussel larvae; however, none of the captured fish had glochidia. On August 18, 2016, a small M. auricularia specimen (110 mm in length) that had recently died was found with its soft tissues still in good condition. This specimen was found at the same locality where R. amarus were captured in the Creuse River. An inspection of the mussel’s gills revealed the presence of five R. amarus embryos (between 8.5 and 9.5 mm in total length) distributed in both the inner and outer demibranchs (Fig 1). On May 8, 2017, two M. auricularia adults (both >150 mm in length) were collected from the same Creuse river locality to obtain glochidia for captive breeding (Soler et al. 2018). Mussels were transported to the laboratory and transferred to 100L aquaria filled with aerated river water. Several days later, after checking for gravidity using a disposable syringe (Soler et al. 2018), one of the mussels ejected seven R. amarus eggs and four embryos at different development stages (embryos were between 5.8 and 9.5 mm in total length).

Figure 1: Two R. amarus embryos (e) in the gills of a M. auricularia specimen from the Creuse River in France

150 This finding constitutes the first record of R. amarus using a Margaritiferidae species as a host. In East Asia, R. sericeus, a sister species of R. amarus, has been reported to use Margaritifera dahurica (Middendorff 1850) (Smith and Hartel 1999, Klishko 2012; Klishko and Bogan 2013) and M. laevis (Haas 1910) (Zhulkov and Nikiforov 1988). Given that the gill anatomy of margaritiferids is different from that of unionids (Soler et al. 2018), this finding supports the idea that European bitterlings can parasitize all native European mussel species regardless of their gill anatomy.

In central European rivers, native mussels have coexisted with bitterlings for hundreds to thousands of years without any apparent decline in populations due to the presence of bitterlings. However, European freshwater mussels are currently in decline, and 12 of the 16 currently recognized species are categorized as Threatened or Near Threatened by the IUCN (Lopes-Lima et al. 2017) (Table 1). Mussel populations that have experienced a drastic decline in density may be particularly impacted by the presence of R. amarus as individual mussels could ultimately host a greater number of bitterling embryos, which may represent an additional stress to these mussels (Van Damme et al. 2007, Prié 2017, Tatoj et al. 2017).

The evidence to date shows that 9 of the 16 European mussel species can serve as host forR. amarus (Table 1). The seven species that have not, as yet, been reported as hosts include

Margaritifera margaritifera (Linnaeus 1758), Unio tumidiformis Castro 1885, U. elongatulus Pfeiffer 1825, U. delphinus Spengler 1793, U. ravoisieri Deshayes 1847, U. gibbus Spengler 1793 and Microcondylaea bonellii (Ferussac 1827). Despite the potential spread of R. amarus through upstream rivers, its preference for slow flowing waters (Carosi et al. 2016) suggests it is unlikely to come into contact with M. margaritifera, another European margaritiferid species also classified as Critically Endangered by the IUCN, which inhabits fast flowing streams. However, special attention should be given to prevent the introduction of the European bitterling to the peninsulas of southern Europe. Despite having lower species richness, these basins serve as habitats for spatially restricted mussel species (including U. tumidiformis, U. elongatulus, U. delphinus, U. ravoisieri, U. gibbus and M. bonellii) and, therefore, are of high conservation priority.

151 1 Table 1: Conservation status of European freshwater mussels and their relationship with R. amarus.

Overlapping distribution with R. Reported as host of R. amarus Species Conservation Status amarus historic range (see text) (References) Anodonta anatina (Linnaeus, 1758) Yes Yes (Reynolds et al. 1997) Least concern (a) A. cygnea (Linnaeus, 1758) Yes Yes (Reynolds et al. 1997) Near threatened (a) Pseudanodonta complanata (Rossmässler, 1835) Yes Yes (Smith et al. 2004) Near threatened (a) Unio pictorum (Linnaeus, 1758) Yes Yes (Balon, 1962) Least concern (a) U. tumidus Retzius, 1788 Yes Yes (Wiepkema, 1961) Least concern (a) U. crassus Philipsson, 1788 Yes Yes (Tatoj et al. 2017) Endangered (b) U. mancus Lamarck, 1819 No Yes (Prié, 2017) * Near threatened (a) Potomida littoralis (Cuvier, 1798) No Yes (Prié, 2017) Endangered (c) Margaritifera auricularia (Spengler, 1793) No Yes (This study) Critically endangered (a) U. delphinus Spengler, 1793 No No Near threatened (a) U. tumidiformis Castro, 1885 No No ** Vulnerable (a) U. gibbus Spengler, 1793 No No Critically endangered (a) U. ravoisieri Deshayes, 1847 No No * Near threatened (a) U. elongatulus Pfeiffer 1825 No No Near threatened (a) Margaritifera margaritifera (Linnaeus, 1758) No No Critically endangered (a) Microcondylaea bonellii (Ferussac, 1827) No No Vulnerable (a) 2 * As Unio elongatulus

3 ** As Unio crassus

4 (a) Cuttelod et al. (2011)

5 (b) Lopes-Lima et al. (2014a)

6 (c) Lopes-Lima et al. (2014b)

152 ACKNOWLEDGEMENTS

This work was supported by the LIFE+ project “LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe”. We thank the anonymous reviewers for their valuable comments and suggestions, and Melinda Modrell, who reviewed the English version of the manuscript.

LITERATURE CITED

Araujo, R. and M. A. Ramos. 2000. A critical revision of the historical distribution of Margaritifera auricularia (Spengler, 1793) (Mollusca: Margaritiferidae) based on museum specimens. Journal of Conchology 37:49–59.

Araujo, R. and M. A. Ramos. 2001. Action Plan for Margaritifera auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Council of Europe Publishing. Nature and environment, 117. Strasbourg.

Balon, E. K. 1962. Note on the number of Danubian bitterlings developmental stages in mussels. Vestnik Ceskoslovenske Spolecnosti Zoologicke 26:250–256.

Bauer, G. 1994. The adaptive value of offspring size among freshwater mussels (Bivalvia; Unioidea). Journal of Animal Ecology 63:933–944.

Bohlen, B., V. Klechtová, N. Bogutskaya and F. Freyhof. 2006. Across Siberia and over Europe: Phylogenetic relationships of the freshwater fish genus Rhodeus in Europe and the phylogenetic position of R. sericeus from the River Amur. Molecular Phylogenetics and Evolution 40:856–865.

Bryja, J., C. Smith, A. Konecný and M. Reichard. 2010 Range-wide population genetic structure of the European bitterling (Rhodeus amarus) based on microsatellite and mitochondrial DNA analysis. Molecular Ecology 19:4708–4722.

Buisson, L., W. Thuiller, S. Lek, P. Lim and G. Grenouillet. 2008. Climate change hastens the turnover of stream fish assemblages. Global Change Biology 14:2232–2248.

153 Carosi, A., L. Ghetti, C. Cauzillo, L. Pompei and M. Lorenzoni. 2016. Occurrence and distribution of exotic fishes in the Tiber River basin (Umbria, central Italy). Journal of Applied Ichthyology 33:274–283.

Confortini, I. 1992. Presenza del rodeo amaro, Rhodeus sericeus (Pallas, 1776) nel Fiume Menago (provincia di Verona) (Pisces, Cyprinidae). Bollettino del Museo Civico di Storia Naturale di Verona 16:329–332.

Cuttelod, A., M. Seddon and E. Neubert. 2011. European Red List of Non-marine. Molluscs. Luxembourg: Publications Office of the European Union.

Douda, K., M. Lopes-Lima, M. Hinzmann, J. Machado, S. Varandas, A. Teixeira and R. Sousa. 2013. Biotic homogenization as a threat to native affiliate species: fish introductions dilute freshwater mussel’s host resources. Diversity and Distributions 19:933–942.

Gehin, J. B., 1868. Révison des Poissons qui vivent dans les cours d’eau et dans les étangns du département de la Moselle. Avec quelques considérations sur le Darwinisme. Metz.

Gensoul, J., 1908. Monographie des Poissons du département de la Saône-et Loire, Autun.

Kadlec, D., A. Simkova, J. Jarkovsky and M. Gelnar. 2003. Parasite communities of freshwater fish under flood conditions. Parasitology Research 89:272–283.

Klishko, O. K., 2012. Some data on reproductive biology of the freshwater mussels (Margaritiferidae, Unionidae) and their relationships with bitterlings (Cyprinidae) in Transbaikalye. The Bulletin of the Russian Far East Malacological Society 15:31–55.

Klishko, O. K. and A. E. Bogan, 2013. The conservation status of the freshwater Pearl Mussel Margarititifera dahurica in Far Eastern Russia. Ellipsaria 15:31–33.

Kottelat, M., and F. Freyhof. 2007. Handbook of European Freshwater Fishes. Kottelat, Cornol and Freyhof, Berlin.

Kozhara, A. V., A. V. Zhulidov, S. Gollasch, M. Przybylski, V. G. Poznyak, D. A. Zhulidov and T. Yu. Gurtovaya. 2007. Range extension and conservation status of the

154 bitterling, Rhodeus sericeus amarus in Russia and adjacent countries. Folia Zoologica 56:97–108.

Lever, C. 1977. The naturalized animals of the British Isles. Hutchinson and Co., London.

Lopes-Lima, M., I. U. Kebapc and D. Van Damme. 2014a. Unio crassus. IUCN red list of threatened species. Version 2015.2. Available at http://www.iucnredlist.org/. Accessed 16.04.2018.

Lopes-Lima, M., V. Prié and M. B. Seddon. 2014b. Potomida littoralis. IUCN red list of threatened species. Version 2015.2. Available at http://www.iucnredlist.org/. Accessed 16.04.2018.

Lopes-Lima, M., R. Sousa, J. Geist, D. C. Aldridge, R. Araujo, J. Bergengren, Y. Bespalaya, E. Bodis, L. Burlakova, D. Van Damme, K. Douda, F. Froufe, D. Georgiev, C. Gumpinger, A. Karatayev, U. Kebapci¸ I. Killeen, J. Lajtner, B. M. Larsen, R. Lauceri, A. Legakis, S. Lois, S. Lundberg, E. Moorkens, G. Motte, K. O. Nagel, P. Ondina, A. Outeiro, M. Paunovic, V. Prie, T. Von Proschwitz, N. Riccardi, M. Rudzite, M. Rudzitis, C. Scheder, M. Seddon, H. Sereflisan, V. Simic, S. Sokolova, K. Stoeckl, J. Taskinen, A. Teixeira, F. Thielen, T. Trichkova, S. Varandas, H. Vicentini, K. Zajac, T. Zajac and S. Zogaris. 2017. Conservation status of freshwater mussels in Europe: state of the art and future challenges. Biological Reviews 92:572–607.

Mills, S. C. and J. D. Reynolds. 2002. Host species preferences by bitterling, Rhodeus sericeus, spawning in freshwater mussels and consequences for offspring survival. Animal Behavior 63:1029–1036.

Mills, S. C., M. I. Taylor and J. D. Reynolds. 2005. Benefits and costs to mussels from ejecting bitterling embryos: a test of the evolutionary equilibrium hypothesis. Animal Behaviour 70:31–37.

Møller, P.R. 2012. Europæisk bitterling. Pages 265–272 in H. Carl and P. R. Møller (editors). Atlas over danske ferskvandsfisk. Statens Naturhistoriske Museum, Denmark

Møller, P.R. and T. Menne. 1998: Bitterling Rhodeus sericeus (Pallas, 1776) – First record from Denmark. Flora og Fauna 104:29–34.

155 Prié, V., 2010. Margaritifera auricularia. The IUCN Red List of Threatened Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN.UK.2010- 4.RLTS.T12798A3381899.en

Prié, V. 2017. Newly overlapping ranges: First records of Potomida littoralis (Cuvier, 1798) infestation by the European Bitterling Rhodeus amarus (Bloch, 1782). Journal of Conchology 42:381–382.

Prié, V., J. Solerm, R. Araujo, X. Cucherat, L. Philippe, N. Patry, B. Adam, N. Legrand, P. Jugé, N. Richard and K. Wantzen. 2018. Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe. Hydrobiologia. https://doi.org/10.1007/s10750-017-3456-0.

Reichard, M., M. Ondrackova, M. Przybylski, H. Liu and C. Smith. 2006. The costs and benefits in an unusual symbiosis: experimental evidence that bitterling fish (Rhodeus sericeus) are parasites of unionid mussels in Europe. Journal of Evolutionary Biology 19:788–796.

Reichard, M., M. Polacik, A. S. Tarkan, R. Spence, O. Gaygusuz, E. Ercan, M. Ondrackova and C. Smith. 2010. The bitterling mussel coevolutionary relationship in areas of recent and ancient sympatry. Evolution 64:3047–3056.

Reichard, M., K. Douda, M. Przybyłski, O. P. Popa, E. Karbanova, K. Matasova, K. Rylkova, M. Polacik, R. Blazek and C. Smith. 2015. Population-specific responses to an invasive species. Proceedings of the Royal Society B 282:20151063.

Reynolds, J. D., V. J. Debuse and D. C. Aldridge. 1997. Host specialization in an unusual symbiosis: European bitterlings spawning in freshwater mussels. Oikos 78:539–545.

Smith, C., M. Reichard, P. Jurajda, and M. Przybylski. 2004. The reproductive ecology of the European bitterling (Rhodeus sericeus). Journal of Zoology 262:107–124.

Smith, C., J. D. Reynolds and W. J. Sutherland. 2000. The population consequences of reproductive decisions. Proceedings of the Royal Society of London B Biological Sciences 267:1327–1334.

156 Smith, C., K. Rippon, A. Douglas and P. Jurajda. 2001. A proximate cue for oviposition site choice in the bitterling (Rhodeus sericeus). Freshwater Biology 46:903–911.

Smith, D.G. and K. E. Hartel. 1999. Margaritiferidae (Mollusca: Unionoida): possible hosts for Rhodeus (Piscies: Cyprinidae). Polskie Archiwum Hydrobiologii 46:277–281.

Soler, J., K. M. Wantzen, P. Jugé and R. Araujo. 2018. Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies 84: 182–189.

Souchon, Y. and L. Tissot. 2012. Synthesis of thermal tolerances of the common freshwater fish species in large Western Europe rivers. Knowledge and Management of Aquatic Ecosystems 405:23–70.

Spence, R. and C. Smith. 2013. Rose bitterling (Rhodeus ocellatus) embryos parasitize freshwater mussels by competing for nutrients and oxygen. Acta Zoologica 94:113–118.

Tatoj, K., A. M. Cmiel, D. Kwasna, A. M. Lipinaska, K. Zajak and T. Zajac. 2017. The endangered thick-shelled river mussel (Unio crassus): a new host species for the European bitterling (Rhodeus amarus). Biodiversity and Conservation 26:1217–1224.

Valenciennes, A., 1829-1848. Histoire naturelle des Poissons. Bertrand Éd., Paris.

Van Damme, D., N. Bogutskaya, R. C. Hoffmann and C. Smith. 2007. The introduction of the European bitterling Rhodeus amarus to west and central Europe. Fish and Fisheries 8:79–106.

Wheeler, A. and P. S. Maintland. 1973. The scarcer freshwater fishes of the British Isles I. Introduced species. Journal of Fish Biology 5:49–68.

Wiepkema, P.R. 1961. An ethological analysis of the reproductive behaviour of the bitterling (Rhodeus amarus Bloch). Archives Néerlandaises de Zoologie 14:103–199.

Zhulkov, A.I. and S. N. Nikiforov. 1998. Some data on the morphology and biology of the bitterling Rhodeus sericeus of the Tym River, Sakhalin. Journal Voprosy Ikhtiologii 28:149–153.

157

158 3.2. Reproductive biology

3.2.1. ARTICLE 3: Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

Résumé

Nous décrivons l'anatomie des branchies de Margaritifera auricularia, y compris une analyse histologique du marsupium, sa capacité d’incubation et la longueur du processus de maturation, fournissant des informations sur la stratégie de reproduction des populations françaises jusqu'ici inconnue. Des ovules et des embryons en développement provenant des populations de la Charente et de la Creuse ont été ont été trouvés en mars et les émissions de glochidies ont eu lieu en avril. Le développement, du premier clivage à la maturité des glochidies, a duré de 25 à 37 jours. Par rapport aux populations espagnoles, la période de reproduction commence quelques semaines plus tard. Le nombre estimé de larves par moule gravide était d'environ 2.000.000. Nous attendons que les résultats présentés ici permettront d'améliorer les projets d'élevage en captivité qui visent à conserver les quelques populations restantes de cette espèce en voie de disparition en Europe

159 .

160 Journal of The Malacological Society of London Molluscan Studies

Journal of Molluscan Studies (2018) 84: 182–189. doi:10.1093/mollus/eyy008 Advance Access publication date: 27 March 2018

Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida: Margaritiferidae)

J. Soler1,4, K. M. Wantzen1,2, P. Jugé3 and R. Araujo4 1Université François Rabelais, UMR 7324 – CITERES, 33 Allée Ferdinand de Lesseps, 37204 Tours cedex 03, France; 2UNESCO River Culture – Fleuves et Patrimoines Chair, Université François Rabelais, UMR 7324 – CITERES, 33 Allée Ferdinand de Lesseps, 37204 Tours cedex 03, France; 3Université François-Rabelais, CETU Elmis Ingénieries, 11 Quai Danton, 37500 Chinon, France; and 4Museo Nacional de Ciencias Naturales – C.S.I.C., c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain

Correspondence: J. Soler; e-mail: [email protected] (Received 9 October 2017; editorial decision 16 January 2018)

ABSTRACT We describe the gill anatomy of Margaritifera auricularia, including a histological analysis of the marsupium, its brooding capacity and the length of the maturation process, providing information on the hitherto unknown reproductive strategy of French populations. Ova and developing embryos from the Charente and Creuse riv- er populations were found in March, with glochidial release occurring in April. Development from first cleav- age to glochidial maturity took 25–37 days. Compared with the Spanish populations, the reproductive period begins some weeks later. The estimated number of larvae per gravid mussel was around 2,000,000. We expect the findings presented here will improve captive breeding projects that aim to conserve the few remaining populations of this endangered species in Europe.

INTRODUCTION most western European rivers at the beginning of the 20th century, nowadays only a few remaining populations of this species persist in Fertilization in freshwater mussels is always internal, with early Spain and France (Araujo & Ramos, 2000a; Cochet, 2001; Prié et al., embryonic and larval development occurring in the gills. Brooding fi 2018). Much of the biological knowledge of M. auricularia comes from has been identi ed as a mechanism for embryonic nourishment as studies of Spanish populations, including studies on sex ratio and well as the protection and retention of the larvae (Sellmer, 1967; gametogenesis (Grande, Araujo & Ramos, 2001), glochidial morph- Wood, 1974; Kat, 1984; Silverman, Kays & Dietz, 1987; Richard, ology (Araujo & Ramos, 1998), glochidial release season (Araujo, Dietz &Silverman, 1991; Tankersley & Dimock, 1992; Tankersley, Bragado & Ramos, 2000), host fish (Araujo, Bragado & Ramos, 1996; Schwartz & Dimock, 2001). Several authors have suggested 2001), range of degree days for glochidial excysment (Araujo, Quirós that this mechanism is an evolutionary adaptation facilitating the & Ramos, 2003), national threats (Araujo & Ramos, 2000b; Araujo transition of bivalves from marine to freshwater habitats by prevent- & Álvarez-Cobelas, 2016) and growth patterns (Nakamura et al., ing larvae from being swept downstream (Davis & Fuller, 1981; Graf 2017). However, except for the brief description provided by Haas & Ó Foighil, 2000; Araujo et al., 2017). (1916, 1924), no studies have characterized the anatomy of the gills Some characteristics related to brooding, including the anatomy and changes in the marsupium during brooding, or embryo develop- of the gills, location and arrangement of glochidia in the brood ment times and brood size in this species. chambers, the degree of swelling of the lamellae and the duration Detailed knowledge of brooding morphology and behaviour of dif- of larval incubation, can also be informative for freshwater mussel ferent populations is important for improving ex situ projects and the taxonomic and conservation studies (Ortmann, 1911; Heard & conservation of this endangered species in Europe. In this paper, we Guckert, 1971; Tankersley, 1996; Graf & Cummings, 2006). describe the anatomy of the gill and present a histological analysis of Freshwater mussels have different strategies in relation to the length the marsupium of M. auricularia.Wealsoprovideinformationon of the brooding period, being either short-term (tachytictic) or long- brooding capacity and the length of the maturation process, revealing term (bradytictic) brooders. Tachytictic species spawn their gametes details about the reproductive strategy of the northern populations in spring and brood the embryos only until glochidia have fully devel- (France) and differences from the southern ones (Spain). oped, whereas bradytictic mussels spawn in late summer and brood their glochidia over winter, releasing them in early spring (Graf & Ó Foighil, 2000). Climate (Sterki, 1903; Ortmann, 1909; Graf, 1997) and synchronization with seasonal host activity (Zale & Neves, 1982) MATERIAL AND METHODS are considered factors that determine brooding patterns. Gill anatomy The giant freshwater pearl mussel Margaritifera auricularia (Spengler, 1793) is considered the most imperilled bivalve species in Europe Observations of the gills of Margaritifera auricularia were made using (Araujo & Ramos, 2001; Prié, 2010). Although likely widespread in specimens collected at the Canal Imperial (Ebro Basin, Spain) and © The Author(s) 2018. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved. 161 For Permissions, please email: [email protected] Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 BROODING IN MARGARITIFERA AURICULARIA

Table 1. Specimens of Margaritifera auricularia studied.

Museum reg. no. Collection date Locality State of gills Histological analysis

MNCN15.07/5187 February 1996 Canal Imperial (Spain) Gravid Horizontal sections MNCN15.07/5188 February 1996 Canal Imperial (Spain) Gravid MNCN15.07/5191 December 1996 Canal Imperial (Spain) Non-gravid Horizontal sections MNCN15.07/7116 February 1996 Canal Imperial (Spain) Gravid Transverse sections MNCN15.07/7076 January 1999 Canal Imperial (Spain) Non-gravid MNCN15.07/7549 February 2001 Canal Imperial (Spain) Gravid

preserved in the malacological collection of the Museo Nacional Table 2. Number of French specimens of Margaritifera auricularia inspected de Ciencias Naturales (MNCN) in Madrid (Table 1). Due the rar- for each population and dates of glochidial release. ity of this species, we did not sacrifice any additional specimens for this study. Three specimens were processed for histological ana- Site Date Specimens Gravid Date of glochidial lyses. Portions of the central part of the outer and inner demi- examined specimens release branchs were excised, gradually dehydrated through an ethanol Creuse 15/03/2015 4 2 (50%) – series, embedded in paraplast and serially sectioned (5–10 μm) with a microtome. The slides were then stained with haematoxy- 12/04/2015 2 2 (100%) – lin–eosin and Heidenhain’s azan and observed under an optical 18/03/2016 5 3 (60%) 18 April 2 May microscope mounted with a digital camera. 26/03/2017 8 4 (50%) 11–12 April 08/05/2017 2 0 (0%) Charente 08/03/2015 10 3 (30%) 2–4 April Collection and maintenance of animals for glochidial maturation and 05/03/2016 10 4 (40%) 8–14 April release observations 21/03/2016 10 7 (70%) 19/03/2017 30 29 (97%) 1–10 April The M. auricularia populations from the Creuse and Charente rivers 21/04/2017 4 0 (0%) (France) were studied between 2015 and 2017. Additionally, we examined the gravidity of Spanish specimens stored at the collection of MNCN that had been collected during different months between 1996 and 2001. Based on this information and the previously reported RESULTS reproductive season of the Spanish populations (Araujo et al., 2000), Gill anatomy and brooding the two French populations were surveyed between March and May (Table 2). During this period, 85 specimens (64 from the Charente The ctenidia of Margaritifera auricularia were composed of inner and and 21 from the Creuse) were captured and labelled by a diver team. outer demibranchs, each with ascending and descending lamellae. Mussels were then held for 1–2 weeks in netted cages in the river to The inner demibranch was always larger than the outer one (Fig. 1). allow larval development to commence or proceed and to limit the Nearly the entire length of the outer demibranch was attached dor- maintenance period in the laboratory. Specimens were regularly sally to the mantle, with 1/9 to 1/4 of its length being free poster- inspected for gravidity and stage of development by observing sam- iorly. The inner demibranch was dorsally connected to the visceral ples of the gills content obtained with the aid of a disposable syringe sac, except for the central and posterior parts. The free posterior under an optical microscope. Mussels from the Charente population ends of the two ctenidia were in contactwitheachotherandwith in 2017 already had developed embryos when captured and were the inner mantle wall, forming the pseudodiaphragm. immediately transferred to the laboratory. The lamellae of each demibranch were intermittently united by In the laboratory, the mussels were maintained at ambient tem- irregularly scattered junctions that slightly formed a diagonal pat- perature (similar to the temperature of the River Creuse) in aqua- tern (Fig. 1). These interlamellar junctions (ILJs) usually consisted ria filled with aerated water from their respective rivers and of extensions of tissue between two adjacent ostia in the direction without sediment. Larval material released by each mussel was of the filament. They were also usually longer in the dorsal and inspected daily under a binocular microscope. central parts of the demibranchs and shorter in the ventral part The duration of egg maturation was estimated by counting the (Fig. 1). Microscopic observations showed that ILJs were formed number of days between the first observation of undivided eggs of connective tissue crossed by muscular fibres attached to the and the presence of mature glochidia (i.e. without egg membranes) base of the gill filaments (Fig. 2A). Within the lamellar tissue of in two specimens from the River Creuse. the demibranchs, the blood vessels, which can be observed macro- To estimate the total number of glochidia harboured by a single scopically, run and branch dorsoventrally (Fig. 2B). The lamellar mussel, we collected all larval material individually released by three tissue of the demibranchs was more developed in the descending specimens in 2016. We diluted and homogenized this material in a lamellae. Water canals, connecting the interior lumen of the demi- known volume of river water, then counted the number of glochidia branchs with the mantle cavity through the ostia, were always pre- in three 1-ml samples in a Sedgewick Rafter counting chamber for sent (Fig. 2C). each specimen. We could not check whether the mussels contained In the demibranchs of non-gravid specimens, the connective tis- more larvae in their gills before being returned to the river or sacri- sues of both the lamellae and the ILJs were sparse, leaving an fice them due to conservation legislation. Therefore, the actual num- irregular internal lumen with numerous epithelial folds (Fig. 2D). ber of glochidia may be underestimated. The muscular fibres of the ILJs were contracted and sinuous, and The size of the glochidia, collected from two individuals per the epithelial cells were compacted, with microvilli facing the French population, was measured under an optical microscope. internal lumen (Fig. 2E). For scanning electron microscopy (SEM), glochidia collected from In gravid individuals, two major changes were observed within the Charente River population were fixed in 70% ethanol, washed the gills. Following the relaxation of the muscular fibres, the ILJs three times with distilled water, cleaned in 5% KOH for 2 h and elongated and unfolded. As a result of this extension, the lamellar washed again five times with distilled water. tissue was compressed, becoming thinner and denser than in the

183 162 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 J. SOLER ET AL.

Figure 1. Right inner and outer demibranchs (right side) of Margaritifera auricularia. Scale bar = 1 cm.

non-gravid gills. Accordingly, the folds were no longer present and In some cases, we observed broods being aborted prior to matur- the entire lumen was stretched, acquiring a more regular shape to ation. These aborted broods consisted of egg masses, embryos at differ- accommodate the larvae (Fig. 2C). Despite the elongation of the ent developmental stages and sometimes glochidia (with and without ILJs, the entire demibranch did not swell in gravid individuals, egg membranes). These were released as white conglutinates and maintaining a similar size as in the non-gravid state. The second appeared as small clumps lacking a well-defined shape that remained major change that occurred during the incubation period was that near the exhalant aperture. However, in some instances, the mussels the epithelial cells of the lamellar tissue and the ILJs expanded, expelled large branch-shaped conglutinates as casts of the demi- becoming cuboidal in shape and full of an unstained material, branchs. The cohesive nature of the egg membranes held this brood which was presumably composed of fluid-filled vacuoles (Fig. 2F). material together. Nevertheless, on some occasions, we also observed a During this phase, the microvilli were not observed. Instead, a mucus matrix within these conglutinates. When development pro- mucoid-like coating was distributed over the cell apices, suggesting ceeded normally, and the majority of larvae were mature, mussels a secretory function. broadcast free mature glochidia suddenly and in large quantities. In the two gravid specimens examined histologically, the outer The estimated numbers of eggs, embryos and glochidia released demibranchs were densely packed with larvae, whereas larval by two Charente mussels, both c. 12 cm in length, were 1,541,575 density was much lower in the inner demibranchs, which had and 2,032,000. A Creuse specimen, measuring 14.5 cm in length, large empty spaces, especially in the dorsal and ventral regions was estimated to have released 1,705,000 larvae (Table 3). (Fig. 3). Accordingly, the ILJs appeared more distended in the The mean length of glochidia of the French populations of inner demibranchs. M. auricularia was 147.17 μm(±SD 4.7; n = 39) and the mean height 142.21 μm(±4.08; n = 40) (Fig. 4). A few minute teeth, – μ μ Glochidial maturation and release measuring 2.3 3.8 m long by 1.2 m wide, were observed along the ventral edge of the shell (Fig. 4). Specimens from the Canal Imperial that were collected during February (1996–2001) consisted of gravid individuals, whereas those collected in early December (1996) and mid-January (2001) DISCUSSION did not. During the 2015–2017 field surveys in France, ova and developing embryos were found throughout the month of March Detailed descriptions of the morphological differences between in the Charente and Creuse populations. In mid-April 2015, glo- marsupial and non-marsupial demibranchs have previously been chidia with egg membranes were found in the River Creuse. No reported for several unionid mussels (Peck, 1877; Lefevre & gravid specimens were found at the end of April in the Charente Curtis, 1910; Ortmann, 1911; Heard & Vail, 1976; Richard et al., or in early May in the Creuse. The percentage of gravid mussels 1991; Tankersley & Dimock, 1992, 1993). The gill morphology of found varied between years; however, in March 2017, almost all several margaritiferid species has also been described (e.g. Haas, inspected specimens in the Charente River were gravid (Table 2). 1910; Ortmann, 1912; Smith, 1988), being characterized by the A total of 19 individuals were recaptured from a previous year (no incomplete fusion of the visceral sac and mantle in the posterior specimens were captured three years in a row) and, of those, five area, and the absence of water tubes. Nevertheless, until now the were gravid in both years. changes in gill morphology associated with brooding have only Mussels taken from the River Charente to the laboratory been reported for Margaritifera margaritifera (Smith, 1979). released glochidia in early April, while those from the Creuse did We show that during gravidity in M. auricularia, the demi- so 7–10 d later (Table 2). Based on our combined field and branchs increase the volume of their internal lumen to harbour laboratory observations of embryonic development, first cleavage larvae by reducing the width of the lamellar connective tissue. to the mature glochidial stage (without egg membrane) in the Apparently, this is controlled by the relaxation of the muscular River Creuse took 31–37 d at an average temperature of 11 °C. fibres of the ILJs, which maximally expand from a contracted

184 163 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 BROODING IN MARGARITIFERA AURICULARIA

Figure 2. Anatomical details of the gills of Margaritifera auricularia. A. Transverse section of gravid outer demibranch showing disposition of muscular fibres within interlamellar junctions. B. Lateral view of inner demibranch showing blood vessels. C. Horizontal section of gravid outer demibranch showing arrangement of tissues during brooding process. D. Horizontal section showing general aspect of non-gravid outer demibranch. E. Horizontal section of non-gravid outer demibranch showing epithelial cells of lamellar tissue. F. Transverse section of gravid outer demibranch showing epithelial cells of lamellar tissue. Abbreviations: bv, blood vessels; e, developing embryos; ec, epithelial cells; ef, epithelial folds; f, filaments; ilj, interlamellar junctions; mf, muscular fibres; mv, microvilli; o, ostia; wc, water canals. Scale bars: A, C = 2 mm; B = 200 μm; D, E = 500 μm; F = 20 μm.

state, such that the epithelia of the internal lumen at the base of Dimock, 1992), may prove to be a phylogenetic constraint of the ILJs stretch towards the gill filaments, thus reducing the width Margaritiferidae (Smith, 1979). of the lamellar connective tissue and elongating the ILJs. As described for M. margaritifera by Smith (1979), the epithelial Therefore, it is adjustments in the linear dimensions of the tissue, cells of both the ILJs and the rest of the internal lumen undergo rather than a stretching action, that are responsible for the slight cell-shape changes during the brooding period, presumably to increase in demibranch width during brooding. This lack of elasti- assume a secretory function. In members of the Unionidae, all of city, which has also been reported for M. margaritifera (Smith, which have continuous septa, the ILJ epithelial cells change shape 1979) but is contrary to reported results in Unionidae (Ortmann, (Lefevre & Curtis, 1912); however, whether other epithelial cells of 1911; Fuller, 1972, 1973; Richard et al., 1991; Tankersley & the lumen also do so is unknown. The extension of the secretory

185 164 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 J. SOLER ET AL.

The function of these secretions is not well understood. Smith (1979) and Baird (2000) proposed that the mucoid secretions may facilitate the release of the glochidial mass by forming a functional viscous coating that reduces resistance along the cell surfaces in M. margaritifera and M. monodonta, respectively. Ortmann (1911) suggested that the secretions from the water tube cells may confer a gelatinous consistency to the egg/embryo masses that form the conglutinates. Based on studies of several Unionidae (Anodontinae and Lampsilini species), some authors have suggested that nutri- ents and calcium concretions could be transferred from females to developing larvae via mucus secreted by cells located in the inter- lamellar septa (Wood, 1974; Tankersley & Dimock, 1992; Tankersley, 1996; Schwartz & Dimock, 2001; McElwain & Bullard, 2014). Like other Margaritiferidae (Harms, 1907; Howard, 1915; Haas, 1916; Murphy, 1942; Smith, 1980), M. auricularia is a tetra- genous species. Although Howard (1915) reported the presence of more eggs and glochidia in the inner demibranchs of one speci- men of M. monodonta, in the analysed specimens of M. auricularia the outer demibranchs showed considerably more embryos than the inner demibranchs. At present, we cannot determine if this is a feature of this species or the result of the specimens we observed not yet being fully gravid prior to fixation. Fecundity of M. auricu- laria was estimated to be around 2,000,000 glochidia per adult based on material collected from three specimens. Although this estimate of fecundity is high compared with other members of Unionoida (Haag, 2013), it is lower than that observed for other margaritiferids, with the exception of M. falcata (Table 3). However, in this latter case, the data were obtained by a volume calculation, suggesting fecundity in M. falcata may have been greatly underestimated. Presumably, for similarly sized mussels, more glochidia can be produced with the same energetic invest- ment if the glochidia are smaller (Bauer, 1994). The apparently lower fecundity of M. auricularia among margaritiferids could be a consequence of glochidial size. Margaritifera auricularia produce glo- chidia with a length range 127–155 μm(Araujo & Ramos, 1998; this study), being the largest among all reported Margaritiferidae (Table 3). The glochidium of M. auricularia, as previously described by Araujo & Ramos (1998) and confirmed in this study, is of the hookless type and lacks larval threads. Rather, it has minute teeth in the shell margin (Araujo & Ramos, 1998), which are likely used to attach to the gills of the host fish. Although there has been some confusion about whether Margaritiferidae are bradytictic or tachytictic (Graf & Ó Foighil, 2000), our observations indicate that M. auricualaria is a short-term brooder as glochidia were released shortly after maturation. Furthermore, the marsupia do not have special anatomical adap- tations to accommodate long-term brooding, such as gill structures to isolate glochidia during development, as has been described for the bradytictic Anodontinae and Lampsilini (Sterki, 1903; Ortmann, 1911; Lefevre & Curtis, 1912). In contrast, the marsu- pium of M. auricularia remains in contact with the external medium through the ostia and the water canals. Based on our observations of M. auricularia, maturation takes approximately one month, similar to European M. margaritifera populations (Harms, 1907; Scheder, Gumpinger & Csar, 2011). Therefore, spawning and fertilization probably occur in early to mid March in French populations and a little earlier, at the end of January to early February, in Spanish ones, with the precise tim- ing being temperature dependent (Wellmann, 1943; Meyers & Figure 3. Transverse sections showing the distributions of larvae in gravid Millemann, 1977; Young & Williams, 1984; Ross, 1992; Hastie & outer (left) and inner (right) demibranchs of Margaritifera auricularia. Young, 2003). Previous knowledge of the reproductive season of Abbreviations: e, developing embryos. Scale bar = 2 mm. M. auricularia is based on Ebro Basin populations, which are gravid in February (see above) and release glochidia in March (Araujo et al., 2000). Field and laboratory observations of the two French function to the entire epithelium of the lumen may be related to populations carried out from 2015 to 2017 showed that mussels having sparse and irregularly distributed ILJs and may be another had developing embryos through March and that the beginning of evolutionary constraint of Margaritiferidae. glochidial release occurred in early to mid April. Taken together,

186 165 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 BROODING IN MARGARITIFERA AURICULARIA

Table 3. Comparison glochidial length and fecundity estimates Margaritifera species.

Species Length glochidia (μm) Adult length Fecundity range (mm)

M. monodonta 55–60, Baird (2000); Howard (1915) 124–190 1,930,000–9,570,000, Baird (2000) M. margaritifera 47,5–71, Harms (1909); Nezlin et al. (1994); Bauer (1994); 66–120 1,030,600–16,851,000, Young & Williams (1984) Pekkarinen & Valovirta (1996) M. laevis 70–90, Awakura (1968) 54–112 440,000–3,660,000, Awakura (1968) M. auricularia 127–155, Araujo & Ramos (1998); this study 119–147 1,541,575–2,032,000, this study M. falcata 50–60, Murphy (1942) 70 1,000,000, Murphy (1942)

AB

Figure 4. Glochidium of Margaritifera auricularia. A. General aspect of glochidium. B. Detail of teeth at ventral edge of shell. Scale bars: A = 50 μm; B = 5 μm.

this indicates a difference in the timing of glochidial release affect reproductive timing in M. falcata (Meyers & Millemann, between Spanish and French populations and, indeed, even 1977) and M. margaritifera (Hastie & Young, 2003). In the latter the between the two French populations, with glochidial release in the authors suggested that annual differences in accumulated degree- more northern River Creuse population delayed by several days. days (sum of daily average temperature) may be a more reliable Although most available information regarding the season of glo- indicator of thermal differences. Accumulated degree-days from chidial release is based on laboratory observations, the evidence six consecutive years (2011–2016) of the three rivers compared in suggests that there is a latitudinal gradient influencing the timing this study indicate a latitudinal gradient in temperature that seems of the reproductive season in this species, similar to that suggested to support this hypothesis (Table 4). in other freshwater mussels populations. For instance, Awakura Despite variations in the period of reproduction among the dif- (1968) reported that the breeding season and glochidial release of ferent populations, M. auricualria begins to release glochidia during a northern Japanese (c. 43°N) M. laevis population started, respect- late winter to early spring, which is earlier than in other margariti- ively, in mid-June and late July. However, Naito (1988) observed ferids. Margaritifera margaritifera usually releases its glochidia from that breeding and glochidial release occurred approximately two late June to early September (Lefevre & Curtis, 1910; Murphy, and a half months earlier at the southernmost distributional limit 1942; Wellmann, 1943; Smith, 1976; Young & Williams, 1984; of the species (c. 34°N). Howard (1915) suggested that differences Ross, 1992), M. dahurica at the end of September and October in reproductive timing may be related to latitudinal differences in (Klishko, 2012), M. laevis from July to August in northern Japan water temperatures. Thermal differences among rivers may also (Awakura, 1968) and from the middle to the end of May in

187 166 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 J. SOLER ET AL.

Table 4. Cumulative thermal degree-days recorded in River Creuse, and other mussel species (Bivalvia: Unionoida) in drift and on fishes in River Charente and Canal Imperial, 2011–2016. an ancient channel of the Ebro River, Spain. Archiv für Hydrobiologie, 148: 147–160. Year Creuse Charente Canal Imperial ARAUJO, R., BRAGADO, D. & RAMOS, M.A. 2001. Identification of the river blenny, Salaria fluviatilis, as a host to the glochidia of 2011 5,552 5,629 5,831 Margaritifera auricularia. Journal of Molluscan Studies, 67: 128–129. 2012 5,168 5,350 5,736 ARAUJO, R., QUIRÓS, M. & RAMOS, M.A. 2003. Laboratory propa- 2013 4,941 5,103 5,218 gation and culturing of juveniles of the endangered freshwater mussel 2014 5,464 5,508 5,703 Margaritifera auricularia (Spengler, 1793). Journal of Conchology, 38:53–60. 2015 5,255 5,441 5,713 ARAUJO, R. & RAMOS, M.A. 1998. Description of the glochidium of 2016 5,033 5,371 5,753 Margaritifera auricularia (Spengler, 1793) (Bivalvia, Unionidae). Philosophical Transactions of the Royal Society of London B, 353: 1553–1559. ARAUJO, R. & RAMOS, M.A. 2000a. A critic revision of the historical distribution of Margaritifera auricularia (Spengler, 1793) (Mollusca: southern populations (Naito, 1988), and M. falcata in May (Meyers Margaritiferidae) based on museum specimens. Journal of Conchology, 37: & Millemann, 1977). The timing of glochidial release is of crucial – fi 49 59. importance in relation to host- sh availability. However, much ARAUJO, R. & RAMOS, M.A. 2000b. Status and conservation of the rel- research is still needed on this topic and will be analysed in future ict giant European freshwater pearl mussel Margaritifera auricularia studies. (Spengler, 1793). Biological Conservation, 96: 233–239. Although the glochidial release behaviour of M. auricularia in ARAUJO, R. & RAMOS, M.A. 2001. Action plan for Margaritifera auricu- natural habitats is unknown, results from a drift-net study by laria. Convention on the Conservation of European Wildlife and Natural Habitats Araujo et al. (2000) indicated that release is a sudden event, with (Bern Convention). Council of Europe Publishing, Nature and environ- the majority of glochidia being released over only 1–2 d, as was ment, no. 117, Strasbourg. also observed in Scottish M. margaritifera populations (Hastie & ARAUJO, R., SCHNEIDER, S., ROE, K.J., ERPENBECK, D. & Young, 2003). Aquaria observations indicated two possible MACHORDOM, A. 2017. The origin and phylogeny of mechanisms for fish infection. The abrupt and massive broadcast- Margaritiferidae, Bivalvia, Unionoida. A synthesis of molecular and ing of free glochidia appears to be the main host-infection strategy fossil data. Zoologica Scripta, 46: 289–307. in M. auricularia. However, we also frequently observed small white AWAKURA, T. 1968. The ecology of parasitic glochidia of the freshwater masses composed of larval material, including glochidia, that were pearl mussel, Margaritifera laevis (Haas). Scientific Reports of the Hokkaido – released but remained near the exhalant aperture, which may be Fish Hatchery, 23:1 21. involved in host attraction. A similar behaviour has been reported BAIRD, M. 2000. Life history and population structure of the spectaclecase mussel, Cumberlandia monodonta (Bivalvia, Margaritiferidae). MSc thesis, for M. monodonta (Baird, 2000) and other freshwater mussels fi (Barnhart et al., 2008)—although it has been proposed to be a Missouri State University, Spring eld, MI. mechanism to deal with respiratory stress caused by hypoxic con- BARNHART, M.C., HAAG, W.R. & ROSTON, W.N. 2008. ditions (Lefevre & Curtis, 1912; Araujo & Ramos, 1998; Aldridge Adaptations to host infection and larval parasitism in Unionoida. Journal of the North American Benthological Society, 27: 370–394. & McIvor, 2003; Haag & Warren, 2003). Further research is required to determine if this behaviour occurs in nature and BAUER, G. 1994. The adaptive value of offspring size among freshwater mussels (Bivalvia: Unionoidea). Journal of Animal Ecology, 63: 933–944. whether it functions in host attraction and infection. COCHET, G. 2001. Redécouverte d’une population vivante de la Grande Mulette, Margaritifera auricularia, sur la Vienne et la Creuse. Recherches Naturalistes en Région Centre, 10:3–16. ACKNOWLEDGEMENTS DAVIS, G.M. & FULLER, S.L.H. 1981. Genetic relationship among This work was conducted within the scope of the LIFE project Recent Unionacea (Bivalvia) of North America. Malacologia, 20: 217–253. ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’. The paper is also issued under the auspices of the FULLER, S.L.H. 1972. Elliptio marsupiobesa, a new fresh-water mussel ‘ — ’ (Mollusca: Bivalvia: Unionidae) from the Cape Fear River, North UNESCO Chair Fleuves et Patrimoine River Culture . Thanks to Carolina. Proceedings of the Academy of Natural Sciences of Philadelphia, 124: the Charente-Maritime and Indre-Loire Prefectures for collection 1–10. permits. We also thank Gemosclera and the Club Chinon Plongé for FULLER, S.L.H. 1973. Fusconaia masoni (Conrad 1834) (Bivalvia: help in sampling, and Carolina Noreña (MNCN-CSIC) for helping Unionacea) in the Atlantic drainage of the southeastern United States. with interpreting histological sections. We are grateful to the Spanish Malacological Review, 6: 105–117. project Potamo Fauna (LIFE12 NAT/ES/001091) for networking. GRAF, D.L. 1997. The effect of breeding period on the biogeography of Water temperature data were provided by the LOGRAMI associ- freshwater mussels (Bivlavia: Unionoidea) in the Minnesota region of ation (River Creuse), Service EAUX CdA La Rochelle (River North America. Occasional Papers on Mollusks, 5: 393–407. Charente) and Confederación Hidrográfica del Ebro (Canal GRAF, D.L. & CUMMINGS, K.S. 2006. Palaeoheterodont diversity Imperial). The figures were compiled by Jesús Muñoz from the pho- (Mollusca: Trigonioida + Unionoida): what we know and what we tography facility of the MNCN. Melinda Modrell reviewed the wish we knew about freshwater mussel evolution. Zoological Journal of the English version of the manuscript. Linnean Society, 148: 343–394. GRAF, D.L. & Ó FOIGHIL, D. 2000. The evolution of brooding charac- ters among the freshwater mussels (Bivalvia: Unionoidea) of North America. Journal of Molluscan Studies, 66: 157–170. REFERENCES GRANDE, C., ARAUJO, R. & RAMOS, M.A. 2001. The gonads of ALDRIDGE, D.C. & MCIVOR, A. 2003. Gill evacuation and release of Margaritifera auricularia (Spengler, 1793) and Margaritifera margaritifera (L., glochidia by Unio pictorum and Unio tumidus (Bivalvia: Unionidae) under 1758) (Bivalvia: Unionoidea). Journal of Molluscan Studies, 67:27–35. thermal and hypoxic stress. Journal of Molluscan Studies, 69:55–59. HAAG, W.R. 2013. The role of fecundity and reproductive effort in defin- ARAUJO, R. & ÁLVAREZ-COBELAS, M. 2016. Influence of flow diver- ing life-history strategies of North American freshwater mussels. sions on Giant freshwater pearl mussel populations in the Ebro River, Biological Reviews, 88: 745–766. Spain. Aquatic Conservation: Marine and Freshwater Ecosystems, 26: 1145–1154. HAAG, W.R. & WARREN, M.L. 2003. Host fishes and infection strat- ARAUJO, R., BRAGADO, D. & RAMOS, M.A. 2000. Occurrence of egies of freshwater mussels in large Mobile Basin streams, USA. Journal glochidia of the endangered Margaritifera auricularia (Spengler, 1793) of the North American Benthological Society, 22:78–91.

188 167 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018 BROODING IN MARGARITIFERA AURICULARIA

HAAS, F. 1910. Pseudunio, neues genus für Unio sinuatus Lam. Nachrichtsblatt PEKKARINEN, M. & VALOVIRTA, I. 1996. Anatomy of the glochidia der Deutschen Malakozoologischen Gesellschaft, 42: 181–183. of the freshwater pearl mussel, Margaritifera margaritifera (L.). Archiv für HAAS, F. 1916. Sobre una concha fluvial interesante (Margaritana auricular- Hydrobiologie, 137: 411–423. ia, Spglr.) y su existencia en España. Boletín de la Sociedad Aragonesa de PRIÉ, V. 2010. Margaritifera auricularia. The IUCN Red List of Threatened Ciencias Naturales, 15:33–44. Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN. HAAS, F. 1924. Anatomische Untersuchungen an europäischen Najaden I. UK.2010-4.RLTS.T12798A3381899.en Archiv für Molluskenkunde, 56:66–82. PRIÉ, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., HARMS, W. 1907. Zur Biologie und Entwicklungsgeschichte der Flußperlmuschel PATRY, N., ADAM, B., LEGRAND, N., JUGÉ, P., RICHARD, N. (Margaritana margaritifera Dupuy). Zoologischer Anzeiger, 31: 814–824. & WANTZEN, K.M. 2018. Margaritifera auricularia (Bivalvia: Unionida), a critically endangered flagship species for challenging HARMS, W. 1909. Postembryonale Entwicklungsgeschichte der Unioniden. downstream ecosystems exploration: results of last 10 years’ dedicated Zoologische Jahrbücher. Abteilung für Anatomie und Ontogenie der Tiere, 28: Hydrobiologia – 325–386. surveys. , 810: 157 175. HASTIE, L.C. & YOUNG, M.R. 2003. Timing of spawning and glochi- RICHARD, P.E., DIETZ, T.H. & SILVERMAN, H. 1991. Structure of Anodonta grandis Ligumia sub- dial release in Scottish freshwater pearl mussel (Margaritifera margaritifera) the gill during reproduction in the unionids , rostrata and Carunculina parva texasensis. Canadian Journal of Zoology, 69: populations. Freshwater Biology, 48: 2107–2117. 1744–1754. HEARD, W.H. & GUCKERT, R.H. 1971. A re-evaluation of the ROSS, H.C.G. 1992. The reproductive biology of the freshwater pearl Recent Unionacea (Pelecypoda) of North America. Malacologia, 10: Margaritifera margaritifera Irish Naturalist 333–355. mussel (L.) in Co Donegal. Journal, 24:43–50. HEARD, W.H. & VAIL, V.A. 1976. Anatomical systematics of Etheria elliptica (Pelecypoda: Mycetopodidae). Malacological Review, 9:15–24. SCHEDER, C., GUMPINGER, C. & CSAR, D. 2011. Application of a five-stage field key for the larval development of the freshwater pearl HOWARD, A.D. 1915. Some exceptional cases of brooding among the mussel (Margaritifera margaritifera Linné, 1758) under different tempera- – Unionidae. Nautilus, 29:4 11. ture conditions. A tool for the approximation of the optimum time for KAT, P.W. 1984. Parasitism and the Unionacea (Bivalvia). Biological host fish infection in captive breeding. Ferrantia, 64:13–22. – Reviews of the Cambridge Philosophical Society, 59: 189 207. SCHWARTZ, M.L. & DIMOCK, R.V. 2001. Ultrastructural evidence KLISHKO, O.K. 2012. Some data on reproductive biology of the fresh- for nutritional exchange between brooding unionid mussels and their water mussels (Margaritiferidae, Unionidae) and their relationships glochidia larvae. Invertebrate Biology, 120: 227–236. with bitterlings (Cyprinidae) in Transbaikalye. Bulletin of the Russian Far SELLMER, G.P. 1967. Functional morphology and ecological life history – East Malacological Society, 15/16:31 55. (in Russian). of the gem clam, Gemma gemma (Eulamellibranchia: Veneridae). LEFEVRE, G. & CURTIS, W.C. 1910. Reproduction and parasitism in Malacologia, 5: 137–223. – the Unionidae. Journal of Experimental Zoology, 9:79 115. SILVERMAN,H.,KAYS,W.T.&DIETZ,T.H.1987.Maternalcalcium LEFEVRE, G. & CURTIS, W.C. 1912. Studies on the reproduction and contribution to glochidial shells in freshwater mussels (Eulamellibranchia: artificial propagation of fresh-water mussels. Bulletin of the United States Unionidae). Journal of Experimental Zoology, 242:137–146. – Bureau of Fisheries, 30: 105 201. SMITH, D.G. 1976. Notes on the biology of Margaritifera margaritifera (Lin.) MCELWAIN, A. & BULLARD, S.A. 2014. Histological atlas of fresh- in central Massachusetts. American Midland Naturalist, 96: 252–256. water mussels (Bivalvia, Unionidae): Villosa nebulosa (Ambleminae: SMITH, D.G. 1979. Marsupial anatomy of the demibranch of Margaritifera Lampsilini), Fusconaia cerina (Ambleminae: Pleurobemini) and Strophitus margaritifera (Lin.) in northeastern North America (Pelecypoda: – connasaugaensis (Unioninae: Anodontini). Malacologia, 57:99 239. Unionacea). Journal of Molluscan Studies, 45:39–44. MEYERS, T.R. & MILLEMANN, R.E. 1977. Glochidiosis of salmonid Margaritifera margaritifera fi SMITH, D.G. 1980. Anatomical studies on and shes. I. Comparative susceptibility to experimental infection with Cumberlandia monodonta (Mollusca: Pelecypoda: Margaritiferidae). Margaritifera margaritifera (L.) (Pelecypoda: Margaritiferidae). Journal of Zoological Journal of the Linnean Society, 69: 257–270. Parasitology, 63: 728–733. SMITH, D.G. 1988. Notes on the biology and morphology of Margaritifera MURPHY, G. 1942. Relationship of the fresh-water mussel to trout in the hembeli (Conrad, 1838) (Unionacea: Margaritiferidae). Nautilus, 102: – Truckee River. California Fish and Game, 28:89 102. 159–163. NAITO, J. 1988. A note on the breeding habit of the fresh-water pearl- STERKI, V. 1903. Notes on the Unionidae and their classification. mussel, Margaritifera laevis Hass (Pisces: Margaritiferidae), in Geihoku- American Naturalist, 37: 103–1013. Cho, Hiroshima Pref., near the southern end in Japan. Miscellaneous TANKERSLEY, R.A. 1996. Effect of larval brooding on the feeding Reports of the Hiwa Museum for Natural History, 27:7–15. physiology of freshwater unionid mussels. Invertebrate Biology, 115: NAKAMURA, K., CUCALA, L., MESTRE, A., MESQUITA-JOANES, 243–255. F., ELBAILE, E., SALINAS, C. & MUÑOZ-YANGUAS, M.A. 2017. Modelling growth in the critically endangered freshwater mussel TANKERSLEY, R.A. & DIMOCK, R.V. 1992. Quantitative analysis of Margaritifera auricularia (Spengler, 1793) in the Ebro basin. Hydrobiologia, the structure and function of the marsupial gills of the freshwater mus- sel Anodonta cataracta. Biological Bulletin, 182: 145–184. 175:1–17. TANKERSLEY, R.A. & DIMOCK, R.V. 1993. The effect of larval NEZLIN, L.P., CUNJAK, R.A., ZOTIN, A.A. & ZIUGANOV, V.V. brooding on the respiratory physiology of the freshwater unionid mus- 1994. Glochidium morphology of the freshwater pearl mussel sel Pyganodon cataracta. American Midland Naturalist, 130: 146–163. (Margaritifera margaritifera) and glochidiosis of Atlantic salmon (Salmo sal- ar): a study by scanning electron microscopy. Canadian Journal of Zoology, WELLMANN, G. 1943. Fischinfektionen mit glochidien der Margaritana – 72:15–21. margaritifera. Zeitschrift fuer Fischerei Band, 41: 385 390. ORTMANN, A.E. 1909. The breeding season of the Unionidae in WOOD, E.M. 1974. Development and morphology of the glochidium lar- – Pennsylvania. Nautilus, 22:91–95, 99–103. va of Anodonta cygnea (Mollusca: Bivalvia). Journal of Zoology, 173:1 13. ORTMANN, A.E. 1911. A monograph of the najads of Pennsylvania. YOUNG, M.R. & WILLIAMS, J.C. 1984. The reproductive biology of Memoirs of the Carnegie Museum, 4: 279–347. the freshwater pearl mussel Margaritifera margaritifera (Linn.) in Scotland I. Field studies. Archiv für Hydrobiologie, : 405–422. ORTMANN, A.E. 1912. Notes upon the families and genera of the 99 najades. Annals of the Carnegie Museum, 8: 222–365. ZALE, A.V. & NEVES, R.J. 1982. Fish host of four species of lampsiline Canadian Journal PECK, R.H. 1877. The minute structure of the gills of lamellibranch mussels (Mollusca: Unionidae) in Big Moccasin Creek. of Zoology – Mollusca. Quarterly Journal of Microscopical Science, 17:43–66. , 60: 2535 2542.

189 168 Downloaded from https://academic.oup.com/mollus/article-abstract/84/2/182/4955204 by CSIC user on 12 July 2018

3.2.2. ARTICLE 4: Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae).

Résumé

La grande mulette (Margaritifera auricularia), est considérée l'espèce de bivalve la plus menacée en Europe. Les connaissances actuelles sur ses poissons hôtes ne peuvent expliquer le recrutement observé récemment dans les bassins versants français. Dans cette note de recherche, sur la base d'approches d'infestation naturelle et artificielle, nous décrivons Gasterosteus aculeatus comme un poisson hôte natif de M. auricularia en France, représentant une avancée solide pour la conservation de l’espèce. Cependant, en raison de la petite taille de G. aculeauts, de son absence dans la Vienne et le faible taux de survie des glochidies observé pendant la période d'infestation, nous pensons qu'une autre espèce de poisson pourrait également être responsable du maintien des populations françaises de M. auricularia. Ainsi, des recherches supplémentaires devraient être consacrées à la recherche d'autres espèces hôtes potentielles pour cette moule d'eau douce en voie de disparition.

169

170 Journal of The Malacological Society of London Molluscan Studies Journal of Molluscan Studies (2018): 1–4. doi:10.1093/mollus/eyy038 Downloaded from https://academic.oup.com/mollus/advance-article-abstract/doi/10.1093/mollus/eyy038/5090260 by CSIC user on 05 September 2018

RESEARCH NOTE Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida: Margaritiferidae)

J. Soler1,2, C. Boisneau1, K.M. Wantzen1,3 and R. Araujo2 1Université de Tours, CNRS UMR 732—CITERES, 33 Allée Ferdinand de Lesseps, 37204 Tours cedex 03, France; 2Museo Nacional de Ciencias Naturales—C.S.I.C. c/José Gutiérrez Abascal 2, 28006 Madrid, Spain; and 3UNESCO River Culture—Fleuves et Patrimoines Chair, Université François Rabelais, UMR 7324—CITERES, 33 Allée Ferdinand de Lesseps, 37204 Tours cedex 03, France

Correspondence: J. Soler; e-mail: [email protected]

The giant freshwater pearl mussel Margaritifera auricularia Therefore, these populations appear to use an unknown host fish (Spengler, 1793) is considered the most imperilled bivalve species species and its determination is the objective of this note. in Europe (Araujo & Ramos, 2001; Prié, 2010). Although likely Given that host fish species for M. auricularia and for the major- widespread in most western European rivers at the beginning of ity of Margaritiferidae species have a close relationship with the the twentieth century, its decline in both abundance and distribu- marine environment, we hypothesized a similar relationship for tion range has been estimated to be over 90% (Prié, 2010) and it this unknown host fish species. Sturgeons are anadromous fishes is now nearly extinct. Only a few populations remain in Spain and the river blenny, although restricted to freshwater, can toler- (Ebro Basin) and France (Loire, Charente, Garonne and Adour ate high salinity levels (Plaut, 1998) and is the only freshwater spe- basins) (Araujo & Ramos, 2000a; Cochet, 2001; Prié et al., 2018; cies of an otherwise entirely marine family. Furthermore, it has Soler et al., 2018). been observed that margaritiferid mussels frequently use anadro- Various causes have likely contributed to the extreme decline of mous host fishes (Curole, Foltz & Brown, 2004; Araujo et al., M. auricularia populations, including pollution, climate change, com- 2017): Margaritifera margaritifera, M. laevis, M. falcata, M. dahurica and mercial exploitation, host fish loss due to exotic fishes and dam con- M. middendorffi only use salmonids (Murphy, 1942; Karna & struction (Araujo & Álvarez-Cobelas, 2016; Prié et al.,2018). Milleman, 1978; Kobayashi & Kondo, 2005; Kondo & Freshwater mussels have a temporary but obligatory parasitic stage Kobayashi, 2005; Klishko & Bogan, 2013). However, the recent in which the larvae (glochidia) attach to the external surface or gill discovery of exclusively freshwater host fishes of the family filaments of their vertebrate hosts (mainly fishes). The glochidia of Hiodontidae for M. monodonta ( Sietman et al., 2017) and Esocidae most freshwater mussels develop successfully into juveniles only on a for M. marrianae and M. hembeli (Fobian et al., 2017; P. Johnson limited number of host fish species (Jansen, Bauer & Zahner-Meike, 2018, personal communication) calls into question this hypothesis, 2001; Barnhart, Haag & Roston, 2008; Strayer, 2008). Therefore, although the salinity tolerances of these fishes are unknown. mussel survival depends not only on habitat conservation but also Freshwater mussel host species can be inferred from natural infes- on the availability of host fishes. Among 21 fish species tested, com- tations of glochidia on fishes, but it is preferable to combine this plete metamorphosis of M. auricularia glochidia into juveniles, which evidence with artificial laboratory infections to determine host suit- takes place after 4–6 weeks in the gills, has only been shown to ability (e.g., Zale & Neves, 1982; Araujo, Gómez & Machordom, occur in four sturgeon species (Acipenser sturio, A. naccarii, A. ruthenus 2005; Taeubert et al., 2012) or by retrieving fully metamorphosed and A. baerii) and the river blenny (Salaria fluviatilis)(Araujo & juveniles from naturally infected fish (Sietman et al., 2017). Ramos, 1998, 2000b; Araujo, Bragado & Ramos, 2001; Altaba & Fish communities in the vicinity of major M. auricularia popula- López, 2001; Araujo, Quirós & Ramos, 2003; López & Altaba, tions from the Charente, Vienne and Creuse rivers were assessed 2005; López et al.,2007; K. Nakamura, E. Elbaile, M.A. Muñoz- for natural infestation by electrofishing boat in April 2016, coin- Yanguas, C. Catalá & C. Salinas, personal communication). ciding with their recently determined glochidial release period Scarce but recent recruitment of M. auricularia has been ( Soler et al., 2018). A total of 613 specimens belonging to 23 fish observed in several French localities ( Prié et al., 2018); however, species were analysed. Although the presence of three-spined none of the known host fishes co-occur with these M. auricularia stickleback (Gasterosteus aculeatus) is known in nearby small tributar- populations. The European sturgeon A. sturio is the only native ies of the Vienne and Creuse rivers, it was not found in the Acipenser species that occurs in France. However, this species was stretches of these two rivers examined in this study. However, in extirpated from most European rivers during the twentieth cen- the Charente River, we found M. auricularia glochidia attached to tury and is now almost extinct. Only one remaining reproductive the gills of 2 of the 29 G. aculeatus specimens examined. This obser- population is known in the Garonne River (France). The river vation represents the first reported natural infestation by M. auricu- blenny is a Mediterranean species, and its distribution range does laria glochidia in the wild. No glochidia of other species of not coincide with that of the French populations of M. auricularia. freshwater mussels were found in the inspected fish.

© The Author(s) 2018. Published by Oxford University Press on behalf of The Malacological Society of London, all rights reserved. 171 For Permissions, please email: [email protected] RESEARCH NOTE

To test the susceptibility of G. aculeatus as a host to M. auricularia, Table 1. Evolution of infestation intensity by Margaritifera auricularia glo- 20 stickleback specimens were captured by electrofishing in a small chidia during the encystment period on Gasterosteus aculeatus and Acipenser stream tributary of the Vienne River near Chinon (France) on 28 baerii. March 2017. This stream was not colonized by M. auricularia. The Species Days after Number Cysts/fish Glochidia animals were transferred to the laboratory in aerated containers. fi Thirty gravid M. auricularia specimens were collected from the infestation of sh loss Charente River on 26 March 2017 and maintained until mature G. aculeatus 12 29 fi Downloaded from https://academic.oup.com/mollus/advance-article-abstract/doi/10.1093/mollus/eyy038/5090260 by CSIC user on 05 September 2018 glochidia emerged. Mussels and sh were kept in separate aquaria 5 1 18 38% filled with aerated water from their respective rivers (without sedi- 10 1 17 41% ment) and at an average water temperature of 12.6 °C and 18 °C, respectively. 18 1 12 59% For induced infestation, glochidia were removed from the exha- 26 1 4 86% lant apertures of the mussels using a pipette and placed in a plastic A. baerii 6 1 2675 bucket filled with aerated water (2,100 glochidia per l in 3.5 l) and – fi fi 28 43 1 2446 9% the 20 sh. After 15 min, sh were transferred to a cylindroconical 29–44 1 2517 6% tank (250 l) equipped with biological and mechanical filters. They 35–50 1 2735 2% were kept in aerated river water at an average temperature of – 19.3 °C and fed red mosquito larvae. Due to their small size, fish 36 51 1 2685 0% had to be sacrificed to verify glochidia encystment in the gills. Fish were anaesthetized then sacrificed at 1, 5, 10, 18 and 26 days after infestation. Encysted glochidia were found in the gills of all inspected speci- mens, showing a gradual loss as metamorphosis progressed (Table 1). An outbreak of the fish parasite Ichthyophthirius sp. occurred 26 days after infestation. Though the fish were treated with malachite green, all died a day later (27 days after infest- ation). Nevertheless, on this day, eight fully transformed live juve- niles were recovered from the gills of two of the fishes (Fig. 1A), thus indicating that the three-spined stickleback is a suitable host for M. auricularia glochidia. Gasterosteus aculeatus (Actinopterygii: Gasterosteidae) is a small fish (6 cm in length) with a large circumarctic and temperate distri- bution. Three-spined sticklebacks are known to inhabit brackish water, and even sea water, and anadromous populations are rela- tively frequent. Their ecological and behavioural characteristics may render them especially vulnerable to infestation by glochidia, since they are benthic feeders and may prey on glochidia (Dartnall & Walkey, 1979). It has been reported as a host for many unionid mussel species distributed throughout Europe, North America and Asia (Table 2) and, according to Lopes-Lima et al. (2017), is the host fish most commonly used by European freshwater mussels. Due to its ability to facilitate the metamorphosis of glochidia of mussels in a variety of genera and tribes, G. aculeatus could be con- sidered a ‘universal host’, similar to other host species of the fam- ilies Fundulidae and Poeciliidae (Haag, 2012). Although further research is required to establish the relative suitability of G. aculeatus, the high glochidial mortality found dur- ing the encystment period suggests that this host would not be ideal for M. auricularia glochidia. This observation is consistent with the low glochidial survival rates usually found in universal hosts (Teutsch, 1997; Haag, 2012; Taeubert et al., 2012). Araujo et al. (2002) reported high survival rates of M. auricularia in the gills of the exotic species A. baerii, the most common host fish used in conservation programs. We re-inspected the preserved fish gills used in that experiment ( Araujo et al., 2002) (Fig. 1B) and determined that glochidial mortality during metamorphosis ran- ged between 0% and 9% (Table 1). Margaritifera auricularia trans- Figure 1. Juvenile of Margaritifera auricularia at the end of metamorphosis formation on the small Spanish native fish Salaria fluviatilis was 318 in the gill of Gasterosteus aculeatus (light microscopy) (A) and Acipenser baerii (scanning electron microscopy) (B). Scale bars: A = 200 μm; B = 1 mm. juveniles per fish ( Araujo et al. 2003), suggesting low mortality dur- ing metamorphosis in this host. Overall our results suggest that, in France, G. aculeatus may serve as a native host for M. auricularia, similar to S. fluviatilis in Spain. Due to the small size of G. aculeauts, its absence in the Vienne Given the inadvisability of introducing exotic species into the River and the low glochidial survival rate observed, we think that French rivers supporting M. auricularia populations, and the diffi- another fish species could also be responsible for the maintenance culty of reintroducing the endangered A. sturio, G. aculeatus could of the French populations of M. auricularia. Thus, further research be used to augment M. auricularia populations by releasing infested should be devoted in order to find other potential host species for fish in river sections where both species occur. this endangered freshwater mussel.

2 172 RESEARCH NOTE

Table 2. Reported mussel species using Gasterosteus aculeatus as host fish.

Mussel species Region Type of infestation Reference

Anodonta beringiana NE Asia and NW America NI Saenko, Shedko & Kholin (2001) NI Cope (1959) Anodonta kennerlyi NW America NI + AI Martel & Lauzon-Guay (2005)

Anodonta nuttalliana NW America AI Maine, Arango & O’Brien (2016) Downloaded from https://academic.oup.com/mollus/advance-article-abstract/doi/10.1093/mollus/eyy038/5090260 by CSIC user on 05 September 2018 Margaritifera falcata NW America NI Karna & Millemann (1978) Pyganodon cataracta NE America NI Lambert & Martel (2012) NI Beaudet (2006) NI Threlfall (1986) Anodonta implicata or NE America NI Wiles (1974) Pyganodon cataracta Elliptio complanata NE America NI Kneeland & Rhymer (2008) Anodonta anatina Europe AI Teutsch (1997) Anodonta cygnea Europe AI Teutsch (1997) NI Dartnall & Wakey (1979) Pseudoanodonta complanata Europe AI Hüby (1988) Unio pictorum Europe AI McIvor (2004) Unio tumidus Europe AI Fleischauer-Rossing (1990) Unio crassus Europe NI Engel (1990) AI Taeubert et al. (2012) Margaritifera auricularia Europe AI This study

Abbreviations: NI, natural infestation; AI, artificial infestation.

ACKNOWLEDGEMENTS (Mollusca: Margaritiferidae) based on museum specimens. Journal of Conchology, 37:49–59. This work was supported by the LIFE+ project ‘LIFE13 BIO/ ’ ARAUJO, R. & RAMOS, M.A. 2000b. Status and conservation of the rel- FR/001162 Conservation of the Giant Pearl Mussel in Europe . ict giant European freshwater pearl mussel Margaritifera auricularia Thanks to the Charente-Maritime and Indre-Loire Prefectures for (Spengler, 1793). Biological Conservation, 96: 233–239. the collection permits. We also thank CETU ELMIS, BIOTOPE, ARAUJO, R. & RAMOS, M.A. 2001. Action plan for Margaritifera auricu- FDAAPPMA 37 and 16, LOGRAMI and ONEMA for their con- laria. Convention on the Conservation of European Wildlife and Natural Habitats tribution to the electrofishing surveys. Thanks to Yann Guerez, (Bern Convention). Nature and environment, No. 117. Council of Europe and Marjolaine Sicot for the capture and maintenance of the Publishing, Strasbourg. three-spined sticklebacks for experimental purposes and to Nina ARAUJO, R., CÁMARA, N. & RAMOS, M.A. 2002. Glochidium meta- Richard, Philippe Jugé and Laure Morisseau for their contribution morphosis in the endangered freshwater mussel Margaritifera auricularia to the project. Thanks also to two anonymous referees and P.D. (Spengler, 1793): a histological and scanning electron microscopy Johnson for their comments. study. Journal of Morphology, 254: 259–265. ARAUJO, R., SCHNEIDER, S., ROE, K.J., ERPENBECK, D. & MACHORDOM, A. 2017. The origin and phylogeny of Margaritiferidae (Bivalvia, Unionoida): a synthesis of molecular and fossil data. Zoologica Scripta, 46: 289–307. REFERENCES BARNHART, M.C., HAAG, W.R. & ROSTON, W.N. 2008. ALTABA, C.R. & LÓPEZ, M.A. 2001. Experimental demonstration of Adaptations to host infection and larval parasitism in Unionoida. viability for the endangered giant pearlmussel Margaritifera auricularia Journal of the North American Benthological Society, 27: 370–394. (Bivalvia: Unionoida) in its natural habitat. Bolleti de la Societat d’Historia BEAUDET, A. 2006. Étude de la dynamique des populations de moules Natural de les Baleares, 44:15–21. d’eau douce (Bivalvia: Unionidea) de deux rivières côtières de l’Est du ARAUJO, R. & ÁLVAREZ-COBELAS, M. 2016. Influence of flow diver- Nouveau-Brunswick, la rivière Kouchibouguac et la rivière sions on giant freshwater pearl mussel populations in the Ebro River, Kouchibouguacis. Mémoires de Maîtrise, UQAR. Spain. Aquatic Conservation—Marine and Freshwater Ecosystems, 26: COCHET, G. 2001. Redécouverte d’une population vivante de la 1145–1154. Grande Mulette, Margaritifera auricularia, sur la Vienne et la Creuse. ARAUJO, R., BRAGADO, D. & RAMOS, M.A. 2001. Identification of Recherches Naturalistes en Région Centre, 10:3–16. the river blenny, Salaria fluviatilis, as a host to the glochidia of COPE, O.B. 1959. New parasite records from stickleback and salmon in Margaritifera auricularia. Journal of Molluscan Studies, 67: 128–129. an Alaska stream. Transactions of the American Microscopical Society, 78: ARAUJO, R., GÓMEZ, I. & MACHORDOM, A. 2005. The identity 157–162. and biology of Unio mancus Lamarck, 1819 (= U. elongatulus) (Bivalvia: CUROLE, J.P., FOLTZ, D.W. & BROWN, K.M. 2004. Extensive allo- Unionidae) in the Iberian Peninsula. Journal of Molluscan Studies, 71: zyme monomorphism in a threatened species of freshwater mussel, 25–31. Margaritifera hembeli Conrad (Bivalvia: Margaritiferidae). Conservation ARAUJO, R., QUIRÓS, M. & RAMOS, M.A. 2003. Laboratory propa- Genetics, 5: 271–278. gation and culturing of juveniles of the endangered freshwater mussel DARTNALL, H.J.G. & WALKEY, M. 1979. The distribution of glochidia Margaritifera auricularia (Spengler, 1793). Journal of Conchology, 38:53–60. of the swan mussels Anodonta cygnea (Mollusca) on the three-spine stickle- ARAUJO, R. & RAMOS, M.A. 1998. Description of the glochidium of back, Gastoresteus aculeatus (Pisces). Journal of Zoology, 189:31–37. Margaritifera auricularia (Spengler, 1793) (Bivalvia, Unionidae). ENGEL, H. 1990. Untersuchungen zur Autökologie von Unio crassus (Philipsson) in Philosophical Transactions of the Royal Society of London B, 353: 1553–1559. Norddeutschland. Dissertation, Universität Hannover. ARAUJO, R. & RAMOS, M.A. 2000a. A critical revision of the historical FLEISCHAUER-ROSSING, S. 1990. Untersuchungen zur Autokologie von distribution of the endangered Margaritifera auricularia (Spengler, 1782) Unio tumidus (Philipsson) und Unio pictorum (Linnaeus) (Bivalvia) unter

3 173 RESEARCH NOTE

besonderer Berucksichtigung der fruhen postparasitaren Phase. PhD thesis, glochidia of the freshwater pearl mussel Margaritifera auricularia. Journal Universitat Hannover. of Molluscan Studies, 73: 207–209. FOBIAN, T.B., BUNTIN, M.L., HOLIFIELD, J.T., TARPLEY, T.A. & MAINE, A., ARANGO, C. & O’BRIEN, C. 2016. Fish associations of the JOHNSON, P.D. 2017. Advancements in culture techniques for California floater (Anodonta californiensis) in the Yakima River basin, Alabama pearlshell (Margaritifera marrianae, R.I. Johnson, 1983) to sup- Washington. Northwest Science, 90: 290–300. port species recovery and reintroduction. In: FMCS 10th Biennial MARTEL, A.L. & LAUZON-GUAY, J.S. 2005. Distribution and density Symposium. Cleveland, OH. of glochidia of the freshwater mussel Anodonta kennerlyi on fish hosts in

HAAG, W.R. 2012. North American freshwater mussels: natural history, ecology, lakes of the temperate rain forest of Vancouver Island. Canadian Journal Downloaded from https://academic.oup.com/mollus/advance-article-abstract/doi/10.1093/mollus/eyy038/5090260 by CSIC user on 05 September 2018 and conservation. Cambridge University Press, New York. of Zoology, 83: 419–431. HÜBY, B. 1988. Zur Entwicklungsbiologie der Fließgewassermuschel MCIVOR, A. 2004. Freshwater mussels as biofilters. PhD thesis, University of Pseudoanodonta complanata. Dissertation, Institut fiir Zoologie, Cambridge. Tierarztliche Hochschule Hannover. MURPHY, G. 1942. Relationship of the fresh water mussel to trout in the JANSEN, W., BAUER, G. & ZAHNER-MEIKE, E. 2001. Glochidial Truckee River. California Fish & Game, 28:89–102. mortality in freshwater mussels. In: Ecology and evolution of the freshwater PLAUT, I. 1998. Comparison of salinity tolerance and osmoregulation in mussels Unionoida (G. Bauer & K. Wachtler, eds), pp. 185–211. two closely related species of blennies from different habitats. Fish Springer-Verlag, Berlin and Heidelberg. Physiology and Biochemistry, 19: 181–188. KARNA, D.W. & MILLEMANN, R.E. 1978. Glochidiosis of salmonid PRIÉ, V., 2010. Margaritifera auricularia. In: The IUCN Red List of Threatened fishes III. Comparative susceptibility to natural infection with Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN. Margaritifera margaritifera (L.) (Pelecypoda: Margaritanidae) and asso- UK.2010-4.RLTS.T12798A3381899.en Journal of Parasitology – ciated histopathology. , 64: 528 537. PRIÉ, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., KLISHKO, O.K. & BOGAN, A.E. 2013. The conservation status of the PATRY, N., ADAM, B., LEGRAND, N., JUGÉ, P., RICHARD, N. & freshwater pearl mussel Margarititifera dahurica in Far Eastern Russia. WANTZEN, K. 2018. Margaritifera auricularia (Bivalvia: Unionida), a Ellipsaria, 15:31–33. critically endangered flagship species for challenging downstream eco- KNEELAND, S.D. & RHYMER, J.M. 2008. Determination of fish host systems exploration: results of last 10 years’ dedicated surveys. use by wild populations of rare freshwater mussels using a molecular Hydrobiologia. https://doi.org/10.1007/s10750-017-3456-0. identification key to identify glochidia. Journal of the North American SAENKO, E.M., SHEDKO, M.B. & KHOLIN, S.K. 2001. Morphology Benthological Society, 27: 150–160. and feature of bionomics of glochidia mollusks of the genus Beringiana KOBAYASHI, O. & KONDO, T. 2005. Difference in host preference (Bivalvia, Unionidae) of Kamchatka and Northern Kuriles. Vestnik between two populations of the freshwater pearl mussel, Margaritifera Zoologii, 35:59–68. (in Russian) margaritifera (Bivalvia, Margaritiferidae) in the Shinano River system, SIETMAN, B., DAVIS, M., HOVE, M., PLETTA, M., WAGNER, T., Japan. Venus, 64:63–70. MARR, S., SECRIST, Z., FREEBURG, M., SCHEUNEMANN, A., KONDO, T. & KOBAYASHI, O. 2005. Revision of the genus KRUPP, K., HAGEMEYER, E., FRANZEN, A., SWANSON, C. & Margaritifera (Bivalvia: Margaritiferidae) of Japan, with description of a SAMPSON, A. 2017. Cumberlandia monodonta—host enigma resolved. new species. Venus, 64: 135–140. Ellipsaria, 19:18–20. LAMBERT, S.D. & MARTEL, A.L. 2012. A summary of glochidia hosts SOLER, J., WANTZEN, K.M., JUGÉ, P. & ARAUJO, R. 2018. for Pyganodon cataracta (Say, 1817) and evidence of additional fish hosts Brooding and glochidia release in Margaritifera auricularia (Spengler, (Fundulus diaphanus and Apeltes quadracus) from Nova Scotia. Ellipsaria, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84: 14:8–11. 182–189. LOPES-LIMA,M.,SOUSA,R.,GEIST,J.,ALDRIDGE,D.C.,ARAUJO, STRAYER, D.L. 2008. Freshwater mussel ecology: a multifactor approach to distri- R., BERGENGREN, J., BESPALAYA, Y., BODIS, E., BURLAKOVA, bution and abundance. University of California Press, London. L.,VANDAMME,D.,DOUDA,K.,FROUFE,E.,GEORGIEV,D., TAEUBERT, J.E., POSADA MARTINEZ, A.M., GUM, B. & GEIST, J. GUMPINGER, C., KARATAYEV, A.,KEBAPCI,U.,KILLEEN,I., 2012. The relationship between endangered thick-shelled river mussel LAJTNER,J.,LARSEN,B.M.,LAUCERI,R.,LEGAKIS,A.,LOIS,S., (Unio crassus) and its host fishes. Biological Conservation, 155:94–103. LUNDBERG,S.,MOORKENS,E.,MOTTE,G.,NAGEL,K.-O., TEUTSCH, S. 1997. Populationsbiologische Untersuchungen an ONDINA,P.,OUTEIRO,A.,PAUNOVIC,M.,PRIE,V.,VON Anodonta: Fortpflanzungsstrategien und Charakterisierung von PROSCHWITZ, T., RICCARDI, N., RUDZITE, M., RUDZITIS, M., Teilpopulationen innerhalb eines Stillgewassers. Diplomarbeit, SCHEDER, C., SEDDON, M., SEREFLISAN, H., SIMIC, V., Lehrstuhl fur Okologie, Morphologie und Evolutionsbiologie, SOKOLOVA, S., STOECKL, K., TASKINEN, J., TEIXEIRA, A., Universitat Freiburg. THIELEN, F., TRICHKOVA, T., VARANDAS, S., VICENTINI, H., ZAJAC, K., ZAJAC, T. & ZOGARIS, S. 2017. Conservation status of THRELFALL, W. 1986. Seasonal occurrence of Anodonta cataracta Say freshwater mussels in Europe: state of the art and future challenges. 1817, glochidia on three-spined sticklebacks, Gasterosteus aculeatus – Biological Reviews, 92:572–607. (Linneaus). Veliger, 29: 231 234. LÓPEZ, M.A. & ALTABA, C.R. 2005. Fish host determination for WILES, M. 1974. The glochidia of certain Unionidae (Mollusca) in Nova fi – Margaritifera auricularia (Bivalvia: Unionoida): results and implications. Scotia and their sh hosts. Canadian Journal of Zoology, 53:33 41. Bollettino Malacologico, 41:89–98. ZALE, A.V. & NEVES, R.J. 1982. Fish hosts of four species of lampsiline LÓPEZ, M.A., ALTABA, C.R., ROUAULT, T. & GISBERT, E. 2007. mussels (Mollusca: Unionidae) in Big Moccasin Creek, Virginia. – The European sturgeon Acipenser sturio is a suitable host for the Canadian Journal of Zoology, 60: 2535 2542.

4 174 3.2.3. ARTICLE 5: An unexpected host for the endangered Giant Freshwater Pearl Mussel Margaritifera auricularia (Spengler, 1793) as a tool against the “native species meltdown” effect.

Résumé

Parmi les nombreuses victimes de la perte de biodiversité mondiale dans l'Anthropocène, la grande mulette Margaritifera auricularia est l'une des plus menacées. Présent à l'origine dans de nombreuses rivières européennes, la grande mulette est une relique désormais limitée à des populations vieillissantes en France et en Espagne, dans lesquelles la reproduction naturelle est presque inexistante.

Comme la plupart des moules unionoïdes, la grande mulette a besoin de poissons hôtes pour le développement de ses larves parasitaires (glochidies). Le déclin de son hôte peut être à l'origine d'un effet de «fusion des espèces indigènes» sur cette moule. L'esturgeon européen (Acipenser sturio), le seul poisson hôte indigène connu de la grande mulette en France, est essentiellement éteint. Par conséquent, notre objectif était d'identifier d'autres hôtes pouvant être responsables des quelques cas de recrutement récent.

L'infestation naturelle de poissons sauvages dans trois rivières françaises a été évaluée afin d'identifier d'autres hôtes potentiels de M. auricularia, tandis que des expériences d'infestation artificielle ont été menées sur la lamproie marine (Petromyzon marinus) et sur le silure européen (Silurus glanis) afin de déterminer leur compatibilité en tant qu'hôtes.

Parmi les 29 espèces de poissons évaluées pour une infestation naturelle, seule l'épinoche (Gasterosteus aculeatus) et l'anguille européenne (Anguilla anguilla) étaient porteurs de glochidies de M. auricularia. Dans les expériences d'infestation artificielle, des moules juvéniles vivantes ont été collectées chez P. marinus et S. glanis. Le nombre de juvéniles prélevés sur un seul spécimen de P. marinus (13.827) suggère que cette espèce est un hôte très efficace. Comme pour les hôtes précédemment connus, les hôtes récemment identifiés semblent également avoir une relation avec les environnements marins.

175 Les résultats suggèrent que P. marinus, en tant qu'hôte alternatif, a joué un rôle clé dans la prévention de l'extinction totale de M. auricularia en France. Ils suggèrent également l'utilisation potentielle de P. marinus dans les stratégies de conservation visant à réintroduire ou à stabiliser les populations de ce mollusque rare.

176

An unexpected host for the endangered Giant Freshwater Pearl Mussel Margaritifera auricularia (Spengler, 1793) as a tool against the “native species meltdown” effect Joaquín Soler 1, 4, Catherine Boisneau 1, Philippe Jugé 2, Nina Richard 2, Yann Guerez 2, Laure Morisseau 1, Karl Matthias Wantzen 1, 3 & Rafael Araujo 4

1 University of Tours, CNRS UMR 7324 – CITERES, France 2 University of Tours, CETU Elmis Ingénieries, France 3 UNESCO Chair River Culture – Fleuves et Patrimoines 4 Museo Nacional de Ciencias Naturales – CSIC, Spain

Corresponding author: Joaquín Soler e-mail: [email protected]

177 ABSTRACT 1. Among the many victims of global biodiversity loss in the Anthropocene, the giant freshwater pearl mussel Margaritifera auricularia (GFPM) is one of the most endangered. Originally occurring in many European rivers, the GFPM is a relict now restricted to a few ageing population in France and Spain in which natural reproduction is almost absent.

2. Like most unionoid mussels, the GFPM needs host fish for the development of their parasitic larvae (glochidia). The decline of its host may be causing a “native species meltdown” effect on this mussel. The European sturgeon (Acipenser sturio), the only known native host fish of the GFPM in France, is essentially extinct. Therefore, our aim was to identify alternative hosts that could be responsible for the few cases of recent recruitment.

3. Natural infestation of wild fishes in three French rivers was assessed to identify potential alternative hosts of M. auricularia, while artificial infestation experiments were conducted on the sea lamprey (Petromyzon marinus) and the European wels (Silurus glanis) to determine their compatibility as hosts.

4. Among the 29 fish species assessed for natural infestation, only the three-spined stickleback (Gasterosteus aculeatus) and the European eel (Anguilla anguilla) carried M. auricularia glochidia. In the artificial infestation experiments, living juvenile mussels were collected from both P. marinus and S. glanis. The number of juveniles collected from a single P. marinus specimen (13,827) suggests this species is a highly efficient host. As with previously known hosts, newly identified ones also appear to have a relationship with marine environments.

5. The present findings suggest that P. marinus, as an alternative host, has played a key role in preventing the total extinction of M. auricularia in France. They also suggest the potential use of P. marinus in conservation strategies aimed at reintroducing or stabilizing populations of this rare mollusc.

Keywords: conservation, freshwater mussels, host fish, lamprey, Petromyzon

178 1. INTRODUCTION

The vast majority of freshwater mussel species (order Unionoida) depend on fishes to host larvae until their metamorphosis into juveniles. Mussel larvae (glochidia) expelled by gravid adults attach to fish fins or gills, where they form cysts that last from weeks to months (Kat, 1984). After encystment, juveniles leave their host and begin their benthic or hyporheic stage of life. As adult mussels are almost sessile, dispersal of these species is largely driven by the mobility of hosts while glochidia are encapsulated in their gills.

Once glochidia are released from the mussels, theoretically they could infest any fish. However, fishes develop immunologic responses that glochidia must be able to overcome in order to be physiologically compatible with their host (Bauer, 1997; Bauer & Wächtler, 2001; Haag, 2012). Moreover, not all physiological hosts may be infested under natural conditions as ecological compatibility between the mussel and the host is also required, particularly in terms of geographic co-occurrence and phenological match (i.e., synchronous occurrence of the fish and the glochidia in the same habitat) (Levine, Lang, & Berg, 2012).

While the glochidia of most freshwater mussels develop successfully until the juvenile stage on only a limited number host species, some mussel species are generalist capable of metamorphosing on numerous hosts (Barnhart, Haag, & Roston, 2008; Jansen, Bauer, & Zahner-Meike, 2001; Strayer, 2008). Although the underlying basis of host specificity is largely unknown, it has been proposed that host preference has a heritable component resulting from co-evolutionary processes (Cummings & Mayer, 1993; Graf & Cummings, 2016; Roe, Simons, & Hartfield, 1997). Thus, to better understand the evolutionary history of unionoids, greater knowledge of host–mussel relationships are needed.

Multiple factors have been identified as responsible for the worldwide decline of freshwater mussels including overharvesting, habitat destruction, pollution, land-use change, and exotic species introductions (Lopes-Lima et al., 2017; Strayer et al., 2004). However, several studies have shown the importance of the availability of appropriate host fishes during the reproductive period for the dispersal and population status of unionoid mussels (Bauer, Hochwald, & Silkenat, 1991; Bogan, 1993; Österling, Greenberg, & Arvidsson, 2008; Vaughn & Taylor, 1999; Watters, 1992). When host fishes disappear, mussel populations decline and finally become extinct (Kat & Davis,

179 1984; Watters, 1995); however, reintroduction of host fish species may result in the recovery of declining mussel populations (Araujo et al., 2015; Smith, 1985).

Thus, the lack of host fish species may cause a “native species meltdown” effect. An “invasional meltdown” hypothesis was originally proposed by Simberloff & Von Holle (1999) to describe the effect of mutually facilitating invasive species that contribute to the decline of the native fauna. Here, the term “native species meltdown” is introduced to describe the decline of one species caused by the decline of other species on which it is dependent by any kind of symbiosis, including mutualism, parasitism, or commensalism. We hypothesize that the giant freshwater pearl mussel Margaritifera auricularia (Spengler, 1973) is experiencing a native species meltdown due to the local extintion of its host, the European sturgeon Acipenser sturio L, 1758.

Margaritifera auricularia is considered one of the rarest and the most imperilled bivalve species in Europe (Araujo & Ramos, 2001; Prié, 2010). The population decline in the last two centuries has been estimated to be over 90% (Prié et al., 2018), and today, it is nearly extinct with only a few remaining populations in Spain (Ebro Basin) and France (Loire, Charente, Garonne and Adour basins) (Araujo & Ramos, 2000a; Cochet, 2001; Prié et al., 2018). Although various causes have been identified at the local scale (Araujo & Álvarez- Cobelas, 2016; Prié et al., 2018; Soler, Boisneau, Wantzen, & Araujo, 2018a), the disappearance of its primary host, A. sturio, is thought to be the one most responsible for the overall decline of M. auricularia in the past (Altaba, 1990; Araujo & Ramos, 2000a; López, Altaba, Rouault, & Gisbert, 2007). Nowadays, A. sturio is virtually extinct and only one remaining reproductive population in Europe is known in the Garonne River (France) (Gesner, Williot, Rochard, Freyhof, & Kottelat, 2010).

Besides A. sturio, only three other sturgeon species (A. naccarii Bonaparte, 1836, A. baerii J. F. Brandt, 1869 and A. ruthenus L, 1758) and the small river blenny Salaria fluviatilis (Asso, 1801) have been identified as host of M. auricularia among 21 species tested (Altaba & López, 2001; Araujo, Bragado, & Ramos, 2001; Araujo, Quirós, & Ramos, 2003; Araujo & Ramos, 1998, 2000b; López & Altaba, 2005; López et al., 2007). However, none of the sturgeon species co-occur with current populations of M. auricularia, and the distribution of S. fluviatilis only overlaps with that of the mussel in the Ebro Basin.

180 Despite the extinction of A. sturio, the age estimations of living specimens of M. auricularia (Nakamura et al., 2018; Nienhuis, 2003; J. Soler, unpublished data) and the recent, albeit weak, recruitment observed at several French and Spanish localities (Nakamura et al., 2018; Prié et al., 2018) indicate that current populations display some level of recruitment, presumably facilitated by alternative hosts. In the Ebro Basin, S. fluviatilis may have prevented the extinction of M. auricularia; however, none of the known hosts of this mussel co-occur with any of the French populations. More recently, the three-spine stickleback (Gasterosteus aculeatus L, 1758) has been identified as a host, indicating it as a potential alternative host fish species in France. However, given its low quality as host and its absence near some recruiting GFPM populations (Soler et al., 2018a), it seems unlikely to be a key alternative host species. French populations appear to use an unknown host species, which further suggest a broader host range for M. auricularia than previously thought.

As previously observed (Soler et al., 2018a), all known host species of M. auricularia share a common feature: they can tolerate some degree of salinity. Sturgeons are anadromous fishes, the three-spined stickleback can inhabit brackish, marine and freshwater and the river blenny, although restricted to freshwater, can tolerate high salinity levels (Plaut, 1998) and belongs to a mainly marine family (Kottelat & Freyhof, 2007). Indeed, although the hosts of many margaritiferid species are still unknown (Lopes-Lima et al., 2018), of the known hosts, the majority share this feature (Soler et al., 2018a and references therein). This pattern suggests that the common ancestor of the Margaritiferidae family likely specialized in using migratory fishes that are able to move between freshwater and marine environments, which could have promoted their dispersal.

Given this context, the hypothesized alternative host species of M. auricularia populations in France is likely salt tolerant. Therefore, in the present study, the “marine ancestor and/or amphidromous migrating” pattern of the host species that may have been responsible for the observed recruitment in these populations was also analysed. One of the most abundant anadromous species co-occurring in French rivers with recruiting populations of M. auricularia is the sea lamprey Petromyzon marinus L, 1758 (Amrein, 2004, 2005; Bach et al., 2016; Dartiguelongue, 2017; Taverny et al., 2005). Lampreys are considered “living fossils” because they have remained largely unaltered for 360 million years (Gess, Coates, & Rubidge, 2006). Thus, these primitive aquatic vertebrates, like sturgeons, could have been the first hosts of the family Margaritiferidae, which originated

181 in the mid-Jurassic (Araujo, Schneider, Roe, Erpenbeck, & Machordom, 2017; Lopes- Lima et al., 2018), and possibly of other freshwater mussel families.

The main objective of this study was to identify alternative hosts of the French populations of M. auricularia. The suitability of the sea lamprey as a host was also investigated, and its potential use in M. auricularia conservation programs are discussed as a tool to offset native species meltdown. Two approaches were followed in this study: an assessment of natural infestation in fishes sampled in the field by electrofishing during the GFPM reproduction period and a laboratory study on the artificial infestation of sea lamprey with glochidia from adult mussels sampled from French rivers. Given the number of known physiological hosts of M. auricularia, namely six fish species belonging to three different families, this mussel may be, paradoxically, a host generalist. For this reason, and given the ubiquity of the exotic species Silurus glanis L, 1758 (European wels) in French and Spanish rivers inhabited by M. auricularia (Elvira & Almodóvar, 2001; Keith, Persat, Feunteun, & Allardi, 2011; Schlumberger, Sagliocco, & Proteau, 2001), this species was also experimentally infested with glochidia to test its potential as a host.

2. METHODS

2.1. Natural infestation assessment

The metamorphosis of M. auricularia glochidia takes 4-6 weeks to complete; therefore, the electrofishing dates were chosen to coincide with the glochidia encystment period (Soler, Wantzen, Jugé, & Araujo, 2018b). During late April and beginning of May of 2016 and 2017, the fish communities of the Creuse, Vienne and Charente rivers were assessed for natural infestation by electrofishing directly downstream of M. auricularia populations. Due to the elevated depths and widths of the river sections, electrofishing was performed from a boat.

Stunned fishes were collected with a dip net and maintained in aerated plastic tanks. Fishes were anesthetized in small containers with a 4% solution of eugenol prior to gill inspection under a binocular microscope. After inspection, most of the captured fishes were released back into the sampling sites. As the gills of the European eel (Anguilla

182 anguilla L, 1758) are not easily accessible, specimens of this species, and any other fish specimens in which infestation could not be clearly assessed in the field, were analysed in the laboratory.

The glochidia of M. auricularia are easily distinguishable from those of other unionid mussels (genera Anodonta and Unio) occurring in these rivers due to their spoon, rather than triangular, shape. However, as encapsulated glochidia of M. auricularia are difficult to distinguish from those of Potomida littoralis (Cuvier, 1798), gill filaments containing encapsulated glochidia were examined under a microscope in the laboratory in order to confirm the species identity.

The electrofishing method in deep waters may limit which species are captured. For this reason, local anglers and regional associations studying migratory fishes (LOGRAMI and Cellule Migrateurs Charente et Seudre) were asked to send frozen heads of migratory species to the laboratory for analysis. These specimens were also captured near the studied M. auricularia populations during the glochidia release period.

2.2. Experimental infestations

In March and April 2017, a total of 20 mussels from the Charente and 16 from the Creuse and Vienne rivers were captured and transported to the laboratory where they were maintained at a mean temperature of 13.5 ºC in aerated aquaria filled with water from their respective rivers and without sediment (see Soler et al., 2018a).

On 26 March 2017, five adult sea lampreys (1.04 kg mean weight ± SD 0.1) were captured from the Loire River near Nantes while they were migrating upstream. They were transported to the laboratory and placed in a 1500 L tank equipped with biological, mechanical and UV filters and filled with aerated river water. Lampreys were not fed as they do not feed during their upstream migration. Water temperature was automatically recorded every 30 minutes to calculate degree-days (DD), i.e. the sum of daily water temperatures (Chezik, Lester, & Venturelli, 2013).

Additionally, five S. glanis (0.6 kg mean weight ± SD 0.2) were captured from the Loire River near Amboise on 23 and 24 April 2017. The specimens were maintained under the same conditions as lampreys except they were fed live Rutilus rutilus (L, 1758) daily until 415 DD post infestation.

183 Due to different release times and the short life span of glochidia, artificial infestations were performed on two dates. On 28 March and 2 April 2017, glochidia originating from two different females from the Charente were used to produce homogeneous suspensions for the artificial infestation of the sea lampreys. Prior to infestation, glochidia viability was tested by observing, under a binocular microscope, the response of a small sampling of glochidia to the addition of NaCl.

On 28 March 2017, four sea lampreys were placed in an infestation bath for 15 min under constant agitation. The infestation concentration of the bath was 10,800 glochidia/L in a total volume of 40 L. On 2 April 2017, the remaining lamprey specimen was infested under the same conditions and glochidia concentration in a total volume of 11.5 L.

To check for changes in encystment rates, one lamprey was anesthetized and sacrificed at 1, 12 and 24 days post infestation (pi). The gills were excised, and the number of cysts were counted under a binocular microscope.

To collect juvenile mussels, approximately 415 DD after infestation, the two remaining lampreys were transferred to a cylindroconical tank (300 L) equipped with biological and mechanical filters. However, one specimen escaped from the tank within 1 day after transfer and died. The animal was recovered and its gills excised for cyst counting (30 days pi). A 100 µm mesh collector was installed in the water circuit and checked daily until no more juveniles were found from the last remaining specimen.

On 23 April 2017, three S. glanis specimens were placed in an infestation bath with a concentration of 8,500 glochidia/L in a total volume of 30 L. Two days later, two other specimens were infested using the same glochidia concentration in a total volume of 10 L. The infestation procedure was as with the lampreys, and likewise, one specimen was sacrificed at 1, 12 and 24 days pi. However, in this case, both of the remaining specimens transferred to the cylindroconical tank died 5 days later and, thus, juveniles (and cysts) could not be further counted.

The research was conducted in compliance of the Directive 2010/62/EU, under permits from the Ministère de la Recherche and the Animal Care Committee (CEEA Val de Loire) of the University of Tours (n°2016121410354629).

184 Table 1: Results of natural infestation assessment by electrofishing in three French rivers inhabited by M. auricularia. The number of specimens inspected for M. auricularia glochidia are given for each fish species, year and river. The number of specimens in which M. auricularia glochidia were found are indicated in parentheses. 2016 2017 Species TOTAL CHARENTE CREUSE CHARENTE CREUSE VIENNE Abramis brama 1 0 0 0 0 1 Alburnoides bipunctatus 0 36 0 16 0 52 Alburnus alburnus 33 15 0 0 0 48 Alosa fallax 2 0 0 0 0 2 Anguilla anguilla 22 (6) 17 20 (8) 10 8 (4) 77 (18) Barbus barbus 0 4 0 30 17 51 Blicca bjoerkna 54 45 0 0 0 99 Carassius sp. 20 4 5 0 0 29 Chondrostoma nasus 0 17 0 1 7 25 Cobitis taenia 0 1 1 0 0 2 Cyprinus carpio 0 0 0 1 0 1 Esox lucius 4 2 4 1 0 11 Gasteroteus aculeatus 29 (2) 0 0 0 0 29 (2) Gobio sp. 2 34 24 2 9 71 Gymnocephalus cernua 4 4 2 5 0 15 Lepomis gibbosus 55 0 0 1 0 56 Leuciscus leuciscus 0 0 0 1 0 1 Liza ramada 6 2 9 1 3 21 Perca fluviatilis 1 3 3 13 6 26 Phoxinus phoxinus 0 1 2 0 2 5 Pseudorasbora parva 0 0 0 6 0 6 Rhodeus amarus 15 1 13 8 2 39 Rutilus rutilus 98 41 0 0 0 139 Salmo trutta 0 0 0 1 0 1 Sander lucioperca 0 1 0 0 0 1 Scardinius erythrophthalmus 8 0 1 0 0 9 Silurus glanis 4 2 0 5 16 27 Squalius cephalus 28 26 12 16 15 97 Tinca tinca 0 0 20 4 1 25

185 3. RESULTS

3.1. Natural infestation

A total of 966 specimens belonging to 29 fish species were captured and their gills inspected for encysted glochidia. No specimens of P. marinus were captured (Table 1). Glochidia of M. auricularia were found only in two species: G. aculeatus (see Soler et al., 2018a) and A. anguilla. In both fish species, glochidia had attached recently and no cysts were observed. Glochidia of other mussel species were not found on the gills of any of the captured fishes. Anguilla anguilla specimens were collected from all studied river sections, and 18 of the 77 inspected specimens from the Charente and Vienne had M. auricularia glochidia in their gills (Table 1). The average size of infested A. anguilla (314 ± 82 mm) was significantly smaller (non-parametric Mann-Whitney test, N = 77, U = 732, p-value = 0.016) than the average size of all collected A. anguilla (405 ± 140 mm).

Two migratory fish species collected near M. auricularia populations in the Charente by regional associations were also inspected: no glochidia were found attached to the gills of the 35 analysed specimen of Alosa fallax Lacèpéde, 1800 and the 2 of Alosa alosa L, 1758.

186

Table 2: Results of Petromyzon marinus artificial infestations 1, 12 and 24-30 days post infestation (pi), and juvenile mussels after excystment. SD= Standard deviation.

1 day pi 12 days pi 24-30 days pi Excystment Temperature Infestation Glochidia Glochidia Glochidia Juveniles Start - end of X ± SD (°C) date n fish n fish n fish n fish per fish per fish per fish per fish excystment (days pi) 3,803 1 28/03/2018 5,804 1 1,226 1 - - - 15.93 ± 1.98 2,195 1 02/04/2018 ------13,827 1 25 - 46 16.67 ± 1.58

187 3.2. Experimental infestations

The glochidia of M. auricularia successfully metamorphosed on the sea lamprey (Figure 1). A total of 13,827 living juvenile mussels were collected from a single P. marinus specimen weighing 1,2 kg (=11.15 juveniles per gram of fish) (Table 2). The first completely metamorphosed and viable juvenile mussels were detected 25 days pi (416 DD), and the last one was recovered 46 days pi (770 DD). Of the total number of juveniles, 92% were collected between 38 and 42 days pi (631-702 DD).

Live juveniles were also obtained from S. glanis, although the total number could not be calculated as these fish died prior to the end of the experiment. Although 177 juveniles were collected from the two fish transferred to the cylindroconical tank, this number seems to be low considering that the specimen sacrificed at 12 days pi had a high number of cysts (10,795 cysts). The two specimens sacrificed at 1 and 24 days had 1,315 and 2,774 cysts respectively. The duration of metamorphosis was the same as in P. marinus.

Figure 1: Encysted glochidia of M. auricularia at the end of metamorphosis (30 days pi) in the gill of P. marinus. (a) General view of the gill; (b) detail view of a cyst. Scale bars: (a) 500 µm; (b) 100 µm.

188

4. DISCUSSION

4.1. Natural and experimental infestation

According to the findings of the infestation experiments, P. marinus is one of the best known physiological hosts of M. auricularia in terms of juvenile production (Soler et al., 2018a). To the best of our knowledge, this study is the first to confirm P. marinus as a host of any freshwater mussel. In fact, several authors have pointed out how surprisingly infrequent lampreys are used as hosts by freshwater mussels despite being widespread and common benthic vertebrates (Haag, 2012; Strayer, 2008).

Although in this study, no P. marinus specimens were captured by electrofishing in any of the studied rivers, likely due to the limitations of electrofishing in deep waters (Moser, Butzerin, & Dey, 2007), the available information on its geographical distribution, phenology, habitat and behaviour suggest that this species may be a good ecological host of M. auricularia.

According to its distribution, the studied rivers, particularly the Vienne and Creuse, are known to harbor large populations of P. marinus (Bach et al., 2016; Dartiguelongue, 2017), and spawning areas of this species have been observed in the vicinity of M. auricularia habitats, at least in the Vienne and Creuse rivers (Portafaix, Senecal, & Baisez, 2015). Indeed, P. marinus is also present in the other French rivers inhabited by recruiting populations of M. auricularia (Amrein 2004, 2005; Prié et al., 2018; Taverny et al., 2005). Furthermore, the historical distribution of M. auricularia coincides with the European distribution range of P. marinus (Figure 2). Thus, as proposed for A. sturio (Altaba, 1990; Araujo & Ramos, 2000a, 2000c), the distribution of P. marinus could explain the distribution of M. auricularia.

189

Figure 2: Distribution of P. marinus (dark shaded) within the M. auricularia geographical range. Distribution of P. marinus modified from Kottelat & Freyhof, (2007) and NatureServe, (2013). Distribution of M. auricularia modified from Prié et al., (2018).

Findings reported by Prié et al. (2018) suggest that M. auricularia is recently extinct in the Rhône and Seine basins but persists in the Ebro, Loire, Charente, Garonne and Adour basins. The extirpation of M. auricularia in the Rhône and Seine basins may be related to the disappearance of A. sturio and P. marinus during the 19th and 20th centuries (Allardi & Keith, 1991 and references therein; Belliard, Boët, & Allardi, 1995). The persistence of P. marinus in the Charente, Loire, and Adour basins may account for the continued presence of GFPM populations in those basins, even after the extirpation of A. sturio. In the Ebro basin, P. marinus was likely abundant until the construction of big dams in the first half of the 20th century (López, Gázquez, Olmo-Vidal, Aprahamian, & Gisbert, 2007; Mateus, Rodríguez-Muñoz, Quintella, Alves, & Almeida, 2012), suggesting that, together with S. fluviatilis, they could have played an important role in maintaining the Spanish populations of the GFPM.

190 Lampreys spend the first four to five years of life living buried in river sediments as filter- feeding larvae called ammocoetes (Hansen et al., 2016; Taverny et al., 2005). They then metamorphose into juveniles and migrate to the sea where they start their hematophagous phase (Youson, 1980). Adult sea lampreys only spend a few months of their lives in rivers implying that the timing of glochidia release is of crucial importance. The migration period of P. marinus, like in other anadromous species, is triggered by rising temperatures. In French rivers, migration occurs from December to late June, peaking in March and April (Taverny & Élie, 2010), which coincides with the glochidia release period of French populations of M. auricularia (Soler et al., 2018b). This period also overlaps with the migration of A. sturio adults from the sea from April to May (CEMAGREF, 1994). Thus, the earlier reproductive period of M. auricularia in relation to other margaritiferids, which is also driven by temperature (Soler et al., 2018b), may be linked to the arrival of anadromous adults to their spawning grounds. Nevertheless, hosts at the juvenile stage appear to be most highly infested with freshwater mussel glochidia (Karna & Millemann, 1978; Modesto et al., 2018; Young & Williams, 1984). Given the large number of ammocoetes present during the reproductive period of M. auricularia, one would expect them to also be infested by glochidia.

As M. auricularia can be hosted by several species of sturgeons, it is quite possible that other lamprey species can also serve as hosts. For this reason, and given their geographical coincidence in France, Lampetra fluviatilis (L, 1758) and L. planeri (Bloch, 1784) should be tested as potential hosts, as originally proposed by Consejo (2016).

Results from the experimental infestation of S. glanis indicate it is a physiological host of M. auricularia. Nevertheless, none of the 27 individuals collected by electrofishing was infested with M. auricularia glochidia, suggesting that it is not a good ecological host. Although its current distribution matches both the current and historic range of M. auricularia, its sympatry is a recent phenomenon. Silurus glanis is native to Central and Eastern European rivers (e.g. Volga, Dnieper, Danube and probably the upper part of the Rhine and some Baltic Sea tributaries) (Triantafyllidis et al., 2002) but was introduced in different areas of Western Europe during the late 19th century due to economic interest (Schlumberger et al., 2001). However, this species was only introduced to Spanish and French rivers, including those inhabited by M. auricularia, between 1974 and 1980 (Elvira & Almodóvar, 2001; Valadou, 2007).

191 The natural infestation assessment of A. anguilla showed that smaller eels were more likely to be infested than larger ones. This finding likely reflects a difference in acquired immunity of the older specimens to glochidia through previous infestations (Barnhart, et al., 2008; Kat, 1984). Interestingly, in infestation experiments of this species with M. auricularia glochidia, no encysted glochidia were found beyond 4 (Araujo et al., 2001) or 9 (Lopez & Altaba 2005) days pi. Given that A. anguilla appears to be an ecological host of M. auricularia, new infestation experiments should be conducted to support the field assessment.

The two Alosa species examined, A. fallax and A. alosa, could also be potential hosts (Consejo, 2016; Prié, Phillipe, & Cochet, 2011). In addition to being anadromous, they ascend the rivers from March to May, coinciding with the glochidia release period of M. auricularia (Dartiguelongue, 2017; Mennesson-Boisneau, Aprahamian, Sabatie, & Cassou-Leins, 2000). Indeed, their spawning habitats match those of M. auricularia, at least in the Creuse River (J. Soler, personal observation). Furthermore, they are the most numerous migratory fish in the Charente River and the second most numerous in the Creuse River (Bach et al., 2016; Dartiguelongue, 2017). Nevertheless, in this study, none of the Alosa specimens were infested with glochidia. The pelagic habits of these species may render them less prone to contact with M. auricularia glochidia (Froese & Pauly, 2018). Despite the difficulty of maintaining these species in captivity, experimental tests with both Alosa species, as has been proposed by Consejo (2016), would reveal their capacity as physiological hosts.

4.2. Generalist vs specialist mussels

Bauer (2001) suggested that host range is the main factor determining host specificity of freshwater mussels. Mussels inhabiting diverse and species-rich habitats tend to be host generalists, while species from specific habitats with a relatively homogeneous fish fauna tend to be host specialists. Accordingly, a host generalist strategy should be more advantageous in environments with a variable and diverse fish community, where glochidial survival is increased by their ability to use a wide range of host species.

Previous studies have suggested that invasive fish species are less suitable as hosts for freshwater mussels than native ones, probably due to a co-evolutionary mechanism of host compatibility between mussels and host fish species (Douda et al., 2013; Salonen,

192 Marjomäki, & Taskinen, 2016; Taeubert, Gum, & Geist, 2012). According to the available data, the vast majority of freshwater mussel species utilizes only native hosts, whereas the glochidia of only a few host generalist species have been proven to successfully metamorphose on non-native fishes (Douda et al., 2017; Huber & Geist, 2017; Modesto et al., 2018; Teixeira et al., 2018).

As mentioned, M. auricularia appears to be, paradoxially, a host generalist in contrast to most other margaritiferid species which use very few species of a single family. The discovery of two new host species (P. marinus and the exotic S. glanis), which increases the number of known physiological host to eight species from five different families, further supports this hypothesis.

4.3. Salinity tolerance of the host

The frequent use of anadromous host fishes by margaritiferid mussels is relatively well documented (Araujo, et al., 2017; Curole, Foltz, & Brown, 2004). In the case of M. auricularia, nearly all known hosts have a notable tolerance to salinity, including P. marinus (Soler et al., 2018a; this study). While S. glanis seems to deviate slightly from this pattern of halotolerance because it mainly occurs in freshwater, it has been observed to enter brackish waters in the Baltic, Black and Mediterranean seas and even to spawn in salt water (Berg, 1964; Frimodt, 1995). Notably, other freshwater fishes that have been identified as likely hosts of M. auricularia, namely Cottus gobio L, 1758 and Gambusia holbrooki Girard, 1859 (Araujo et al., 2003; López & Altaba, 2005; M. A. López, pers. comm.) also display some tolerance to salt (Kottelat & Freyhof, 2007; Nordlie & Mirandi, 1996).

This preference may have arisen because it was advantageous for mussel dispersion. Araujo et al. (2017) suggested that the current distribution of margaritiferids can be explained by the capacity of their hosts to successfully bridge continents or to circumvent hostile land areas that act as barriers for dispersal by going through environments with reduced salinities. If freshwater mussels developed an affinity for euryhaline hosts during the first stages of their evolution, it is possible that both sturgeons and lampreys have since been used as hosts. Freshwater mussels are an ancient group that evolved from marine bivalves during the Triassic (Haas, 1969; Watters, 2001). Thus, if parasitism

193 evolved during the colonization of freshwater environments, the ancestral unionoid likely used euryhaline hosts.

According to a phylogeny of the order Unionoida, Margaritiferidae can be the basal family within the order (Strayer, 2008). Given this context, we hypothesize that present- day margaritiferids retained the trait of using euryhaline hosts from their unionoid ancestors. Further research is required to support this hypothesis including the identification of the host fishes of mussel species most closely related to M. auricularia, such as M. marocana and M. homsensis. Likewise, it would be interesting to test whether P. marinus could be a functional host of these species considering that, although outside its main distribution range, P. marinus has been cited in areas currently inhabited by both M. marocana and M. homsensis (Bilecenoglu, Taskavak, Mater, & Kaya, 2002; Boutellier, 1918; Çevik, Ergüden, & Tekelioğlu, 2010; Dollfus, 1955; Furnestin et al., 1958). Indeed, the ability of lampreys to host other margaritiferids should be investigated generally, and in particular, in M. margaritifera given the high degree of sympatry reported in Central European rivers (Geist, 2005).

4.4. Implications for conservation

All confirmed host species of M. auricularia appear to be good physiological hosts except G. aculeatus (Soler et al., 2018a). However, G. aculeatus is the only host that has been found infested with M. auricularia glochidia both in the wild and in artificial infections (Soler et al., 2018a). Information on the geographical distribution, habitat requirements, phenology and behaviour of the physiological hosts suggests that most could also be considered ecological hosts except S. glanis and the exotic sturgeons A. baeri and A. ruthenus.

Silurus glanis is an invasive species, and although it has a wide dietary spectrum, it has been reported to prey on anadromous species, including P. marinus, Salmo salar L, 1758 and Alosa spp., in some rivers like the Garonne (Boulêtreau, Gaillagot, Carry, Tétard, & De Oliveira, 2018; Cucherousset et al., 2018; Guillerault, Bouletreau, Iribar, Valentini, & Santoul, 2017; Syväranta et al., 2009). Its presence in regulated rivers may also be contributing to significant changes in the native fish fauna (Gavioli et al., 2018), particularly by comprising the threatened S. fluviatilis whose decline has been associated with predation by exotic fishes (Carol, 2007; Doadrio, 2001). However, the abundance

194 and wide distribution of S. glanis in the vicinity of GFPM populations make it an easy fish to obtain. Although not conclusive, the findings presented here suggest that this species has a high capacity to transform large quantities of glochidia. If so, this fish could be a better alternative to A. baeri for obtaining juveniles in the laboratory.

Given the difficulty of reintroducing the endangered A. sturio, only S. fluviatilis, G. aculeatus and P. marinus can be used in the application of a simplified, cost effective technique to boost M. auricularia populations in the wild. This technique, which has been successfully used for other species (Altmüller & Dettmer, 2006; Araujo, Feo, Pou, & Campos, 2015; Carey, Jones, Butler, & Hallerman, 2015), consists of catching host species in the vicinity of existing mussel populations, infesting them with glochidia, and then immediately releasing them back into the wild in order to optimize the window of recruitment of this European heritage species. In France and Spain, P. marinus is considered threatened and, in some areas, has disappeared or become increasingly rare mainly due to habitat loss related to dam construction, habitat disruption and overfishing (Hansen et al., 2016; Mateus et al., 2012). This type of measure would favour the reinforcement of not only M. auricularia populations but also those of P. marinus by expanding their spawning habitats. In addition, Limm & Power (2011) reported that Pacific lamprey (Petromyzon tridentatus Richardson, 1836) larvae grow faster when found near Margaritifera falcata (Gouls, 1850) beds, where mussels capture, concentrate, and deposit food near their burrows. A similar potential mutualistic relationship between M. auricularia and P. marinus should be investigated in future studies.

5. Conclusions

The findings reported here suggest that M. auricularia is a host generalist with a preference for halotolerant host species. This pattern could have arisen early in its evolution because of the dispersion advantage that diadromous species putatively confer. However, anthropogenic causes, such as dam construction and water quality and habitat deterioration, have led to a severe decline of most migratory fish species in Europe during the last two centuries (Hansen et al., 2016; Limburg & Waldman, 2009; Mateus et al., 2012). Although other causes may have contributed to the overall decline of M. auricularia, the loss of hosts seems to be a main factor.

195 In addition, the present findings suggest that P. marinus, as an alternative host, has played a key role in preventing the total extinction of M. auricularia following the disappearance of the European sturgeon. Furthermore, the potential use of P. marinus to reinforce M. auricularia populations along its entire geographic range may counteract this native species’ meltdown.

ACKNOWLEDGEMENTS

This work was conducted within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’. This paper has been produced under the auspices of the UNESCO Chair “Fleuves et Patrimoine / River Culture”. Thanks to the Charente-Maritime and Indre et-Loire Prefectures for sampling permits. We also thank the club Chinon Plongée for help in mussel sampling. We are grateful to all the institutions that made the fishing surveys possible: LOGRAMI, Cellule Migrateur Charente, BIOTOPE, the regional and local services of the French Biodiversity Agency (AFB 16, 17, 37, 86) and some anglers associations (FDAAPPMA 16, 17, 37). Figure 2 was prepared by Victoria González-Garzón from the Computer Biogeography Laboratory of the MNCN-CSIC. Pictures were compiled by Jesús Muñoz from the photography facility of the MNCN-CSIC. Melinda Modrell reviewed the English version of the manuscript.

REFERENCES

Allardi, J., & Keith, P. (1991). Atlas préliminaire des poissons d’eau douce de France (Patrimoines naturels; 4). Paris, Muséum national d'Histoire naturelle.

Altaba, C. R. (1990). The last known population of the freshwater mussel Margaritifera auricularia (Bivalvia, Unionoida): A conservation priority. Biological Conservation, 52, 271–286.

Altaba, C. R., & López, M.A. (2001). Experimental demonstration of viability for the endangered Giant Pearl Mussel Margaritifera auricularia (Bivalvia: Unionoida) in its natural habitat. Bolleti de la Societat d’Historia Natural de les Baleares, 44, 15–21.

196 Altmüller, R. & Dettmer, R. (2006). Erfolgreiche Artenschutzmaßnahmen für die Flussperlmuschel Margaritifera margaritifera L. durch Reduzierung von unnatürlichen Feinsedimentfrachten - Erfahrungen im Rahmen des Lutterprojekts. Informationsdienst Naturschutz Niedersachsen, 26, 192–204.

Amrein, S. (2004). Carte de répartition de la lamproie marine sur le bassin de l’Adour Tranche 2/3 : Adour et affluents rive gauche. Pau: Association pour la restauration des poissons migrateurs sur le bassin de l’Adour.

Amrein, S. (2005). Carte de répartition de la lamproie marine sur le bassin de l’Adour Tranche 3/3 : Adour et affluents rive droite. Pau: Association pour la restauration des poissons migrateurs sur le bassin de l’Adour.

Araujo, R., & Álvarez-Cobelas, M. (2016). Influence of flow diversions on Giant Freshwater Pearl Mussel Populations in the Ebro River, Spain. Aquatic Conservation - Marine and Freshwater Ecosystems, 26, 1145–1154.

Araujo, R., Bragado, D., & Ramos, M. A. (2001). Identification of the river blenny, Salaria fluviatilis, as a host to the glochidia of Margaritifera auricularia. Journal of Molluscan Studies, 67, 128–129.

Araujo, R., Feo, C., Pou, Q., & Campos, M. (2015). Conservation of two endangered European freshwater mussels (Bivalvia: Unionidae): A three-year, semi-natural breeding experiment. The Nautilus, 129, 126–135.

Araujo, R., Quirós, M., & Ramos, M.A. (2003). Laboratory propagation and culture of juveniles of the endangered freshwater mussel Margaritifera auricularia (Spengler, 1793). Journal of Conchology, 38, 53–60.

Araujo, R., & Ramos, M.A. (1998). Description of the glochidium of Margaritifera auricularia (Spengler, 1793) (Bivalvia, Unionidae). Philosophical Transactions of the Royal Society of London B, 353, 1553–1559.

Araujo, R., & Ramos, M. A. (2000a). A critical revision of the historical distribution of the endangered Margaritifera auricularia (Spengler, 1782) (Mollusca: Margaritiferidae) based on Museum specimens. Journal of Conchology, 37, 49–59

197 Araujo, R., & Ramos, M. A. (2000b). Status and conservation of the relict giant European freshwater pearl mussel Margaritifera auricularia (Spengler, 1793). Biological Conservation, 96, 233–239.

Araujo, R., & Ramos, M. A. (2000c). Life History data on the virtually unknown Margaritifera auricularia. In Bauer, G., & Wächtler, K. (Eds.), Ecological Studies: Ecology and Evolution of the Freshwater Mussels Unionoida (pp. 143–154), Berlin- Heidelberg, Germany: Springer-Verlag.

Araujo, R., & Ramos, M. A. (2001). Action Plan for Margaritifera auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Nature and environment, No. 117. Strasbourg: Council of Europe Publishing.

Araujo, R., Schneider, S., Roe, K. J., Erpenbeck, D., & Machordom, A. (2017). The origin and phylogeny of Margaritiferidae (Bivalvia, Unionoida): A synthesis of molecular and fossil data. Zoologica Scripta, 46, 289–307.

Bach, J. M., Parouty, T., Leon, C., Senecal, A., Portafaix, P., Barault, A., Rautureau, C., Rimbert L., Duperray A., & Baisez A. (2016). Volet scientifique: suivis biologiques des populations de poissons grands migrateurs, année 2016. Rapport d’activité du Programme de Recherches Appliquées en Faveur des Poissons Migrateurs. Poitiers: Association LOGRAMI.

Barnhart, M. C., Haag, W. R., & Roston, W. N. (2008). Adaptations to host infection and larval parasitism in Unionoida. Journal of the North American Benthological Society, 27, 370–394.

Bauer, G. (1997). Host relationships at reversed generation times: Margaritifera (Bivalvia) and salmonids. Ecological Studies, 130, 69–79.

Bauer, G. (2001). Framework and driving forces for the evolution of naiad life histories. In Bauer, G., & Wächtler, K. (Eds.), Ecological Studies: Ecology and Evolution of the Freshwater Mussels Unionoida (pp. 233–256), Berlin-Heidelberg, Germany: Springer- Verlag.

198 Bauer, G., Hochwald, S., & Silkenat, W. (1991). Spatial distribution of freshwater mussels: the role of host fish and metabolic rate. Freshwater Biology, 26, 377–386.

Bauer, G., & Wächtler, K. (2001). Ecology and evolution of the Naiads. In Bauer, G., & Wächtler, K. (Eds.), Ecological Studies: Ecology and Evolution of the Freshwater Mussels Unionoida (pp. 383–388), Berlin-Heidelberg, Germany: Springer-Verlag.

Belliard, J, Boët, P, & Allardi, J. 1995. Evolution à long terme du peuplement piscicole du bassin de la Seine. Bulletin Français de la Pêche et de la Pisciculture, 337: 83–91.

Berg, L.S. (1964). Freshwater fishes of the U.S.S.R. and adjacent countries. Volume 2, 4th edition. Jerusalem: Israel Program for Scientific Translations Ltd.

Bilecenoglu, M., Taskavak, E., Mater, S., & Kaya, S. M. (2002). Checklist of the marine fishes of Turkey. Zootaxa, 113, 1–194.

Bogan, A. E. (1993). Freshwater bivalve extinctions (Mollusca: Unionoida): A search for causes. American Zoologist, 33, 599–609.

Boulêtreau, S., Gaillagot, A., Carry, L., Tétard, S., & De Oliveira, E. 2018. Adult Atlantic salmon have a new freshwater predator. PLoS ONE, 1–12.

Boutellier, P. (1918). La péche maritime au Maroc (Casablanca, Fédala, Rabat). Rabat : G. Mercié & Co.

Carey, C. S., Jones, J. W., Butler, R. S., & Hallerman E. M. (2015). Restoring the endangered oyster mussel (Epioblasma capsaeformis) to the upper Clinch River, Virginia: an evaluation of population restoration techniques. Restoration Ecology, 23, 447–454.

Carol, J. (2007). Ecology of an invasive fish (Silurus glanis) in Catalan reservoirs. (PhD thesis). University of Girona, Spain.

CEMAGREF. (1994). Opération de sauvegarde de l'esturgeon européen Acipenser sturio. Bordeaux: Centre d’Etude du Mécanisme Agricole du Génie Rural des Eaux et Forêts.

199 Çevik, C., Ergüden, D., & Tekelioğlu, N. (2010). Confirming the presence of sea lamprey (Petromyzon marinus Linnaeus, 1758) in the Levantine Sea (Petromyzoniformes: Petromyzonidae). Zoology in the Middle East, 49, 107–108.

Chezik, K. A., Lester, N. P., & Venturelli, P. A. (2013). Fish growth and degree-days I: selecting a base temperature for a within-population study. Canadian Journal of Fisheries and Aquatic Sciences, 71, 47–55.

Cochet, G. (2001). Redécouverte d’une population vivante de la Grande Mulette, Margaritifera auricularia, sur la Vienne et la Creuse. Recherches Naturalistes en Région Centre, 10, 3–16.

Consejo, J. 2016. Investigaciones sobre los posibles peces hospedadores de los gloquidios de la almeja de río Margaritifera auricularia (Mollusca, Bivalvia, Margaritiferidae). (Final undergraduate thesis). Universidad Autónoma de Madrid, Spain.

Cucherousset, J., Horky, P., Slavík, O., Ovidio, M., Arlinghaus, R., Boulêtreau, S., Britton, R., García-Berthou, E., & Santoul, F. 2018. Ecology, behaviour and management of the European catfish. Reviews in Fish Biology and Fisheries, 28: 177–190.

Cummings, K. S., & Mayer, C. A. (1993). Distribution and host species of the federally endangered freshwater mussel, Potamilus capax (Green, 1832) in the Lower Wabash River, Illinois and Indiana. Illinois Natural History Survey, Center for Biodiversity, Technical Report, 1–29.

Curole, J. P., Foltz, D. W., & Brown, K. M. (2004). Extensive allozyme monomorphism in a threatened species of freshwater mussel, Margaritifera hembeli Conrad (Bivalvia: Margaritiferidae). Conservation Genetics, 5, 271–278.

Dartiguelongue, J. (2017). Contrôle du fonctionnement de la passe à poissons installée à Crouin (16) sur la Charente. Suivi de l’activité ichtyologique en 2016. Services et Conseils en Environment Aquatique. Toulouse: Services et Conseils en Environment Aquatique.

Doadrio, I. (2001). Atlas y libro rojo de los peces continentales de España. Madrid: Dirección General de Conservación de la Naturaleza.

200 Dollfus, R. P. (1955). Première contribution à l’établissement d’un fichier ichtyologique du Maroc atlantique. Travaux de l’Institute Scientifique Chérifien, Série Zoologie, 6, 1- 226.

Douda, K., Lopes-Lima, M., Hinzmann, M., Machado, J., Varandas, S., Teixeira, A., & Sousa, R., (2013). Biotic homogenization as a threat to native affiliate species: fish introductions dilute freshwater mussel's host resources. Diversity and Distributions, 19, 933–942.

Douda, K., Liu, H.-Z., Yu, D., Rouchet, R., Liu, F., Tang, Q.-Y., Methling, C., Smith, C. & Reichard, M. 2017. The role of local adaptation in shaping fish-mussel coevolution. Freshwater Biology, 62:1858–1868.

Elvira, B., & Almodóvar, A. (2001). Freshwater fish introductions in Spain: facts and figures at the beginning of the 21st century. Journal of Fish Biology, 59, 323–331.

Frimodt, C. (1995). Multilingual illustrated guide to the world's commercial coldwater fish. Oxford, UK: Fishing News Books.

Froese, R. & Pauly, D. (2018). FishBase. World Wide Web electronic publication. www.fishbase.org, version (06/2018).

Furnestin, J., Dardignac, J., Maurin, C., Vincent, A., Coupé, R., & Boutière, H. (1958). Donnees nouvelles sur les poissons du Maroc atlantique. Revue des travaux de l'Institut des pêches maritimes, 22, 383–493.

Gavioli, A., M. Mancini, M. Milardi, V. Aschonitis, E. Racchetti, P. Viaroli, and G. Castaldelli. (2018). Exotic species, rather than low flow, negatively affect native fish in the Oglio River, Northern Italy. River Research and Applications, DOI: 10.1002/rra.3324.

Geist J. (2005). Conservation genetics and ecology of European freshwater pearl mussels (Margaritifera margaritifera L.). (PhD thesis). Technischen Universität Münich, Germany.

Gessner, J., Williot, P., Rochard, E., Freyhof, J., & Kottelat, M. (2010). Acipenser sturio. The IUCN Red List of Threatened Species 2010: e.T230A13040963.

201 http://dx.doi.org/10.2305/IUCN.UK.2010-1.RLTS.T230A13040963.en. Downloaded on 28 June 2018.

Gess, R. W., Coates, M. I., & Rubidge, B. S. (2006). A lamprey from the Devonian period of South Africa. Nature, 446, 981–984.

Graf, D. L., & Cummings, K. S. (2016). Palaeoheterodont diversity (Mollusca: Trigonioida + Unionoida): what we know and what we wish we knew about freshwater mussel evolution. Zoological Journal of the Linnean Society, 148, 343–394.

Guillerault, N., Bouletreau, S., Iribar, A., Valentini, A., & Santoul, F. (2017). Application of DNA metabarcoding on faeces to identify European catfish Silurus glanis diet. Journal of Fish Biology, 90, 2214–2219.

Haag, W. R. (2012). North American freshwater mussels. Natural History, Ecology and Conservation. New York: Cambridge University Press.

Haas, F. (1969). Superfamily Unionacea. In Moore, R. C. (Ed), Treatise on invertebrate paleontology (Part N, Vol. 1 pp. N411–N470). Lawrence: Geological Society of America and the University of Kansas.

Hansen, M. J., Madenjian, C. P., Slade, J. W., Steeves, T. B., Almeida, P. R., & Quintella, B. R. (2016). Population ecology of the sea lamprey (Petromyzon marinus) as an invasive species in the Laurentian Great Lakes and an imperiled species in Europe. Reviews in Fish Biology and Fisheries, 26, 509–535

Huber, V., & Geist, J. (2017). Glochidial development of the freshwater swan mussel (Anodonta cygnea, Linnaeus 1758) on native and invasive fish species. Biological Conservation, 209, 230–238.

NatureServe, 2013. Petromyzon marinus. The IUCN Red List of Threatened Species 2013: e.T16781A18229984. http://dx.doi.org/10.2305/IUCN.UK.2013-1.RLTS. T16781A18229984.en. Downloaded 4 June 2018.

Jansen, W., Bauer, G., & Zahner-Meike, E. (2001). Glochidial mortality in freshwater mussels. In Bauer, G., & Wächtler, K. (Eds.), Ecological Studies: Ecology and Evolution

202 of the Freshwater Mussels Unionoida (pp. 185–211), Berlin-Heidelberg, Germany: Springer-Verlag.

Jatteau, Ph., & Fraty, R. (2012). Etude de la tolérance à l’hypoxie des juvéniles de grande alose (Alosa alosa). Bordeaux: Institut National de Recherche en Sciences et Technologies pour l’Environnement et l’Agriculture.

Karna, D. W., & Millemann, R. E. (1978). Glochidiosis of salmonid fishes III. Comparative susceptibility to natural infection with Margaritifera margaritifera (L.) (Pelecypoda: Margaritanidae) and associated histopathology. The Journal of Parasitology, 64, 528–537.

Kat, P. W. (1984). Parasitism and the Unionacea (Bivalvia). Biological Review, 59, 189– 207.

Kat, P. W., & Davis, G. M. (1984). Molecular genetics of peripheral, populations of Nova Scotian Unionidae (Mollusca:Bivalvia). Biological Journal of the Linnean Society, 22, 157–185.

Keith P., Persat H., Feunteun E., & Allardi, J. (2011). Les poissons d’eau douce de France. Paris: Museum National d’Histoire Naturelle et Biotope.

Kottelat, M., & Freyhof, J. (2007). Handbook of European freshwater fishes. Berlin: Kottelat and Freyhof.

Levine T. D., Lang, B. K., & Berg, D. J. (2012). Physiological and ecological hosts of Popenaias popeii (Bivalvia: Unionidae): laboratory studies identify more hosts than field studies. Freswater Biology, 57, 1954–1864.

Limburg, K. E., & Waldman, J. R. (2009). Dramatic declines in north Atlantic diadromous fishes. BioScience, 59, 955–965.

Limm, M. P. & Power, M. E. (2011). Effect of the western pearlshell mussel Margaritifera falcata on Pacific lamprey Lampetra tridentata and ecosystem processes. Oikos, 120, 1076–1082.

Lopes-Lima, M., Bolotov, I. N., Do, V. T., Aldridge D. C., Fonsecab, M. M., Ganh, H. M., Gofarov, A. V., Kondakov, A. V., Prié, V., Sousa, R., Varandas, S., Vikhrev, I. V.,

203 Teixeira, A., Wu, R. W., Wu, X., Zieritz, A., Froufe, E., & Bogan, A. E. (2018). Expansion and systematics redefinition of the most threatened freshwater mussel family, the Margaritiferidae. Molecular Phylogenetics and Evolution, 127, 98–118.

Lopes-Lima, M., Sousa, R., Geist, J., Aldridge, D.C., Araujo, R., Bergengren, J., … & Zogaris, S. (2017). Conservation status of freshwater mussels in Europe: State of the art and future challenges. Biological Reviews, 92, 572–607.

López, M. A. & Altaba, C. R. (2005). Fish host determination for Margaritifera auricularia (Bivalvia: Unionoida): Results and implications. Bolletino Malacologico, 41, 89–98.

López, M. A., Altaba, C.R., Rouault, T. & Gisbert, E. (2007). The European sturgeon Acipenser sturio is a suitable host for the glochidia of the freshwater pearl mussel Margaritifera auricularia. Journal of Molluscan Studies, 73, 207–209.

López, M. A., Gázquez, N., Olmo-Vidal, J. M., Aprahamian, M. W., & Gisbert, E. (2007). The presence of anadromous twaite shad (Alosa fallax) in the Ebro River (western Mediterranean, Spain): An indicator of the population’s recovery? Journal of Applied Ichthyology, 23, 163–166.

Mateus, C. S., Rodríguez-Muñoz, R., Quintella, B. R., Alves, M. J., & Almeida, P. R. (2012). Lampreys of the Iberian Peninsula: Distribution, population status and conservation. Endangered Species Research, 16, 183–198.

Mennesson-Boisneau, C., Aprahamian, M. W., Sabatie, M. R., & Cassou-Leins, J. J. (2000). Remontée migratoire des adultes. In Biologie, écologie et exploitation des aloses (Alosa sp.) de l'Atlantique Est et de la Méditerranée occidentale (pp. 55-73). Paris : INRA, CEMAGREF.

Modesto, V., Ilarri, M., Souza, A.T., Lopes-Lima, M., Douda, K., Clavero, M., & Sousa, R. (2018). Fish and mussels: Importance of fish for freshwater mussel conservation. Fish and Fisheries, 19, 244–259.

Moser M. L., Butzerin, J. M., & Dey, D. B. (2007). Capture and collection of lampreys: The state of the science. Reviews in Fish Biology and Fisheries, 17, 45–56.

204 Nakamura, K., Cucala, L., Mestre, A., Mesquita-Joanes, F., Elbaile, E., Salinas, C., & Muñoz-Yanguas, M. A. (2018). Modelling growth in the critically endangered freshwater mussel Margaritifera auricularia (Spengler, 1793) in the Ebro basin. Hydrobiologia, 810: 375–391.

Nienhuis, J. A. J. H., (2003). The rediscovery of Spengler’s freshwater pearlmussel Pseudunio auricularius (Spengler, 1793) (Bivalvia, Unionoidea, Margaritiferidae) in two river systems in France, with an analysis of some factors causing its decline. Basteria, 67, 67–86.

Nordlie, F. G., & Mirandi, A. (1996). Salinity relationships in a freshwater population of eastern mosquitofish. Journal of Fish Biology, 49, 1226–1232.

Österling, M. E., Greenberg, L.A., & Arvidsson, B. A. (2008). Relationship of biotic and abiotic factors to recruitment patterns in Margaritifera margaritifera. Biological Conservation, 141, 1365–1370.

Plaut, I. (1998). Comparison of salinity tolerance and osmoregulation in two closely related species of blennies from different habitats. Fish Physiology and Biochemistry, 19, 181–188.

Portafaix, P., Senecal, A. & Baisez, A. (2015). Evaluation de la phase juvénile de la lamproie marine (Petromyzon marinus) sur le bassin de la Vienne. Rapport ammocètes bassin de la Vienne. Poitiers: Association LOGRAMI.

Prié, V., (2010). Margaritifera auricularia. The IUCN Red List of Threatened Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN.UK.2010- 4.RLTS.T12798A3381899.en

Prié, V., Philippe, L., & Cochet, G. 2011. Plan National d’Actions en faveur de la Grande Mulette Margaritifera auricularia – 2012-2017. Paris, Ministère de l’Ecologie, du Développement Durable, des Transports et du Logement.

Prié, V., Soler, J., Araujo, R., Cucherat, X., Philippe, L., Patry, N., Adam, B., Legrand, N., Jugé, P., Richard, N., & Wantzen, K. (2018). Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe. Hydrobiologia, 810, 157–175.

205 Roe, K. J., Simons, A. M., & Hartfield, P. (1997). Identification of a fish host of the inflated heelsplitter Potamilus inflatus (Bivalvia: Unionidae) with a description of its glochidium. American Midland Naturalist, 138, 48–54.

Salonen, J.K., Marjomäki, T.J., & Taskinen, J., (2016). An alien fish threatens an endangered parasitic bivalve: the relationship between brook trout (Salvelinus fontinalis) and freshwater pearl mussel (Margaritifera margaritifera) in northern Europe. Aquatic Conservation: Marine and Freshwater Ecosystems, 26, 1130–1144.

Schlumberger, O., Sagliocco, M., & Proteau, J. P. (2001). Biogéographie du silure glane (Silurus glanis): Causes hydrographiques, climatiques et anthropiques. Bulletin Français de la Peche et de la Pisciculture, 357, 533–547

Simberloff, D., & Von Holle, B. (1999). Positive interactions of nonindigenous species: Invasional meltdown? Biological Invasions, 1, 21–32.

Smith, D. G. (1985). Recent range expansion of the freshwater mussel Anodonta implicata and its relationship to clupeid fish restoration in the Connecticut river system. Freshwater Invertebrate Biology, 4, 105–108.

Soler, J., Boisneau, C., Wantzen, K. M., & Araujo, R. (2018a). Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies. doi:10.1093/mollus/eyy038

Soler, J., Wantzen, K. M., Jugé, P., & Araujo, R. (2018b). Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84, 182–189.

Strayer, D. L. (2008). Freshwater mussel ecology. A Multifactor Approach to Distribution and Abundance. Berkeley, University of California Press.

Strayer, D. L., Downing, J. A., Haag, W. R., King, T. L., Layzer, J. B., Newton, T. J. & Nichols, S. J. (2004). Changing perspectives on pearly mussels, North America’s most imperiled animals. BioScience, 54, 429–439.

206 Syväranta, J., Cucherousset, J., Kopp, D., Martino, A., Céréghino, R. & Santoul, F. (2009). Contribution of anadromous fish to the diet of European catfish in a large river system. Naturwissenschaften, 96, 631–635.

Taeubert, J. E., Gum, B., & Geist, J., (2012). Host-specificity of the endangered thick- shelled river mussel (Unio crassus, Philipsson 1788) and implications for conservation. Aquatic Conservation: Marine and Freshwater Ecosystems, 22, 36–46.

Taverny, C., & Elie, P. (2010). Les lamproies en Europe de l’Ouest. Écophases, especes et habitats. Paris, Edition Quae.

Taverny, C., Urdaci, M., Elie, A.-M., Beaulaton, L., Ortusi, I., Daverat, F., & Elie, P. (2005). Biologie, écologie et pêche des lamproies migratrices (Agnathes amphihalins). Rapport no. 99. Bordeaux: Centre d’Etude du Mécanisme Agricole du Génie Rural des Eaux et Forêts.

Teixeira, A., Dias, A. R., Miranda, F., Lopes-Lima, M., Varandas, S., Froufe, E., Filipe, A. F., Beja, P., & Sousa, R. (2018). Can invasive alien fish species act as effective hosts of native freshwater mussels (Unionidae) in Iberia? 2018 Iberian Congress of Ichtiology, Faro, Portugal

Triantafyllidis, A., Krieg, F., Cottin, C., Abatzopoulos, T. J., Triantaphyllidis, C., & Guyomard, R. (2002). Genetic structure and phylogeography of European catfish (Silurus glanis) populations. Molecular Ecology, 11, 1039–1055.

Valadou, B. (2007). Le silure glane (Silurus glanis, L.) en France. Evolution de son aire de répartition et prédiction de son extension. Paris: Institut de recherche et de Développement, Conseil Supérieur de la Pêche Protection des Milieux.

Vaughn, C. C., & Taylor, C. M. (1999). Impoundments and the decline of freshwater mussels: a case study of an extinction gradient. Conservation Biology, 13, 912–920.

Watters, G. T. (1992). Unionids, fishes, and the species–area curve. Journal of Biogeography, 19, 481–490.

Watters, G. T. (2001). The evolution of the Unionacea in North America, and its implications for the worldwide fauna. In Bauer, G., & Wächtler, K. (Eds.), Ecological

207 Studies: Ecology and Evolution of the Freshwater Mussels Unionoida (pp. 281–307), Berlin-Heidelberg, Germany: Springer-Verlag.

Watters, G. T. (1995). Small dams as barriers to freshwater mussels (Bivalvia, Unionoida) and their hosts. Biological Conservation, 75, 79–85.

Young, M.R. & Williams, J.C. (1984). The reproductive biology of the freshwater pearl mussel Margaritifera margaritifera (Linn.) in Scotland I. Field studies. Archiv für Hydrobiologie, 99, 405–422.

Youson, J. H. (1980). Morphology and physiology of lamprey metamorphosis. Canadian Journal of Fisheries and Aquatic Sciences, 37, 1687−1710.

208 3.3. Research on the early stages of the European margaritiferids

3.3.1. ARTICLE 6: Who wins in the “weaning” process? Juvenile feeding morphology of two freshwater mussel species.

Résumé

Le déclin global des moules d'eau douce peut être partiellement attribué à leur cycle de vie complexe. Leur survie du glochide à l’âge adulte est comme une longue course d’obstacles, la mortalité juvénile étant un point critique. Des mortalités massives peu après l’entrée dans un état juvénile ont été rapportés chez des populations sauvages et captives, affaiblissant ainsi la population efficace de bivalves. Un phénomène similaire se produit lors de la métamorphose dans les populations naturelles et en écloserie de bivalves marins juvéniles. Sur la base d’une analyse morphologique utilisant la microscopie électronique à balayage de juvéniles nouvellement formés des espèces d’eau douce Margaritifera margaritifera (L.) (Margaritiféridae) et Unio mancus Lamarck (Unionidae), nous montrons qu’une deuxième métamorphose, consistant en des changements morphologiques drastiques, se produit et conduit à une alimentation suspensivore à la place de l'alimentation par le pied cilié. Nous émettons l'hypothèse que l'alimentation en suspension chez ces deux espèces s'améliore grâce au développement progressif de plusieurs caractéristiques morphologiques, y compris le contact entre les cils des filaments postérieurs des branchies internes, la réflexion des branchies internes, l'apparition de la rainure ventrale cténidienne et la formation des palpes labiaux. Indépendamment de la présence de nourriture disponible, un mode d’alimentation suspensivore remplace l’alimentation dépositivores, et les juvéniles incapables de réussir leur transition morphologique ou de s’adapter aux changements alimentaires périssent probablement.

209 210 Received: 17 May 2017 | Revised: 25 July 2017 | Accepted: 9 August 2017 DOI: 10.1002/jmor.20748

RESEARCH ARTICLE

Who wins in the weaning process? Juvenile feeding morphology of two freshwater mussel species

Rafael Araujo1 | Miquel Campos2 | Carles Feo2 | Catuxa Varela3 | Joaquín Soler1 | Paz Ondina3

1Department of Biodiversity and Evolutive Biology, Museo Nacional de Ciencias Abstract Naturales-CSIC, Jose Gutierrez Abascal 2, The global decline of freshwater mussels can be partially attributed to their complex life cycle. Madrid 28006, Spain Their survival from glochidium to adulthood is like a long obstacle race, with juvenile mortality as a 2 Consorci de l’Estany, Plaça dels Estudis 2, key critical point. Mass mortality shortly after entering into a juvenile state has been reported in Banyoles, Girona 17820, Spain both wild and captive populations, thus weakening the effective bivalve population. A similar 3Department of Zoology, Faculty of phenomenon occurs during metamorphosis in natural and hatchery populations of juvenile marine Veterinary Science, University of Santiago de Compostela, Lugo 27002, Spain bivalves. Based on a morphological analysis using scanning electron microscopy of newly formed juveniles of the freshwater species Margaritifera margaritifera (L.) (Margaritiferidae) and Unio Correspondence mancus Lamarck (Unionidae), we show that a second metamorphosis, consisting of drastic morpho- Rafael Araujo, Museo Nacional de Ciencias logical changes, occurs that leads to suspension feeding in place of deposit feeding by the ciliated Naturales-CSIC, Jose Gutierrez Abascal 2, 28006 Madrid, Spain. foot. We hypothesize that suspension feeding in these two species improves due to a gradual Email: [email protected] development of several morphological features including the contact between cilia of the inner gill posterior filaments, the inner gill reflection, the appearance of the ctenidial ventral groove and the Funding information European Commission formation of the pedal palps. Regardless of the presence of available food, a suspension feeding mode replaces deposit feeding, and juveniles unable to successfully transition morphologically or adapt to the feeding changes likely perish.

KEYWORDS biodiversity conservation, development, juvenile mortality, metamorphosis

1 | INTRODUCTION unionids feed directly on the living microbial and algal components of fine organic material on the sediment surface (Raikow & Hamilton, The majority of bivalves are rhythmic filter and/or deposit feeders 2001). Other authors have hypothesized that the food required by two (Morton, 1973). Adult freshwater bivalves of the Unionoida are gener- species of Margaritifera comes from a healthy rhizosphere, whereas ally considered suspension feeders (phytoplankton, zooplankton and eutrophication of the immediate environment was responsible for the particulate detritus) or deposit feeders (particulate detritus in sedi- absence of food (Howard, Cuffey, & Solomon, 2005; Hruska, 1999; ment); however, some authors have stated that deposit-feeding Negishi, Katsuki, Kume, Nagayama, & Kayaba, 2014). Indeed, seasonal bivalves are never found in fresh water (Nicol, 1984). Various authors differences in feeding and metabolism have also been reported for dif- have reported that the diet of some freshwater mussels consisted of ferent mussel species (Baker & Hornbach, 2001; Strayer, 2008; suspended bacteria, fungal spores, dissolved or sediment organic Vaughn, Nichols, & Spooner, 2008; Vaughn, Spooner, & Galbraith, matter (Vaughn & Hakenkamp, 2001), or clearing coliform bacteria 2007; Yasuno et al., 2014). Despite having much information, a good (Silverman et al., 1997). River and lake unionids always ingest bacterial understanding of the food sources and feeding mechanisms of fresh- carbon instead of algal carbon, and are not always primary consumers water mussels (Nichols & Garling, 2002; Raikow & Hamilton, 2001; or omnivores (Nichols & Garling, 2000). Nevertheless, the importance Strayer, 2008; Vaughn & Hakenkamp, 2001; Vaughn et al., 2008; of algae in the diet is derived from the phytosterol, and cholesterol and Welker & Walz, 1998) is still lacking, hindered by the abstract concept vitamin B12 are also key dietary items (Nichols & Garling, 2000). Some of detritus (Coker, Shira, Clark, & Howard, 1921; Howard, 1922;

Journal of Morphology. 2017;1–13. wileyonlinelibrary.com/journal/jmor VC 2017 Wiley Periodicals, Inc. | 1 211 2 | ARAUJO ET AL.

Lefevre & Curtis, 1912), a seemingly ‘catch all’ term for less clearly 2007; Lasee, 1991; Lima, Kovitvadhi, Kovitvadhi, & Machado, 2006; defined food sources. Schartum, Mortensen, Pittman, & Jakobsen, 2016; Trump, 2010). This Little is generally known about adult feeding mode of the Sphaerii- process would be similar to the metamorphosis observed during settle- dae, which may share common size and feeding mechanisms with juve- ment in marine juvenile bivalves (Cannuel, Beninger, McCombie, & nile freshwater mussels of other families (Yeager, Cherry, & Neves, Boudry, 2009; Helm, Bourne, & Lovatelli, 2004; O’Foighil, Kingzett, 1994). Several adult sphaeriid species have been maintained with bac- O’foighil, & Bourne, 1990; Reid, McMahon, O’foighil, & Finnigan, terial suspensions. However, whether they are deposit feeders, using 1992). the foot for particle collection, or suspension feeders, filtering the over- The little information we know about feeding modes in juvenile lying water, is unclear. Sphaeriids are highly mobile, and adopt an infau- freshwater mussels of the families Unionidae and Margaritiferidae nal position with the dorsal surface facing downwards, from just below stems from rearing captive cultures. Recently emerged juveniles appear the sediment surface to a depth of several centimetres (Holopainen, to first pedal feed, then switch to deposit and suspension feeding (e.g., 1985; Meier-Brook, 1969), indicating that the overlying water drawn bacteria, bacteria-sized particles, and algae in interstitial water) (Yeager into the mantle cavity by the mantle aperture is not inhaled (Lopez & et al., 1994) or pedal deposit feeding (e.g., algae and detritus but not Holopainen, 1987). Indeed, in these species, the short siphon does not bacteria) before finally becoming algae filter feeders (Beck & Neves, reach the sediment surface and pumping of the overlying water is low 2003; Gatenby, Neves, & Parker, 1996; Gatenby, Parker, & Neves (Efford & Tsumura, 1973), indicating they are infaunal suspension 1997). However, captive breeding of freshwater mussels for conserva- feeders siphoning interstitial water and/or filtering microbes from the tion are still hindered by the lack of detailed knowledge of the diet and sediment (Holopainen & Hanski, 1979). Glucose uptake only accounts feeding mode of juveniles, with few studies available describing for a small proportion of the metabolic demands of Pisidium casertanum changes in their feeding morphology during breeding (Kovitvadhi et al., (Poli); therefore, utilization of dissolved organic matter from interstitial 2007; Lasee, 1991; Lima et al., 2006; Schartum et al., 2016; Trump, water is likely a less important feeding mode in sphaeriids (Efford & 2010). Tsumura, 1973). Nevertheless, some authors have claimed sphaeriids Based on the knowledge gained to date and information from deposit feed, using the ciliated foot to create inhalant currents that col- marine bivalve aquaculture, some authors have developed controlled lect food, including microbes (Burky, 1983; Hornbach, Way, Wissing, & rearing systems, providing extra food in the form of algae (Araujo, Burky, 1984; Mackie & Qadri, 1978; Mitropolskii, 1966). In contrast, Quiros, & Ramos, 2003; Barnhart, 2006; Beck & Neves, 2003; other authors suggest this mechanism is not an important feeding Eversole, Stuart, & Brume, 2008; Eybe, Thielen, Bohn, & Sures, 2013; mode for sphaeriids (Lopez & Holopainen, 1987), concluding that inter- Gatenby et al., 1996, 1997; Gatenby et al., 2003; Gum, Lange, & Geist, stitial suspension feeding on bacteria is the primary mode and that the 2011; Guyot, 2005; Henley, Zimmerman, Neves, & Kidd, 2001; Hudson small size of these bivalves is likely the result of selection on this feed- & Isom, 1984; Jones, Mair, & Neves, 2005; Kovitvadhi et al., 2006, ing mode. Finally, food choice may also differ between genera: plant 2007; Kovitvadhi, Kovitvadhi, Sawangwong, & Machado 2008; Liberty, detritus or herbivory appears to be more important for Sphaerium,and 2004; Liberty, Ostby, & Neves, 2007; Malo, 2012; OBeirn, Neves, & microbes for Pisidium (Holopainen & Lopez, 1989). Steg, 1998; Schmidt & Vandre, 2010; Thielen, 2011). Freshwater mussels (Unionoida) have a complex life cycle which In this study, using information from the literature and our personal involves a temporary but obligatory parasitic stage in which the larvae experience in freshwater mussel captive care, the anatomical feeding (glochidia) attach to the external surface or gill filaments of their verte- organs of M. margaritifera and U. mancus juveniles were studied. We brate hosts (mainly fish). To ultimately reproduce, they have to survive investigated the high level of early juvenile mortality observed in natural a long obstacle race in which more than the 99% of offspring die (and captive) habitats, in which both pedal and filter feeding occur (Bauer, 1991). These organisms must first come into contact with their (Araujo, Feo, Pou, & Campos, 2015). Using juveniles of two species of host fish during the glochidium stage, survive metamorphosis in the endangered freshwater mussels belonging to two different families, fish’s gills until juvenile stages and then detach in a suitable settlement Margaritiferidae and Unionidae, we aimed to test two hypotheses. Our habitat for further maturation. Despite reaching this stage, juveniles first hypothesis is to check if young juveniles experience another meta- are still fragile and experience high mortality, although the reasons morphosis that causes a change from deposit feeding by the ciliated (winter survival, ecological problems, absence of food) are unclear foot to suspension feeding. Our second hypothesis is that, regardless (Archambault, Cope, & Kwak, 2014 ASTM, 2013; Augspurger, Dwyer, of the availability of food, many juveniles at this stage are unable to Ingersoll, & Kane, 2007; Strayer & Malcom, 2012). We hypothesize successfully transition feeding modes, thus resulting in high mortality. that this critical juvenile stage can be related to a second metamorpho- sis event that occurs just after emergence as juveniles. Due that 2 | MATERIAL AND METHODS deposit feeders use the foot for particle collection, but suspension feeders filter the overlying water, this second metamorphosis results in Juveniles of Margaritifera margaritifera (Linnaeus, 1758) were collected drastic morphological changes that, regardless of the presence of from infections of 01 Salmo salar L., in a mussel rearing facility in available food, substitutes deposit feeding for suspension feeding Galicia (NW Spain) with glochidia specimens from the Arnego River (Kovitvadhi, Kovitvadhi, Sawangwong, Thongpan, & Machado, 2006, (Ulla basin). The metamorphosis process in M. margaritifera lasts

212 ARAUJO ET AL. | 3 between six and nine months. After six months, the water temperature 3.1 | Margaritifera margaritifera of the infected fish was increased incrementally from 10 to 18 8C The mean length of newly emerged juveniles was 350 mm(Figure1aand (Eybe, Thielen, Bohn, & Sures, 2015), and juveniles emerged between Tables 1 and 2). The most conspicuous characters were the ciliated foot, 29th March and 8th April 2015. Four hundred juveniles were main- with a marked ventral groove, the ciliated border of the mantle (Figure tained in a box filled with 475 ml of river water and 25 ml of detritus 1b–d) and the presence of three filaments and one posterior developing without substrate. Juveniles were fed algae once weekly during a water gill bud (Figure 1a). The ventral pedal groove is the byssus groove with exchange. The algae consisted of 120 ml of Shellfish Diet 1800 the byssus pit (the posterior) and the byssus disc pit (the anterior) at either (Isochrysis sp., Pavlova sp., Thalossiosira weissflogii (Grunow) and end (Figure 1b). Frontal, lateral and latero-frontal cirri were observed in Tetraselmis sp., with a diameter of 4–20 mm) and 200 mlofNannochlorop- the gill filaments, which connected adjacent and opposing filaments. The sis sp. (1.5–2 mm)suspendedin10Lofriverwater(Eybeetal.,2013; outer demibranchs of the internal gill were clearly developing (Figure 1e); Scheder, Lerchegger, Jung, Csar, & Gumpinger, 2014). The boxes were however, they did not bend before day 130, in specimens greater than kept in a conditioning cabinet at a constant temperature of 17 8C. The 1 mm. There were also two diagonal rows of long cilia on the inner side algal diet was doubled after the first month and tripled after six months. of the posterior mantle wall (Figure 1f). Juveniles of Unio mancus (Lamarck, 1819) were collected from At 10 days (430 mm) post-emergence, these features are main- infections of Barbus meridionalis Risso, with glochidia specimens from tained, and a shell ring had grown around the original shell. At this time Banyoles Lake (Spain); they emerged on 12 June 2015. The metamor- point, there were many more gill filaments, with greater connections phosis process in U. mancus lasts between 7 and 28 days (Araujo et al., between them, resulting in a basket-like structure having an oval- 2015). Juveniles were then maintained in a PVC cylinder measuring shaped orifice posteriorly surrounded with long cilia (Figure 1g). Diago- 16 cm in diameter and 17 cm in height (total volume capacity of 2.2 L) nal rows of long cilia on the inner mantle wall were present. m without sediment but with a 200 m mesh at the bottom. This cylinder By 30 days (580 mm), there were five gill filaments covered with cilia 3 3 was suspended within another receptacle (33 46 44 cm) filled and cirri (Figure 1h) and one developing posterior gill bud. We observed with 70 L of water from Banyoles Lake (Spain). This culture was fed the anterior adductor muscle and a ciliated mouth (Figure 2a), but the three times weekly with 2 mg/L of frozen Easy Reefs (1/3 Nannochlor- labial palps were not yet present. The foot and the mantle border were – m 1 – opsis gaditana Lubian 2 4 m 1/3 Tetraselmis chuii Butcher 12 14 heavily ciliated. m 1 – m m 1/3 Phaedactylum tricornutum Bohlin 2 6 m). The outer recep- The cilia around the mantle margin began to diminish between 40 tacle was artificially aerated, and daily, the inner cylinder was moved and 80 days (700–800 mm), although there were ciliary tufts on the up and down five times first to empty the water through the mesh and margin (Figure 2b). The cilia on the foot and on the inner mantle also then to homogenize the inner cylinder water with the surrounding began to disappear. The labial palps and the ventral groove of the gills water. Water exchanges of 100% were performed every two days. were still absent. Between 5 and 10 juveniles of each species were sacrificed every The labial palps were present at 150 days (1.4 mm; Figure 2c). At 10 days until day 90 and then once a month until day 360. Prior to 180 days, 13 filaments were present on each side of the gill, but the fixation in 2.5% glutaraldehyde (2–24 hr), specimens were first relaxed ventral gill groove was still not present (Figure 2d). Ciliary tufts were by slowly adding MgCl2 until the valves opened. If the valves did not present on the posterior inner mantle roof above the ulterior exhalant open, one valve was broken before dehydration. Specimens were then aperture. cleaned with PBS buffer and dehydrated through a graded ethanol At day 210 (1.8 mm), the gill had 21 filaments, the ventral groove series (30–60 min each in 30%, 50%, 70%, 90%, 96%, 100%, and between the inner and outer lamellae was already marked (Figure 2e,f) 100% ethanol). Specimens were stored in 100% ethanol until graded and the labial palps were highly discernible (Figure 2f,g). The palps into solutions of 100% ethanol and hexamethyldisilazane (HMDS) were folded, and the inner sides ciliated (Figure 2g). The cilia and the (15–30 min each in 2:1; 1:1; 1:2 ethanol:HMDS). ventral groove of the foot were shorter than at 100 days (Figure 2e). Scanning electron microscopy analyses focused on the following Two pseudodiaphragms were well-developed at the posterior mantle features: (a) Gill buds and inner gill growth. (b) Inner gill reflection and margin, increasing the separation between the gill chambers (Figure ventral groove. (c) Mouth. (d) Labial palps. (e) Mantle cilia (border and 2h). The area around the inhalant aperture was darkly pigmented in live inner side). (f) Cilia and ventral groove on the foot. (g) Gill buds and the animals at 200 days. More cilia were present on the mantle margin appearance of the external gill. near the pseudodiaphragm that in other places of the mantle border. Additional observations were made in living specimens using a The first gill bud of the external gill was present at 210 days. dissecting scope. At day 240 (2 mm), the cilia of the mouth were shorter than at younger stages. Cilia were present on the pseudodiaphragm (Figure 3a) as were external gill filaments (Figure 3b). In adults, these filaments, 3 | RESULTS along with the pseudodiaphragm, will form the gill chambers. At 300 days (2.1 mm), the developing external gill was apparent and possessed After one year of culture, the juvenile mortality of Margaritifera margaritifera bent filaments (Figure 3c). The byssal pit on the ventral foot and ciliary was about 90%, while that of Unio mancus was about 40%. tufts on the posterior roof of the mantle were both present (Figure 3d),

213 4 | ARAUJO ET AL.

FIGURE 1 Margaritifera margaritifera, scanning electron micrographs images of different ontogenetical stages. (a) Ventral view. Anterior part is down (0 days). (b) Ciliated foot with the byssal ventral groove and the two byssus pits (arrows). Anterior part is down (0 days). (c) Cilia at the mantle border (0 days). (d) Detail of the cilia at the mantle border (0 days). (e) Gill filaments (0 days). (f) Diagonal row of cilia at the inner mantle wall (0 days). (g) Inner gill filaments, gill buds (arrow) and long cilia at the posterior mantle margins. Anterior part is to the left (10 days). (h) Cilia and cirri in the inner gill filaments (30 days)

and the papillae of the inhalant aperture were discernible. The pseudo- were covered with cilia (Figure 3f), the external gill was better diaphragm and the borders of the inhalant and exhalant apertures were developed and the papillae on the inhalant aperture were more visible. both ciliated. Ciliary tufts on the mantle border were evident, leaving At no time point were muscular connections between gill filaments circular marks (Figure 3e). At day 360 (2.2 mm), the folded labial palps observed.

214 ARAUJO ET AL. | 5

TABLE 1 Mean length (mm and mm) of juvenile freshwater mussel species at specified days as reported in this study and by other authors

1 day 10 days 30 days 60 days 80–100 days 150 days 180 days 210 days 300 days 360 days mm mm mm mm mm mm mm mm mm mm

M. margaritiferaa 350 430 580 780 0.8–1 1.4 1.6 1.8 2.1 2.2

M. margaritiferab – 460 ––– –––––

M. margaritiferac – ––– – 0.4–0.9 ––––

M. margaritiferad –––600–700 0.7–0.9 0.7–0.9 ––0.7–0.9 1.1–1.2

M. margaritiferae 300–480 – –– –––––

U. mancusa 260 450 750 1100 1.2–1.4 1.7 1.9 2 2.2 3

V. irisf 300–400 ––– 0.8–1 1.7–2.3 ––––

L. ventricosag 227 350 530 889 – –––––

U. imbecillish 320 ––800 – 2.1 ––––

H. myersianai 150 200 – 1800 2.5 24 –––85 aThis article. bMalo (2012). cSchmidt and Vandre (2010). dLavictoire et al. (2016). eSchartum et al. (2016). fGatenby et al. (1997). gLasee (1991). hTrump (2010). iKovitvadhi et al. (2007, 2008).

3.2 | Unio mancus mantle (Figure 4e), rather than a pseudodiaphragm as in margaritiferids. The developing external gill also already had bent filaments. The mean length of newly emerged juveniles was 260 mm(Tables1 The papillae of the inhalant siphon were clearly recognizable at and 2). The ciliated mouth of U. mancus was much larger than that of 345 days (2.8–3 mm; Figure 4f), resembling spheres on the inner man- M. margaritifera; however, the foot, mantle margin (Figure 4a–c), and tle border (Figure 4g). By this time, the overall appearance of the juve- gills were very similar in the two species at this initial time point. nile was similar to that of the adult, although with unfused mantle At 60 days (1.1 mm), the outer demibranch was reflected, and the borders (Figure 4h). The inner sides of the labial palps had ciliary folds ventral groove of the inner gill had formed, and at 70 days, primordial (Figure 5a) and a flat external surface with a ciliary flange (Figure 5b). labial palps with cilia were present (Figure 4d). The mantle border was also covered with cilia, although these were shorter than those Also at 345 days, 45 inner gill filaments with complete cilia and cirri observed at younger stages. were present and connected by horizontal muscular bridges (Figure 5c, At 120 days (1.6 mm), the inner gill had 18 filaments. The external d). The foot was covered with short cilia, and at 395 days (3.8 mm), the gill became evident at 180 days (1.9 mm) and was associated with the ventral groove only remained in the posterior part (Figure 5e). The posterior part of the inner gill, where a separation between the upper exhalant siphon was completely developed (Figure 5f), but the ventral and lower gill chambers will eventually form on the inner wall of the borders of the inhalant siphon were not yet connected.

TABLE 2 Days of appearance of feeding structures in juvenile freshwater mussel species reported in this study and by other authors

Inner gill cirri Basket Inner gill bent Ventral groove Labial palps Outer gill Inhalant papillae

M. margaritiferaa 0 10 130 180 180 240 300

M. margaritiferae ––1.1–2.2 mm* 2.2–4.5 mm* – 4.5 mm* –

U. mancusa 01040–60 60 70 180 300

V. irisf – 14 –– ––After 272

L. ventricosag 0 – After 56 – 2 ––

U. imbecillish 07–28 113 – 3 130 175

H. myersianai 0 10 30 After 50 – 90 50

For references, see in Table 1. *No data about age.

215 6 | ARAUJO ET AL.

FIGURE 2 Margaritifera margaritifera, scanning electron micrographs images of different ontogenetical stages. (a) Detail of the ciliated mouth (arrow) (30 days). (b) Cilia at the mantle margin (80 days). (c) The foot and the labial palps (arrows). Anterior part is to the right (150 days). (d) Connection between the left and right inner gill filaments. Anterior part is to the right (180 days). (e) Ventral view. See the labial palps. Anterior part is up (210 days). (f) Ventral groove at the inner gill (210 days). (g) Outer surface of the right labial palps. (h) Ventral view of the posterior part. Pseudodiaphragm (arrow) and posterior filaments of the inner gill. (210 days)

4 | DISCUSSION character (Reid et al., 1992; Stasek, 1963). This feeding method has been observed in some juvenile and small adult marine and freshwater Pedal feeding is likely the first method used in young juvenile-staged bivalves [e.g., Macoma balthica (L.), Corbicula fluminea (Muller) and Kur- benthic bivalves as an evolutionary constraint of a plesiomorphic tiella bidentata (Montagu)] (Caddy, 1969; Reid et al., 1992; Schartum

216 ARAUJO ET AL. | 7

FIGURE 3 Margaritifera margaritifera, scanning electron micrographs images of different ontogenetical stages. (a) Ciliated pseudodiaphragm (240 days). (b) Inner right gill and the beginning of the outer gill (arrow). Anterior part is to the right (240 days). (c) Lateral view. See the labial palps (arrow) and the mouth everted. Anterior part is to the left (300 days). (d) Ciliary tufts at the posterior mantle roof (300 days). (e) Cilia and ciliary marks at the mantle margin (330 days). (f) Ciliary folds at the inner surface of the labial palps (360 days) et al., 2016). The transition to suspension feeding, at least in phyto- Based on the observation of the living specimens, newly emerged plankton fed juvenile freshwater mussels or in the wild, seems to be a juveniles of both M. margaritifera and U. mancus first feed by pedal gradual process (Beck & Neves, 2003; Lasee, 1991; Trump, 2010; feeding using the cilia of the foot and of the mantle margin, similar to Yeager et al., 1994) related to the development of new organs, particu- what has been observed in other post-metamorphosed juveniles larly of two basic bivalve feeding organs, the gills and the labial palps (Kovitvadhi et al., 2007; Lasee, 1991; Trump, 2010; Uthaiwan, Chatch- (Galbraith, Frazier, Allison, & Vaughn, 2009; Kellogg, 1915; Schartum avalvanich, Noparatnaraporn, & Machado, 2001). These observations, et al., 2016; Stasek, 1963). Once these two structures are well- using dissection scopes and analyzing scanning electron micrographs, developed, definitive filter feeding is initiated. Our results are in agree- suggest that the initiation of suspension feeding begins when the cilia ment with other authors (Schartum et al., 2016; Trump, 2010): the ana- of the posterior parts of the inner gill filaments join to form a virtual tomical development of feeding structures is a result of an overall hole for water current. In effect, the gill filaments build a basket with a increase in an animal’s size rather than strictly as a function of age dorsal and a ventral chamber with a ciliated posterior opening. This (Tables 1 and 2). opening allows water to enter, facilitated by the mantle and gill cilia,

217 8 | ARAUJO ET AL.

FIGURE 4 Unio mancus, scanning electron micrographs images of different ontogenetical stages. (a) Ventral view. See the big mouth (arrow) and the gill filaments and buds. Anterior part is up (0 days). (b) Cilia at the inner gill filaments. (c) Cilia at the mantle margin. (d) Labial palps in formation (arrows) (70 days). (e) Formation of the outer gill (arrow). Anterior part is to the right (180 days). (f) Exhalant siphon papillae and posterior inner and outer gill filaments (345 days). (g) Birth of the siphon papillae (345 days). (h) Ventral view. Anterior part is to the right (395 days) similar to the ulterior inhalant aperture (or siphon) in the adult. In this two-inner gill demibranchs, which direct food into the mouth, and as new mode of suspension feeding, a posterior-anterior current, opposite suspension feeding improves, the cilia on the foot become shorter and to the one used for pedal feeding (i.e., anterior-posterior), is generated. less numerous. Furthermore, the function of this basket improves after the formation In M. margaritifera and U. mancus,asinLampsilis ventricosa (Barnes) of the labial palps and the corresponding ventral groove between the (Lasee, 1991), we did not find evidence of an elaboration of the byssal

218 ARAUJO ET AL. | 9

FIGURE 5 Unio mancus, scanning electron micrographs images of different ontogenetical stages. (a) Ciliated folds in the inner side of the labial palps (345 days). (b) Outer side of a labial palp. See the cilia at the margin (arrow). (c) Filaments in the inner gill. See the connections between filaments (395 days). (d) Filaments in the inner gill. See the connections between filaments (395 days). (e) Foot, inner gills and labial palps. Anterior part is to the right (395 days). (f) Exhalant (above) and inhalant siphons (395 days) thread; nevertheless, in both species, the two byssus pits on the ventral 2010). Rejected food likely reaches the outside current by the cilia pedal groove are present in newly metamorphosed juveniles, whereas forming diagonal rows on the inner mantle wall and those on the pos- in L. ventricosa, the byssal complex forms between 4 and 8 weeks post- terior mantle roof and anterior mantle margin (Beninger, Veniot, & metamorphosis (Lasee, 1991). Although pedal feeding is facilitated by Poussart, 1999). In Villosa iris (Lea) juveniles 14 days post- the ciliated foot and mouth, initial suspension feeding is initiated by metamorphosis, pedal and filtering feeding are both utilized but always currents made by the lateral filaments on the posterior mantle margin using the pedal cilia and not the posterior apertures (Yeager et al., and by the cilia from the left and right gills forming the gill basket 1994). These authors concluded that the cilia of the foot create the fil- (Gatenby et al., 1996; Lasee, 1991; Schartum et al., 2016; Trump, tering current, although in this case, the foot does not extend as it

219 10 | ARAUJO ET AL. does during pedal (deposit) feeding (Yeager et al., 1994). However, as days and 5–10 pairs at 28 days (Lasee, 1991). For U. imbecillis, the data observed in 500–800 mm larvae of the marine bivalve Pecten maximus are unclear: one study reported 23 pairs of filaments in a 74-day old, (L.), particle capture is likely inefficient (Beninger, Dwiono, & Le Pen- 5.1 mm juvenile (Hudson & Isom, 1984), while another reported 23 fila- nec, 1994). In 60-day-old Hyriopsis myersiana (Lea) juveniles, both the ments in an 158-day old, 2.1 mm juvenile (Trump, 2010). The outer ciliated foot and the cilia on the mantle and gills are used to transport demibranch, which appears to be the last structure to form prior to phytoplankton (Kovitvadhi et al., 2006, 2008). The development of definitive filter feeding, forms when species reach approximately 2 mm inner gill frontal cilia and laterofrontal cirri is important to form the gill in shell length. Different species reach this length at different times: 90 basket for initial filter feeding (Schartum et al., 2016). In L. ventricosa, days in H. myersiana (Kovitvadhi et al., 2007), 130 days in U. imbecillis H. myersiana and Utterbackia imbecillis (Say), the laterofrontal cirri were (Trump, 2010), 240 days in M. margaritifera (this study) [although it has always present in recently emerged juveniles, (Kovitvadhi et al., 2007; also been reported to occur at 4.5 mm (Schartum et al., 2016)], and Lasee, 1991; Trump, 2010; Wright, 1995). The broadening of the distal 180 days in U. mancus (this study). However, in another M. margariti- ends of the gill filaments between 7 and 28 days in U. imbecillis (Trump, fera study, it has been reported that the inner gill begins to bend at 16 2010) and at 56 days (890 mm) in L. ventricosa (Lasee, 1991) sufficiently months and the complete gills (inner and outer) are formed at three forms a basket for particle capture during gill bud and reflected gill years (Lavictoire, Moorkens, Ramsey, Sinclair, & Sweeting, 2015). stages. During metamorphosis in P. maximus juveniles (300–900 mmof Notably, we observed an important difference in inner gill develop- shell length), a similar process of gill basket formation occurs while the ment between M. margaritifera and U. mancus, likely explained as a velum (larval feeding organ) disappears; but in this case, the inner gill character difference between families: muscular junctions between fila- has not yet reflected (Beninger et al., 1994). Nevertheless, formation of ments were clearly observed at 345 days in U. mancus but were com- both the inner demibranch ventral groove and the pedal palps is pletely absent in M. margaritifera. required for successful filter feeding (Kellogg, 1915). Once these struc- The papillae of the inhalant aperture were present at 330 days tures are well developed, and filter feeding becomes the main (or only) (2.2 mm) in M. margaritifera, although at this time, the flange, which will feeding mechanism, juvenile mortality likely decreases. Based on our form the pseudodiaphragm, was already present in the internal wall of observations here, filter feeding occurs around 180 days (1.6 mm) in M. the mantle. The papillae of the inhalant siphon of U. mancus were pres- margaritifera with the formation of the labial palp ridges and cilia and ent at 300 days (2.2 mm). In comparison, papillae were observed at the ventral groove of the inner gill. However, other authors have 100 days in H. myersiana (Kovitvadhi et al., 2007), 175 days (3 mm) in reported this feeding change occurs later (2.2 mm) in this species U. imbecillis (Trump, 2010), and 272 days in V. iris (Gatenby et al., (Schartum et al., 2016). In U. mancus, the palp primordia and the transi- 1996). tion to filter feeding occurs at 70 days (1 mm). The initial small folds of Of the freshwater mussels compared (Table 1), H. myersiana is the the labial palps were present at two days in L. ventricosa (Lasee, 1991) fastest growing species, reaching 1.8 mm in 60 days, 24 mm in 150 and three days in U. imbecillis (Trump, 2010), but in the latter species, days and 85 mm in one year. The feeding organs of this species also the ridges did not appear until 130 days (Trump, 2010). develop more rapidly, having well-developed gills and siphons by 50 We have observed the reflection of the outer demibranch of the days post-emergence (Table 2). However, for the other Unionidae spe- inner gill at 130 days (1.2 mm) in M. margaritifera and between 40 and cies mentioned (Table 2), these organs developed at similar times. In 60 days (0.8–1mm)inU. mancus. In these species, there is a clear rela- the case of M. margaritifera, the only member of the Margaritiferidae tionship between effective filtering mechanisms and inner gill reflec- studied, these organs developed much more slowly. Interestingly tion, as has also been hypothesized in other freshwater (Herbers, although, cultured M. margaritifera juveniles at 360 days had a mean 1914; Kovitvadhi et al., 2007; Trump, 2010) and marine (Beninger shell length twice that of wild populations (Lavictoire, Moorkens, Ram- et al., 1994; Cannuel et al., 2009) species. In M. margaritifera, the inner sey, Sinclair, & Sweeting, 2016). During the winter, wild populations gills of juveniles between 0.8 and 1.1 mm in length initially have I- stop growing and thus are delayed, whereas our cultured juveniles shaped filaments, bending to a final V-shaped form in juveniles were grown at a constant temperature of 17 8C. However, ctenidial between 1.1 and 4.5 mm, whereas the outer gill filaments are already development is not related to temperature (Schartum et al., 2016). bent from initial growth (Schartum et al., 2016). Inner gill reflection Freshwater mussels are one of the most imperilled animals in the occurs at 7 weeks in Anodonta cygnea (L.) (Herbers, 1914), 30 days in world, and one of the most fragile stage in a mussel’s life is as recently H. myersiana (Kovitvadhi et al., 2007) and 113 days in U. imbecillis (con- emerged juveniles (Archambault et al., 2014; ASTM, 2013; Augspurger sisting of 15–18 filaments; Trump, 2010). However, by 56 days in L. et al., 2007; Strayer & Malcom, 2012). Based on these facts, and to ventricosa, gill reflection had still not been observed (Lasee, 1991). The improve artificial rearing of these molluscs, a better understanding of increase in the number of gill filaments during growth improves filter the natural diet and the development of the feeding organs of newly feeding. In our study, M. margaritifera juveniles had 4–6filaments50 emerged juveniles is crucial (Kovitvadhi et al., 2006, 2007; Lasee, days post-emergence, 8–12 up to 120 days, and more than 20 at 210 1991; Lima et al., 2006; Schartum et al., 2016; Trump, 2010; Tucker, days. One filament is added for every 123 mm of shell length (Schartum 1927; Uthaiwan et al., 2001). Given the high level of juvenile mortality et al., 2016). In U. mancus, 18 filaments were present at 120 days and reported for U. mancus and M. margaritifera in natural habitats with no 45 at 345 days. In L. ventricosa, there were four pairs of filaments at 21 food issues (Araujo et al., 2015; Eybe et al., 2015; Hastie & Young,

220 ARAUJO ET AL. | 11

2003), early juvenile mortality can be due to the inability to successfully freshwater mussel (Unionidae) early life stages. Environ. Toxicology – transition feeding modes, facilitated by changing anatomical structures. and Chemistry, 26(10), 2025 2028. The transition from pedal feeding to filter feeding occurs around 150– Baker, S. M., & Hornbach, D. J. (2001). Seasonal metabolism and bio- chemical composition of two unionid mussels, Actinonaias ligamentina 200 days post-emergence in M. margaritifera and around 70 days in U. and Amblema plicata. Journal of Molluscan Studies, 67, 407–416. mancus, after juveniles are greater than 1 mm in length, which coin- Barnhart, M. C. (2006). Buckets of muckets: A compact system for rear- cides with the timing of reported high mortality. Once this feeding met- ing juvenile freshwater mussels. Aquaculture, 254, 227–233. amorphosis is complete, juvenile mortality decreases. Bauer, G. (1991). Plasticity in life History Traits of the Freshwater Pearl In our cultures of M. margaritifera and U. mancus, juveniles were Mussel- Consequences for the danger of extinction and for conserva- maintained for more than one year. Based on the anatomical observa- tion measures. In A. Seitz & V. Loeschicke (Eds.), Species conservation: A population-biological approach (pp. 103–120). Basel, Switzerland: tions reported here, juveniles of both species first feed with the foot Birkhäuser Verlag. and then by filtering algae or another food source. Future studies will Beck, K., & Neves, R. J. (2003). An evaluation of selective feeding by next focus on determining the actual food type ingested during these three age-groups of the Rainbow mussel Villosa iris. North American two distinct feeding stages. Journal of Aquaculture, 65, 203–209. Beninger, P. G., Dwiono, S. A. P., & Le Pennec, M. (1994). Early develop- ACKNOWLEDGMENTS ment of the gill and implications for feeding in Pecten maximus (Bival- via: Pectinidae). Marine Biology, 119, 405–412. The culture of M. margaritifera was funded by the LIFE-Margal Ulla Beninger, P. G., Veniot, A., & Poussart, Y. (1999). Principles of pseudofe- (LIFE09 NAT/ES/000514) and rearing was achieved at the Veral ces rejection on the bivalve mantle: integration in particle processing. rearing facility (Consellería de Medio Ambiente, Xunta de Galicia). Marine Ecology Progress Series, 178, 259–269. We are grateful to S. Latas and P. Caballero (Consellería de Medio Burky, A. J. (1983). Physiological ecology of freshwater bivalves. In W. – Ambiente, Xunta de Galicia) and Dr. R. Mascato for constantly sup- D. Russell-Hunter (Ed.), The mollusca (Vol. 6, pp. 281 327). New York, NY: Academic Press. porting the Galician FPM conservation project. We also thank E. Caddy, J. F. (1969). Development of mantle organs, feeding, and locomo- Corral, R. Ocharan and P. Castrillo for their active assistance of the tion in postlarval Macoma balthica (L.) (Lamellibranchiata). Canadian rearing work. Furthermore, we wish to thank Dr. F. Thielen and his Journal of Zoology, 47, 609–617. team for their encouragement and support of mussel culturing. The Cannuel, R., Beninger, P. G., McCombie, H., & Boudry, P. (2009). Gill rearing of U. mancus was carried out at the Laboratory of breeding development and its functional and evolutionary implications in the of freshwater mussels of the lake of Banyoles (Consorci de l’Estany) blue mussel Mytilus edulis (Bivalvia: Mytilidae). Biological Bulletin, 217, 173–188. under the LIFE-Potamo Fauna (LIFE12 NAT/ES/001091). We are Coker, R. E., Shira, A. F., Clark, H. W., & Howard, A. D. (1921). Natural grateful to Dr. Q. Pou, I. Camos, R. Casadevall and G. Dalmau for history and propagation of fres-water mussels. Bulletin of the U. S. their active assistance of the rearing work. This study was made Bureau of Fisheries, 37,77–181. possible by the networking of the two aforementioned LIFE projects Efford, I. A., & Tsumura, K. (1973). Uptake of disolved glucose and gly- co-financed by the European Commission. A. Jorge and L. Tormo cine by Pisidium, a freshwater bivalve. Canadian Journal of Zoology, – made the scanning photographs and J. Munoz~ made the plates. 51, 825 832. Eversole, A. G., Stuart, K. R., & Brume, D. E. (2008). Effect of temperature and phytoplankton concentration of partitioned aquaculture system water on ORCID freshwater mussel filtration. Aquaculture Research, 39, 1691–1696. Rafael Araujo http://orcid.org/0000-0001-5842-666X Eybe, T., Thielen, F., Bohn, T., & Sures, B. (2013). The first millimetre- rearing juveniles freshwater pearl mussels (Margaritifera margaritifera L.) in plastic boxes. Aquatic Conservation: Marine & Freshwater Ecosys- REFERENCES tems, 23, 964–975. Araujo, R., Feo, C., Pou, Q., & Campos, M. (2015). Conservation of two Eybe, T., Thielen, F., Bohn, T., & Sures, B. (2015). Influence of the endangered European freshwater mussels. A three year semi-natural excystment time on the breeding success of juvenile freshwater pearl breeding experiment. The Nautilus, 129(3), 126–135. mussels (Margaritifera margaritifera). Aquatic Conservation: Marine & Araujo, R., Quiros, M., & Ramos, M. A. (2003). Laboratory propagation and Freshwater Ecosystems, 25,21–30. culture of juveniles of the endangered freshwater mussel Margaritifera Galbraith, H. S., Frazier, S. E., Allison, B., & Vaughn, C. C. (2009). Com- auricularia (Spengler, 1793). Journal of Conchology, 38,53–60. parison of gill surface morphology across a guild of suspension- Archambault, J. M., Cope, W. G., & Kwak, T. J. (2014). Survival and feeding unionid bivalves. Journal of Molluscan Studies, 75, 103–107. behaviour of juvenile unionid mussels exposed to thermal stress and Gatenby, C. M., Neves, R. J., & Parker, B. C. (1996). Influence of sedi- dewatering in the presence of a sediment temperature gradient. ment and algal food on cultured juvenile freshwater mussels. Journal Freshwater Biology, 59, 601–613. of the North American Benthological Society, 15(4), 597–609. ASTM E2455–06 (American Society for Testing and Materials) (2013). Gatenby, C. M., Orcutt, D. M., Kreeger, D. A., Parker, B. C., Jones, V. A., Standard Guide for Conducting Laboratory Toxicity Tests with Fresh- & Neves, R. J. (2003). Biochemical composition of three algal species water Mussels. West Conshohocken, PA: ASTM International. proposed as food for captive freshwater mussels. Journal of Applied Retrieved from www.astm.org Phycology, 15,1–11. Augspurger, T., Dwyer, F. J., Ingersoll, C. G., & Kane, C. M. (2007). Gatenby, C. M., Parker, B. C., & Neves, R. J. (1997). Growth and survival Advances and opportunities in assessing contaminant sensitivity of of juvenile rainbow mussels, Villosa iris (lea, 1829) (Bivalvia:

221 12 | ARAUJO ET AL.

Unionidae), reared on algal diets and sediment. American Malacological Kovitvadhi, S., Kovitvadhi, U., Sawangwong, P., Thongpan, A., & Bulletin, 14(1), 57–66. Machado, J. (2006). Optimization of diet and culture environment for Gum, B., Lange, M., & Geist, J. (2011). A critical reflection on the success larvae and juvenile freshwater pearl mussels, Hyriopsis (Limnoscapha) of rearing and culturing juvenile freshwater mussels with a focus on myersiana Lea, 1856. Invertebrate Reproduction and Development, 49, – the endangered freshwater pearl mussel (Margaritifera margaritifera 61 70. L.). Aquatic Conservation: Marine & Freshwater Ecosystems, 21, 743– Kovitvadhi, S., Kovitvadhi, U., Sawangwong, P., Thongpan, A., & 751. Machado, J. (2007). Morphological development of the juvenile Guyot, J. A. (2005). Restoration of the endangered Cumberland elktoe through to the adult in the freshwater pearl mussel Hyriopsis (Limno- (Alasmidonta atropurpurea) and Cumberland bean (Villosa trabalis) scapha) myersiana under artificial culture. Invertebrate Development, (Bivalvia: Unionidae) in the Big South Fork National River and Recreation 50, 207–218. Area, Tennessee and Kentucky (Master of Science Thesis). Department Lasee, B. A. (1991). Histological and ultrastructural studies of larval and of Fisheries and Wildlife Sciences: Virginia Polytechnic Institute and juvenile Lampsilis (Bivalvia) from the Upper Mississippi River (PhD Dis- State University. sertation). Iowa State University, Ames, IA. Hastie, L. C., & Young, M. R. (2003). Conservation of the freshwater pearl Lavictoire, L., Moorkens, E., Ramsey, A., Sinclair, W., & Sweeting, R. A. mussel I: captive breeding techniques. Conserving Natura 2000 Rivers (2015, November). Investigations into feeding structures of juvenile Ecology Series No. 2, English Nature, Peterborough, England. freshwater pearl mussels (Margaritifera margaritifera) through scan- Helm, M. M., Bourne, N., & Lovatelli, A. (2004). Cultivo de bivalvos en ning electron microscopy (p. 55). Paper presented at: 2nd Interna- criadero. Un manual practico . FAO Fisheries Technical Paper 471. tional Seminar. Rearing of unionoid mussels. Clarvaux, Luxembourg. Roma: FAO. Lavictoire, L., Moorkens, E., Ramsey, A. D., Sinclair, W., & Sweeting, R. A. Henley, W. F., Zimmerman, L. L., Neves, R. J., & Kidd, M. R. (2001). (2016). Effects of substrate size and cleaning regime on growth and Design and evaluation of recirculating water systems for mainte- survival of captive-bred juvenile freshwater pearl mussels, Margariti- nance and propagation of freshwater mussels. North American Journal fera margaritifera (Linnaeus, 1758). Hydrobiologia, 766,89–102. of Aquaculture, 63, 144–155. Lefevre, G., & Curtis, W. C. (1912). Studies on the reproduction and arti- Herbers, K. (1914). Entwickungsgeschichte von Anodonta cellensis Schrot.€ ficial propagation of fresh-water mussels. Bulletin of the Bureau of Zeitschrift fur€ Wissenschaftliche Zoologie, 108,1–174. Fisheries, 30, 105–201. Holopainen, I. (1985). Feeding biology of Pisidiidae (Bivalvia) with special Liberty, A. (2004). An Evaluation of the Survival and Growth of Juvenile emphasis on functional morphology of the digestive tract. Lammi and Adult Freshwater Mussels at the Aquatic Wildlife Conservation Cen- Notes, 12,5–9. ter (AWCC), Marion, Virginia (Master of Science Thesis). Virginia Poly- technic Institute and State University, Blacksburg, VA. Holopainen, I. J., & Hanski, I. (1979). Annual energy flow in populations of two Pisidium species (Bivalvia, Spaeriidae), with discussion on pos- Liberty, A., Ostby, B. J., & Neves, R. (2007). Determining a suitable size sible competition between them. Archiv fur€ Hydrobiologie, 86, 338– and sampling frequency for rearing juvenile Rainbow mussels Villosa 354. iris. North American Journal of Aquaculture, 69,44–52. Holopainen, I. J., & Lopez, G. R. (1989). Functional anatomy and histol- Lima, P., Kovitvadhi, U., Kovitvadhi, S., & Machado, J. (2006). In vitro cul- ogy of the digestive tract of fingernail clams (Sphaeriidae, Bivalvia). ture of glochidia from the freshwater mussel Anodonta cygnea. Inver- Annales Zoologici Fennici, 26,61–72. tebrate Biology, 125(1), 34–44. Hornbach, D. J., Way, C. M., Wissing, T. E., & Burky, A. J. (1984). Effects Lopez, G. R., & Holopainen, I. J. (1987). Interstitial suspension-feeding by of particle concentration and season on the filtration rates of the Pisidium spp. (Pisidiidae: Bivalvia): a new guild in the lentic benthos? freshwater clam, Sphaerium striatinum Lamarck (Bivalvia: Pisididae). American Malacological Bulletin, 5(1), 21–30. Hydrobiologia, 108,83–96. Mackie, G. L., & Qadri, S. U. (1978). Effects of substratum on growth Howard, A. D. (1922). Experiments in the culture of fresh-water mussels. and reproduction of Musculium securis (Bivalvia: Sphaeriidae). The Bulletin of the U. S. Bureau of Fisheries, 38,63–89. Nautilus, 92, 135–144. Howard, J. K., Cuffey, K. M., & Solomon, M. (2005). Toward using Mar- Malo, D. (2012). Mise en production et optimisation de l’alimentation de garitifera falcata as an indicator of base level nitrogen and carbon iso- l’esp.ce Margaritifera margaritifera du Massif armoricain en elevage tope ratios: insights from two California Coast Range rivers. intensif. Memoire de fin d’etude Pour l’obtention du diplome de Hydrobiologia, 541, 229–236. Licence Professionnelle Aquaculture et Gestion Durable de son Hruska, J. (1999). Nahrungsanspruche€ der Flussperlmuschel und deren Environnement. halbnaturliche€ Aufzucht in der Tschechischen Republik. Heldia, 4,69– Meier-Brook, C. (1969). Substrate relations in some Pisidium species 79. (Eulamellibranchiata: Sphaeriidae). Malacologia, 9(1), 121–125. Hudson, R. G., & Isom, B. G. (1984). Rearing juveniles of the freshwater Mitropolskii, V. I. (1966). On mechanisms of filtration and nutrition of mussels (Unionidae) in a laboratory setting. The Nautilus, 98, 129– sphaeriids (Mollusca, Lamellibranchiata). Trudy Instituta Biologii Vnu- 135. trennykh Vodnany Akademiia Nauka SSSR, 12, 129–133. Jones, J. W., Mair, R., & Neves, R. (2005). Factors affecting survival and Morton, B. (1973). A new theory of feeding and digestion in the filter- growth of juvenile freshwater mussels cultured in recirculating aqua- feeding lamellibranchia. Malacologia, 14,63–79. – culture systems. North American Journal of Aquaculture, 67, 210 220. Negishi, J. N., Katsuki, K., Kume, M., Nagayama, S., & Kayaba, Y. (2014). Kellogg, J. L. (1915). Ciliary mechanisms of Lamellibranchs with descrip- Terrestrialization alters organic matter dynamics and habitat quality tion of anatomy. Journal of Morphology, 26, 625–701. for freshwater mussels (Unionoida) in floodplain backwaters. Fresh- – Kovitvadhi, S., Kovitvadhi, U., Sawangwong, P., & Machado, J. (2008). A water Biology, 59, 1026 1038. laboratory-scale recirculating aquaculture system for juveniles of Nichols, S. J., & Garling, D. (2000). Food-web dynamics and trophic-level freshwater pearl mussel Hyriopsis (Limnoscapha) myersiana (Lea, interactions in a multispecies community of freshwater unionids. 1856). Aquaculture, 275, 169–177. Canadian Journal of Zoology, 78, 871–882.

222 ARAUJO ET AL. | 13

Nichols, S. J., & Garling, D. (2002). Evaluation of substitute diets for live Trump, M. P. (2010). Feeding morphology of juvenile Utterbackia imbecillis algae in the captive maintenance of adult and subadult unionidae. (Bivalvia: Unionidae) (Master Degree). Department of Biology: Wake Journal of Shellfish Research, 21(2), 875–881. Forest University. Nicol, D. (1984). Constraints to adaptive radiation in deposit-feeding Tucker, M. E. (1927). Morphology of the glochidium and juvenile of the pelecypods. The Nautilus, 98(4), 165–166. mussel Anodonta imbecillis. Transactions of the American Microscopical OBeirn, F. X., Neves, R. J., & Steg, M. B. (1998). Survival and growth of Society, 46, 286–293. juvenile freshwater mussels (Unionidae) in a recirculating aquaculture Uthaiwan, K., Chatchavalvanich, K., Noparatnaraporn, N., & Machado, J. system. American Malacological Bulletin, 14, 165–171. (2001). Scanning electron microscopy of glochidia and juveniles of O’Foighil, D., Kingzett, B., O’foighil, G., & Bourne, N. (1990). Growth and the freshwater mussel, Hyriopsis myersiana. Invertebrate Reproduction survival of Japanese scallops, Patinopecten yessoensis, in nursery cul- and Development, 40(2–3), 143–151. ture. Journal of Shellfish Research, 9, 135–144. Vaughn, C. C., & Hakenkamp, C.,C. (2001). The functional role of bur- Raikow, D. F., & Hamilton, S. K. (2001). Bivalve diets in a midwestern U. rowing bivalves in freshwater ecosystems. Freshwater Biology, 46, S. stream: a stable isotope enrichment study. Limnology & Oceanogra- 1431–1446. phy, 46(3), 514–522. Vaughn, C. C., Nichols, S. J., & Spooner, D. E. (2008). Community and Reid,R.G.B.,McMahon,R.F.,O’foighil, D., & Finnigan, R. (1992). Anterior foodweb ecology of freshwater mussels. Journal of the North Ameri- inhalant currents and pedal feeding in Bivalves. The Veliger, 35(2), 93–104. can Benthological Society, 27(2), 409–423. Schartum, E., Mortensen, S., Pittman, K., & Jakobsen, P. J. (2016). From Vaughn, C. C., Spooner, D. E., & Galbraith, H. S. (2007). Context-depend- pedal to filter feeding: ctenidial organogenesis and implications for ent species identity effects within a functional group of filter-feeding feeding in the postlarval freshwater pearl mussel Margaritifera mar- bivalves. Ecology, 88(7), 1654–1662. – garitifera (Linnaeus, 1758). Journal of Molluscan Studies, 83(1), 36 42. Welker, M., & Walz, N. (1998). Can mussels control the plancton in riv- https://doi.org/10.1093/mollus/eyw037) ers? A planktological approach applying a Lagrangian sampling strat- Scheder, C., Lerchegger, B., Jung, M., Csar, D., & Gumpinger, C. (2014). egy. Limnology & Oceanography, 43(5), 753–762. Practical experience in the rearing of freshwater pearl mussels (Margar- Wright, A. H. (1995). Particle selection and feeding morphology of in vitro ’ itifera margaritifera): advantages of a work saving infection approach, and in vivo transformed juvenile Uterbackia imbecillis (Master Thesis). – survival, and growth of early life stages. Hydrobiologia, 735, 203 212. Wake Forest University, Winston-Salem, North Carolina. Schmidt, C., & Vandre, R. (2010). Ten years of experience in the rearing of Yasuno, N., Shindo, K., Takagi, Y., Kanaya, G., Shikano, S., Fujimoto, Y., young freshwater pearl mussels (Margaritifera margaritifera). Aquatic Shimada, T., ... Kikuchi, E. (2014). Ontogenetic changes in the – Conservation: Marine and Freshwater Ecosystems, 20, 735 747. trophic position of a freshwater Unionidae mussel. Fundamental & Silverman, H., Nichols, S. J., Cherry, J. S., Achberger, E., Lynn, J. W., & Applied Limnology, 184(4), 341–349. Dietz, T. H. (1997). Clearance of laboratory-cultured bacteria by Yeager, M. M., Cherry, D. S., & Neves, R. J. (1994). Feeding and freshwater bivalves: differences between lentic and lotic unionids. burrowing behavior of juvenile rainbow mussels, Villosa iris (Bivalvia: Canadian Journal of Zoology , 75, 1857–1866. Unionidae). Journal of the North American Benthological Society, 13(2), Stasek, C. R. (1963). Synopsis and discussion of the association of cteni- 217–222. dia and labial palps in the bivalved mollusca. The Veliger, 6(2), 91–97. Strayer, D. L. (2008). Freshwater Mussel Ecology. Freshwater Ecology Series, Vol. 1. Berkeley and Los Angeles, California: University of Cal- How to cite this article: Araujo R, Campos M, Feo C, Varela C, ifornia Press. Soler J, Ondina P. Who wins in the weaning process? Juvenile Strayer, D. L., & Malcom, H. M. (2012). Causes of recruitment failure in feeding morphology of two freshwater mussel species. Journal freshwater mussel populations in southeastern New York. Ecological – Applications, 22(6), 1780–1790. of Morphology. 2017;00:000 000. https://doi.org/10.1002/ Thielen, F. (2011). Rearing of unionoid mussels (with special emphasis on the jmor.20748 Freshwater Pearl Mussel Margaritifera margaritifera). Ferrantia, 64:5–66.

223

224 3.3.2. SCIENTIFIC REPORT 2: First experiences of captive breeding of juveniles of Margaritifera auricularia in France.

Résumé

En raison de son état de conservation critique, il semble que la protection des populations existantes de M. auricularia ne soit pas suffisante pour préserver l’espèce pour l’avenir, car la plupart des populations sont trop petites et présentent un recrutement trop faible. Ainsi, les plans d'action français, espagnol et européen en faveur de M. auricularia concluent à la nécessité d'un renforcement artificiel des populations. Entre 2015 et 2017, nous avons mené les premières expériences d'élevage en captivité de juvéniles de M. auricularia en France. L’objective de ce travail était d'identifier les systèmes d'élevage qui maximisent les taux de survie et de croissance des juvéniles de M. auricularia au cours de leurs premiers mois de vie. Des juvéniles récemment excystés obtenus chaque année ont été collectés et transférés vers différents systèmes d'élevage sur la base d'études antérieures: i) un système semi-naturel extérieur de circulation alimenté constamment par de l’eau naturelle, ii) un système intérieur de recirculation d'eau et iii) des chambres de culture statiques. Plusieurs sources de nourriture ont été testées, notamment les eaux naturelles, le phytoplancton, les détritus, la protéine animale et les bactéries. De plus, l'effet du sédiment a été exploré.

Après infestation, le nombre total de juvéniles libérés par le poisson hôte pour être utilisés dans les différents systèmes au cours des expériences de 2015-2017 était de 38.033. La libération des juvéniles par les poissons hôtes s'est produit entre les jours 30 et 67 après l'infection (IP), en fonction de la température de l'eau (représentant un minimum de 531 et un maximum de 1.071 degrés-jours). Parmi les systèmes d'élevage testés, dans les systèmes i et ii utilisés respectivement en 2015 et 2016, aucun juvénile n'a été maintenu en vie pendant plus de 1,5 mois. Les taux de survie des juvéniles de M. auricularia élevés dans le système iii testé en 2017 étaient plus élevés et nous ont permis de formuler les observations suivantes: a) la présence de substrat dans les systèmes d'élevage semble être d'une grande importance car elle facilite le nettoyage des valves; b) la présence de détritus semble améliorer les taux de survie; c) les algues semblent fournir une combinaison de nutriments adaptés au développement des juvéniles.

225

226 INTRODUCTION

Freshwater mussels are amongst the most threatened invertebrates in the world (Bogan, 1993; Araujo & Ramos, 2000; Young et al., 2001; Lydeard et al., 2004; Strayer et al., 2004; Strayer, 2008). Multiple factors have been identified as responsible for the worldwide decline of these animals, including overharvesting, habitat destruction, pollution, land-use change, and exotic species introductions (Strayer et al., 2004; Lopes- Lima et al., 2017). Different conservation strategies have been employed in order to protect the remaining freshwater mussel populations, including reinforcement of declining populations with adult mussels from healthy populations, creation of protected areas, releasing large numbers of host fish infected with larvae (glochidia), and cultivating and restocking juveniles (Ziuganov et al., 1994). At least in some cases, this last strategy seems to be the only feasible option for restoring severely depleted populations (Araujo, et al., 2003; Preston et al., 2007; Bolland et al., 2010; Araujo et al., 2015; Patterson et al., 2018). Given the worldwide decline in freshwater mussel populations observed in recent times, artificial rearing programs and initiatives have largely increased over the last decades (Strayer et al., 2004; Gum et al., 2011; Patterson et al., 2018).

The reproductive strategy of freshwater mussels involves an obligatory parasitic stage, in which the larvae (glochidia) attach to the external surface of a suitable host and metamorphose into free-living juveniles (Lefevre & Curtis, 1912; Kat, 1984). Thus, the first step in artificial propagation programs involve harvesting glochidia for host fish infestation under controlled conditions. After the complete development of the mussel larvae, the next stage consist in rearing the juveniles. Although recently there has been a considerable increase in publications concerning the captive breeding of freshwater mussels for propagation, high juvenile mortality during the first months after excystment is still considered the bottleneck in freshwater bivalves rearing (Nichols & Garling, 2002; Jones et al., 2005; Eybe et al., 2013; Schartum et al., 2016; Lavictoire et al., 2018).

In Europe, most of the experience on freshwater mussel propagation comes from studies on Margaritifera margaritifera and two different approaches have been generally used for the artificial breeding of their juveniles differing in the intensity of mussel care required (Gum et al., 2011). The first strategy is based on the method developed by Hruška (1999, 2001) in the Czech Republic. This methodology include a laboratory pre-

227 culture of the juveniles in small containers fed with organic detrital suspension (also known as detritus boxes), and their posterior transfer into cages or containers placed in natural rivers or semi-natural flow channels. On the other hand, Preston et al. (2007) developed a less labour-intensive method based on a semi-natural flow-through system supplied by river water with juvenile mussels excysting from the host fish directly into the gravel substratum of a hatchery raceway.

Multiple variations of both strategies have been developed in the last decades including raceway upweller, pulsed flow-through system, recirculating systems, etc. (see Mair, 2018). The suitability of each method may vary depending on the different culture requirements of each mussel species (i.e., food, flow, substrate, etc.) and even on the developmental stage of the species. Furthermore, every propagation facility has its own characteristics in terms of water quality supplied, temperatures, etc., and therefore, a culture system that works well at one facility may not be appropriate at another (Mair, 2018).

Despite the rearing system employed, the overall success in the juvenile culturing seems to be strongly determined by the survival rates during the first months of life. It has been observed that once the juveniles reach approximately 1 mm in length, mortality usually decreases (Mair, 2018; Araujo et al., 2018). Although an important part of the effort devoted in rearing young juveniles is the identification of optimal diets that maximizes growth and survival during this critical stage, the nutritional requirements, specifically for juveniles, remain largely unknown (Eybe et al., 2013; Schartum et al., 2016; Araujo et al., 2018; Lavictoire et al., 2018; Mair, 2018).

Successful results in rearing juveniles have been obtained by providing the water, food, and sediment present in the mussels’ natural ecosystems (Lefevre & Curtis, 1912; Coker et al., 1921; Howard, 1922; Hruska, 1999; Araujo et al., 2015). Hruska (1999) hypothesized that the food required by M. margaritifera juveniles comes from a healthy rhizosphere reaching the aquatic environment by underground flows. Detritus, which is widely available in sediments of aquatic systems, represents a complex matrix which often contains cellulose from terrestrial and aquatic sources, chitin from fungus, rotifers and zooplankton, and it is colonized by bacterial, fungal and protistan fauna (Nichols et al., 2001).

228 Nevertheless, numerous studies have successfully used algae as the main food source for captive juvenile mussels (e.g. Hudson & Isom, 1984, Gatenby et al., 1996; Beck & Neves, 2003; Kovitvadhi et al., 2006; Liberty et al., 2007; Eybe et al., 2013). Algae appear to provide key nutrients such as vitamins and phytosterols. Other authors have tested enriched diets by adding animal protein to improve the amino acid supply (Eybe et al., 2013).

Margaritifera auricularia is considered one of the most threatened bivalves in Europe and only a few remaining populations still survive in France and Spain (Araujo & Ramos, 2001; Prié et al., 2011, 2018; in Section 3.1.1 of this thesis). Due to its critical conservation status, it seems that protecting the existing populations of this freshwater mussel is not sufficient to safeguard the species for the future, as most populations are too small and show too little recruitment to be self-sustainable. Thus, the French National Action Plan for this species (Prié et al., 2011), the Spanish Recuperation Plan (MARM, 2009) and the European Action plan for Pearl Mussels (Araujo & Ramos, 2001) independently come to the same conclusion that artificial reinforcement is needed, in order to foster populations of this European heritage species.

Until now, there have been very few researches focused on the breeding of Margaritifera auricularia juveniles. Araujo et al. (2003) obtained juveniles that survive several weeks in a nutrient laden media recreated in tanks filled with well and green water, and natural sediments from the Ebro River. More recently, Nakamura et al. (2018a) obtained about 50 juveniles kept alive from the cohorts of the years 2014, 2015, 2016 and 2017 using detritus boxes (Nakamura et al., pers. comm.). The largest juvenile exceeds 2.5 cm shell length and the first experimental systems are being prepared in semi-captivity to test whether these juveniles, born and fed in captivity, are able to start feeding on their own under semi-natural conditions (Nakamura et al., pers. comm.).

The aim of the present work was to identify rearing systems that we have tested in order to maximize the survival and growth rates of Margaritifera auricularia juveniles during their first months of life.

229 METHODS

The captive breeding of Margaritifera auricularia was conducted between 2015 and 2017 within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’. As a main goal of the project included the artificial rearing of M. auricularia, a breeding laboratory was built in Chinon (France), 200 m from the Vienne River (Loire basin) where one of the studied populations of adult M. auricularia lives (see section 3.1.2 of this thesis).

The Chinon breeding facility was composed of a laboratory and two container (mobile) modules. The laboratory was equipped with stereomicroscopes, a conditioning cabinet and standard devices to observe excysted juveniles, to transfer them into the rearing systems, to prepare food solutions, etc. The container modules, with an external dimensions of 600 cm (length) × 230 cm (depth) × 220 cm (height), are fully acclimatized and have water reservoirs equipped with ultraviolet and biological filters. One of the modules was prepared for keeping adult and juvenile mussels and the other for maintaining the fish before, during and after infestation. They were equipped with all necessary elements of an animal-testing facility, including an entry zone where shoes can be changed or sterilized, cupboard for laboratory coats, sterilisable laboratory tables for the infestation procedure, aquaria and cylindroconics for keeping fish and gaining excysted juveniles, the recirculating system for rearing the juveniles, filtering devices, cupboard for safe storage of chemicals, food, refrigerator, freezer, etc.

As the laboratory was not yet operational in 2015, when the first generation of juveniles was obtained, the first experiments of captive breeding took place at the Naiad Breeding Laboratory of Banyoles (Girona, Spain), in the framework of a inter LIFE networking collaboration. This field station is located 500 m from the Banyoles Lake (Ter basin) and receives a constant supply of lake water (see Araujo et al., 2015).

1. Source of juvenile mussels

Between 2015 and 2017, 70 Margaritifera auricularia specimens were collected in spring from the Charente and Creuse rivers by scuba diving (Table 1) and transported to the laboratory in refrigerated coolers. In the laboratory, the mussels were maintained at

230 ambient temperature in aquaria filled with aerated water from their respective rivers and without sediment. As was said in the chapter 3.2.1 (Soler et al., 2018a), liberation of the glochidia occurred between April 1 and May 2. Mature glochidia were collected with pipettes for fish infestation and then mussels were returned to their natural habitat. Prior to infestation, glochidia were checked for viability by observing their response after the addition of NaCl to a small aliquot under a stereomicroscope.

For induced infestations, specimens of Acipenser baerii Brandt, 1869 were purchased from fish farms and acclimated to laboratory conditions in aerated indoor tanks equipped with biological and mechanical filters. The number of specimens of fish and naiads involved was different over the three years of the study (Table 1). All fish used measured 11-19 cm in total length, except 4 big sturgeons of the same species (ca. 35 cm long) that were used in 2017. The infestation process was similar to the one used in Soler et al. (2018b in Section 3.2.2 of this thesis). Small sturgeons were infested by immersion in glochidia baths for 15 minutes under constant agitation by aeration. Big sturgeons were infested by pipetting the glochidia directly on the fish gills. Between 100 and 900 glochidia per fish gram was used in all the experiments, except for 20 small sturgeons infested in 2016 with 4,000 glochidia per fish gram, which died few days after the infestation. Infested fish were then kept in aerated conical tanks (180 L) equipped with biological and mechanical filters. Fish were fed every day until 3 days before the release of juveniles in order to reduce the amount of fish faeces and facilitate mussel collection. The number of degree days expected for the completion of metamorphosis was based on the literature (Araujo & Ramos, 2000; Araujo et al. 2001, 2002 and 2003).

Table 1. Margaritifera auricularia specimens and number of fish used per year for juvenile production.

Mussels Year Fish Charente Creuse 2015 10 0 74 2016 20 5 200 2017 30 5 25

Two days before juvenile release, a 100 µm mesh collector was installed in the water circuit to retrieve them. The juveniles retained in the collection system were daily

231 transferred to Petri dishes and there, only those deemed alive were recovered with a 100 µl micropipette under a stereomicroscope (Figure 1).

Figure 1. Juvenile recuperation process. Juvenile collection system (above), daily collection under stereomicroscope (bottom left) and freshly excysted Margaritifera auricularia juveniles (bottom right).

2. Rearing experiments

The collection system was checked every day until no more juveniles were found. Young individuals of Margaritifera auricularia were collected within a maximum period of 24 h after excystment from the host fish and transferred to a rearing system. Different culture systems and food sources were tested each year, as follows.

232

2.1. Experiments of 2015

These experiments were performed at the Banyoles naiad laboratory in Girona (Spain), in one outdoor canal fed by a continuous supply of water from the Banyoles Lake; no extra food was administered. Juveniles were introduced in 3 different areas of the canal with substrates of different granulometry (fine, medium and coarse) (Figure 2) in one outdoor canal of the lab and fed by a continuous supply of water from the Banyoles Lake; no extra food was supplied.

Figure 2. Outdoor channel of the naiad laboratory (Banyoles, Spain) with continuous supply of lake water (left) and different sediment granulometry of the substrate where the juveniles of Margaritifera auricularia were installed in 2015 (right).

2.2. Experiments of 2016

The 2016 experiments were carried out in the Chinon naiad laboratory. Juveniles were placed equitably in 4 rearing system consisting of large 200 l troughs filled with a 2–3 cm layer of substrate. A continuous water current was created by a pump (Figures 3 and 4) and every week, 30% of the water in the troughs was renewed. Two of these channels were divided in 2 parts to separate the juveniles from the Creuse River and those of the Charente (Figure 4).

233

Figure 3. Schematic draw of the breeding troughs used in 2016 experiments. Above, side view; below, top view (A: water return pipe fed by a pump; B: water; C: river substrate; D: plastic trough; E: grill holding the substrate to one side; F: trough purging system). (modified from Beaume et al., 2016).

Figure 4. Pictures of the rearing systems used at Chinon laboratory in 2016.

In each of the rearing systems, the following 4 different treatments were applied:

1) Natural water from the Vienne River (without filtration) + natural sediment.

2) Natural water from the Vienne River (without filtration) + sieved sediment.

3) Natural water from the Vienne River filtered and UV sterilized + natural sediment + algae.

234 4) Natural water from the Vienne River filtered and UV sterilized + sieved sediment + algae.

Natural water for the four experiments came from the Vienne River. Natural sediment means sediment from the river without sieving (Figure 5A). Sieved sediment means natural sediment from the river but sifted by a sieve of 2 mm and then sterilized by heating at 80ºC during 24 h (Figure 5B). The water from the Vienne River was filtered through a 90 µm mesh filter and then sterilized with UV light. Algae means a mixture of commercially-sourced (ReedMariculture©) microalgae with a combination of Shellfish Diet® (1.4 ml) + Nanno 3600 (0.6 ml) that where supplied weekly (Figure 6).

Figure 5. River sediment used in the rearing systems in 2016. A: natural or not sieved, B: sifted sediment.

Figure 6. Commercial algae used in the rearing systems in 2016.

235 2.3. Experiments of 2017

The experiments were also carried at the Chinon naiad laboratory. Obtained juveniles were transferred to circular glass boxes with water from the Vienne River. Each box (14 cm in diameter) was loosely closed with a cover to allow air exchange and stored at a constant temperature in a conditioning cabinet (Liebherr WK201, Germany) of 18°C for a period of 110 days (Figure 7).

Figure 7. Circular glass boxes stored in a conditioning cabinet.

Each box contained 200 young mussels in 400 ml of river water from the River Vienne that was previously filtered through 90 µm, treated with biomechanical and ultraviolet light filters and kept at 18ºC. The mussels were fed with different food diets, 3 times per week during water exchange, that consisted of algae (A), detritus (H), algae + detritus (S), egg yolk (E) and bacteria (B) (Table 2). See below for the diets preparation. Each of these diet treatments consisted of three replicas (i.e., 12 boxes in total). In order to facilitate the control of growth and survival of juveniles initially, we did not add substrate in the boxes. However, after the high mortalities observed in the first weeks of the experiment, a thin layer of 3 mm thick of fine substrate (80-650 µm) was added 3 weeks after the beginning of the experiment.

Additionally, 4 boxes of different dimensions containing a variable number of juveniles (Table 2) were used to test a combination diet of algae and detritus in presence of sediment. These boxes contained a thin layer (3 mm thick) of coarser substrate (300-1,300 µm particle size), except for one of the boxes (S1), in which fine substrate (80-650 µm) was used (Table 2).

Initially it was planned to carry out the cleaning of the boxes twice a week. However, it was observed that bacterial growths were occurring both on the surface of the water, as a whitish film, and on the bottom of the boxes, as brownish masses, which endangered the

236 survival of the juveniles. Therefore, at 3 weeks after starting the experiment we decided to perform this cleaning 3 times per week. In the boxes that did not have sediment during the first weeks, cleaning was done by transferring the juveniles to a 100 µm filter and then to a new clean box with renewed water and the corresponding food dose. The boxes were washed with diluted bleach and then rinsed thoroughly with river water. In the boxes containing sediment, the content of the box was slightly stirred (by gentle circular movements) and the supernatant was passed through a 100 µm mesh filter. Immediately, the box was filled with river water. After repeating this operation three times, the material retained in the filter was transferred with the aid of a wash bottle to a Petri dish and the juveniles recovered with a 1 ml micropipette under the stereomicroscope. The rest of the sediment that remained in the box was transferred to the 100 µm filter and washed with 500 ml of river water applied with moderate pressure, with the help of a wash bottle. Once clean, the sediment was transferred to a new clean box with renewed water and the corresponding food dose.

Table 2. Distribution of juveniles in the different culture treatments in 2017 experiments.* indicates that sediment was added 3 weeks after the beginning of the experiment.

Box Box Number of Substrate volumen Box type Diet Code juveniles size (µm) (ml) A1 200 400 80-650 * Circular (14 cm diameter) Algae A2 200 400 80-650 * Circular (14 cm diameter) Algae A3 200 400 80-650 * Circular (14 cm diameter) Algae B1 200 400 80-650 * Circular (14 cm diameter) Bacteria B2 200 400 80-650 * Circular (14 cm diameter) Bacteria B3 200 400 80-650 * Circular (14 cm diameter) Bacteria H1 200 400 80-650 * Circular (14 cm diameter) Detritus H2 200 400 80-650 * Circular (14 cm diameter) Detritus H3 200 400 80-650 * Circular (14 cm diameter) Detritus E1 205 400 80-650 * Circular (14 cm diameter) Egg E2 200 400 80-650 * Circular (14 cm diameter) Egg E3 198 400 80-650 * Circular (14 cm diameter) Egg S1 203 700 80-650 Square (12 × 12 cm) Algae + Detritus S3 474 400 300-1,300 Circular (14 cm diameter) Algae + Detritus S4 82 400 300-1,300 Circular (14 cm diameter) Algae + Detritus S5 241 400 300-1,300 Circular (14 cm diameter) Algae + Detritus

237 2.3.1. Diets preparation

Detritus: The objective was to create a detritus from riparian tree leaves. Two liters of dry leaves of Alnus glutinosa, Salix alba and Fraxinus angustifolia were incubated in a 6 liters plastic tray filled with river water during 10-15 days at room temperature. The content was mixed on a daily basis and used during one week. After one week, a fresh new detritus volume was prepared. The mixture of water and leaves were filtered through a 60 μm sieve and 8 ml of the resulting filtrate was added per river water liter in each replica.

Bacteria: The aim was to culture aerobic microorganisms living in the river bed sediments. For this, sediment was collected from the Vienne river at Sauvegrain, where Margaritifera auricularia specimens smaller than 10 cm where previously detected. Before two hours after collection, 100 ml of this sediment was mixed by hand over 5 minutes in 900 ml of river water. The solution was filtered by 250 and 63 µm mesh sieves and the final extract was collected to inoculate the culture medium. This culture medium was composed of 750 ml of river water, 1 g of confectioner's glucose, 1 g of yeast extract,

2 g of CaCO3 and 20 ml of commercial universal fertilizer with micronutrients. A volume of 250 ml of the sediment extract was added to the culture medium in a 2 l plastic bottle arranged vertically. In order to aerate and homogenize the culture, an air diffuser stone ballasted and connected to an air pump was introduced into the bottle. A small volume of the culture was extracted and reserved in a beaker, and then filtered by 50 µm. A volume of 3.2 ml of the resulting solution was applied as food in each replica.

Egg: In order to obtain this diet, a chicken egg was boiled during 10 minutes. Once cold, 1 g of yolk was collected and mixed with 500 ml of mineral water with the aid of an electric mixer during 5 minutes for the preparation of a freezing broth at a 2,000 mg/l concentration. The desired concentration of egg yolk in the rearing boxes was 6.5 mg/l and is based on the work of Nichols & Garling (2002). To obtain that concentration, 1.3 ml of broth was added to each replica.

Algae: A mixture of commercial marine algae was used: Nanno 3600 and Shellfishdiet 1800® (Reed Mariculture Inc., Campbell, California, USA). Nanno 3600 consists of Nannochloropsis sp. with a diameter of 1–2 μm, and Shellfishdiet 1800® is a mixture of different algae (Isochrysis sp., Pavlova sp., Thalassiosira weissflogii, Tetraselmis sp.)

238 with a diameter of 4–20 μm. The algae dosage per replica was: 10 μl/l of Shellfish diet 1800® and 10 μl/l of Nanno 3600.

Algae + detritus: In this treatment, a mixture of algae and detritus was used. The dosage was the sum of the same quantities described above for each individual treatment.

3. Growth and survival assessment

In order to control the growth and survival of the juveniles reared in 2015 and 2016, all the juveniles (in 2015) or a proportion of them (2016) were placed in a sector delimited by visual marks on the channels. For revisions conducted during the first months of the experiments, a sediment sample from the delimited sectors was obtained from the breeding systems, placed on a Petri dish and inspected under the stereomicroscope. Nakamura et al. (2018b), reported that the mean length of 5 juveniles 60 weeks old, was 3.23 mm. Thus, for revisions made at one year or a little after the juvenile excystment, the sediment was sieved through a 500 µm mesh in the 2015 experiment and through a 2 mm mesh in the 2016 experiment. Both were then inspected with magnifier glasses.

In the experiment with different diets made in 2017, juveniles were counted and measured every 15 days and the dead specimens removed. When living juveniles were found, they were photographed under a stereomicroscope coupled to a camera and their lengths measured using an image processing software (Leica Aplication Suite 3.2.0).

RESULTS AND DISCUSSION

Juvenile production

In 2015-2017, the release of glochidia at the laboratory occurred between April 1 and May 2 (Soler et al., 2018a). Fish mortality was negligible during the experiment in 2015 (only one fish died) and it did not exist in 2017. However, in 2016, 63 fish (31.5%) died during the experiments by glochidia overdose or by the high NH4 and NO2 concentrations observed in the water during metamorphosis. The release of juveniles from the host fish occurred between days 30 and 67 post infection (PI), depending on the water

239 temperature (representing a minimum of 531 and a maximum of 1,071 degree days) (Table 3). After infestation of the sturgeons, the total number of juveniles released by the host fish for use in the different systems was 38,033.

Table 3. Number of juveniles of Margaritifera auricularia obtained per infection and year using 11-19 cm Acipenser baerii except for + (ca 35 cm A. baerii). * indicates that excystment probably started before the juvenile collection (see text).

N Start - end of excystment Temperature Year Juveniles fish Days PI Degree days X ± SD (°C) 2015 24 4,530 37-45 544-704 15.7 (±2.95) 24 1,910 38-46 568-738 15.9 (±2.9) 25 410 36-44 559-719 16.4 (±2.8) Total 73 6,850 - - - 2016 10 3,117 45-61 577-971 15.9 (±2.1) 17 4,822 43-63 645-1012 16 (±2.2) 11 4,248 45-63 673-999 15.8 (±2.1) 20 2,845 42-59 668-980 16.6 (±2.2) 19 5,213 45-67 671-1071 15.9 (±2.2) 20 2,117 46-63 689-1000 15.8 (±2.2) 20 2,361 43-62 636-980 15.8 (±2.3) 20 2,992 38-58 571-935 16.1 (±2.2) Total 137 27,715 - - - 2017 4+ 1,433 30-45 531-810 18 (±0.8) 7 1,020 32-39 596*-729 18.6 (±0.4) 7 87 32-39 614*-747 19.2 (±0.4) 7 928 32-39 603*-736 18.8 (±0.6) Total 25 3,468 - - -

As previously observed in Margaritifera auricularia and other freshwater mussel species (e.g., Araujo et al., 2003, 2015), the release of juveniles from the host is a gradual process, with a peak around the middle of the excystment period. In our experiment of 2017 with the small sturgeons, the juvenile collection system was probably installed in the tank after the initiation of the juvenile release, since the number of collected juveniles continuously decreased until no more juveniles were retrieved. Nevertheless, in this year the encystment period was the shortest observed (30-32 days PI) in our experiments, likely due to the constant and relatively high water temperatures (18ºC). In 2016, the excystment period was longer and delayed and required more degree days when compared to that of 2015, even though the mean temperatures were similar. However, as shown in Figure 8, lower water temperatures were more frequent during the encystment period in 2016.

240 These data suggest that the velocity at which glochidia metamorphose is temperature dependent and it may be reduced with lower temperatures.

Figure 8. Water temperature distribution during the experiments of metamorphosis in 2015- 2017.

Perhaps one of the most extreme examples in which this effect has been observed is that of Margaritifera margaritifera. In this species, the metamorphosis can last from 20-60 days at high temperatures, up to about 11 months when temperatures are low and glochidia pass a diapause stage in which there is a cessation of growth and development (Taubert et al., 2013; Murzina et al., 2017). Previous studies also showed that the duration of metamorphosis of M. auricularia glochidia in Acipenser baeri varies, depending on the temperature. Araujo & Ramos (2000) reported that metamorphosis required 700 degree days (30 days PI) at 23-24°C. Araujo et al. (2002) found that this process is slower (1,100 degree days and 65 days PI) with lower temperatures (16-17°C). In addition, Araujo et al. (2003) reported unsuccessful metamorphosis (glochidia were sloughed several days PI) when temperatures were 24ºC, suggesting that this process is interrupted with high temperatures. Thus, degree days calculation as the sum of daily average temperature, may not reflect this effect and should include a correction factor for temperature. However, further investigations are required to stablish threshold temperature values at which metamorphosis is reduced or interrupted. This is important since a better understanding of the degree days required and the timing of the excystment period is essential in order to minimize the laborious control for the presence of juveniles and the period of fasting of infested fish. Furthermore, this information may help to

241 optimize and standardize artificial breeding of M. auricularia and to avoid the loss of juvenile mussels due to delayed collection (Hastie & Young 2003; Taubert et al., 2013).

Rearing experiments

1. Experiments of 2015

During 2015 and the first months of 2016 some sediment samples were carefully obtained from the delimited area of the outdoor channel at the Banyoles naiad laboratory in Girona (Spain) where the juveniles of Margaritifera auricularia were installed (Figure 9). Samples were analysed under a stereomicroscope but no juveniles were found. One year after excystment, in June 2016, we sampled the complete areas where the juveniles were introduced. This sampling was made by pumping the first 5 cm of sediment in each of the 3 areas of the channel (Figure 10). All the pumped sediment was then filtered through a 500 µm sieveand checked with and without stereomicroscope. In order to get the living juveniles, we did not looked for the small empty shells of dead juveniles. All the obtained sediment was checked during 3 days by trained and experienced people (Figure 11). No juveniles were ever found. Although several factors could explain these results (see below), it has to be noted that both water and sediment used in 2015 were from the Banyoles lake, with physico-chemical features very different to the ones of the original habitat from M. auricularia.

Figure 9. Careful sampling of the area of coarse sediment at the canal with Margaritifera auricularia juveniles during 2015 and the first months of 2016.

242

Figure 10. The area of fine sediment with Margaritifera auricularia juveniles before and after being sampled in June 2016.

Figure 11. Plastic trays with sediment of the canal and sampling looking for the Margaritifera auricularia juveniles.

2. Experiments of 2016

The first revision of the rearing systems was made on July 18th 2016 when the juveniles had approximately one month and a half of life (Figure 12). In this sampling only systems containing sieved sediment (systems 2 and 4) were checked. The mean size of juveniles from system 2 (natural water of the Vienne River without filtration + sieved sediment) was 274.6 μm in length (n= 18), corresponding to a growth increment of 90%. The average size in system 4 (Water of the Vienne River filtered and UV sterilized + sieved

243 sediment + algae) was 310.6 μm in length (n= 16) (Figure 12) representing a growth increment of 115%. These results suggest that during these first 65 days of captive breeding the growth was greater in the treatment with algae.

Figure 12. Margaritifera auricularia juveniles from system 4 measured 45 days after excystment.

The next revision was made the 29th of August 2016, when the juveniles were around 3 months old. On that occasion, no juveniles were found alive. Similarly, in the following revisions, carried out on October 2016 and January 2017, no living juveniles were found in any of the breeding systems.

Despite these results, the 4 breeding systems were maintained regularly. On May 2017, system 2 (with sieved sediment) was drained and all the sediment was screened through a 2 mm sieve. All material superior to that size was analyzed but no juveniles of Margaritifera auricularia were found. The remaining systems were emptied and analyzed in September 2017. System 4, also with sieved sediment, was processed in the same way as system 2, giving the same results. In the other two systems, containing not sifted river substrates, all the material was screened through a series of sieves of 1 cm, 4 mm and 2 mm. Material greater than 1 cm was washed with river water under pressure and the detached material was passed to white plastic trays for review under hand magnifiers. The rest of the material, sieved between 2 and 4 mm, was also passed to white plastic trays and analyzed under hand magnifiers. No juveniles of M. auricularia were found.

244 3. Experiments of 2017

Three days after the excystment, most of the juveniles bred in the Bacteria treatment died. Given how early and how fast these juveniles died, it does not seem that it was due to a lack of food, but that the most probable cause is that the culture medium of bacteria presented inadequate physico-chemical conditions. The 190 surviving juveniles were retrieved and transferred to box S5 with algae + detritus.

In all the treatments except that of Algae + Detritus, a high mortality was observed in the first weeks of the experiment when the sediment was not already added. All the juveniles had organic adhesions in the shells, apparently yellow-brown bacterial growths (Figure 13A), which hindered the correct opening and closing of the valves. However, in the treatment of Algae + Detritus, in which unlike the other treatments the boxes had a layer of substrate from the beginning of the treatment, this phenomenon was not observed and the valves of the juveniles were completely clean (Figures 13 C and D).

For this reason, after 3 weeks, it was decided to add substrate to all the boxes of the Algae, Detritus and Egg treatments (boxes named A, H and E respectively). Following the substrate addition, it was observed that the valves of the juveniles were gradually cleaner (Figure 13B), and the mortality stabilized.

In the Detritus treatment (H boxes), it was observed that the bacterial growths attached to the valves of the juveniles were much less evident (Figure 14) and, accordingly, the mortality in those first weeks, before the adding of sediment, was less marked, especially in the H1 box (Figure 15). This difference could be due to the fact that detritus is a source of food less rich in nutrients and organic matter than the rest, so that bacteria have less nutritive substrate for their development.

245

Figure 13. Influence of rearing Margaritifera auricularia juveniles with and without sediments. A) 26 days old living juvenile reared without sediment from box A3. B) 60 days old living juvenile from box A3 after adding sediments 3 weeks after the beginning of the experiment. C, D) 32 and 62 old, respectively living juveniles from box S1 containing sediments from the beginning of the experiment.

Figure 14. 25 days old juvenile from box H1.

246 Detritus 100% 90% 80% 70% 60% 50% H1 40% H2 Survival Survival (%) 30% H3 20% 10% 0% 0 20 40 60 80 100 120 Days

Figure 15. Survival of Margaritifera auricularia juveniles fed with Detritus.

Nakamura et al. (2015) also observed that Margaritifera auricularia juveniles cultured without sediment become covered with organic matter mixed with fungal hyphae and bacteria, which led to high mortality rates week after week. Other authors have pointed out the importance of substrate in rearing different freshwater mussel species. In this sense, Jones et al. (2005) showed that when using algal diets, survival and growth of juvenile the oyster mussels Epioblasma capsaeformis (Lea, 1834) and the rainbow mussel Villosa iris (Lea, 1829), were greatly enhanced by the addition of fine sediments to the culture dishes, and that growth and survival of juvenile mussels cultured with algae, but without sediment, was so poor that long-term culture failed. Other studies (Gatenby et al., 1996; O’Beirn et al., 1998) also showed that sediment significantly increases culture success of juvenile freshwater mussels. In Gatemby et al. (1997), the best growth rates of Villosa iris juveniles after 140 days (specimens of 1,747 µm in length and 30% survival) were obtained using a three algae diet (Neochloris oleoabundans, Phaeodactylum tricornutum and Bracteacoccus grandis) with fine sediment. Although the survival rate was lower, they also reached 140 days using only fine sediment. Jones et al. (2005) concluded that sediment and associated microflora probably serve several important functions, such as aiding in digestion, providing a nutritional resource of both organic and inorganic nutrients, protecting from predators, facilitating feeding orientation, and enhancing hygiene.

247 The treatment Algae + Detritus (S boxes) achieved the highest survival rates with a 34% after 110 days reached in box S1 (Figure 16). This could be related to the fact that in this treatment, sediments were added from the beginning of the experiments and the juvenile mortalities were less important than in the other treatments where the sediment was added later. Nevertheless, an event of high mortality was observed at 60 days that affected boxes S3, S4 and S5 (Figure 16). The cause of this sudden mortality is unknown, and it does not seem to be related to the treatment itself, since in box S1 there was no increase in mortality at that time. The great speed at which this mortality event occurred may suggest the presence of a pathogenic agent.

Algae + detritus 100% 90% 80% 70% 60% S1 50% S3 40% Survival Survival (%) 30% S4 20% S5 10% 0% 0 20 40 60 80 100 120 Days

Figure 16. Survival of Margaritifera auricularia juveniles fed with Algae + Detritus.

Survival in Detritus treatment was high compared to the Egg and Algae (Figures 17 and 18) except for the H2 box (Figure 15). This higher survival could be related, as mentioned before, with the lower growth of bacterial adhesions in the juvenile valves. Thus, it seems that detritus favours good conditions for the maintenance of juveniles. Similar results were achieved by Nakamura et al. (2015) in a Margaritifera auricularia diet experiment, where the most successful treatment was Algae + Detritus, followed by the treatment with only Detritus. Eybe et al. (2013), rearing M. margaritifera juveniles in small boxes, observed that when detritus was added to the boxes, levels of nitrite and ammonium were reduced by more than 50% compared with the initial value within 8 days. Without detritus, ion concentrations increased noticeably (ammonium >50%, nitrite >150%), probably explaining higher mortality rates (Eybe et al., 2013).

248 Egg 100% 90% 80% 70% 60% 50% E1 40% E2 Survival Survival (%) 30% E3 20% 10% 0% 0 20 40 60 80 100 120 Days

Figure 17. Survival of Margaritifera auricularia juveniles fed with Egg.

Algae 100% 90% 80% 70% 60% 50% A1 40% A2 Survival Survival (%) 30% A3 20% 10% 0% 0 20 40 60 80 100 120 Days

Figure 18. Survival of Margaritifera auricularia juveniles fed with Algae.

The highest mean growths at the end of the experiment were observed in the treatments containing algae (Algae, and Algae + Detritus treatments) (Figures 19-22). However, a high intra-group variability was found.

The largest individual at the end of the experiment was obtained in the box S1 of the combined treatment (Algae + Detritus), measuring 617 µm in length at 105 days of age (Figure 23). The juveniles of box S4 reached an average length of 561 µm, while in boxes S5 and S3 they reached 422 and 334 µm respectively (Figure 19). These differences in

249 growth seem to be related to the number of juveniles in each box at the beginning of the experiment (Table 1). In this way, in the boxes in which there were fewer juveniles, they had more available food which would translate into greater growth. Similarly, Eybe et al. (2013) found that the number of mussels (200, 300 or 400) maintained in a box did not influence the survival rate of juvenile M. margaritifera. However, they found that growth rates were significantly higher when juveniles were maintained in lower densities.

In the Egg treatment, growth rates comparable to those reached in the treatments that included algae were obtained up to 75 days after the excystment (Figure 20). This suggests that the egg yolk is potentially a food resource with which juveniles can develop. However, all the specimens died around day 75, so it was not possible to make further comparisons at the end of the experiment.

Algae + Detritus 600

500

400 S1 300

S3 Size Size (µm) 200 S4 100 S5

0 0 20 40 60 80 100 120 Age (days)

Figure 19. Mean growth of Margaritifera auricularia juveniles fed with Algae + Detritus.

250 Egg 400 350 300 250 200 E1

Size Size (µm) 150 E2 100 E3 50 0 0 20 40 60 80 Age (days)

Figure 20. Mean growth of Margaritifera auricularia juveniles fed with Egg.

Algae 600

500

400

300 A1

Size Size (µm) A2 200 A3 100

0 0 20 40 60 80 100 120 Age (days)

Figure 21. Mean growth of Margaritifera auricularia juveniles fed with Algae.

251 Detritus 350 300 250 200 H1 150

Size Size (µm) H2 100 H3 50 0 0 20 40 60 80 100 120 Age (days)

Figure 22. Mean growth of Margaritifera auricularia juveniles fed with Detritus.

Figure 23. Larger specimen recorded during the experiment from box S1 measuring 617.41 µm at 105 days of age.

252 CONCLUSIONS

The following conclusions can be derived from the results obtained in 2017:

 The presence of substrate in the rearing systems seems to be of great importance since it facilitates the cleaning of the valves. In fact, the best results, both in survival and growth, were obtained in the treatment in which the sediment was added since the beginning of the experiment (Algae + Detritus).  The survival was greater in the Algae + Detritus treatment followed by the Detritus treatment, which seems to indicate that the detritus contributes to maintain good physico-chemical conditions in the culture medium.  The highest growth rates were found in the treatments in which algae were used as a food source (Algae and Algae + Detritus treatments), suggesting that the algae provide a combination of nutrients suitable for the development of juveniles. Nevertheless, in the Egg treatment, growth rates comparable to those reached in the treatments that included algae were obtained up to 75 days after the excystment. This suggests that the egg yolk is potentially a food resource with which juveniles can develop. However, all the specimens died around day 75, so it was not possible to make further comparisons at the end of the experiment.

The analysis of the overall results obtained between 2015 and 2017 may allow us establish further remarks that could help for improving our knowledge on rearing Margaritifera auricularia juveniles. Of the rearing systems tested, in both the outdoor semi-natural flow-through system supplied by wild water and the indoor recirculating water system, used in 2015 and 2016 respectively, no juveniles were maintained alive for more than 1.5 months. The results obtained in 2015 could be partially due to the utilization of water and sediment that were not original from M. auricularia habitats. Nevertheless, in the experiment of 2016, also unsuccessful, water and sediment were obtained from the same habitats where M. auricularia lives; therefore, other factors should be identified in order to improve this rearing system. Given the critical status conservation of M. auricularia, this should be a priority since this approach allows producing large number of juveniles with a relatively low investment (Araujo et al., 2015; Beaume et al., 2016). In this sense, it would be interesting to test at the Chinon propagation facility a seminatural rearing system consisting in channels filled with natural sediment and supplied with a constant flow of river water.

253 It has to be noted that the same system tested in 2015 in the Banyoles laboratory, has been successfully used in rearing Unio mancus Lamarck, 1819 and U. ravoisieri Deshayes, 1847, with survival rates up to 5% in one year (M. Campos, personal communication). On the other hand, the system tested in 2016 was inspired on the successful results in rearing Margaritifera margaritifera juveniles obtained by the LIFE project “Conservation de la moule perlière d’eau douce du Massif armoricain” in France (Beaume et al., 2016). Nevertheless, each mussel species may have different culture requirements, so a culture system that works well for one species may not work for another (Mair, 2018). This is also exemplified by recent experiences conducted in May 2018 at the Banyoles propagation facility. Recently excysted M. auricularia juveniles obtained in the Aragón propagation facility were cultured in a raceway upweller system, which has been successfully used for rearing U. mancus juveniles. Despite the 90% survival rates obtained for U. mancus after 100 days, 100% of the juveniles of M. auricularia died after 45 days of culture (M. Campos, personal communication).

In the static culture chambers system tested in 2017 the maximal survival rate after 110 days was 34%. Although the survival rates were still low, these results seems to indicate that rearing Margaritifera auricularia juveniles in small boxes can yield better results. Nakamura et al. (2018a) reported similar results, indicating that after a decade of unsuccessful experiences in rearing M. auricularia in Aragón (Spain), they were able to maintain juveniles alive beyond the first month when they stated to use small boxes as a culture system. They obtained about 50 juveniles kept alive from the cohorts of 2014, 2015, 2016 and 2017 (Nakamura, personal communication). A key factor for these promising results seems to be related to the production of large number of juveniles. Between 2014 and 2017, the mean number of M. auricularia juveniles used in the Aragon facility for captive breeding was around 425,000 per year (Nakamura et al., 2018). Nevertheless, our results and those obtained in Aragón indicate that the survival rates using this methodology are still too low for a sustainable, large-scale reintroduction of M. auricularia in the wild, and therefore, further research is required to improve survival rates.

Similar research efforts have led to the significant improvements observed in the field of freshwater mussel propagation since 2010. Artificial rearing has evolved from releasing hundreds of thousands of small, two-month-old juvenile mussels (less than 5 mm in length) to releasing thousands of large subadult mussels (Mair, 2018). Similarly, the

254 cultivation of M. margaritifera juveniles has improved greatly in recent years. Except for some exceptions, such as the few juveniles obtained by Hrušca in the 80s (see above), until less than 5 years ago the experiments with M. margaritifera were normally conducted to obtain juveniles of one millimeter, since the mortalities were always very high (see Eybe et al., 2013, and references therein).

REFERENCES

ARAUJO, R., CÁMARA N. & RAMOS, M. A. 2002. Glochidium metamorphosis in the endangered freshwater mussel Margaritifera auricularia (Spengler, 1793): A histological and scanning electron microscopy study. Journal of Morphology, 254: 259-265.

ARAUJO, R., CAMPOS, M., FEO, C., VARELA, C., SOLER, J. & ONDINA, P. 2018. Who wins in the weaning process? Juvenile feeding morphology of two freshwater mussel species. Journal of Morphology, 279(1): 4-16.

ARAUJO, R., FEO, C., POU, Q. & CAMPOS, M. 2015. Conservation of two endangered European freshwater mussels (Bivalvia: Unionidae): A three-year, semi-natural breeding experiment. The Nautilus, 129: 126-135.

ARAUJO, R., QUIRÓS. M. & RAMOS, M. A. 2003. Laboratory propagation and culture of juveniles of the endangered freshwater mussel Margaritifera auricularia (Spengler, 1793). Journal of Conchology, 38: 53-61.

ARAUJO, R. & RAMOS, M. A. 2000. Status and conservation of the giant European freshwater pearl mussel (Margaritifera auricularia) (Spengler, 1793) (Bivalvia: Unionoidea). Biological Conservation, 96: 233-239.

ARAUJO, R., & RAMOS, M. A. 2001. Action Plan for Margaritifera auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Council of Europe Publishing. Nature and environment, No. 117. Strasbourg.

BEAUME, N., PASCO, P. Y., RIBEIRO, M., DURY, P., ROSTAGNAT, L. & BEAUFILS, B., 2016. Conserver la moule perlière d’eau douce dans nos rivières. Recueil

255 d’expérience du programme LIFE+ Nature « Conservation de la moule perlière d’eau douce du Massif armoricain » 2010-2016. Bretagne Vivante / CPIE des Collines / FFPPMA 29 / SIAES / PNRNM. Brest. 72 pp.

BECK, K. & NEVES, R. J. 2003. An evaluation of selective feeding by three age-groups of the Rainbow mussel Villosa iris. North American Journal of Aquaculture, 65: 203- 209.

BOGAN, A. E. 1993. Freshwater bivalve extinctions (Mollusca: Unionoida): a search for causes. American Zoologist, 33: 599-609.

BOLLAND, J. D., BRACKEN, L. J., MARIN, R. & LUCAS, M. C. 2010. A protocol for stocking hatchery reared freshwater pearl mussel Margaritifera margaritifera. Aquatic Conservation: Marine and Freshwater Ecosystems, 20: 695-707.

COKER, R. E., SHIRA, A. F., CLARK, H. W. & HOWARD, A. D. 1921. Natural history and propagation of freshwater mussels. Bulletin of the United States Bureau of Fisheries, 37: 77-181.

EYBE T., THIELEN F., BOHN T. & SURES B. 2013. The first millimeter: Rearing juvenile freshwater pearl mussels (Margaritifera margaritifera L.) in plastic boxes. Aquatic Conservation: Marine and Freshwater Ecosystems, 23: 964-975.

GATENBY, C. M., NEVES, R. J. & PARKER, B. C. 1996. Influence of sediment and algal food on cultured juvenile freshwater mussels. Journal of the North American Benthological Society, 15: 597-600.

GUM, B., LANGE, M. & GEIST, J. 2011. A critical reflection on the success of rearing and culturing juvenile freshwater mussels with a focus on the endangered freshwater pearl mussel (Margaritifera margaritifera L.). Aquatic Conservation: Marine and Freshwater Ecosystems, 21: 743-751.

HASTIE, L. C., YOUNG, M. R. 2003. Conservation of the Freshwater Pearl Mussel I: Captive Breeding Techniques. Conserving Natura 2000 Rivers Conservation Techniques Series No. 2. English Nature, Peterborough.

256 HOWARD, A. D. 1922. Experiments in the culture of fresh-water mussels. Bulletin of the U. S. Bureau of Fisheries, 38: 63-89.

HRUŠKA, J. 1999. Nahrungsansprüche der Flussperlmuschel und deren halbnatürliche Aufzucht in der Tschechischen Republik. Heldia, 4: 69-80.

HRUŠKA J. 2001. Experience of semi-natural breeding program of freshwater pearl mussel in the Czech Republic. In: Die Flussperlmuschel in Europa: Bestandssituation und Schutzmaßnahmen. Kongressband. WWA Hof, Albert-Ludwigs Universität: Freiburg.

HUDSON, R. G. & ISOM, B. G. 1984. Rearing juveniles of the freshwater mussels (Unionidae) in a laboratory setting. Nautilus, 98: 129-135.

KOVITVADHI, S., KOVITVADHI, U., SAWANGWONG, P., THONGPAN, A. & MACHADO, J. 2006. Optimization of diet and culture environment for larvae and juvenile freshwater pearl mussels, Hyriopsis (Limnoscapha) myersiana Lea, 1856. Invertebrate Reproduction and Development, 49: 61-70.

JONES, J. W., MAIR, R. A. & NEVES, R. J. 2005. Factors affecting survival and growth of juvenile freshwater mussels cultured in recirculating aquaculture systems. North American Journal of Aquaculture, 67: 210-220.

KAT, P. W. 1984. Parasitim and the Unionacea (Bivalvia). Biological Review, 59: 189- 207.

LANGE, M. & SELHEIM, H. 2011. Growing factors of juvenile freshwater pearl mussels and their characteristics in selected pearl mussel habitats in Saxony (Germany). Ferrantia, 64: 30-37.

LAVICTOIRE, L., RAMSEY, A. D., MOORKENS, E. A., SOUCH, G. & BARNHART, M. C. 2018. Ontogeny of juvenile freshwater pearl mussels, Margaritifera margaritifera (Bivalvia Margaritiferidae). PlosOne, doi:10.17632/k7m8p7bvjb.2

LEFEVRE, G. & CURTIS, W. 1912. Studies on the reproduction and artificial propagation of fresh-water mussels. Bulletin of the Bureau of Fisheries, 30: 105-201.

257 LIBERTY, A., OSTBY, B. J. & NEVES, R. 2007. Determining a suitable size and sampling frecuency for reearing juvenile Rainbow mussels Villosa iris. North American Journal of Aquaculture, 69: 44-52.

LOPES-LIMA, M., SOUSA, R., GEIST, J., ALDRIDGE, D.C., ARAUJO, R., BERGENGREN, J., BESPALAYA, Y., BODIS, E., BURLAKOVA, L., VAN DAMME, D., DOUDA, K., FROUFE, E., GEORGIEV, D., GUMPINGER, C., KARATAYEV, A., KEBAPCI¸ U., KILLEEN, I., LAJTNER, J., LARSEN, B.M., LAUCERI, R., LEGAKIS, A., LOIS, S., LUNDBERG, S., MOORKENS, E., MOTTE, G., NAGEL, K.-O., ONDINA, P., OUTEIRO, A., PAUNOVIC, M., PRIE, V., VON PROSCHWITZ, T., RICCARDI, N., RUDZITE, M., RUDZITIS, M., SCHEDER, C., SEDDON, M., SEREFLISAN, H., SIMIC, V., SOKOLOVA, S., STOECKL, K., TASKINEN, J., TEIXEIRA, A., THIELEN, F., TRICHKOVA, T., VARANDAS, S., VICENTINI, H., ZAJAC, K., ZAJAC T., & ZOGARIS, S. 2017. Conservation status of freshwater mussels in Europe: state of the art and future challenges. Biological Reviews, 92: 572-607.

LYDEARD, C., COWIE, R. H., PONDER, W. F., BOGAN, A. E., BOUCHET, P., CLARK, S. A., CUMMINGS, K. S., FREST, T. J., GARGOMINY, O., HERBERT, D. G., HERSHLER, R., PEREZ, K. E., ROTH, B., SEDDON, M., STRONG, E. E. & THOMPSON, F. G. 2004. The global decline of nonmarine molluscs. Bioscience, 54: 321-330.

MAIR, R. A. 2018. Juvenile mussel culture. In: Freshwater mussel propagation for restoration. In: Patterson, M. A., Mair, R. A., Eckert, N. L., Gatenby, C. M., Brady, T., Jones, J. W., Simmons, B. R. & Devers, J. eds.), pp. 159-222. Cambridge University Press, Cambridge, United Kingdom.

MARM. 2009. Estrategia Nacional para la conservación de la almeja de río (Margaritifera auricularia) en España. Ministerio de Medio Ambiente y Medio Rural y Marino, Madrid. 39 pp.

MURZINA, S. A., IESHKO, E. P. & ZOTIN, A. A. 2017. The freshwater pearl mussel Margaritifera margaritifera L.: Metamorphosis, growth, and development dynamics of encysted glochidia. Biology Bulletin, 44: 6-13.

258 NAKAMURA, K., CUCALA, L., MESTRE, A., MESQUITA-JOANES, F., ELBAILE, E., SALINAS, C. & MUÑOZ-YANGUAS, M. A. 2018b. Modelling growth in the critically endangered freshwater mussel Margaritifera auricularia (Spengler, 1793) in the Ebro basin. Hydrobiologia, 810: 375-391. Doi: 1010071075001731039.

NAKAMURA, K., GUERRERO, J., ALCÁNTARA, M., MUÑOZ, M. A. & ELBAILE, E. 2018a. Tiempos de incertidumbre para la náyade Margaritifera auricularia. Quercus, 383: 16-24.

NAKAMURA, K., MUÑOZ, M. A., RUIZ-ZARZUELA, I., BLAS, I., ELBAILE, E., SALINAS, C. 2015. The first year of Margaritifera auricularia (Spengler, 1793) Breeding in captivity in Aragon, Spain. In: Clervaux, Luxembourg, Conference: 2 International Seminar Rearing of unionoid mussels, DOI: 10.13140/RG.2.2.12440.49921.

NICHOLS, S. J., ALLEN, J., WALKER, G., YOKOYAMA, M. & GARLING, D. 2001. Lack of surface-associated microorganisms in a mixed species community of freshwater Unionidae. Journal of Shellfish Research, 20: 329-335.

NICHOLS, S. J. & GARLING, D. 2002. Evaluation of substitute diets for live algae in the captive maintenance of adult and subadult Unionidae. Journal of Shellfish Research, 21: 875-881.

O’BEIRN, F. X., NEVES, R. J. & STEG, M. B. 1998. Survival and growth of juvenile freshwater mussels (Unionidae) in a recirculating aquaculture system. American Malacological Bulletin, 14: 165-171.

PATTERSON, M. A., JONES, J. W. & GATENBY, C. M. 2018. Why propagate freshwater mussels? In: Freshwater mussel propagation for restoration (M. A. Patterson, R. A. Mair, N. L. Eckert, C. M. Gatenby, T. Brady, J. W. Jones, B. R. Simmons & J. Devers, eds.), pp. 1-24. Cambridge University Press, Cambridge. 320 pp.

PRESTON, S. J., KEYS, A. & ROBERTS, D. 2007. Culturing freshwater pearl mussel Margaritifera margaritifera: a breakthrough in the conservation of an endangered species. Aquatic Conservation: Marine and Freshwater Ecosystems, 17: 539-549.

259 PRIE, V., PHILIPPE, L. & COCHET, G. 2011. Plan National d’Actions en faveur de la Grande Mulette Margaritifera auricularia – 2012-2017. Ministère de l’écologie, du développement durable, des transports et du logement, Paris.

PRIE, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., PATRY, N., ADAM, B., LEGRAND, N., JUGE, P., RICHARD, N., & WANTZEN, K. 2018. Challenging exploration of troubled waters: a decade of surveys of the giant freshwater pearl mussel Margaritifera auricularia in Europe. Hydrobiologia, 810: 157-175.

SCHARTUM, E., MORTENSEN, S., PITTMAN, K. & JAKOBSEN, P. J. 2016. From pedal to filter feeding: ctenidial organogenesis and implications for feeding in the postlarval freshwater pearl mussel Margaritifera margaritifera (Linnaeus, 1758), Journal of Molluscan Studies, doi: 10.1093/mollus/eyw037.

SOLER, J., BOISNEAU, C., WANTZEN, K. M., & ARAUJO, R. 2018b. Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 1-4. doi:10.1093/mollus/eyy038.

SOLER, J., WANTZEN, K. M., JUGE, P., & ARAUJO, R. 2018a. Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84: 182-189.

STRAYER, D. L. 2008. Freshwater mussel ecology. A Multifactor Approach to Distribution and Abundance. Freshwater Ecology Series, University of California Press, Berkeley. 204 pages.

STRAYER, D. L., DOWNING, J. A., HAAG, W. R., KING, T. L., LAYZER, J. B., NEWTON, T. J. & NICHOLS, S. J. 2004. Changing perspectives on pearly mussels, North America’s most imperiled animals. BioScience, 54: 429-439.

TAEUBERT, J. E., GUM, B. & GEIST, J. 2013. Variable development and excystment of freshwater pearl mussel (Margaritifera margaritifera) at constant temperature. Limnologica – Ecology and Management of Inland Waters, 43: 319-322.

YOUNG, M., COSGROVE, P. & HASTIE, L. 2001. The extent of, and causes for, the decline of a highly threatned naiad: Margaritifera margaritifera. Pages 337–357 in G.

260 Bauer, and K. Wächtler, editors. Ecology and Evolutionary Biology of the freshwater mussels Unionoidea. Springer Verlag, Berlin.

ZIUGANOV, V., ZOTIN, A., NEZLIN, L. & TRETIAKOV, V. 1994. The freshwater pearly mussels and their relationships with salmonid fish. Vniro Publishing House: Moscow.

261

262

General Discussion

263 264 4. GENERAL DISCUSSION

As reflected in the previous sections, the studies included in this thesis provide an improvement of the knowledge of Margaritifera auricularia in diverse aspects of its biology, ecology and biogeography that allow addressing the conservation of the species with new tools and from a broader perspective.

From a chronological point of view in the ontogeny of the species, the incubation process of the embryos until the formation of the glochidia and the morphological changes that occurs in the mussel gills have been studied for the first time. For this, and in order to avoid the sacrifice of new specimens, we used material of M. auricularia preserved at the Museo Nacional de Ciencias Naturales in Madrid, Spain. On the other hand, the fertility of two French populations was verified and an estimation of the fecundity of the species is given for the first time. Likewise, the way in which glochidia are released has been described, discussing their implications in the host infestation strategy. The hitherto unknown glochidia release period in the French populations has been defined and compared to that of the Spanish populations, which has made it possible to detect a latitudinal gradient related to the water temperature. Likewise, three new host fish species (Gasterosteus aculeatus, Petromyzon marinus and the exotic Silurus glanis) have been identified, and their availability in terms of geographic distribution and phenology has been related to that of the M. auricularia glochidia in nature. The discovery of these new host species is an outstanding tool that can be used in the conservation of the species and also it adds new elements to better understand the causes of the decline of the species. Regarding the metamorphosis of the glochidia in the gills of the fish, data on its duration under different temperature regimes are provided, which can be very useful in future conservation programs. Likewise, morphological changes facilitating a transition of the feeding mode in the early stages of juvenile growth, which seem to be largely responsible for the high mortalities they experience, were studied in two other species of freshwater mussels.

From an ecological and conservation perspective, the historical and current distribution of the species was studied and the information about the conservation status of the currently known worldwide populations was updated. In this regard, an in-depth study of the two most important French populations was carried out, which allowed to more accurately describe their conservation status from the point of view of their distribution, abundance and population

265 structure, as well as their temporal evolution in the short and medium term. This study also allowed to improve the knowledge about the ecological requirements of the species, showing that Margaritifera auricularia tolerates a wide range of salinities and calcium concentrations than previously thought. On the other hand, a new potential threat was discovered for the species derived from the recent colonization of its habitats by Rhodeus amarus, a fish that parasitizes freshwater mussels. Finally, the first experiences of captive breeding of M. auricularia in France were carried out, the results of which, despite not allowing for the moment the large-scale reintroduction of the species, provides valuable information for future experiences.

This chapter presents an integrated discussion of the main results of this thesis, regarding the status, reproductive biology and the early life stages of M. auricularia. Finally, the implications of the study for the species conservation are presented and discussed.

4.1 Current status of the populations

The historical review has confirmed that Margaritifera auricularia was once present as far as the Thames in England and Netherlands and Germany, where fossil specimens have been found and studied (Araujo & Ramos, 2001). On historical times, we found museum records (recent shells) from the Rhine in France or Germany (precise location being unknown), the Seine and the Rhône in France, the Po in Italy and the Tagus in Spain, where the species is now believed to be extinct (Araujo & Ramos, 2001).

Only three populations were known worldwide before 2007. Intensive surveys in the last decade covering a total of 2,500 km of rivers in France and Spain allowed for the rediscovery of nine more populations. Today, M. auricularia is considered restricted to five watersheds: from north to south the Loire watershed (two close populations in the Vienne and Creuse Rivers), the Charente watershed, the Garonne watershed (two very isolated populations, in the Dronne and Save Rivers), the Adour watershed (at least three isolated populations, one in the Adour itself, one in the Luy and one in the Arros) and the Ebro River (four populations, three in channels and a small one remaining in the Ebro itself). Given the magnitude of the efforts allocated to surveying the species in its historical range, we now believe that there are very few chances to rediscover unnoticed populations (except maybe in north-east France). This current distribution represents a range contraction of about a 90% in the last two centuries, as previously estimated by Prié et al. (2014).

266 In Spain, the main population, with 5,000 live specimens, lives in the Canal Imperial in Aragón (Ebro basin, Spain), although there have been some recent mortalities. A new population was recently discovered in the Quinto Ditch, with 25 live specimens. The populations on the Canal de Tauste still host several live specimens and the one in the lower Ebro River is today practically testimonial. Nevertheless, the situation of Margaritifera auricularia in Spain has worsened alarmingly in recent years. Since 2013, mass mortality events have been observed in the Canal Imperial de Aragón, where more than 90% of the Spanish population is concentrated. During this short period, more than 35% of the specimens have died (Nakamura et al., 2018) and if this mortality rate is maintained, it is likely that the species will be completely extinct in Spain in the very short term.

In France, it has been estimated that the Vienne and Creuse Rivers (Loire watershed) harbor about 1,000 individuals (Cochet, 2001a), of which we have found 264. In the Garonne watershed, three sites with a few tens of live specimens were discovered in the Dronne River, and only five live specimens were observed in the Save River. In The Adour drainage, the total population is estimated to be about 300 specimens in the total length of the Adour mainstream, about 150 specimens on the Luy tributary, and about 200 individuals on the 54 km stretch of favorable habitat of the Arros River. The worldwide biggest population lives in the Charente River with an estimated 100,000 individuals. Recruitment is very scarce in all populations but living specimens estimated to be less than 10-15 years old have been found in the Ebro in Spain and in the Vienne, Charente, Dronne and Adour rivers in France.

In addition to a general description of the status of the known populations of Margaritifera auricularia, a more detailed study was carried out in the populations of the Charente, Vienne and Creuse rivers, considered the most important ones in France (Cochet, 2002; Prié et al. 2008). Shell length–frequency distributions indicated that the Charente population is ageing and the recruitment is almost inexistent, which is in agreement with previously reported results by Nienhuis (2003), Prié et al. (2008) and Prié (2010). In this river, although the comparison with previous studies should be interpreted with caution, the results suggest that the population of M. auricularia seems to have remained relatively stable in the last 6-9 years.

Conversely, in the Vienne and Creuse rivers, the results obtained in our study differ markedly from previous results by Cochet (2001) and Nienhuis (2003). Evidences of a recent recruitment were found, constituting the most notable event of Margaritifera auricularia recent reproduction reported so far worldwide. We observed that the sizes of living specimens ranged

267 between 8 and 18 cm, with a 34% of the individuals belonging to size classes smaller than 12 cm. A preliminary study of the age of the individuals, based on external and internal growth rings counting of the shells of this population, suggests that the 10-11 cm size class corresponds to an age of 15-20 years (Nakamura et al., 20018; Soler, unpublished data). This implies that, after a period without effective reproduction, an important recruitment event occurred between the late 1990s and the beginning of 2000’s. Despite these encouraging results, a comparative study on the abundances of this population suggest a 45% decline over the last 10 years which is in accordance with an estimated annual mortality rate ranging between 1 and 13%. Therefore, despite the important recent recruitment observed, the results obtained here show that if conservation measures are not addressed urgently, the species may be locally extinct in the near future.

4.2. Reproductive biology

Given that one of the most important problems affecting the conservation of Margaritifera auricularia is the lack of reproduction in natural populations, expanding the knowledge on this topic was one of the main objectives of this thesis.

Moreover, although all Unionoidea share a common type of life cycle, their reproductive mode may differ considerably. This is evidenced by the variation of such important traits as glochidial size, fertility, host range, site of attachment to the host and developmental mode (i.e., duration of the parasitic stage, growth on the host) (Bauer, 1994).

4.2.1. Brooding and glochidia release in Margaritifera auricularia

Although previous studies have greatly improved the knowledge on the reproductive biology of M. auricularia, except for the brief description provided by Haas (1916, 1924), no studies have characterized the anatomy of the gills and changes in the marsupium during brooding, or embryo developmental times and brood size in this species. This information is basic to better understand other biological traits and is valuable for taxonomic and systematic studies.

One of the key morphological features used for the diagnosis of the Margaritiferidae is the structure of the gills, which are characterized by the incomplete fusion of the gills to the visceral sac and mantle in the posterior area, and the absence of water tubes (Ortmann, 1912; Smith, 1988). Based on specimens collected at the Canal Imperial (Ebro Basin, Spain) and preserved in the malacological collection of the Museo Nacional de Ciencias Naturales (MNCN) in

268 Madrid (Spain), we described the anatomy of the gills of M. auricularia, corroborating the presence of these basic features and giving further details that may be useful in more advanced comparative studies. Regarding the anatomical changes associated to brooding in M. auricularia, we observed that during gravidity, the demibranchs increase the volume of their internal lumen to harbour larvae by reducing the width of the lamellar connective tissue. It was observed that adjustments in the linear dimensions of the tissue, rather than a stretching action, are responsible for the slight increase in demibranch width during brooding. This lack of elasticity, which has also been reported for M. margaritifera (Smith, 1979) but is contrary to reported results in the the family Unionidae (Ortmann, 1911; Fuller, 1972, 1973; Richard et al., 1991; Tankersley & Dimock, 1992), may prove to be a specific anatomical character of the family Margaritiferidae (Smith, 1979).

As described for M. margaritifera by Smith (1979), the epithelial cells of both the interlamellar junctions (ILJs) and the rest of the internal lumen undergo cell-shape changes during the brooding period, presumably to assume a secretory function. In members of the Unionidae, all of which have continuous septa, the ILJ epithelial cells change shape (Lefevre & Curtis, 1912); however, whether other epithelial cells of the lumen also do so is unknown. The extension of the secretory function to the entire epithelium of the gill lumen may be related to having sparse and irregularly distributed ILJs and may be another anatomical character of Margaritiferidae. The function of these secretions is not well understood. Based on studies of several Unionidae (Anodontinae and Lampsilini species), some authors have suggested that nutrients and calcium concretions could be transferred from females to developing larvae via mucus secreted by cells located in the interlamellar septa (Wood, 1974; Tankersley & Dimock, 1992; Tankersley, 1996; Schwartz & Dimock, 2001; McElwain & Bullard, 2014).

Like other Margaritiferidae (Harms, 1907; Howard, 1915; Haas, 1916; Murphy, 1942; Smith, 1980), M. auricularia is a tetragenous species (it broods its eggs and glochidia in the four demibranchs). Although Howard (1915) reported the presence of more eggs and glochidia in the inner demibranchs of one specimen of M. monodonta, in the analysed specimens of M. auricularia the outer demibranchs showed considerably more embryos than the inner demibranchs. At present, we cannot determine if this is a feature of this species or the result of the specimens we observed not yet being fully gravid prior to fixation. Further research on this subject may be useful in order to elucidate the evolution of tetrageny in Margaritiferidae and determine whether it is a plesiomorphic or derived condition.

269 Although there has been some confusion about whether Margaritiferidae are bradytictic or tachytictic (Graf & Ó Foighil, 2000), our observations indicate that M. auricularia is a short- term brooder as glochidia were released shortly after maturation. Furthermore, the marsupia do not have special anatomical adaptations to accommodate long-term brooding, such as gill structures to isolate glochidia during development as has been described for the bradytictic fubfamily Anodontinae and the tribe Lampsilini (Sterki, 1903; Ortmann, 1911; Lefevre & Curtis, 1912). In contrast, the marsupium of M. auricularia remains in contact with the external medium through the ostia and the water canals.

Based on our combined field and laboratory observations of embryonic development, first cleavage to the mature glochidial stage (without egg membrane) in the Creuse River took 31– 37 days at an average temperature of 11°C. This time period required for brood maturation is similar to that observed in several European M. margaritifera populations (Harms, 1907; Scheder et al., 2011).

Fecundity of M. auricularia was estimated to be around 2,000,000 glochidia per adult, based on material collected from three specimens. This information is important from a conservation perspective since it allows to estimate the number of gravid mussels necessary to carry out artificial infestations within a propagation program, and it may be useful for future comparative studies. Furthermore, it also improves our knowledge about the life-history traits of the species. Although this estimate of fecundity is high compared with other members of Unionoida (Haag, 2013), it is lower than that observed for other margaritiferids. Presumably, for similarly sized mussels, more glochidia can be produced with the same energetic investment if the glochidia are smaller (Bauer, 1994). The apparently lower fecundity of M. auricularia among margaritiferids could be a consequence of glochidial size. As we have found, M. auricularia produce glochidia with a length range 127-155 μm (Araujo & Ramos, 1998), being the largest among all reported Margaritiferidae. The glochidium of M. auricularia, as previously described by Araujo & Ramos (1998) and confirmed in our study, is of the hookless type and lacks larval threads. Rather, it has minute teeth in the shell margin (Araujo & Ramos, 1998), which are used to attach to the gills of the host fish.

4.2.2. Identification of the reproductive period of French populations

Although the reproductive period of the Spanish populations has been previously identified (Araujo et al., 2000), this information was still unknown for French populations. Actually, it

270 has never been verified so far whether the French populations are fertile. This basic knowledge is essential in order to establish conservation measures. Collecting gravid females at the right time of the year is essential to carry out captive breeding programs. Collection should occur when the brooded larvae are mature in order to avoid large maintenance periods of adult mussels in the laboratory. In addition, the larvae of short-term brooders (tachytictic) species are generally mature for a very short period, so missing the adequate moment can cause the failure of the whole year of the breeding program. Furthermore, the timing of glochidial release is of crucial importance in relation to host-fish availability, and may give cues to identify the host species that are being used in the wild.

Previous knowledge of the reproductive season of Margaritifera auricularia was based on Ebro Basin populations, which are gravid in February and release glochidia in March (Araujo et al., 2000). Field and laboratory observations of the Charente and Creuse rivers populations, carried out from 2015 to 2017, confirmed the fertility of the specimens and showed that mussels had developing embryos through March, and that the beginning of glochidial release occurs in early to mid-April. Taken together, this indicates a difference in the timing of glochidial release between Spanish and French populations and, indeed, even between the two studied French populations, with glochidial release in the more northern River Creuse population delayed by several days.

Although most available information regarding the season of glochidial release is based on laboratory observations, the evidence suggests that there is a latitudinal gradient influencing the timing of the reproductive season in this species, similar to that suggested in other freshwater mussels populations. Howard (1915) suggested that differences in reproductive timing may be related to latitudinal differences in water temperatures. Thermal differences among rivers seems to be responsible for differences in reproductive timing in other margaritiferids such as Margaritifera laevis (Awakura, 1968; Naito, 1988), M. falcata (Meyers & Millemann, 1977) and M. margaritifera (Hastie & Young, 2003). In the latter species, the authors suggested that annual differences in accumulated degree days (sum of daily average temperature) may be a more reliable indicator of thermal differences. Accumulated degree days from six consecutive years (2011-2016) of the three rivers compared in our study (Ebro, Charente and Creuse), indicated a latitudinal gradient in temperature that seems to support this hypothesis.

Despite variations in the period of reproduction among the different populations, Margaritifera auricularia begins to release glochidia during late winter to early spring, which is earlier than

271 in other margaritiferids. As discussed below, the timing of glochidial release in M. auricularia seems to be related to host fish availability.

Although the glochidial release behaviour of M. auricularia in natural habitats is unknown, results from a drift-net study by Araujo et al. (2000) indicated that release is a sudden event, with the majority of glochidia being released over only 1-2 days, as was also observed in Scottish M. margaritifera populations (Hastie & Young, 2003). Aquaria observations indicated two possible mechanisms for fish infection. The abrupt and massive broadcasting of free glochidia appears to be the main host-infection strategy in M. auricularia. However, we also frequently observed small white masses composed of larval material, including glochidia, that were released but remained near the exhalant aperture, which may be involved in host attraction. A similar behaviour has been reported for M. monodonta (Baird, 2000) and other freshwater mussels (Barnhart et al., 2008), although it has been proposed to be a mechanism to deal with respiratory stress caused by hypoxic conditions (Lefevre & Curtis, 1912; Araujo & Ramos, 1998; Aldridge & McIvor, 2003; Haag & Warren, 2003). Further research is required to determine if this behaviour occurs in nature and whether it functions in host attraction and infection.

4.2.3. Identification of host fish species for French populations of Margaritifera auricularia

Until the beginning of this thesis, Acipenser sturio Linnaeus, 1758 was the only known host fish whose distribution coincided with that of Margaritifera auricularia French populations (Altaba, 1990; Araujo & Ramos, 2000, 2001; López et al., 2007). However, this fish suffered dramatic population declines in the late 19th and early 20th centuries, and nowadays is virtually extinct with only one remaining reproductive population in the Garonne River (France) (Gesner et al., 2010). Thus, the previous knowledge on the M. auricularia host fish could not explain the recent recruitment observed in some French populations, suggesting that these populations are using an unknown host species.

As we have demonstrated, all the previously identified host species of M. auricularia have a close relationship with the marine environment (see Section 3.2.2 of this thesis). Therefore, we hypothesized a similar relationship for the unknown host species. In order to identify this mysterious host, a double approach was followed, including the assessment of natural infestation in the wild and laboratory studies on artificial infestations.

272 The fish communities of the Creuse, Vienne and Charente rivers were assessed for natural infestation by electrofishing directly downstream of Margaritifera auricularia populations. Based on the previously identified reproductive period of these populations (see Section 3.2.1 of this thesis), the electrofishing dates were chosen in order to coincide with the glochidial encystment period. During the electrofishig surveys, we analyzed 966 individuals belonging to 29 fish species. Glochidia of M. auricularia were found only in two species: Gasterosteus aculeatus Linnaeus, 1758 and Anguilla anguilla Linnaeus, 1758. This last species has been previously tested for experimental infestation with M. auricularia glochidia in two independent studies, but no encysted glochidia were found beyond one week post infestation (Araujo et al., 2001; López & Altaba, 2005).

In 2017, we tested Gasterosteus aculeatus for experimental infestation, and 27 days post- infestation fully transformed live juveniles were recovered, indicating that the species is a suitable host for M. auricularia glochidia. Nevertheless, we observed a high glochidia loss during the encystment period. This seems to be frequent in universal hosts, as we consider G. aculeatus, since this species has been reported as a host for 14 unionid mussel species distributed throughout Europe, North America and Asia, and is the host fish most commonly used by European freshwater mussels (Lopes-Lima et al., 2017). In addition to the high glochidial loss during the encystment period, G. aculeatus is not present in the Creuse and Vienne Rivers. Therefore, we suspected that another host species could also be responsible for the maintenance of the French populations of M. auricularia.

One of the most abundant anadromous species in French rivers co-occurring with still living populations of Margaritifera auricularia is the sea lamprey Petromyzon marinus Linnaeus, 1758. We tested this species in 2018 and we found that the glochidia of M. auricularia successfully metamorphosed on the sea lamprey. A total of 13,827 living juvenile mussels were collected from a single P. marinus of 1,200 g weight. To the best of our knowledge, our study is the first to confirm P. marinus as a host of any freshwater mussel. In fact, several authors have pointed out how surprisingly infrequent lampreys are used as hosts by freshwater mussels despite being widespread and common benthic vertebrates (Haag, 2012; Strayer, 2008).

Although we did not capture any P. marinus specimen during the electrofishing surveys, the available information on its geographical distribution, phenology and habitat suggest that it could be a good ecological host for M. auricularia. Regarding its distribution, the studied rivers, particularly the Vienne and Creuse, are known to harbor large populations of P. marinus (Bach

273 et al., 2016; Dartiguelongue, 2017), and spawning areas of this species have been observed in the vicinity of M. auricularia habitats, at least in the Vienne and Creuse rivers (Portafaix et al., 2015). Indeed, P. marinus is also present in the other French rivers inhabited by recruiting populations of M. auricularia (Amrein 2004, 2005; Prié et al., 2018; Taverny et al., 2005). Furthermore, the historical distribution of M. auricularia coincides with the European distribution range of P. marinus (Kottelat & Freyhof, 2007; NatureServe, 2013). Thus, as proposed for Acipenser sturio (Altaba, 1990; Araujo & Ramos, 2000a, 2000c), the distribution of P. marinus could explain the distribution of M. auricularia.

Lampreys spend the first four to five years of life living buried in river sediments as filter- feeding larvae called ammocoetes (Hansen et al., 2016; Taverny et al., 2005). They then metamorphose into juveniles and migrate to the sea where they start their hematophagous phase (Youson, 1980). Adult sea lampreys only spend a few months of their lives in rivers implying that the timing of glochidia release is of crucial importance. The migration period of P. marinus, like in other anadromous species, is triggered by rising temperatures. In French rivers, migration occurs from December to late June, peaking in March and April (Taverny & Élie, 2010), which coincides with the glochidia release period of French populations of M. auricularia (see Section 3.2.1 of this thesis). This period also overlaps with the migration of A. sturio adults from the sea from April to May (CEMAGREF, 1994). Thus, the earlier reproductive period of M. auricularia in relation to other margaritiferids, which is also driven by temperature (see Section 3.2.1 of this thesis), may be linked to the arrival of anadromous adults to their spawning grounds. Nevertheless, juveniles seem to be the most important stage as host for all freshwater mussels, including other margaritiferids (Karna & Millemann, 1978; Modesto et al., 2018; Young & Williams, 1984). Given the large number of ammocoetes present during the reproductive period of M. auricularia, one would expect them to also be infested by glochidia.

The discovery of the sea lamprey as a host species for M. auricularia extends the number of physiological hosts to seven species belonging to four different families, suggesting that M. auricularia could exhibit a more host generalist behavior than previously expected. Previous studies suggest that invasive fish species have a lower suitability as hosts for freshwater mussels than native ones, probably due to a co-evolutionary mechanism of host compatibility between mussels and host fish species (Taeubert et al., 2012; Douda et al., 2013; Salonen et al., 2016). According to the available data, the vast majority of freshwater mussel species utilizes only native hosts whereas the glochidia of a few host generalist species have been proven to

274 successfully metamorphose on non-native fishes (Huber & Geist, 2017; Modesto et al., 2018; Teixeira et al., 2018).

Given the ubiquity of the exotic Silurus glanis Linnaeus, 1758 along the giant pearl mussel populations in France and in Spain, we decided to test this species for experimental infestations in 2018. After 24 days post-infestation some juveniles were collected indicating that S. glanis is a physiological host of M. auricularia glochidia. Nevertheless, we did not found M. auricularia glochidia attached to any of the 27 specimens inspected during our electrofishing surveys, suggesting that S. glanis is not a good ecological host.

The results obtained with S. glanis seems to support the idea that M. auricularia has a host generalist behavior, in contrast to the other species of the family Margaritiferidae for which their hosts are known, which generally use very few species of a single family. As a narrow range of host species usually implies a larger risk of extinction, our results seem to be a paradox given that the species is critically endangered.

As we have already mentioned, nearly all currently known Margaritifera auricularia hosts have a notable tolerance to salinity (Curole et al., 2004; Araujo et al., 2017). Sturgeons and Petromyzon marinus are anadromous, Gasterosteus aculeatus can inhabit both brackish or freshwater and anadromous populations are relatively frequent, and Salaria fluviatils (Asso, 1801), one of the host fish of M. auricularia in Spain, although restricted to freshwater, can tolerate high salinity levels (Plaut, 1998) and belongs to large family of mostly marine species (Kottelat & Freyhof, 2007). While S. glanis seems to deviate slightly from this pattern of halotolerance because it mainly occurs in freshwater, it has been observed to enter brackish waters in the Baltic, Black and Mediterranean seas and even to spawn in salt water (Berg, 1964; Frimodt, 1995).

Margaritifera margaritifera, M. laevis, M. falcata, M. dahurica and M. middendorffi only use salmonids as host (Murphy, 1942; Karna & Milleman, 1978; Kobayashi & Kondo, 2005; Kondo & Kobayashi, 2005; Klishko & Bogan, 2013). However, the recent discovery of exclusively freshwater host fishes of the family Hiodontidae for Margaritifera monodonta (Sietman et al., 2017) and Esocidae for M. marrianae and M. hembeli (Fobian et al., 2017; P. Johnson, 2018, personal communication), calls into question this hypothesis, although the salinity tolerances of these fishes are unknown. This preference may have arisen because it was advantageous for mussel dispersion. Araujo et al. (2017) suggested that the current distribution of margaritiferids

275 can be explained by the capacity of their hosts to successfully bridge continents, or to circumvent hostile land areas that act as barriers for dispersal, by going through environments with reduced salinities. If freshwater mussels developed an affinity for euryhaline hosts during the first stages of their evolution, it is possible that both sturgeons and lampreys have since been used as hosts. Freshwater mussels are an ancient group that evolved from marine bivalves during the Triassic (Haas, 1969; Watters, 2001). Thus, if parasitism evolved during the colonization of freshwater environments, the ancestral unionoid likely used euryhaline hosts. According to a phylogeny of the order Unionoida, Margaritiferidae can be the basal family within the order (Strayer, 2008). Given this context, we hypothesize that present-day margaritiferids retained the trait of using euryhaline hosts from their unionoid ancestors.

4.3 Early stages of Margaritifera auricularia

Due to its critical conservation status, it seems that protecting the existing populations of this freshwater mussel is not sufficient to safeguard the species for the future, as most populations are too small and show too little recruitment to be self-sustainable. Thus, the French National Action Plan for this species (Prié et al., 2011), the Spanish Recuperation Plan (MARM, 2009) and the European Action plan for Pearl Mussels (Araujo & Ramos, 2001) independently come to the same conclusion that artificial reinforcement is needed, in order to foster populations of this European heritage species.

Between 2015 and 2017 we conducted the first experiences of captive breeding of juveniles of M. auricularia in France within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’. The aim of this work was to identify rearing methods that maximize the survival and growth rates of M. auricularia juveniles during their first months of life. Recently excysted juveniles obtained each year were collected and transferred to different rearing systems based on previous studies (Eybe et al., 2013; Araujo et al., 2015; Beaume et al., 2016): 1) an outdoor semi-natural flow-through system supplied by wild water, 2) an indoor recirculating water system and 3) static culture chambers. Several food sources were tested, including wild water, phytoplankton, detritus, animal protein and bacteria. In addition, the effect of the sediment was explored.

After infestation, the total number of juveniles released by the host fish for use in the different systems during the 2015-2017 experiments was 38,033. The release of juveniles from the host fish occurred between days 30 and 67 post-infection (PI) depending on the water temperature

276 (representing a minimum of 531 and a maximum of 1,071 degree days). Compared with the results of 2015 and 2017, in 2016 the excystment period was longer and delayed, and required more degree days likely due to the lower water temperatures during the encystment period. These data suggest that the velocity at which glochidia metamorphose is temperature dependent and it may be reduced with lower temperatures. This effect was previously noted by Araujo et al. (2002). In addition, Araujo et al. (2003) reported unsuccessful metamorphosis (glochidia were sloughed several days PI) when temperatures were 24ºC, suggesting that this process is interrupted with high temperatures. Thus, degree days calculation as the sum of daily average temperature, may not reflect these effect and should include a correction factor for temperature. However, further investigations are required for stablishing threshold temperature values at which metamorphosis is reduced or interrupted. This is important since a better understanding of the degree days required and the timing of the excystment period, is essential in order to minimize the laborious control for the presence of juveniles and the period of fasting of infested fish. Furthermore, this information may help to optimize and standardize artificial breeding of M. auricularia and to avoid the loss of juvenile mussels due to delayed collection (Hastie & Young, 2003; Taubert et al., 2013).

Of the rearing systems tested, in both the outdoor semi-natural flow-through system supplied by wild water and the indoor recirculating water system, used in 2015 and 2016 respectively, no juveniles were maintained alive for more than 1.5 months. The bad results obtained in 2015 could be partially due to the utilization of water and sediment that were not original from M. auricularia habitats. Nevertheless, in 2016 those elements were obtained from the same habitats where M. auricularia is found, and therefore there were other factors involved in the high mortality rates found in these rearing system. Given the critical status conservation of M. auricularia, this should be a priority since this approach allows producing large number of juveniles with a relatively low resources investment (Araujo et al., 2015; Beaume et al., 2016).

The survival rates of Margaritifera auricularia juveniles, reared in the static culture chambers system tested in 2017, were higher and allowed us to make further observations. After three days of excystment, most of the juveniles bred in the Bacteria treatment died, probably due to inadequate physico-chemical conditions of the culture medium. In all the treatments without sediment, a high mortality was observed in the first weeks of the experiment. These juveniles had organic adhesions in the shells, apparently yellow-brown bacterial growths, which hindered the correct opening and closing of the valves. However, in the treatment of Algae + Detritus, in which unlike the other treatments the boxes had a layer of substrate, this organic adhesions were

277 not observed, and the valves of the juveniles were completely clean. Substrate was added to all the boxes and it was observed that gradually the juveniles were cleaner, and the mortality stabilized. This result indicates that the presence of substrate in the crops seems to be of great importance as has also been reported in other studies (Gatenby et al., 1996; O’Beirn et al., 1998;

Jones et al., 2005; Nakamura et al., 2015).

The survival was greater in the Algae + Detritus treatment followed by the Detritus treatment, which seems to indicate that the detritus is of great importance for the maintenance of good physico-chemical conditions in the culture medium. The highest growth rates occurred in the treatments in which algae were used as a food source (Algae and Algae + Detritus treatments), suggesting that the algae provide a combination of nutrients suitable for the development of juveniles. However, it must kept in mind that in the Algae + Detritus treatment, sediments were added from the beginning of the experiments, and the juvenile mortalities were less important than in the other treatments where the sediment was added later. The largest individual at the end of the experiment was obtained in the combined treatment (Algae + Detritus), measuring 617 µm in length at 105 days of age. Nevertheless, in the Egg treatment, growth rates comparable to those reached in the treatments that included algae were obtained up to 75 days after the excystment. This suggests that the egg yolk is potentially a food resource with which juveniles can develop. However, all the specimens died around day 75, so it was not possible to make further comparisons at the end of the experiment. Finally, the combination of Algae + Detritus seems to be the best of the diets tested, since it maximizes the growth and survival of juveniles.

In the static culture chambers system tested in 2017 the maximal survival rate after 110 days was 34%. Although the survival rates were still low, these results seems to indicate that rearing M. auricularia juveniles in small boxes can yield better results. Nakamura et al. (2018) reported similar results, indicating that after a decade of unsuccessful experiences in rearing M. auricularia in Aragón (Spain), they were able to maintain juveniles alive beyond the first month when they stated to use small boxes as a culture system. They obtained about 50 juveniles kept alive from the cohorts of 2014, 2015, 2016 and 2017 (Nakamura, personal communication). A key factor for these promising results seems to be related to the production of large number of juveniles. Between 2014 and 2017, the mean number of M. auricularia juveniles used in the Aragon facility for captive breeding was ca 425,000 per year (Nakamura et al., 2018). Nevertheless, our results and those obtained in Aragón indicate that the survival rates using this

278 methodology are still too low for a sustainable, large-scale reintroduction of M. auricularia in the wild, and therefore further research is required to obtain better survival rates.

In order to improve breeding protocols and increase juvenile survival, it is important to improve our knowledge on early life ontogeny. Juveniles may have different requirements according to their feeding mode (Henley et al., 2001) and mortality may increase when developmental changes occur (Fitt et al., 1984; Beninger et al., 1994; Cannuel & Beninger, 2006) due to inability to meet energetic demands during morphogenesis (Veniot et al., 2003). The high mortality rates of juveniles of Margaritifera auricularia found in the first months of culture may be related to their ontogenetic development, as it seems to happen in other species of freshwater mussels. Based on a morphological analysis of newly formed juveniles of M. margaritifera (Margaritiferidae) and Unio mancus (Unionidae), we propose that a second metamorphosis, consisting of drastic morphological changes occurs, leading to suspension feeding in place of deposit feeding by the ciliated foot (see section 3.3.1 of this thesis). The suspension feeding in these two species improves due to a gradual development of several morphological features, like the contact between cilia of the inner gill posterior filaments, the inner gill reflection, the appearance of the ctenidial ventral groove and the formation of the labial palps. Regardless of the presence of available food, a suspension feeding mode replaces deposit feeding, and juveniles unable to successfully transition morphologically or adapt to the feeding changes likely perish. The transition from pedal feeding to filter feeding occurs around 150-200 days post-emergence in M. margaritifera and around 70 days in U. mancus, after juveniles are greater than 1 mm in length, which coincides with the timing of high mortality (see section 3.3.1 of this thesis). Once this feeding metamorphosis is complete, juvenile mortality decreases. Of course, these morphological changes do not occur at the same time in all freshwater mussels (Patterson et al., 2018) but we suppose a similar pattern in M. auricularia.

Nevertheless, although it is advantageous to reach the adult stage as soon as possible in order to avoid high mortality rates during early stages, species with small glochidia tend to metamorphose into small young mussels, which take a long time to grow up (Bauer, 1994). Selection for large young mussels should be particularly intense in habitats where juvenile growth rates are low (Sibly & Calow, 1986), as is the case of M. margaritifera, which inhabits in oligotrophic streams (Bauer et al., 1992). Based on the period required for the newborn juveniles to attain the first millimetre, the morphological changes required for the transition from pedal to filter feeding may be slower in M. auricularia than in the other two species

279 studied. The newly excysted juvenile of M. auricularia barely attain 200 µm in length, and according to the available data (Nakamura et al., 2015, 2018), it does not reach the first millimetre until approximately 185 days of age. Although complete studies are not yet available, the results reported in Nakamura et al. (2015) suggest that mortality in captive-bred M. auricularia stabilized after 175-190 days, which seems to confirm our hypothesis. In contrast, the juveniles of M. margaritifera, with an initial size of ca 350 µm, reach the first millimetre at around 100 days old. On the other hand, recently excysted juveniles of Unio mancus measure 260 µm in length and reach the first millimetre with approximately 50 days of age (Figure 1).

If this relationship between the size and the development of morphological structures related to the mode of feeding is fulfilled, the slower development of M. auricularia during the first growth phases could explain in part the apparent great difficulty involved in artificially breeding this species (see below).

3,5

3

2,5

2

1,5 Length(mm) 1

0,5

0 0 100 200 300 400 Age (days) M. margaritifera U. mancus M. auricularia

Figure 20. Mean length (mm) of juvenile freshwater mussel species reared with commercial algae at specified days. Data source: M. margaritifera and U. mancus (Araujo et al., 2017); M. auricularia (first 210 days from Nakamura et al., 2015; rest from Nakamura et al., 2018).

It has to be noted that the same system of channels with natural water and sediment tested to breed juveniles of M. auricularia in the Banyoles laboratory (Spain) in 2015, has been successfully used in rearing Unio mancus and U. ravoisieri, with survival rates up to 5% in 1 year (M. Campos, personal communication). On the other hand, the system tested in 2016 was

280 inspired on the successful results in rearing M. margaritifera juveniles obtained by the LIFE project “Conservation de la moule perlière d’eau douce du Massif armoricain” in France (Beaume et al., 2016). Nevertheless, each mussel species may have different culture requirements, so a culture system that works well for one species may not work for another (Mair, 2018). This is also exemplified by the recent experiences conducted in May 2018 at the Banyoles propagation facility. Recently excysted M. auricularia juveniles obtained in the Aragón propagation facility, were cultured in a raceway upweller system, which has been successfully used for rearing U. mancus juveniles. Despite the 90% survival rates obtained for U. mancus after 100 days, 100% of the juveniles of M. auricularia died after 45 days of culture (M. Campos, personal communication).

On the other hand, the great mortality of juveniles of M. auricularia observed in the static culture chambers (detritus boxes), may also be related to the particularities imposed on this system by the small size of the juveniles. The smaller starting size of the juveniles of M. auricularia, and the longer time required to grow up to the first millimetre, implies that finer mesh sizes have to be used when cleaning the rearing systems (detritus boxes) during periodic water exchanges. These meshes are more susceptible to clogging with fine particles both included in the water and originated in the culture systems (food remains, bacterial growths, etc.). Interstitial spaces can become clogged by fine particles (Buddensiek, 1995; Brim et al., 1999; Geist & Auerswald, 2007) and anoxic, which may cause juvenile mortalities (Dimock & Wright, 1993). Nevertheless, further research is required in order to test these hypotheses.

4.4. Implications for conservation

4.4.1 Threats

Multiple factors have been identified as responsible for the decline of Margaritifera auricularia, including habitat destruction, commercial overharvesting, water and sediment pollution, and host fish loss due to exotic fishes and changes of the natural flow and flood regime by dam construction (Altaba, 1990; Álvarez-Halcón, 1998; Araujo & Ramos, 2000a; Gomez & Araujo, 2005; López et al., 2007; Prié et al., 2011; Araujo & Álvarez-Cobelas, 2016).

4.4.1.1 Host fish availability

The overall decline during historic times of Margaritifera auricularia has been related to the disappearance of Acipenser sturio (Altaba, 1990; Araujo & Ramos, 2000a; López et al., 2007).

281 Two facts support this hypothesis: the historic distribution of A. sturio matches with the original geographic range of M. auricularia (Araujo & Ramos 2000a, c) and both species have suffered dramatic population declines in the late 19th and early 20th centuries.

Our findings (see Section 3.1.1 of this thesis) suggest that M. auricularia is recently extinct in the Rhône and Seine basins, but persists in the Ebro, Loire, Charente, Garonne and Adour basins. The extirpation of M. auricularia in the Rhône and Seine basins may be related to the disappearance of Acipenser sturio and Petromyzon marinus during the 19th and 20th centuries (Allardi & Keith, 1991, and references therein; Belliard et al., 1995). The persistence of P. marinus in the Charente, Loire and Adour basins, may account for the continued presence of M. auricularia populations in those basins, even after the local extinction of A. sturio. In the Ebro basin, P. marinus was likely abundant until the construction of big dams in the first half of the 20th century (López et al., 2007; Mateus et al., 2012), suggesting that together with Salaria fluviatilis, they could have played an important role in maintaining the Spanish populations of M. auricularia.

Host abundance may also modulate freshwater mussel abundance (Haag & Warren, 1998; Mulcrone, 2004; Stoeckl et al., 2015). Although further research is required, the lack of recruitment during the last decades, and its recently weak recovery observed in several French M. auricularia populations (see Sections 3.1.1 and 3.1.2. of this thesis), seems to be linked to P. marinus population trends in Europe. Based on data from the Garonne River and other three European rivers, Beaulaton et al. (2008) reported that the abundance of European P. marinus populations experienced a 60% decline from 1973 to the end of the 1990s, followed by an increase, leading to levels similar to those of the 1960s by the beginning of the 2000s. This trend seems to match with our observations on the structure and age estimations of the M. auricularia Vienne-Creuse population, which suggested that after a period without effective reproduction, an important recruitment event occurred between the late 1990s and the beginning of 2000’s. Furthermore, the reduced recruitment in the Charente population could be related to the low abundance of P. marinus, which is 10 times lower than in the Vienne and Creuse rivers (Bach et al., 2016; Dartiguelongue, 2017).

Although other causes may have contributed to the overall decline of M. auricularia, the loss of hosts seems to be a main factor. Based on the “invasional meltdown” hypothesis proposed by Simberloff & Von Holle (1999), we introduced the term “native species meltdown” to describe the decline of one species caused by the decline of other species on which it is

282 dependent by any kind of symbiosis, including mutualism, parasitism, or commensalism. We hypothesize that the giant freshwater pearl mussel M. auricularia is experiencing a native species meltdown due to the local extinction of its primary hosts A. sturio and P. marinus in Europe, and the decline of S. fluviatilis in the Ebro basin.

4.4.1.2 Water Quality

Water pollution has been evoked as a potential cause for the overall decline of M. auricularia (Prié et al., 2011) despite the lack of data to corroborate it. Nevertheless, in our study of the Charente and Vienne-Creuse populations, we observed that the highest densities of mussels were found in urban areas or near wastewater discharges, suggesting that M. auricularia seems capable of tolerating a certain degree of water pollution, which is in accordance with previous observations by Prié (2010). On the other hand, the water quality data of the studied rivers indicate that M. auricularia can tolerate moderate levels of nutrients as reflected by the mean values of sulphate and nitrate concentrations in the Charente River. This is corroborated by the data that Araujo & Ramos (2000b) reported for the Canal Imperial (Ebro basin, Spain), where concentrations of up to 373 mg/l of SO4 were recorded during low waters. Furthermore, our observation highlighted the wide tolerance of M. auricularia to different calcium concentrations and water conductivity values. Araujo & Ramos (2000b) defined the species as a hardwater species living in subsaline waters, since the calcium levels in the Canal Imperial ranged between 114 and 163 mg/l, and the conductivity varied between 1,077 and 1,547 µS/cm. However, the Vienne and Creuse data, where the average calcium concentrations are 37.2 and 41.4 mg/l, with conductivity values of 304.2 and 287.3 µS/cm, respectively, indicate that M. auricularia can inhabit at a wide range of calcium concentrations and salinities.

Nevertheless, further ecotoxicological studies are required in order to investigate tolerance limits of M. auricularia to different pollutant concentrations and, especially, in their early stages of life.

4.4.1.3 Invasive species

Introduction of invasive species is a relatively recent threat to freshwater mussels, but has the potential to become one of the major concerns for their conservation. The effects of invasive bivalve species such as the zebra mussel, Dreissena polymorpha (Pallas, 1771), and the Asian clam, Corbicula fluminea (Müller, 1774) on Margaritifera auricularia populations have not been well studied, but it could be of major concern given the filtering capacities of those small

283 bivalves (Bogan, 1993; Parker et al., 1998; Ricciardi et al., 1998; Neves, 1999; Yeager et al., 1999; Burlakova et al., 2000; Lydeard et al., 2004; Strayer, 2006). In this sense, Haag (2018) suggested that the effects of the colonization of C. fluminea could be responsible for the recent enigmatic declines of freshwater mussels in North America.

Corbicula fluminea seems to have colonized virtually all habitats where M. auricularia still occurs. In the Vienne River, it first appeared in the late 90's, and since then it has spread throughout the basin and has reached high densities, which have been related to a significant decrease in phytoplankton (Hesse et al., 2015). In the Ebro River it first appeared in 1993 (Alvarez-Halcón, pers. comm.), although in the sections where M. auricularia lives, it seems that the colonization occurred in 2011 (Nakamura, pers. comm.). Future investigations are necessary to establish the impact of C. fluminea on the populations of M. auricularia.

Introduced fish species may also be an important threat to freshwater mussels by significantly contributing to the disappearance of native fish faunas, which are essential as hosts for unionoids. Furthermore, we have identified a new potential threat to M. auricularia, derived from the recent colonization of their habitats by an exotic fish species. The presence of the European bitterling, Rhodeus amarus (Bloch, 1782), was detected during electrofishing surveys that were carried out to identify host fish of M. auricularia in the Vienne, Creuse and Charente rivers. This cyprinid fish is known for its unusual life cycle characterized by its obligatory symbiosis of spawning in freshwater mussels (Smith et al., 2004).

The presence of bitterling embryos in mussels causes a fitness cost for the hosts: they compete with the host for oxygen (Smith et al., 2001) and reduce water circulation over the mussel’s gills, potentially affecting their ability to filter feed (Mills et al., 2005). These effects may account for the significantly reduced growth rates of mussels infected with bitterling embryos (Reichard et al., 2006). In freshwater mussels, size is positively correlated with fecundity (Bauer, 1994); therefore, a reduction in growth represents a fitness cost for mussels.

Although Rhodeus amarus is considered a native species over much of its present-day range in Europe, recent studies have demonstrated that it expanded relatively recently (centuries to millennia before the present day) from the Black Sea region into central and western Europe (Bohlen et al., 2006, Van Damme et al., 2007, Bryja et al., 2010). During its expansion, R. amarus successfully used mussel species present in its historic range as hosts such as Anodonta anatina (Linnaeus, 1758), A. cygnea (Linnaeus, 1758), Pseudanodonta complanata

284 (Rossmässler, 1835), Unio pictorum (Linnaeus, 1758), U. tumidus Retzius, 1788 (Wiepkema, 1961, Balon, 1962, Reynolds et al., 1997; Smith et al., 2000; Smith et al., 2004) and U. crassus Philipsson, 1788 (Reichard et al., 2010; Tatoj et al., 2017). Compared with unionid populations in the Pontic region, native mussel populations across continental Europe have limited adaptations to avoid or eject R. amarus eggs, probably due to a shorter duration of sympatry (Reichard et al., 2010).

Following its dramatic decline in abundance from 1960 to 1980, R. amarus was declared an endangered species in central and western Europe and included in Appendix II of the Habitat Directive (92/43/EEC). Nevertheless, since 1980, the distribution of R. amarus has expanded in many parts of Europe, particularly in eastern Europe where a considerable and rapid increase in abundance has also been observed (Kozhara et al., 2007). In France, this species was restricted to only the northeastern basins during the 19th century (Valenciennes, 1848; Gehin, 1868; Gensoul, 1908); however, nowadays it has spread nearly throughout the entire country, especially in the southwest, where it is considered invasive (Kottelat & Freyhof, 2007). Two explanations have been proposed for the recent expansion of R. amarus: anthropogenic activities, including its introduction by anglers, aquarists and artificial connections of waterway systems, and climate change (Kozhara et al., 2007, Van Damme et al., 2007).

The recent expansion of R. amarus into new habitats and new geographical areas has led to its contact with different freshwater mussel species. Following its recent expansion through Western Europe, R. amarus was observed to ovoposit in Unio mancus Lamarck, 1819 and Potomida littoralis (Cuvier, 1798) in French rivers (Prié, 2017). The extensive expansion of this bitterling species, and the reduction in fitness that it could cause, may be problematic for mussel species at risk of extinction as in the case of M. auricularia.

Field and laboratory observations permitted to verify for the first time that R. amarus uses M. auricularia as a host of their eggs and embryos. Given that the gill anatomy of margaritiferids is different from that of unionids (see Section 3.2.1 of this thesis), this finding supports the idea that European bitterlings can parasitize all native European mussel species regardless of their gill anatomy.

In central European rivers, native mussels have coexisted with bitterlings for hundreds to thousands of years without any apparent decline in populations due to the presence of bitterlings. However, European freshwater mussels are currently in decline, and 12 of the 16

285 currently recognized species are categorized as Threatened or Near Threatened by the IUCN (Lopes-Lima et al., 2017). Mussel populations that have experienced a drastic decline in density may be particularly impacted by the presence of R. amarus, as individual mussels could ultimately host a greater number of bitterling embryos, which may represent an additional stress (Van Damme et al., 2007, Prié, 2017, Tatoj et al., 2017).

4.4.2. Identified hosts as a tool for conservation

All confirmed host species of Margaritifera auricularia appear to be good physiological hosts except Gasterosteus aculeatus (see Section 3.2.2 of this thesis). However, G. aculeatus is the only host that has been found infested with M. auricularia glochidia in the wild (see Section 3.2.2 of this thesis). Information on the geographical distribution, habitat requirements, phenology and behaviour of the physiological hosts suggests that most of them could also be considered ecological hosts, except Silurus glanis and the exotic sturgeons Acipenser baeri and A. ruthenus.

Silurus glanis is an invasive species, and although it has a wide dietary spectrum, it has been reported to prey on anadromous species, including Petromyzon marinus, Salmo salar Linnaeus, 1758 and Alosa spp., in some rivers like the Garonne (Boulêtreau et al., 2018; Cucherousset et al., 2018; Guillerault et al., 2017; Syväranta et al., 2009). Its presence in regulated rivers may also be contributing to significant changes in the native fish fauna (Gavioli et al., 2018), particularly on the threatened Salaria fluviatilis whose decline in Spain has been associated with predation by exotic fishes (Carol, 2007; Doadrio, 2001). However, the abundance and wide distribution of S. glanis in the vicinity of M. auricularia populations make it an easy fish to obtain. Although not conclusive, the findings presented here (see Section 3.2.3 of this thesis), suggest that this species has a high capacity to transform large quantities of glochidia. If so, this fish could be a good alternative to Acipenser baeri for obtaining juveniles in the laboratory.

Given the difficulty of reintroducing the endangered Acipenser sturio, only Salaria fluviatilis, Gasterosteus aculeatus and Petromyzon marinus can be used in the application of a simplified, cost effective technique to boost Margaritifera auricularia populations in the wild. This technique, which has been successfully used for other species (Altmüller & Dettmer, 2006; Araujo et al., 2015; Carey et al., 2015), consists of catching host species in the vicinity of existing mussel populations, infesting them with glochidia, and then immediately releasing them back into the wild in order to optimize the window of recruitment of this European

286 heritage species. In France and Spain, P. marinus is considered threatened and, in some areas, has disappeared or become increasingly rare mainly due to habitat loss related to dam construction, habitat disruption and overfishing (Hansen et al., 2016; Mateus et al., 2012). This type of measure would favour the reinforcement of not only M. auricularia populations, but also those of P. marinus by expanding their spawning habitats. In addition, Limm & Power (2011) reported that Pacific lamprey (Petromyzon tridentatus Richardson, 1836) larvae grow faster when found near Margaritifera falcata (Goulds, 1850) beds, where mussels capture, concentrate, and deposit food near their burrows. A similar potential mutualistic relationship between M. auricularia and P. marinus should be investigated in future studies.

287

288

Conclusions

289 290 5. CONCLUSIONS

The general goal of this thesis was to improve the knowledge on the conservation status of the Margaritifera auricularia French populations, the reproductive biology and early life stages of the species, allowing to better understand the degree and causes of its decline and to propose conservation measures. To do so, four aspects were focused: the evolution of the conservation status of the species, as inferred by its currently known populations, bibliographic references and museum collections; the brooding process, including the brood size and the glochidia release behaviour and timing; the parasitic stage of the mussels and the identification of new host species, and the early stages of freshwater mussels and the artificial rearing of M. auricularia. The main findings of this study were as follows:

1) On historical times, M. auricularia was present in the Rhine in France and Germany, the Seine and the Rhône in France, the Po in Italy and the Tagus in Spain, where the species is now believed to be extinct. Today, Margaritifera auricularia is considered restricted to five watersheds in France and Spain (Loire, Charente, Garonne, Adour and Ebro), representing a range contraction of about a 90% in the last two centuries. Recruitment is very scarce in all populations although evidences of an important recent recruitment were found in the Vienne and Creuse rivers, constituting the most notable event of M. auricularia recent reproduction reported so far. 2) The current lack of recruitment does not seem to be related to the infertility of the specimens. Ova and developing embryos were found throughout the month of March in the Charente and Creuse populations, and the beginning of glochidial release occurred in early to mid-April. Compared with the Spanish populations, the reproductive period begins some weeks later, likely due to differences in water temperature. Development from first cleavage to glochidial maturity took 25-37 days and the estimated number of larvae per gravid mussel was around 2,000,000. 3) The discovery of three new host species (Gasterosteus aculeatus, Petromyzon marinus and the exotic Silurus glanis) increases the number of known physiological host to eight species from five different families, indicating that the mussel has a more host generalist behavior than previously expected. This is the first time that a lamprey has been tested and used as a possible host for a freshwater bivalve belonging to the Unionoida Order. The potential use of Petromyzon marinus to reinforce M. auricularia populations along

291 its entire geographic range may counteract the decline of the species. All currently known Margaritifera auricularia hosts have a notable tolerance to salinity. This pattern could have arisen early in its evolution because of the dispersion advantage that diadromous species potentially confer. However, anthropogenic causes, such as dam construction and degradation of water quality and habitat, have led to a severe decline of most migratory fish species in Europe during the last two centuries. Although other causes may have contributed to the overall decline of Margaritifera auricularia, the loss of hosts seems to be a main factor. 4) The utilization of Margaritifera auricularia by Rhodeus amarus, a cyprinid fish that spawns in the mantle cavity of freshwater mussels resulting in a fitness cost for the mussels, was discovered. The extensive expansion of this bitterling species, and the reduction in fitness that it could cause on mussels, may be problematic for species facing extinction risk, especially in areas inhabited by spatially restricted mussel species. 5) The juvenile breeding of Margaritifera auricularia is a difficult task and although the survival results are still very low, it seems that the use of an initial laboratory phase with the newborn juveniles reared in detritus boxes can offer more successful results than the direct cultivation of the juveniles into raceways. As it seems to happen in other species of freshwater mussels, the high mortality rates of juveniles of Margaritifera auricularia found in the first months of culture may be related to their ontogenetic development, characterized by the initial small size of the juveniles and the slow growth until reaching the first millimeter.

Nowadays Margaritifera auricularia has a restricted distribution and their populations are ageing, showing scarce recruitment and evidences of severe decline. Of particular concern is its situation in the Canal Imperial de Aragón (Ebro basin, Spain), where mass mortality events have been observed since 2013, causing the death of more than 35% of the specimens. The causes of these mortalities are unknown, so future research is required to clarify them and thus be able to have tools to conserve the last Mediterranean population of the species and prevent similar situations in French populations.

It seems that the decline during historic times of Margaritifera auricularia is related to the extirpation and local trends of its so far identified primary hosts: Acipenser sturio and Petromyzon marinus. The sea lamprey has likely played a key role in preventing the total extinction of Margaritifera auricularia in some basins following the disappearance of the European sturgeon. Therefore, the potential use of Petromyzon marinus to reinforce M.

292 auricularia populations along its entire geographic range may counteract this native species’ meltdown.

The characteristics of the parasitic stage of Margaritifera auricularia, especially the ease of obtaining large numbers of juveniles in captivity in a fast and efficient way, makes this an adequate strategy to complement the reinforcement of the remaining population. However, given the difficulty of keeping juveniles alive in the laboratory, the reintroduction of juveniles raised to a size of several centimeters is not yet viable and future efforts are required to improve this technique.

The present study contributed to improve the knowledge on the conservation status, reproductive biology and early life stages of Margaritifera auricularia and to identify some key biological factors important for its future conservation. However, more research is required to complement some of the results presented here:

 The number of glochidia per gravid individual was calculated with 3 specimens. It would be interesting to study more specimens and see if there are differences between populations.  It would be necessary to study the evolution of the tetrageny in the phylogeny of the family Margaritiferidae.  With regard to host fish, a detailed study should be carried out comparing the suitability of each of the identified hosts. It would also be necessary to test other species that have been identified as potentially host, such as Alosa alosa, Alosa fallax, Anguilla anguilla, Lampetra fluviatilis, Lampetra planeri, Cottus gobio and Gambusia holbrooki.  Further research is required to improve the breeding techniques and obtain better survival rates of juveniles of M. auricularia for its large-scale reintroduction into the wild.

Since the successful reproduction of freshwater mussels depends on the completion of their whole life cycle, it is fundamental to consider the threats and problems potentially affecting all the stages of the life cycle. Therefore, beyond the framework of this thesis, further research is required in aspects such as habitat conditions, food web ecology and ecotoxicology in order to evaluate bottlenecks for Margaritifera auricularia populations. Likewise, conservation genetics is an important field for analyzing the mussel genetic diversity needed to preserve rare alleles, and to avoid interferences with evolutionary adaptations in mussel artificial rearing programs.

293 294 6. CONCLUSIONES

El objetivo general de esta tesis ha sido mejorar el conocimiento sobre el estado de conservación, la biología reproductiva y las primeras fases de vida de Margaritifera auricularia, lo que puede permitir comprender mejor el grado y las causas de su declive y proponer medidas de conservación. Para ello, me centré en cuatro aspectos: la evolución del estado de conservación de la especie, inferida a través del estudio las poblaciones actualmente conocidas, referencias bibliográficas y colecciones de museos; el proceso de incubación, incluyendo el número de larvas incubadas y el comportamiento y periodo de liberación de gloquidios; la fase parasitaria de esta especie y la identificación de nuevas especies hospedadoras, y las primeras fases de vida de las almejas de agua dulce y la cría artificial de M. auricularia. Los principales hallazgos de este estudio fueron los siguientes: 1) En épocas históricas, Margaritifera auricularia estuvo presente en el Rhin en Francia y Alemania, el Sena y el Ródano en Francia, el Po en Italia y el Tajo en España, donde se cree que la especie está actualmente extinguida. Hoy en día, M. auricularia se considera restringida a cinco cuencas hidrográficas en Francia y España (Loira, Charente, Garona, Adur y Ebro), lo que representa una contracción de su distribución de aproximadamente un 90% en los dos últimos siglos. El reclutamiento es muy escaso en todas las poblaciones, aunque durante el presente estudio se encontraron evidencias de un importante reclutamiento reciente en los ríos Vienne y Creuse, lo que constituye el evento de reproducción reciente de M. auricularia más notable conocido hasta la fecha. 2) La actual falta de reclutamiento no parece estar relacionada con la infertilidad de los especímenes. Se encontraron óvulos y embriones en desarrollo durante todo el mes de marzo en las poblaciones de Charente y Creuse y el inicio de la liberación de gloquidios ocurrió de principios a mediados de abril. Comparado con las poblaciones españolas, el período reproductivo comienza unas semanas más tarde, probablemente debido a las diferencias en la temperatura del agua. El desarrollo desde la primera división embrionaria hasta la madurez de los gloquidios duró entre 25 y 37 días y el número estimado de larvas por ejemplar grávido fue de alrededor de 2.000.000. 3) El descubrimiento de tres nuevas especies de hospedadores (Gasterosteus aculeatus, Petromyzon marinus y el exótico Silurus glanis) aumenta el número de hospedadores fisiológicos conocidos a ocho especies de cinco familias diferentes, lo que indica que M.

295 auricularia tiene un comportamiento más generalista de lo que se esperaba. Es la primera vez que se experimenta y se utiliza una lamprea como posible hospedador de un bivalvo de agua dulce del Orden Unionoida. El uso potencial de Petromyzon marinus para reforzar las poblaciones de M. auricularia en todo su rango geográfico puede ayudar a contrarrestar el declive de la especie. Todos los hospedadores de Margaritifera auricularia conocidos actualmente tienen una notable tolerancia a la salinidad. Esta relación podría haber surgido gracias a la ventaja potencial que confieren las especies diádromas para la dispersión de los moluscos, lo que podría haberse mantenido en esta especie como un constraint evolutivo. Sin embargo, nuevas amenazas, como la construcción de presas y la degradación de la calidad del agua y el hábitat, han provocado un grave declive de la mayoría de las especies de peces migradores en Europa durante los últimos dos siglos. Aunque el declive general de M. auricularia puede deberse a numerosas causas, la pérdida de especies hospedadoras parece haber sido un factor principal. 4) En esta tesis también se descubrió la utilización de M. auricularia por Rhodeus amarus, un pez ciprínido que pone sus huevos en la cavidad del manto de las náyades provocando un coste en fitness para las náyades. La gran expansión de R. amarus y la reducción de fitness que podría causar en las almejas de agua dulce, puede poner en peligro la supervivencia de algunas especies en riesgo de extinción, especialmente las que tienen una distribución más restringida. 5) La cría en cautividad de juveniles de Margaritifera auricularia es una tarea difícil y, aunque los resultados de supervivencia son todavía muy bajos, parece que el uso de una fase inicial de laboratorio donde los juveniles recién nacidos se críen en cajas de detritus, puede ofrecer resultados más exitosos que el cultivo directo de juveniles en diferentes tipos de canales de cultivo. Como parece ocurrir en otras especies de almejas de agua dulce, las altas tasas de mortalidad de los juveniles de M. auricularia encontradas en los primeros meses de cultivo pueden estar relacionadas con su desarrollo ontogenético, caracterizado por el pequeño tamaño inicial de los juveniles y su lento crecimiento hasta alcanzar el primer milímetro.

En la actualidad Margaritifera auricularia tiene una distribución restringida y sus envejecidas poblaciones muestran un reclutamiento escaso y evidencias de un grave declive. Su situación en el Canal Imperial de Aragón (cuenca del Ebro, España) es especialmente preocupante, donde desde 2013 se vienen observando eventos de mortalidad masiva que han causado la muerte de más del 35% de los ejemplares. Las causas de estas mortalidades son actualmente desconocidas, por lo que se requieren investigaciones futuras para aclararlas y así poder tener herramientas

296 para conservar la última población mediterránea de la especie y prevenir situaciones similares en las poblaciones francesas. Parece que el declive en tiempos históricos de Margaritifera auricularia está relacionado con las extinciones y tendencias locales de sus principales hospedadores identificados hasta ahora: Acipenser sturio y Petromyzon marinus. En algunas cuencas, la lamprea marina probablemente ha desempeñado un papel clave en prevenir la extinción total de Margaritifera auricularia después de la desaparición del esturión europeo. Por lo tanto, el uso potencial de Petromyzon marinus para reforzar las poblaciones de M. auricularia a lo largo de todo su rango geográfico puede ayudar a contrarrestar el declive esta especie. Las características de la fase parasitaria de Margaritifera auricularia, especialmente la facilidad de obtener grandes cantidades de juveniles en cautividad de manera rápida y eficiente, hacen de esta una estrategia adecuada para complementar el refuerzo de las poblaciones que todavía subsisten en la actualidad. Sin embargo, dada la dificultad de mantener vivos a los juveniles en el laboratorio, la reintroducción de juveniles criados hasta un tamaño de varios centímetros todavía no es viable y se requieren esfuerzos futuros para mejorar esta técnica. El presente estudio contribuye a mejorar el conocimiento sobre el estado de conservación, la biología reproductiva y las primeras fases de vida de Margaritifera auricularia y a identificar algunos factores biológicos importantes para su conservación. Sin embargo, se requiere más investigación para complementar algunos de los resultados aquí presentados:  El número de gloquidios por individuo grávido se calculó con 3 ejemplares. Sería interesante estudiar más especímenes y ver si hay diferencias entre las distintas poblaciones.  Sería necesario estudiar la evolución de la tetragenia en la filogenia de la familia Margaritiferidae.  Con respecto a los peces hospedadores, se debería realizar un estudio detallado comparando la idoneidad de cada uno de los hospedadores identificados. También sería necesario probar otras especies que han sido identificadas como potencialmente hospedadoras, como Alosa alosa, Alosa fallax, Anguilla anguilla, Lampetra fluviatilis, Lampetra planeri, Cottus gobio y Gambusia holbrooki.  Se requiere seguir investigando en la mejora de las técnicas de reproducción artificial y obtener mejores tasas de supervivencia de los juveniles de M. auricularia que permitan su reintroducción a gran escala en la naturaleza.

297 Dado que el éxito de la reproducción de las náyades depende de completar todo su ciclo de vida, es fundamental considerar las amenazas y los problemas que pueden afectar a todas las etapas de dicho ciclo. Por lo tanto, más allá del marco de esta tesis, se requieren investigaciones adicionales en aspectos como las condiciones del hábitat, la ecología de las redes tróficas y la ecotoxicología para evaluar los cuellos de botella para las poblaciones de Margaritifera auricularia. Del mismo modo, la genética de la conservación es un campo importante para analizar la diversidad genética de las náyades necesaria para poder preservar alelos raros y para evitar posibles pérdidas de adaptaciones evolutivas en los programas de cría artificial.

298 7. BIBLIOGRAPHY

ALDRIDGE, D. C. 1999. The morphology, growth and reproduction of Unionidae (Bivalvia) in a fenland waterway. Journal of Molluscan Studies, 65: 47–60.

ALDRIDGE, D. C. 2000. The impacts of dredging and weed cutting on a population of freshwater mussels (Bivalvia: Unionidae). Biological Conservation, 95: 247–257.

ALDRIDGE, D. C. & McIVOR, A. 2003. Gill evacuation and release of glochidia by Unio pictorum and Unio tumidus (Bivalvia: Unionidae) under thermal and hypoxic stress. Journal of Molluscan Studies, 69: 55–59.

ALLARDI, J. & KEITH, P. 1991. Atlas préliminaire des poissons d’eau douce de France. Patrimoines naturels; 4. Muséum national d'Histoire naturelle, Paris.

ALTABA, C. R. 1990. The last known population of the freshwater mussel Margaritifera auricularia (Bivalvia, Unionoida): A conservation priority. Biological Conservation, 52: 271– 286.

ALTABA, C. R. 1997. Al limit de l'extincio: Margaritifera auricularia (Bivalvia: Unionidae). Bulleti de la Institucio Catalana d'Historia Natural, 65: 137–148.

ALTABA, C. R. 2001. Demografía, hábitats y ciclo vital de Margaritifera auricularia (Mollusca: Unionidae) en el curso inferior del Ebro. p. e. m. d. M. A. IMEDEA, Ministerio de Medio Ambiente, Madrid.

ALTABA, C. R. & LÓPEZ, M. A. 2001. Experimental demonstration of viability for the endangered Giant Pearl Mussel Margaritifera auricularia (Bivalvia: Unionoida) in its natural habitat. Bolleti de la Societat d’Historia Natural de les Baleares, 44: 15–21.

ALTMÜLLER, R. & DETTMER, R. 2006. Erfolgreiche Artenschutzmaßnahmen für die Flussperlmuschel Margaritifera margaritifera L. durch Reduzierung von unnatürlichen Feinsedimentfrachten - Erfahrungen im Rahmen des Lutterprojekts. Informationsdienst Naturschutz Niedersachsen, 26: 192–204.

299 ÁLVAREZ HALCÓN, R. M. 1998. La industria del nácar de Margaritifera auricularia en Aragón y la gestión ambiental. Temas de Antropología Aragonesa, 8: 113–212.

AMREIN, S. 2004. Carte de répartition de la lamproie marine sur le bassin de l’Adour Tranche 2/3: Adour et affluents rive gauche. MIGRADOUR, Pau.

AMREIN, S. 2005. Carte de répartition de la lamproie marine sur le bassin de l’Adour Tranche 3/3: Adour et affluents rive droite. MIGRADOUR, Pau.

ANDRÉS, D. & ORTUÑO, A. 2014. Margaritifera auricularia (Spengler, 1793): primeras evidencias sobre el antiguo cuaternario de Yecla (Murcia). Athene, 24: 21–37.

ANSELIN, L. 1996. The Moran Scatterplot as an ESDA Tool to Assess Local Instability in Spatial Association. In: Spatial Analytical Perspectives on GIS in Environmental and Socio- Economic Sciences (M. Fischer, H. Scholten & D. Unwin, eds.), pp. 111–125. Taylor & Francis, London.

ANSELIN, L., SYABRI, I. & KHO, Y. 2006. GeoDa: An introduction to spatial data analysis. Geographical Analysis, 38: 5–22.

ANTHONY, J. L., KESLER, D. H., DOWNING, W. L. & DOWNING, J. A. 2001. Length- specific growth rates in freshwater mussels (Bivalvia: Unionidae): extreme longevity or generalized growth cessation? Freshwater Biology, 46: 1349–1359.

ARAUJO, R. 2006. Margaritifera auricularia (Spengler, 1793). In: Libro rojo de los invertebrados de España (Verdú & Galante, eds.), pp. 304–306. Dirección general para la Biodiversidad; Ministerio de Medio Ambiente, Madrid.

ARAUJO, R. 2012. Margaritifera auricularia. In: Bases ecológicas preliminares para la conservación de las especies de interés comunitario en España: Invertebrados. Ministerio de Agricultura, Alimentación y Medio Ambiente, Madrid.

ARAUJO, R. & ÁLVAREZ-COBELAS, M. 2016. Influence of flow diversions on Giant Freshwater Pearl Mussel Populations in the Ebro River, Spain. Aquatic Conservation: Marine and Freshwater Ecosystems, 26: 1145–1154.

ARAUJO, R., BRAGADO, D. & RAMOS, M. A. 2000. Occurence of glochidia of the endangered Margaritifera auricularia (Spengler, 1793) and other mussel species (Bivalvia:

300 Unionoidea) in drift and on fishes in an ancient channel of the Ebro River, Spain. Archiv für Hydrobiologie, 148: 147–160.

ARAUJO, R., BRAGADO, D. & RAMOS, M. A. 2001. Identification of the river blenny, Salaria fluviatilis, as a host to the glochidia of Margaritifera auricularia. Journal of Molluscan Studies, 67: 128–129.

ARAUJO, R., BUCKLEY, D., NAGEL, K.-O. & MACHORDOM, A. 2016. Potomida littoralis (Bivalvia, Unionidae) evolutionary history: slow evolution or recent speciation? Zoological Journal of the Linnean Society DOI: 10.1111/zoj.12470.

ARAUJO, R., CÁMARA, N. & RAMOS, M. A. 2002. Glochidium metamorphosis in the endangered freshwater mussel Margaritifera auricularia (Spengler, 1793): A histological and scanning electron microscopy study. Journal of Morphology, 254: 259–265.

ARAUJO, R., CAMPOS, M., FEO, C., VARELA, C., SOLER, J. & ONDINA, P. 2018. Who wins in the weaning process? Juvenile feeding morphology of two freshwater mussel species. Journal of Morphology, 279: 4–16.

ARAUJO, R., FEO, C., POU, Q. & CAMPOS, M. 2015. Conservation of two endangered European freshwater mussels (Bivalvia: Unionidae): A three-year, semi-natural breeding experiment. The Nautilus, 129: 126–135.

ARAUJO, R., GÓMEZ, I. & MACHORDOM, A. 2005. The identity and biology of Unio mancus Lamarck, 1819 (= U. elongatulus) (Bivalvia: Unionidae) in the Iberian Peninsula. Journal of Molluscan Studies, 71: 25–31.

ARAUJO, R., MADEIRA, M. J. & AYALA, I. 2007. Estudio del estado actual de Margaritifera auricularia en las aguas del río Ebro. Confederación Hidrográfica del Ebro, Zaragoza.

ARAUJO, R., MADEIRA, M. J. & AYALA, I. 2009. Estudio del estado actual de conservación de Margaritifera auricularia en las aguas del río Ebro. Informe final. CSIC-ACEBI, Madrid.

ARAUJO, R. & MORENO, R. 1999. Antigua distribución de Margaritifera auricularia (Spengler) (Bivalvia: Margaritiferidae) en la Península Ibérica. Iberus, 17: 127–136.

301 ARAUJO, R., QUIRÓS. M. & RAMOS, M. A. 2003. Laboratory propagation and culture of juveniles of the endangered freshwater mussel Margaritifera auricularia (Spengler, 1793). Journal of Conchology, 38: 53–60.

ARAUJO, R. & RAMOS, M. A. 1998. Description of the glochidium of Margaritifera auricularia (Spengler, 1793) (Bivalvia, Unionoidea). Royal Society Philosophical Transactions Biological Sciences, 353: 1553–1559.

ARAUJO, R. & RAMOS, M. A. 1998. Margaritifera auricularia (Unionoidea, Margaritiferidae), the giant freshwater pearl mussel rediscovered in Spain. Graellsia, 54: 129– 130.

ARAUJO, R. & RAMOS, M. A. 2000. Status and conservation of the giant European freshwater pearl mussel (Margaritifera auricularia) (Spengler, 1793) (Bivalvia: Unionoidea). Biological Conservation, 96: 233–239.

ARAUJO, R. & RAMOS, M. A. 2000. A critical revision of the historical distribution of the endangered Margaritifera auricularia (Spengler, 1793) (Mollusca: Margaritiferidae) based on Museum specimens. Journal of Conchology, 37: 49–59.

ARAUJO, R. & RAMOS, M. A. 2001. Action Plan for Margaritifera auricularia. Convention on the Conservation of European Wildlife and Natural Habitats (Bern Convention). Council of Europe Publishing, Nature and environment, no. 117, Strasbourg.

ARAUJO, R. & RAMOS, M. A. 2001. Life-history data on the virtually unknown Margaritifera auricularia. In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 143–152. Springer-Verlag, Berlin Heidelberg.

ARAUJO, R., REIS, J., MACHORDOM, A., TOLEDO, C., MADEIRA, M. J., GÓMEZ, I., VELASCO, J. C., MORALES, J., BAREA, J. M., ONDINA, P. & AYALA, I. 2009. Las náyades de la península Ibérica. Iberus, 27: 7–72.

ARAUJO, R., SCHNEIDER, S., ROE, K. J., ERPENBECK, D. & MACHORDOM, A. 2017. The origin and phylogeny of Margaritiferidae (Bivalvia, Unionoida): a synthesis of molecular and fossil data. Zoologica Scripta, 46: 289–307.

302 ARAUJO, R., TOLEDO, C., VAN DAMME, D., GHAMIZI, M. & MACHORDOM, A. 2009. Margaritifera marocana (Pallary, 1918) a valid current species inhabiting the Moroccan rivers. Journal of Molluscan Studies, 75: 95–101.

ARCHAMBAULT, J.-M., COPE, W. G. & KWAK, T. J. 2014. Survival and behaviour of juvenile unionid mussels exposed to thermal stress and dewatering in the presence of a sediment temperature gradient. Freshwater Biology, 59: 601–613.

ARENDT, J. D. 1997. Adaptive intrinsic growth rates: An integration across taxa. Quarterly Review of Biology, 72: 149–177.

ASTM E2455–06 (American Society for Testing and Materials). 2013. Standard Guide for Conducting Laboratory Toxicity Tests with Freshwater Mussels. ASTM International, West Conshohocken, PA. Retrieved from www.astm.org.

AUGSPURGER, T., DWYER, F. J., INGERSOLL, C. G. & KANE, C. M. 2007. Advances and opportunities in assessing contaminant sensitivity of freshwater mussel (Unionidae) early life stages. Environmental Toxicology and Chemistry, 26: 2025–2028.

AWAKURA, T. 1968. The ecology of parasitic glochidia of the freshwater pearl mussel, Margaritifera laevis (Haas). Scientific Reports of the Hokkaido Fish Hatchery, 23: 1–21.

BACH, J. M., PAROUTY, T., LEON, C., SENECAL, A., PORTAFAIX, P., BARAULT, A., RAUTUREAU, C., RIMBERT L., DUPERRAY A. & BAISEZ, A. 2016. Volet scientifique: suivis biologiques des populations de poissons grands migrateurs, anne 2016. Rapport d’activité du Programme de Recherches Appliquées en Faveur des Poissons Migrateurs. Association LOGRAMI, Poitiers.

BAIRD, M. 2000. Life history and population structure of the spectaclecase mussel, Cumberlandia monodonta (Bivalvia, Margaritiferidae). MSc thesis, Missouri State University, Springfield, MI.

BAKER, S. M. & HORNBACH, D. J. 2001. Seasonal metabolism and biochemical composition of two unionid mussels, Actinonaias ligamentina and Amblema plicata. Journal of Molluscan Studies, 67: 407–416.

303 BALON, E. K. 1962. Note on the number of Danubian bitterlings developmental stages in mussels. Vestnik Ceskoslovenske Spolecnosti Zoologicke, 26: 250–256.

BARNHART, M. C. 2006. Buckets of muckets: A compact system for rearing juvenile freshwater mussels. Aquaculture, 254: 227–233.

BARNHART, M. C., HAAG, W. R. & ROSTON, W. N. 2008. Adaptations to host infection and larval parasitism in Unionoida. Journal of the North American Benthological Society, 27: 370–394.

BAUER, G. 1987. Reproductive strategy of the freshwater pearl mussel Margaritifera margaritifera. Journal of Animal Ecology, 56: 691–704.

BAUER, G. 1988. Threats to the freshwater pearl mussel Margaritifera margaritifera L. in central Europe. Biological Conservation, 45: 239–253.

BAUER, G. 1991. Plasticity in life-History Traits of the Freshwater Pearl Mussel - Consequences for the danger of extinction and for conservation measures. In: Species conservation: A population-biological approach (A. Seitz & V. Loeschcke, eds.), pp. 103–120. Birkhäusen-Verlag, Basel.

BAUER, G. 1992. Variation in the life span and size of the freshwater pearl mussel. Journal of Animal Ecology, 61: 425–436.

BAUER, G. 1994. The adaptive value of offspring size among freshwater mussels (Bivalvia: Unionoidea). Journal of Animal Ecology, 63: 933–944.

BAUER, G. 1997. Host relationships at reversed generation times: Margaritifera (Bivalvia) and salmonids. Ecological Studies, 130: 69–79.

BAUER, G. 1998. Allocation policy of female freshwater mussels. Oecologia, 117: 90–94.

BAUER, G. 2001. Factors Affecting Naiad Occurrence and Abundance. In: Ecology and Evolutionary Biology of the Freshwater Mussels Unionoidea (G. Bauer & K. Wächtler, eds.) pp 155–162. Springer-Verlag, Berlin Heidelberg.

304 BAUER, G. 2001. Framework and driving forces for the evolution of naiad life histories. In: Ecology and Evolutionary Biology of the Freshwater Mussels Unionoidea (G. Bauer & K. Wächtler, eds.), pp. 233–255. Springer-Verlag, Berlin Heidelberg.

BAUER, G., HOCHWALD, S. & SILKENAT, W. 1991. Spatial distribution of freshwater mussels: the role of host fish and metabolic rate. Freshwater Biology, 26: 377–386.

BAUER, G. & WÄCHTLER, K. 2001. Ecology and evolution of the Naiads. In: Ecology and Evolution of the Freshwater Mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 383-388. Springer-Verlag, Berlin .Heidelberg.

BEAUDET, A. 2006. Étude de la dynamique des populations de moules d’eau douce (Bivalvia : Unionidea) de deux rivières côtières de l’Est du Nouveau-Brunswick, la rivière Kouchibouguac et la rivière Kouchibouguacis. Mémoires de Maîtrise, UQAR, Université du Québec à Rimouski.

BEAUME, N., PASCO P. Y., RIBEIRO, M., DURY, P., ROSTAGNAT, L. & BEAUFILS, B. 2016. Conserver la moule perlière d’eau douce dans nos rivières. Recueil d’expérience du programme LIFE+ Nature « Conservation de la moule perlière d’eau douce du Massif armoricain » 2010-2016. Bretagne Vivante, CPIE des Collines, FFPPMA 29, SIAES, PNRNM, Brest.

BECK, K. & NEVES, R. J. 2003. An evaluation of selective feeding by three age-groups of the Rainbow mussel Villosa iris. North American Journal of Aquaculture, 65: 203–209.

BELLIARD, J., BOËT, P. & ALLARDI, J. 1995. Evolution à long terme du peuplement piscicole du bassin de la Seine. Bulletin Français de la Pêche et de la Pisciculture, 337: 83–91.

BENINGER, P. G., DWIONO, S. A. P. & LE PENNEC, M. 1994. Early development of the gill and implications for feeding in Pecten maximus (Bivalvia:Pectinidae). Marine Biology, 119: 405–412.

BENINGER, P. G., VENIOT, A. & POUSSART, Y. 1999. Principles of pseudofeces rejection on the bivalve mantle: integration in particle processing. Marine Ecology Progress Series, 178: 259–269.

305 BERG, L. S. 1964. Freshwater fishes of the U.S.S.R. and adjacent countries. 4th edition. Israel Program for Scientific Translations Ltd, Jerusalem. (Russian version published 1949).

BERRIE, A. & BOIZE, B. 1985. The fish hosts of Unio glochidia in the River Thames. Verhandlungen Internationale Vereinigung für theoretische und angewandte Limnologie, 22: 2712–2716.

BILECENOGLU, M., TASKAVAK, E., MATER, S. & KAYA, S. M. 2002. Checklist of the marine fishes of Turkey. Zootaxa, 113: 1–194.

BOGAN, A. E. 1993. Freshwater bivalve extinctions (Mollusca: Unionoida): a search for causes. American Zoologist, 33: 599–609.

BOGAN, E. 2008. Global diversity of freshwater mussels (Mollusca, Bivalvia) in freshwater. Hydrobiologia, 595: 139–147.

BOHLEN, B., KLECHTOVÁ, V, BOGUTSKAYA, N. & FREYHOF, F. 2006. Across Siberia and over Europe: Phylogenetic relationships of the freshwater fish genus Rhodeus in Europe and the phylogenetic position of R. sericeus from the River Amur. Molecular Phylogenetics and Evolution, 40: 856–865.

BOLLAND, J. D., BRACKEN, L. J., MARIN, R. & LUCAS, M. C. 2010. A protocol for stocking hatchery reared freshwater pearl mussel Margaritifera margaritifera. Aquatic Conservation: Marine and Freshwater Ecosystems, 20: 695–707.

BOLOTOV, I. N., VIKHREV, I. V., BESPALAYA, Y. V., GOFAROV, M. Y., KONDAKOV, A. V., KONOPLEVA, E. S., BOLOTOV, N. N. & LYUBAS, A. A. 2016. Multi-locus fossil- calibrated phylogeny, biogeography and a subgeneric revision of the Margaritiferidae (Mollusca: Bivalvia: Unionoida). Molecular Phylogenetics and Evolution, 103: 104–121.

BONNEMÈRE, L. 1901. Les mollusques des eaux douces de France et leurs perles. Publié par Institut international de bibliographie scientifique, Paris.

BÖSSNECK, U., KLEEMANN, R. & BUTTSTEDT, L. 2006. Die Große Flussperlmuschel (Pseudunio auricularius Spengler, 1793) in Mitteldeutschland: Neue Befunde zur historischen und fossilen Verbreitung (Bivalvia: Margaritiferidae). Malakologische Abhandlungen, 24: 141–156.

306 BOULÊTREAU, S., GAILLAGOT, A., CARRY, L., TÉTARD, S. & OLIVEIRA, E. 2018. Adult Atlantic salmon have a new freshwater predator. PLoS ONE, 13 DOI:10.1371/journal.pone.0196046

BOURGUIGNAT, J. R. 1883. Aperçu sur les Unionidæ de la Péninsule Italique.

BOUTELLIER, P. 1918. La pêche maritime au Maroc (Casablanca, Fédala, Rabat). G. Mercié & Co., Rabat.

BRADY, T. & GATENBY, C. 2018. Host species identification, acquisition and captive care. In: Freshwater mussel propagation for restoration (M. A. Patterson, R. A. Mair, N. L. Eckert, C. M. Gatenby, T. Brady, J. W. Jones, B. R. Simmons & J. Devers, eds.), pp. 58–81. Cambridge University Press, Cambridge.

BRIM BOX, J. & MOSSA, J. 1999. Sediment, land use, and freshwater mussels: prospects and problems. Journal of the North American Benthological Society, 18: 99–117.

BRYJA, J., SMITH, C., KONECNÝ, A. & REICHARD, M. 2010 Range-wide population genetic structure of the European bitterling (Rhodeus amarus) based on microsatellite and mitochondrial DNA analysis. Molecular Ecology, 19: 4708–4722.

BUDDENSIEK, V. 1995. The culture of juvenile freshwater pearl mussels Margaritifera margaritifera L. in cages: A contribution to conservation programmes and the knowledge of habitat requirements. Biological Conservation, 74: 33–40.

BUDDENSIEK, V., ENGEL, H., FLEISCHAUER-RÖSSING, S. & WÄCHTLER, K. 1993. Studies on the chemistry of interstitial water taken from defined horizons in the fine sediments of bivalve habitats in several northern German lowland waters II: Microhabitats of Margaritifera margaritifera L., Unio crassus (Philipsson) and Unio tumidus Philipsson. Archives für Hydrobiology, 127: 151–166.

BUHAY, J. E., SERB, J. M., DEAN, C. R., PARHAM, Q. & LYDEARD, C. 2002. Conservation genetics of two endangered unionid bivalve species, Epioblasma florentina walkeri and E. capsaeformis (Unionidae: Lampsilini). Journal of Molluscan Studies, 68: 385– 391.

307 BUISSON, L., THUILLER, W., LEK, S., LIM, P. & GRENOUILLET, G. 2008. Climate change hastens the turnover of stream fish assemblages. Global Change Biology, 14: 2232– 2248.

BURKY, A. J. 1983. Physiological ecology of freshwater bivalves. In: The mollusca (W. D. Russell-Hunter, ed.), pp. 281–327. Academic Press, New York.

BURLAKOVA, L. E., KARATAYEV, A. Y. & PADILLA, D. K. 2000. The impact of Dreissena polymorpha (Pallas) invasion on unionid bivalves. International Review of Hydrobiology, 85: 529–541.

CADDY, J. F. 1969. Development of mantle organs, feeding, and locomotion in postlarval Macoma balthica (L.) (Lamellibranchiata). Canadian Journal of Zoology, 47: 609–617.

CANNUEL, R. & BENINGER, P. G. 2006. Gill development, functional and evolutionary implications in the Pacific oyster Crassostrea gigas (Bivalvia: Ostreidae). Marine Biology, 149: 547–563.

CANNUEL, R., BENINGER, P. G., McCOMBIE, H. & BOUDRY, P. 2009. Gill development and its functional and evolutionary implications in the blue mussel Mytilus edulis (Bivalvia: Mytilidae). Biological Bulletin, 217: 173–188.

CAREY, C. S., JONES, J. W., BUTLER, R. S. & HALLERMAN E. M. 2015. Restoring the endangered oyster mussel (Epioblasma capsaeformis) to the upper Clinch River, Virginia: an evaluation of population restoration techniques. Restoration Ecology, 23: 447–454.

CAROL, J. 2007. Ecology of an invasive fish (Silurus glanis) in Catalan reservoirs. PhD thesis, University of Girona, Spain.

CAROSI, A., GHETTI, L., CAUZILLO, C., POMPEI, L. & LORENZONI, M. 2016. Occurrence and distribution of exotic fishes in the Tiber River basin (Umbria, central Italy). Journal of Applied Ichthyology, 33: 274–283.

CARRIZO, S. F., SMITH, K. G. & DARWALL, R. T. 2013. Progress towards a global assessment of the status of freshwater fishes (Pisces) for the IUCN Red List: application to conservation programmes in zoos and aquariums. International Zoo Yearbook, 47: 46–64.

308 CEMAGREF. 1994. Operation de sauvegarde de l'esturgeon europeen Acipenser sturio: 1–23. Centre national du machinisme agricole, du génie rural, des eaux et des forêts (CEMAGREF), Bordeaux.

ÇEVIK, C., ERGÜDEN, D. & TEKELIOĞLU, N. 2010. Confirming the Presence of Sea Lamprey (Petromyzon marinus Linnaeus, 1758) in the Levantine Sea (Petromyzoniformes: Petromyzonidae). Zoology in the Middle East, 49: 107–108.

CHEZIK, K. A., LESTER, N. P. & VENTURELLI, P. A. 2013. Fish growth and degree-days I: selecting a base temperature for a within-population study. Canadian Journal of Fisheries and Aquatic Sciences, 71: 47–55.

CLAASSEN, C. 1994. Washboards, pigtoes, and muckets: Historic musseling in the Mississippi watershed. Historical Archaeology, 28: 1–145.

COCHET, G. 2001. Redécouverte d’une population vivante de la Grande Mulette, Margaritifera auricularia, sur la Vienne et la Creuse. Recherches Naturalistes en Région Centre, 10: 3–16.

COCHET, G. 2001. Margaritifera auricularia sur le cours de la Vienne. Bilan des connaissances, impact des aménagements, gestion et protection. DIREN Centre.

COCHET, G. 2002. La Grande Mulette (Margaritifera auricularia) dans la Vienne et la Creuse. Rapport et atlas. DIREN Centre & DIREN Poitou-Charente.

COCHET, G. 2004. La Moule perlière et les nayades de France. Histoire d'une sauvegarde. Christian Bouchardy, Nohanent.

COCHET, G. 2006. La Grande Mulette (Margaritifera auricularia) dans la Vienne au niveau du tracé de la LGV SEA. Réseau Ferré de France.

COKER, R. E., SHIRA, A. F., CLARK, H. W. & HOWARD, A. D. 1921. Natural history and propagation of fresh-water mussels. Bulletin of the United States Bureau of Fisheries, 37: 77– 181.

CONFORTINI, I. 1992. Presenza del rodeo amaro, Rhodeus sericeus (Pallas, 1776) nel Fiume Menago (provincia di Verona) (Pisces, Cyprinidae). Bollettino del Museo Civico di Storia Naturale di Verona, 16: 329–332.

309 CONSEJO, J. 2016. Investigaciones sobre los posibles peces hospedadores de los gloquidios de la almeja de río Margaritifera auricularia (Mollusca, Bivalvia, Margaritiferidae). Final undergraduate thesis, Universidad Autónoma de Madrid, Spain.

COPE, O. B. 1959. New parasite records from stickleback and salmon in an Alaska stream. Transactions of the American Microscopical Society, 78: 157–162.

COX, L. R. 1969. General features of bivalvia. In: Treatise on invertebrate paleontology part N, Mollusca 6, Bivalvia (R. C. Moore, ed.), N2–N129. Geological Society of America,New York and University of Kansas, Lawrence.

CUCHERAT, X. 2009. Mise en place d’un plan de conservation pour les mollusques inscrits aux annexes de la Directive « Habitats-Faune-Flore » et protégés au titre de l’arrêté du 23 avril 2007 en Picardie – Phase I-. DREAL Picardie – Biotope.

CUCHERAT, X., JAMIER, M., PHILIPPE, L., PRIÉ, V., SIMON, L. & TAPKO, N. 2011. Mise en place d’un plan de conservation pour les mollusques inscrits aux annexes de la Directive «Habitats-Faune-Flore» et protégés au titre de l’arrêté du 23 avril 2007 – Phase II- . DREAL Picardie - Biotope.

CUCHERAT, X. & PRIÉ, V. 2011. Étude d’impact du projet de curage du canal de la Somme entre Sailly-Laurette et Sormont (Département de la Somme) - Recherches spécifiques de la Grande Mulette et de la Mulette épaisse. Conseil Général de la Somme – Biotope.

CUCHERAT, X., PRIÉ., V. & TAPKO, N. 2011. Étude d’impact du projet de curage du canal de la Somme entre Sailly-Laurette et Sormont (Département de la Somme) - Recherches spécifiques de la Grande Mulette et de la Mulette épaisse. Conseil Général de la Somme – Biotope.

CUCHEROUSSET, J., HORKY, P., SLAVÍK, O., OVIDIO, M., ARLINGHAUS, R., BOULÊTREAU, S., BRITTON, R., GARCÍA-BERTHOU, E. & SANTOUL, F. 2018. Ecology, behaviour and management of the European catfish. Reviews in Fish Biology and Fisheries, 28: 177–190.

CUMMINGS, K. S. & MAYER, C. A. 1993. Distribution and host species of the federally endangered freshwater mussel, Potamilus capax (Green, 1832) in the Lower Wabash River, Illinois and Indiana. Technical report, Illinois Natural History Survey, Center for Biodiversity.

310 CUROLE, J. P., FOLTZ, D. W. & BROWN, K. M. 2004. Extensive allozyme monomorphism in a threatened species of freshwater mussel, Margaritifera hembeli Conrad (Bivalvia: Margaritiferidae). Conservation Genetics, 5: 271–278.

CUROLE, J. P. & KOCHER, T. D. 2002. Ancient sex-specific extension of the Cytochrome c Oxidase II Gene in bivalves and the fidelity of doubly-uniparental inheritance. Molecular Biology and Evolution, 19: 1323–1328.

CUTTELOD, A., SEDDON, M. &. NEUBERT, E. 2011. European Red List of Non-marine Molluscs. Publications Office of the European Union, Luxembourg.

DARTIGUELONGUE, J. 2017. Contrôle du fonctionnement de la passe à poissons installée à Crouin (16) sur la Charente. Suivi de l’activité ichtyologique en 2016. Rapport SCEA (Services et Conseils en Environnement Aquatique) pour CMCS (Cellule migrateurs Charente et Seudre), Toulouse.

DARTNALL, H. J. G. & WALKEY, M. 1979. The distribution of glochidia of the swan mussels Anodonta cygnea (Mollusca) on the three-spine stickleback, Gastoresteus aculeatus (Pisces). Journal of Zoology, 189: 31–37.

DAVIS, G. M. & FULLER, S. L. H. 1981. Genetic relationship among Recent Unionacea (Bivalvia) of North America. Malacologia, 20: 217–253.

DIMOCK, R. V. & WRIGHT, A. H. 1993. Sensitivity of juvenile freshwater mussels to hypoxic, thermal and acid stress. Journal of the Elisha Mitchell Scientific Society, 109: 183– 192.

DOADRIO, I. 2001. Atlas y libro rojo de los peces continentales de España. Dirección General de Conservación de la Naturaleza, Museo Nacional de Ciencias Naturales, Madrid.

DOHOGNE, 2008. Recherche de la Moule perlière (Margaritifera margaritifera) et de la Grande mulette (Margaritifera auricularia) dans le département de l’Indre. Rapport d’étude, Indre nature.

DOLLFUS, R. P. 1955. Première contribution à l’établissement d’un fichier ichtyologique du Maroc atlantique. Travaux de l´Institut. Scientifique, Chérifien, Série Zoologie, 6, 1-226.

311 DOUDA, K., LIU, H.-Z., YU, D., ROUCHET, R., LIU, F., TANG, Q.-Y., METHLING, C., SMITH, C. & REICHARD, M. 2017. The role of local adaptation in shaping fish-mussel coevolution. Freshwater Biology, DOI 10.1111/fwb.13026

DOUDA, K., LOPES-LIMA, M., HINZMANN, M., MACHADO, J., VARANDAS, S., TEIXEIRA, A. & SOUSA, R. 2013. Biotic homogenization as a threat to native affiliate species: fish introductions dilute freshwater mussel’s host resources. Diversity and Distributions, 19: 933–942.

DUDGEON, D., ARTHINGTON, A. H., GESSNER, M. O., KAWABATA, Z.-I., KNOWLER, D. J., LEVEQUE, C., NAIMAN, R. J., PRIEUR-RICHARD, A.-H., SOTO, D. M., STIASSNY, L. J. & SULLIVAN, C. A. 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biological Reviews, 81: 163–182.

EFFORD, I. A. & TSUMURA, K. 1973. Uptake of disolved glucose and glycine by Pisidium, a freshwater bivalve. Canadian Journal of Zoology, 51: 825–832.

ELVIRA, B. & ALMODÓVAR, A. 2001. Freshwater fish introductions in Spain: facts and figures at the beginning of the 21st century. Journal of Fish Biology, 59: 323–331.

ENGEL, H. 1990. Untersuchungen zur Autökologie von Unio crassus (Philipsson) in Norddeutschland. Dissertation, Universität Hannover.

EVERSOLE, A. G., STUART, K. R. & BRUME, D. E. 2008. Effect of temperature and phytoplankton concentration of partitioned aquaculture system water on freshwater mussel filtration. Aquaculture Research, 39: 1691–1696.

EYBE, T., THIELEN, F., BOHN, T. & SURES, B. 2013. The first millimeter rearing juveniles freshwater pearl mussels (Margaritifera margaritifera L.) in plastic boxes. Aquatic Conservation: Marine and Freshwater Ecosystems, 23: 964–975.

EYBE, T., THIELEN, F., BOHN, T. & SURES, B. 2015. Influence of the excystment time on the breeding success of juvenile freshwater pearl mussels (Margaritifera margaritifera). Aquatic Conservation: Marine and Freshwater Ecosystems, 25: 21–30.

FITT, W. K., FISHER, C. R. & TRENCH, R. K. 1984. Larval biology of Tridacnid clams. Aquaculture, 39: 181–195.

312 FLEISCHAUER-ROSSING, S. 1990. Untersuchungen zur Autokologie von Unio tumidus (Philipsson) und Unio pictorum (Linnaeus) (Bivalvia) unter besonderer Berucksichtigung der fruhen postparasitaren Phase. PhD thesis, Universitat Hannover.

FOBIAN, T., B., BUNTIN, M. L., HOLIFIELD, J. T., TARPLEY, T. A. & JOHNSON, P. D. 2017. Advancements in culture techniques for Alabama pearlshell (Margaritifera marrianae, R. I. Johnson, 1983) to support species recovery and reintroduction. In: FMCS 10th Biennial Symposium. Cleveland, OH.

FRIMODT, C. 1995. Multilingual illustrated guide to the world's commercial coldwater fish. Fishing News Books, Osney Mead, Oxford.

FROESE, R. & PAULY, D. 2018. FishBase. World Wide Web electronic publication. www.fishbase.org, version (06/2018).

FROUFE, E., PRIÉ, V., FARIA, J., GHAMIZI, M., GONÇALVES, D. V., EMRE GÜRLEK, M., KARAOUZAS, I., KEBAPÇI, Ü., SEREFLISAN, H., SOBRAL, C., SOUSA, R., TEIXEIRA, A., VARANDAS, S., ZOGARIS, S. & LOPES-LIMA, M. 2016. Phylogeny, phylogeography, and evolution in the Mediterranean region: News from a freshwater mussel (Potomida, Unionida). Molecular Phylogenetics and Evolution, 100: 322–332.

FULLER, S. L. H. 1972. Elliptio marsupiobesa, a new fresh-water mussel (Mollusca: Bivalvia: Unionidae) from the Cape Fear River, North Carolina. Proceedings of the Academy of Natural Sciences of Philadelphia, 124: 1–10.

FULLER, S. L. H. 1973. Fusconaia masoni (Conrad 1834) (Bivalvia: Unionacea) in the Atlantic drainage of the southeastern United States. Malacological Review, 6: 105–117.

FURNESTIN, J., DARDIGNAC, J., MAURIN, C., VINCENT, A., COUPE, R. & BOUTIÈRE, H. 1958. Données nouvelles sur les poissons du Maroc atlantique. Revue des travaux de l'Institut des pêches maritimes, 22: 383–493.

GALBRAITH, H. S., FRAZIER, S. E., ALLISON, B. & VAUGHN, C. C. 2009. Comparison of gill surface morphology across a guild of suspension feeding unionid bivalves. Journal of Molluscan Studies, 75: 103–107.

313 GATENBY, C. M., NEVES, R. J. & PARKER, B. C. 1996. Influence of sediment and algal food on cultured juvenile freshwater mussels. Journal of the North American Benthological Society, 15: 597–609.

GATENBY, C. M., ORCUTT, D. M., KREEGER, D. A., PARKER, B. C., JONES, V. A. & NEVES, R. J. 2003. Biochemical composition of three algal species proposed as food for captive freshwater mussels. Journal of Applied Phycology, 15: 1–11.

GATENBY, C. M., PARKER, B. C. & NEVES, R. J. 1997. Growth and survival of juvenile rainbow mussels, Villosa iris (Lea, 1829) (Bivalvia: Unionidae), reared on algal diets and sediment. American Malacological Bulletin, 14: 57–66.

GAVIOLI, A., MANCINI, M., MILARDI, M., ASCHONITIS, V., RACCHETTI, E., VIAROLI, P. & CASTALDELLI, G. 2018. Exotic species, rather than low flow, negatively affect native fish in the Oglio River, Northern Italy. River Research and Application, DOI: 10.1002/rra.3324

GEHIN, J. B. 1868. Révision des Poissons qui vivent dans les cours d’eau et dans les étangs du département de la Moselle. Avec quelques considérations sur le Darwinisme. Metz.

GEIST J. 2005. Conservation Genetics and Ecology of European Freshwater Pearl Mussels (Margaritifera margaritifera L.). Dissertation PhD, Technischen Universität Münich.

GEIST, J. & AUERSWALD, K. 2007. Physicochemical stream bed characteristics and recruitmentof the freshwater pearl mussel (Margaritiferamargaritifera). Freshwater Biology, 52: 2299–2316.

GEIST, J. & KUEHN, R. 2005. Genetic diversity and differentiation of central European freshwater pearl mussel (Margaritifera margaritifera L.) populations: implications for conservation and management. Molecular Ecology, 14: 425–439.

GEIST, J., SÖDERBERG, H., KARLBERG, A. & KUEHN, R. 2010. Drainage- independent genetic structure and high genetic diversity of endangered freshwater pearl mussels (Margaritifera margaritifera) in northern Europe. Conservation Genetics, 11: 1339–1350.

GENSOUL, J. 1908. Monographie des Poissons du département de la Saône-et Loire. Autun.

314 GESS, R. W., COATES, M. I. & RUBIDGE, B. S. 2006. A lamprey from the Devonian period of South Africa. Nature, 446: 981–984.

GESSNER, J., WILLIOT, P., ROCHARD, E., FREYHOF, J. & KOTTELAT, M. 2010. Acipenser sturio. The IUCN Red List of Threatened Species 2010: e.T230A13040963. http://dx.doi.org/10.2305/IUCN.UK.2010-1.RLTS.T230A13040963.en. Downloaded on 28 June 2018.

GIRIBET, G. 2008. Bivalvia. In: Phylogeny and evolution of the Mollusca (W. F. Ponder & D. R. Lindberg, eds.), pp.105–141. University of California Press, Berkeley.

GÓMEZ, I. & ARAUJO, R. 2008. Channels and ditches as the last shelter for freshwater mussels: the case of Margaritifera auricularia and other naiads inhabiting the mid Ebro River Basin, Spain. Aquatic Conservation: Marine and Freshwater Ecosystems, 18: 658–670.

GOOD, S. C. 1998. Freshwater bivalve fauna of the Late Triassic (Carnian-Norian) Chinle, Dockum, and Dolores Formations in the Southwest United States. In: Bivalves: an Eon of Evolution – Paleobiological Studies Honoring Norman D. Newell (P. A. Johnston & J. W. Haggart, eds.), pp. 223–249. University of Calgary Press, Calgary.

GRAF, D. L. 1997. The effect of breeding period on the biogeography of freshwater mussels (Bivlavia: Unionoidea) in the Minnesota region of North America. Occasional papers on Mollusks, Havard, 5: 393–407.

GRAF, D. L. & CUMMINGS, K. S. 2006. Palaeoheterodont diversity (Mollusca: Trigonioida + Unionoida): what we know and what we wish we knew about freshwater mussel evolution. Zoological Journal of the Linnean Society, 148: 343–394.

GRAF, D. L. & CUMMINGS, K. S. 2007. Review of the systematics and global diversity of freshwater mussel specie (Bivalvia: Unionoida). Journal of Molluscan Studies, 73: 291–314.

GRAF, D. L. & Ó FOIGHIL, D. 2000. The evolution of brooding characters among the freshwater mussels (Bivalvia: Unionoidea) of North America. Journal of Molluscan Studies, 66: 157–170.

315 GRANDE, C., ARAUJO, R. & RAMOS, M. A. 2001. The gonads of Margaritifera auricularia (Spengler, 1793) and M. margaritifera (Linnaeus, 1758) (Bivalvia: Unionidae). Journal of Molluscan Studies, 67: 27–35.

GUARNERI, I., POPA, O. P., GOLA, L., KAMBURSKA, L., LAUCERI, R., LOPES-LIMA, M., POPA, L. O. & RICCARDI, N. 2014. A morphometric and genetic comparison of Sinanodonta woodiana (Lea, 1834) populations: does shape really matter? Aquatic Invasions,9: 183–194.

GUILLERAULT, N., BOULETREAU, S., IRIBAR, A., VALENTINI, A. & SANTOUL, F. 2017. Application of DNA metabarcoding on faeces to identify European catfish Silurus glanis diet. Journal of Fish Biology, 90: 2214–2219.

GUM, B., LANGE, M. & GEIST, J. 2011. A critical reflection on the success of rearing and culturing juvenile freshwater mussels with a focus on the endangered freshwater pearl mussel (Margaritifera margaritifera L.). Aquatic Conservation: Marine and Freshwater Ecosystems, 21: 743–751.

GUYOT, J. A. 2005. Restoration of the endangered Cumberland elktoe (Alasmidonta atropurpurea) and Cumberland bean (Villosa trabalis) (Bivalvia: Unionidae) in the Big South Fork National River and Recreation Area, Tennessee and Kentucky. MSc thesis, Department of Fisheries and Wildlife Sciences, Virginia Polytechnic Institute and State University.

HAAG, W. R. 2012. North American freshwater mussels. Natural History, Ecology and Conservation. Cambridge University Press, New York.

HAAG, W. R. 2013. The role of fecundity and reproductive effort in defining life-history strategies of North American freshwater mussels. Biological Reviews, 88: 745–766.

HAAG, W. R. & COMMENS-CARSON, A. M. 2008. Testing the assumption of annual shell ring deposition in freshwater mussels Canadian. Journal of Fisheries and Aquatic Science, 65: 493–508.

HAAG, W. R. & WARREN, M. L. 1997. Host fishes and reproductive biology of 6 freshwater mussel species from the Mobile basin, USA. Journal of the North American Benthological Society, 16: 576–585.

316 HAAG, W. R. & WARREN, M. L. 1998. Role of ecological factors and reproductive strategies in structuring freshwater mussel communities. Canadian Journal of Fisheries and Aquatic Sciences, 55: 297–306.

HAAG, W. R. & WARREN, M. L. 2003. Host fishes and infection strategies of freshwater mussels in large Mobile Basin streams, USA. Journal of the North American Benthological Society, 22: 78–91.

HAAG, W. R., WARREN, M. L. & SHILLINGSFORD, M. 1999. Host fishes and host- attracting behavior of Lampsilis altilis and Villosa vibex (Bivalvia: Unionidae). American Midland Naturalist, 141: 149–157.

HAAS, F. 1910. Pseudunio, neues genus für Unio sinuatus Lam. Nachrichtsblatt der Deutschen Malakozoologischen Gesellschaft, 42: 181–183.

HAAS, F. 1910. On Unio, Margaritana, Pseudanodonta and their occurrence in the Thames Valley. Proceedings of the Malacological Society of London, 9: 106–112.

HAAS, F. 1916. Sobre una concha fluvial interesante (Margaritana auricularia, Spglr.) y su existencia en España. Boletín de la Sociedad Aragonesa de Ciencias Naturales, 15: 33–44.

HAAS, F. 1917. Estudios sobre la Náyades del Ebro. Boletín de la Sociedad Aragonesa de Ciencias Naturales, 16: 71–82.

HAAS, F. 1924. Anatomische Untersuchungen an europäischen Najaden I. Archiv für Molluskenkunde, 56: 66–82.

HAAS, F. 1969. Superfamily Unionacea. In: Treatise on invertebrate paleontology, Part N, Mollusca, 6 1: Bivalvia (R. C. Moore, ed.), N411–N470. Geological Society of America and University of Kansas Press, Lawrence.

HALL T. A. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium. Series, 41, 95–98.

HANSEN, M. J., MADENJIAN, C. P., SLADE, J. W., STEEVES, T. B., ALMEIDA, P. R. & QUINTELLA, B. R. 2016. Population ecology of the sea lamprey (Petromyzon marinus) as an invasive species in the Laurentian Great Lakes and an imperiled species in Europe. Reviews in Fish Biology and Fisheries, 26: 509–535

317 HARMS, W. 1907. Zur Biologie und Entwicklungsgeschichte der Flußperlmuschel (Margaritana margaritifera Dupuy). Zoologischer Anzeiger, 31: 814–824.

HARMS, W. 1909. Postembryonale Entwicklungsgeschichte der Unioniden. Zoologische Jahrbücher. Abteilung für Anatomie und Ontogenie der Tiere, 28: 325–386.

HASTIE, L. C. & YOUNG, M. R. 2003. Conservation of the freshwater pearl mussel I: captive breeding techniques. Conserving Natura 2000, Rivers Ecology Series No. 2, English Nature, Peterboroug.

HASTIE, L. C. & YOUNG, M. R. 2003. Timing of spawning and glochidial release in Scottish freshwater pearl mussel (Margaritifera margaritifera) populations. Freshwater Biology, 48: 2107–2117.

HASTIE, L. C., YOUNG, M. R., BOON, P. J., COSGROVE, P. J. & HENNINGER, B. 2000. Sizes, densities and age structures of Scottish Margaritifera margaritifera (L.) populations. Aquatic Conservation: Marine and freshwater ecosystems, 10: 229–247.

HASZPRUNAR, G. et al. 2008. Relationships of higher molluscan taxa. In: Phylogeny and evolution of the Mollusca (W. F. Ponder & D. R. Lindberg, eds.), pp. 19–32. Univeristy of California Press, Berkeley.

HEARD, W. H. 1998. Brooding patterns in freshwater mussels. Malacological Review, 31: 105–122.

HEARD, W. H. & GUCKERT, R. H. 1971. A re-evaluation of the Recent Unionacea (Pelecypoda) of North America. Malacologia, 10: 333–355.

HEARD, W. H. & VAIL, V. A. 1976. Anatomical systematics of Etheria elliptica (Pelecypoda: Mycetopodidae). Malacological Review, 9: 15-24.

HELM, M. M., BOURNE, N. & LOVATELLI, A. 2004. Cultivo de bivalvos en criadero. Un manual práctico. FAO Fisheries Technical Paper 471, Roma.

HENLEY, W. F., ZIMMERMAN, L. L., NEVES, R. J. & KIDD, M. R. 2001. Design and evaluation of recirculating water systems for maintenance and propagation of freshwater mussels. North American Journal of Aquaculture, 63: 144–155.

318 HERBERS, K. 1914. Entwickungsgeschichte von Anodonta cellensis Schröt. Zeitschrift für Wissenschaftliche Zoologie, 108: 1–174.

HESSE, A. S., BÉRENGER, M., de VANNOISE, V. & MANGOT, S. 2015. Historique de propagation de la corbicule, Corbicula fluminea (Bivalvia: Corbiculidae), en région Centre – Val de Loire (France). MalaCo, 11: 6–13.

HOCHWALD, S. 1988. Untersuchungen zur Populationsokologie und Fortpflanzungsbiologie der Bachmuschel Unio crassus Phil. 1788. Diplomarbeit, Lehrstuhl Tierokologie I, Universitat Bayreuth.

HOCHWALD, S. 2001. Plasticity of life-history traits in Unio crassus. In: Ecology and Evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 127–141. Springer-Verlag, Berlin Heidelberg.

HOLOPAINEN, I. 1985. Feeding biology of Pisidiidae (Bivalvia) with special emphasis on functional morphology of the digestive tract. Lammi Notes, 12: 5–9.

HOLOPAINEN, I. J. & HANSKI, I. 1979. Annual energy flow in populations of two Pisidium species (Bivalvia, Spaeriidae), with discussion on possible competition between them. Archiv für Hydrobiologie, 86: 338–354.

HOLOPAINEN, I. J. & LOPEZ, G. R. 1989. Functional anatomy and histology of the digestive tract of fingernail clams (Sphaeriidae, Bivalvia). Annales Zoologici Fennici, 26: 61–72.

HORNBACH, D. J., WAY, C. M., WISSING, T. E. & BURKY, A. J. 1984. Effects of particle concentration and season on the filtration rates of the freshwater clam, Sphaerium striatinum Lamarck (Bivalvia: Pisididae). Hydrobiologia, 108: 83–96.

HOWARD, A.D. 1915. Some exceptional cases of brooding among the Unionidae. Nautilus, 29: 4–11.

HOWARD, A. D. 1922. Experiments in the culture of fresh-water mussels. Bulletin of the United States Bureau of Fisheries, 38: 63–89.

HOWARD, J. K., CUFFEY, K. M. & SOLOMON, M. 2005. Toward using Margaritifera falcata as an indicator of base level nitrogen and carbon isotope ratios: insights from two California Coast Range rivers. Hydrobiologia, 541: 229–236.

319 HRUŠKA, J. 1999. Nahrungsansprüche der Flussperlmuschel und deren halbnatürliche Aufzucht in der Tschechischen Republik. Heldia, 4: 69–79.

HRUŠKA J. 2001. Experience of semi-natural breeding program of freshwater pearl mussel in the Czech Republic. In: Die Flussperlmuschel in Europa: Bestandssituation und Schutzmaßnahmen. Kongressband. WWA Hof, Albert-Ludwigs Universität: Freiburg.

HUBER, V. & GEIST, J. 2017. Glochidial development of the freshwater swan mussel (Anodonta cygnea, Linnaeus 1758) on native and invasive fish species. Biological Conservation, 209: 230–238.

HÜBY, B. 1988. Zur Entwicklungsbiologie der FlieBgewassermuschel Pseudoanodonta complanata. Dissertation, Institut für Zoologie, Tierarztliche Hochschule Hannover.

HUDSON, R. G. & ISOM, B. G. 1984. Rearing juveniles of the freshwater mussels (Unionidae) in a laboratory setting. The Nautilus, 98: 129–135.

HUFF, S. W., CAMPBELL, D., GUSTAFSON, D. L., LYDEARD, C., ALTABA, C. R. & GIRIBET, G. 2004. Investigations into the phylogenetic relationships of freshwater pearl mussels (Bivalvia: Margaritiferidae) based on molecular data: implications for their taxonomy and biogeography. Journal of Molluscan Studies, 70: 379–388.

HURLBERT, A. J. et al. 2007. Life and death: Metabolic rate, membrane composition, and life span of animals. Physiological Reviews, 87: 1175–1213.

JACKSON, R. B., CARPENTER, S. R., DAHM, C. N., McKNIGHT, D. M., NAIMAN, R. J., POSTEL, S. L. & RUNNING, S. W. 2001. Water in a changing world. Ecological Applications, 11: 1027–1045.

JANSEN, W., BAUER, G. & ZAHNER-MEIKE, E. 2001. Glochidial mortality in freshwater mussels. In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 185–211. Springer-Verlag, Berlin Heidelberg.

JATTEAU, PH. & FRATY, R. 2012. Étude de la tolérance à l’hypoxie des juvéniles de grande alose (Alosa alosa). IRSTEA, Bordeaux.

320 JONES, J. W., HALLERMAN, E. M. & NEVES, R. J. 2006. Genetic management guidelines for captive propagation of freshwater mussels (Unionoidea). Journal of Shellfish Research, 25: 527–535.

JONES, J. W., MAIR, R. & NEVES, R. 2005. Factors affecting survival and growth of juvenile freshwater mussels cultured in recirculating aquaculture systems. North American Journal of Aquaculture, 67: 210–220.

KADLEC, D., SIMKOVA, A., JARKOVSKY, J. &. GELNAR, M. 2003. Parasite communities of freshwater fish under flood conditions. Parasitology Research, 89: 272–283.

KARNA, D. W. & MILLEMANN, R. E. 1978. Glochidiosis of salmonid fishes III. Comparative susceptibility to natural infection with Margaritifera margaritifera (L.) (Pelecypoda: Margaritanidae) and associated histopathology. Journal of Parasitology, 64: 528– 537.

KAT, P. W. 1983. Sexual selection and simultaneous hermaphroditism among the Unionidae (Bivalvia: Mollusca). Journal of Zoology (London), 201: 395–416.

KAT, P.W. 1984. Parasitism and the Unionacea (Bivalvia). Biological Reviews of the Cambridge Philosophical Society. 59: 189–207.

KAT, P. W. & DAVIS, G. M. 1984. Molecular genetics of peripheral, populations of Nova Scotian Unionidae (Mollusca: Bivalvia). Biological Journal of the Linnean Society, 22: 157– 185.

KAYS, W. T., SILVERMAN, H. & DIETZ, T. H. 1990. Water channels and water canals in the gill of the freshwater mussel, Ligumia subrostrata: ultrastructure and histochemistry. Journal of Experimental Zoology, 254: 256–269.

KEITH, P., PERSAT, H., FEUNTEUN E. & ALLARDI, J. 2011. Les poissons d’eau douce de France. Collection Inventaires et Biodiversité. Museum National d’Histoire Naturelle et Biotope, Paris.

KELLOGG, J. L. 1915. Ciliary mechanisms of Lamellibranchs with description of anatomy. Journal of Morphology, 26: 625–701.

321 KILLEEN, I. J., OLIVER, P. G. & FOWLES, A. P. 1998. The loss of a freshwater pearl mussel (Margaritifera margaritifera) population in NW Wales. Journal of Conchology Special Publication, 2: 245–250.

KLISHKO, O. K. 2012. Some data on reproductive biology of the freshwater mussels (Margaritiferidae, Unionidae) and their relationships with bitterlings (Cyprinidae) in Transbaikalye. Bulletin of the Russian Far East Malacological Society, 15/16: 31–55. (in Russian).

KLISHKO, O. K & BOGAN, A. E. 2013. The conservation status of the freshwater pearl mussel Margarititifera dahurica in Far Eastern Russia. Ellipsaria, 15: 31–33.

KNEELAND, S. D. & RHYMER, J. M. 2008. Determination of fish host use by wild populations of rare freshwater mussels using a molecular identification key to identify glochidia. Journal of the North American Benthological Society, 27: 150–160.

KOBAYASHI, O. & KONDO, T. 2005. Difference in host preference between two populations of the freshwater pearl mussel, Margaritifera margaritifera (Bivalvia, Margaritiferidae) in the Shinano River system, Japan. Venus, 64: 63–70.

KONDO, T. & KOBAYASHI, O. 2005. Revision of the genus Margaritifera (Bivalvia: Margaritiferidae) of Japan, with description of a new species. Venus, 64: 135–140.

KOTTELAT, M. &. FREYHOF, F. 2007. Handbook of European Freshwater Fishes. (Kottelat, & Freyhof, eds.). Publications Kottelat, Cornol, CH.

KOVITVADHI, S., KOVITVADHI, U., SAWANGWONG, P. & MACHADO, J. 2008. A laboratory-scale recirculating aquaculture system for juveniles of freshwater pearl mussel Hyriopsis (Limnoscapha) myersiana (Lea, 1856). Aquaculture, 275: 169–177.

KOVITVADHI, S., KOVITVADHI, U., SAWANGWONG, P., THONGPAN, A. & MACHADO, J. 2006. Optimization of diet and culture environment for larvae and juvenile freshwater pearl mussels, Hyriopsis (Limnoscapha) myersiana Lea, 1856. Invertebrate Reproduction and Development, 49: 61–70.

KOVITVADHI, S., KOVITVADHI, U., SAWANGWONG, P., THONGPAN, A. & MACHADO, J. 2007. Morphological development of the juvenile through to the adult in the

322 freshwater pearl mussel Hyriopsis (Limnoscapha) myersiana under artificial culture. Invertebrate Development, 50: 207–218.

KOZHARA, A. V., ZHULIDOV, A. V., GOLLASCH, S., PRZYBYLSKI, M., POZNYAK, V. G., ZHULIDOV, D. A. &. YU. GURTOVAYA, T. 2007. Range extension and conservation status of the bitterling, Rhodeus sericeus amarus in Russia and adjacent countries. Folia Zoologica, 56: 97–108.

KUNZ, G. F. 1898. A brief story of the gathering of fresh-water pearls in the United States. Bulletin of the United States. Fish Commission, 17: 321–330.

LAMARCK, J. B. 1819. Histoire naturelle des animaux sans vertèbres, …6(1). Verdière, Deterville & chez l’auteur, Paris.

LAMBERT, S. D. & MARTEL, A. L. 2012. A summary of glochidia hosts for Pyganodon cataracta (Say, 1817) and evidence of additional fish hosts (Fundulus diaphanus and Apeltes quadracus) from Nova Scotia. Ellipsaria, 14: 8–11.

LANGE, M. & SELHEIM, H. 2011. Growing factors of juvenile freshwater pearl mussels and their characteristics in selected pearl mussel habitats in Saxony (Germany). Ferrantia, 64: 30– 37.

LASEE, B. A. 1991. Histological and ultrastructural studies of larval and juvenile Lampsilis (Bivalvia) from the Upper Mississippi River. PhD Dissertation, Iowa State University, Ames.

LAVICTOIRE, L., MOORKENS, E., RAMSEY, A., SINCLAIR, W. & SWEETING, R. A. 2015. Investigations into feeding structures of juvenile freshwater pearl mussels (Margaritifera margaritifera) through scanning electron microscopy. In: Second International on the rearing of unionoid mussels. Clervaux, Luxembourg.

LAVICTOIRE, L., MOORKENS, E., RAMSEY, A. D., SINCLAIR, W. & SWEETING, R. A. 2016. Effects of substrate size and cleaning regime on growth and survival of captive-bred juvenile freshwater pearl mussels, Margaritifera margaritifera (Linnaeus, 1758). Hydrobiologia, 766: 89–102.

323 LAVICTOIRE, L., RAMSEY, A. D., MOORKENS, E. A., SOUCH, G. & BARNHART, M. C. 2018. Ontogeny of juvenile freshwater pearl mussels, Margaritifera margaritifera (Bivalvia Margaritiferidae). PlosOne, doi:10.17632/k7m8p7bvjb.2

LAYZER, J. B., GORDON, M. E. & ANDERSON, R. M. 1993. Mussels: the forgotten fauna of regulated rivers. A case study of the Caney Fork River. Regulated Rivers: Research and Management, 8: 63–71.

LE BLOCH, F., HENRY, R., DICEV, G., FLAMANT, N., SIBLET, S., VACHER, A., LOUFTI, E., MONNIER, E., SAUZON, T., VANEL, Q., CAMUS, M., MONTAGNE, S., LOISEAU, J., CLEVENOT, P., MOLLARD, D., GIBEAU, C., BOUYON, H., VRIGNAUD, S., GENOUD, D. & MOUTHON, J. 2014. Project de mise à grand gabarit de la Seine entre Bray-sur-Seine (77) et Nogent-sur-Seine (10). Voies Navigables de France – Écosphère – Hydrosphère.

LEFEVRE, G. & CURTIS, W. 1910. Reproduction and parasitism in the Unionidae. Journal of Experimental Zoology, 9: 79–115.

LEFEVRE, G. & CURTIS, W. C. 1912. Studies on the reproduction and artificial propagation of fresh-water mussels. Bulletin of the United States Bureau of Fisheries, 30: 105–201.

LEPAGE, M. & ROCHARD, E. 1995. Threatened fishes of the world: Acipenser sturio Linnaeus, 1758 (Acipenseridae). Environmental Biology of Fishes, 43: 28.

LEVER, C. 1977. The naturalized animals of the British Isles. Hutchinson & Co., London.

LEVINE T. D., LANG, B. K. & BERG, D. J. 2012. Physiological and ecological hosts of Popenaias popeii (Bivalvia: Unionidae): laboratory studies identify more hosts than field studies. Freshwater Biology, 57: 1954–1864.

LIBERTY, A. 2004. An Evaluation of the Survival and Growth of Juvenile and Adult Freshwater Mussels at the Aquatic Wildlife Conservation Center (AWCC), Marion, Virginia. MSc thesis, Virginia Polytechnic Institute and State University, Blacksburg, VA.

LIBERTY, A., OSTBY, B. J. & NEVES, R. 2007. Determining a suitable size and sampling frequency for rearing juvenile Rainbow mussels Villosa iris. North American Journal of Aquaculture, 69: 44–52.

324 LIMA, P., KOVITVADHI, U., KOVITVADHI, S. & MACHADO, J. 2006. In vitro culture of glochidia from the freshwater mussel Anodonta cygnea. Invertebrate Biology, 125: 34–44.

LIMBURG, K. E. & WALDMAN, J. R. 2009. Dramatic declines in north Atlantic diadromous fishes. BioScience, 59: 955–965.

LIMM, M. P. & POWER, M. E. 2011. Effect of the western pearlshell mussel Margaritifera falcata on Pacific lamprey Lampetra tridentata and ecosystem processes. Oikos, 120: 1076– 1082.

LIU, S., WROBLEWSKI, J. S. & SCHNEIDER, D. C. 2017. Comparison of three methods for aging the Iceland Scallop Chlamys islandica.Journal of Shellfish Research, 36: 607–614.

LLORENTE, L., ROSELLÓ, E., MORALES, A., LIESAU, C., DAZA, A. & ARAUJO, R. 2015. Las náyades del calcolítico del camino de las Yeseras. In: La investigación arqueomalacológica en la Península Ibérica: nuevas aportaciones (I. Gutiérrez, D. Cuenca & M. R. González, eds.), pp. 125–134. Nadir Ediciones, Santander.

LOPES-LIMA, M., BOLOTOV, I. N., DO, V. T., ALDRIDGE, D. C., FONSECAB, M. M., GANH, H. M., GOFAROV, A. V., KONDAKOV, A. V., PRIÉ, V., SOUSA, R., VARANDAS, S., VIKHREV, I. V., TEIXEIRA, A., WU, R. W., WU, X., ZIERITZ, A., FROUFE, E. & BOGAN, A. E. 2018. Expansion and systematics redefinition of the most threatened freshwater mussel family, the Margaritiferidae. Molecular Phylogenetics and Evolution, 127: 98–118.

LOPES-LIMA, M., HINZMANN, M., TEIXEIRA, A., VARANDAS, S., MACHADO, J., SOUSA, R. & FROUFE, E. 2016. The Strange Case of the Tetragenous Anodonta anatina. Journal of Experimental Zoology, 325A: 52–56.

LOPES-LIMA, M., KEBAPC, I. U. &. VAN DAMME, D. 2014. Unio crassus. IUCN red list of threatened species. Version 2015.2. Available at http://www.iucnredlist.org/. Accessed 16.04.2018.

LOPES-LIMA, M., PRIÉ, V. & SEDDON, M. B. 2014. Potomida littoralis. IUCN red list of threatened species. Version 2015.2. Available at http://www.iucnredlist.org/. Accessed 16.04.2018.

325 LOPES-LIMA, M., SOUSA, R., GEIST, J., ALDRIDGE, D. C., ARAUJO, R., BERGENGREN, J., BESPALAYA, Y., BODIS, E., BURLAKOVA, L., VAN DAMME, D., DOUDA, K., FROUFE, E., GEORGIEV, D., GUMPINGER, C., KARATAYEV, A., KEBAPCI¸ U., KILLEEN, I., LAJTNER, J., LARSEN, B. M., LAUCERI, R., LEGAKIS, A., LOIS, S., LUNDBERG, S., MOORKENS, E., MOTTE, G., NAGEL, K. O., ONDINA, P., OUTEIRO, A., PAUNOVIC, M., PRIÉ, V., VON PROSCHWITZ, T., RICCARDI, N., RUDZITE, M., RUDZITIS, M., SCHEDER, C., SEDDON, M., SEREFLISAN, H., SIMIC, V., SOKOLOVA, S., STOECKL, K., TASKINEN, J., TEIXEIRA, A., THIELEN, F., TRICHKOVA, T., VARANDAS, S., VICENTINI, H., ZAJAC, K., ZAJAC T. & ZOGARIS, S. 2017. Conservation status of freshwater mussels in Europe: state of the art and future challenges. Biological Reviews, 92: 572–607.

LOPEZ, G. R. & HOLOPAINEN, I. J. 1987. Interstitial suspension-feeding by Pisidium spp. (Pisidiidae: Bivalvia): a new guild in the lentic benthos? American Malacological Bulletin, 5: 21–30.

LÓPEZ, M. A. & ALTABA, C. R. 2005. Fish host determination for Margaritifera auricularia (Bivalvia: Unionoida): Results and implications. Bolletino Malacologico, 41: 89–98.

LÓPEZ, M. A., ALTABA, C. R., ROUAULT, T. & GISBERT, E. 2007. The European sturgeon Acipenser sturio is a suitable host for the glochidia of the freshwater pearl mussel Margaritifera auricularia. Journal of Molluscan Studies, 73: 207–209.

LÓPEZ, M. A., GÁZQUEZ, N., OLMO-VIDAL, J. M., APRAHAMIAN, M. W. & GISBERT, E. 2007. The presence of anadromous twaite shad (Alosa fallax) in the Ebro River (western Mediterranean, Spain): An indicator of the population’s recovery? Journal of Applied Ichthyology, 23: 163–166.

LYDEARD, C., COWIE, R. H., PONDER, W. F., BOGAN, A. E., BOUCHET, P., CLARK, S. A., CUMMINGS, K. S., FREST, T. J., GARGOMINY, O., HERBERT, D. G., HERSHLER, R., PEREZ, K. E., ROTH, B., SEDDON, M., STRONG, E. E. & THOMPSON, F. G. 2004. The global decline of nonmarine molluscs. Bioscience, 54: 321–330.

MACHORDOM, A., ARAUJO, R., ERPENBECK, D. & RAMOS, M.-A. 2003. Phylogeography and conservation genetics of endangered European Margaritiferidae (Bivalvia: Unionoidea). Biological Journal of the Linnean Society, 78: 235–252.

326 MACKIE, G. L. & QADRI, S. U. 1978. Effects of substratum on growth and reproduction of Musculium securis (Bivalvia: Sphaeriidae). The Nautilus, 92: 135–144.

MAINE, A., ARANGO, C. & O’BRIEN, C. 2016. Fish associations of the California floater (Anodonta californiensis) in the Yakima River basin, Washington. Northwest Science, 90: 290– 300.

MAIR, R. A. 2018. Juvenile mussel culture. In: Freshwater mussel propagation for restoration (M. A. Patterson, R. A. Mair, N. L. Eckert, C. M. Gatenby, T. Brady, J. W. Jones, B. R. Simmons & J. Devers, eds.), pp. 159–222. Cambridge University Press, Cambridge.

MALMQVIST, B. & RUNDLE, S. 2002. Threats to the running water ecosystems of the world. Environmental Conservation, 29: 134–153.

MALO, D. 2012. Mise en production et optimisation de l’alimentation de l’espace Margaritifera margaritifera du Massif armoricain en elevage intensif. Memoire de fin d’étude Pour l’obtention du diplome de Licence Professionnelle Aquaculture et Gestion Durable de son Environnement.

MARM. 2009. Estrategia Nacional para la conservación de la almeja de río (Margaritifera auricularia) en España. Ministerio de Medio Ambiente y Medio Rural y Marino, Madrid.

MARTEL, A. L. & LAUZON-GUAY, J. S. 2005. Distribution and density of glochidia of the freshwater mussel Anodonta kennerlyi on fish hosts in lakes of the temperate rain forest of Vancouver Island. Canadian Journal of Zoology, 83: 419–431.

MATEUS, C. S., RODRÍGUEZ-MUÑOZ, R., QUINTELLA, B. R., ALVES, M. J. & ALMEIDA, P. R. 2012. Lampreys of the Iberian Peninsula: Distribution, population status and conservation. Endangered Species Research, 16: 183–198.

McELWAIN, A. & BULLARD, S. A. 2014. Histological atlas of freshwater mussels (Bivalvia, Unionidae): Villosa nebulosa (Ambleminae: Lampsilini), Fusconaia cerina (Ambleminae: Pleurobemini) and Strophitus connasaugaensis (Unioninae: Anodontini). Malacologia, 57: 99−239.

McIVOR, A. 2004. Freshwater mussels as biofilters. PhD thesis, University of Cambridge.

327 McIVOR, A. L. & ALDRIDGE, D. C. 2007. The reproductive biology of the depressed river mussel, Pseudanodonta complanata (Bivalvia: Unionidae), with implications for its conservation. Journal of Molluscan Studies, 73: 259–266.

MEIER-BROOK, C. 1969. Substrate relations in some Pisidium species (Eulamellibranchiata: Sphaeriidae). Malacologia, 9: 121–125.

MENNESSON-BOISNEAU, C., APRAHAMIAN, M. W., SABATIE, M. R. & CASSOU- LEINS, J. J. 2000. Remontée migratoire des adultes. In: Biologie, écologie et exploitation des aloses (Alosa sp.) de l'Atlantique Est et de la Méditerranée occidentale, pp. 55-73. INRA - CEMAGREF, INRA Editions, Versailles.

METCALFE, N. B. & MONAGHAN, P. 2003. Growth versus lifespan: perspectives from evolutionary ecology. Experimental Gerontology, 38: 935–940.

MEYERS, T. R. & MILLEMANN, R. E. 1977. Glochidiosis of salmonid fishes. I. Comparative susceptibility to experimental infection with Margaritifera margaritifera (L.) (Pelecypoda: Margaritiferidae). Journal of Parasitology, 63: 728–733.

MILLS, S. C. & REYNOLDS, J. D. 2002. Host species preferences by bitterling, Rhodeus sericeus, spawning in freshwater mussels and consequences for offspring survival. Animal Behavior, 63: 1029–1036.

MILLS, S. C., TAYLOR, M. I. &. REYNOLDS, J. D. 2005. Benefits and costs to mussels from ejecting bitterling embryos: a test of the evolutionary equilibrium hypothesis. Animal Behaviour, 70: 31–37.

MITROPOLSKII, V. I. 1966. On mechanisms of filtration and nutrition of sphaeriids (Mollusca, Lamellibranchiata). Trudy Instituta Biologii Vnutrennykh Vodnany Akademiia Nauka SSSR, 12: 129–133.

MODESTO, V., ILARRI, M., SOUZA, A. T., LOPES-LIMA, M., DOUDA, K., CLAVERO, M. & SOUSA, R. 2018. Fish and mussels: Importance of fish for freshwater mussel conservation. Fish and Fisheries, 19: 244–259.

MØLLER, P.R. 2012. Europæisk bitterling. In: Atlas over danske ferskvandsfisk (H. Carl & P. R. Møller, eds.), pp. 265–272. Statens Naturhistoriske Museum, Denmark.

328 MØLLER, P. R. &. MENNE, T. 1998: Bitterling Rhodeus sericeus (Pallas, 1776) – First record from Denmark. Flora og Fauna, 104: 29–34.

MOOLENBEEK, R. G., 2000. Een levende Pseudunio auricularius (Spengler, 1793). in een zijrivier van de Loire. Correspondentiebladvan de Nederlandse Malacologische Vereniging, 317: 132–133.

MORRISON, J. P. E. 1942. Preliminary report on mollusks found in the shell mounds of the Pickwick Landing basin in the Tennessee River valley. Bureau of American Ethnology Bulletin, 129: 337–392.

MORTON, B. 1973. A new theory of feeding and digestion in the filterfeeding lamellibranchia. Malacologia, 14: 63–79.

MOSER M. L., BUTZERIN, J. M. & DEY, D. B. 2007. Capture and collection of lampreys: The state of the science. Reviews in Fish Biology and Fisheries, 17: 45–56.

MOUTHON, J. & DAUFRESNE, M. 2006. Effects of the 2003 heatwave and climatic warming on mollusc communities of the Saône: a large lowland river and its two main tributaries (France). Global Change Biology, 12: 441–449.

MULCRONE, R. S. 2004. Incorporating habitat characteristics and fish hosts to predict freshwater mussel (Bivalvia: Unionidae) distributions in the Lake Erie drainage, southeastern Michigan. Ph.D. dissertation, University of Michigan, Ann Arbor, MI.

MUÑOZ-YANGUAS, M. A, CATALÁ, C., ELBAILE, E., SALINAS, C. & NAKAMURA, K. 2011. First assessment of the population effects of the maintenance channel works in the naiads of the Canal Imperial de Aragón. In: EUROMALACOL. Vitoria, Spain.

MURPHY, G. 1942. Relationship of the fresh-water mussel to trout in the Truckee River. California Fish and Game, 28: 89–102.

MURZINA, S. A., IESHKO, E. P. & ZOTIN, A. A. 2017. The freshwater pearl mussel Margaritifera margaritifera L.: Metamorphosis, growth, and development dynamics of encysted glochidia. Biology Bulletin, 44: 6–13.

329 NAITO, J. 1988. A note on the breeding habit of the fresh-water pearl-mussel, Margaritifera laevis Hass (Pisces: Margaritiferidae), in Geihoku-Cho, Hiroshima Pref., near the southern end in Japan. Miscellaneous Reports of the Hiwa Museum for Natural History, 27: 7–15.

NAKAMURA, K., CAÑETE, J., VIJUESCA, D., SORRIBAS, V., GUILLEN, N., SOSA, C., ARMENGOL, X., GINES, E., SOUSA, R. & MESQUITA-JOANES, F. 2018. Sensitivity of Margaritifera auricularia (Spengler, 1793) to heavy metals and ammonium: first approximation. In: First Freshwater Mollusk Conservation Society Meeting in Europe. Verbania, Italy.

NAKAMURA, K., CUCALA, L., MESTRE, A., MESQUITA-JOANES, F., ELBAILE, E., SALINAS, C. & MUÑOZ-YANGUAS, M. A. 2018. Modelling growth in the critically endangered freshwater mussel Margaritifera auricularia (Spengler, 1793) in the Ebro basin. Hydrobiologia, 810: 375–391. DOI:1010071075001731039

NAKAMURA, K., ELBAILE, E., MUÑOZ-YANGUAS, M. A., CATALÁ, C. & SALINAS, C. 2012. Captive breeding of the endangered peal mussel Margaritifera auricularia (Spengler, 1793). Large scale laboratory production of juveniles. In: International Meeting on Conservation and Biology of Freshwater Bivalves. Bragança, Portugal.

NAKAMURA, K., ELBAILE, E., MUÑOZ-YANGUAS, M. A., CATALÁ, C. & SALINAS, C. 2012. Size-specific growth pattern of fresh water mussel Margaritifera auricularia in the Ebro river channels (Spain). In: International Meeting on Biology and Conservation of Freshwater Bivalves. Bragança, Portugal.

NAKAMURA, K. & GUERRERO, J. 2008. Margaritifera auricularia, un difícil reto de conservación. Quercus, 265: 20–25.

NAKAMURA, K., GUERRERO, J., ALCÁNTARA, M., MUÑOZ, M. A. & ELBAILE, E. 2018. Tiempos de incertidumbre para la náyade Margaritifera auricularia. Quercus, 383: 16– 24.

NAKAMURA, K., MUÑOZ, M. A., RUIZ-ZARZUELA, I., BLAS, I., ELBAILE, E., SALINAS, C. 2015. The first year of Margaritifera auricularia (Spengler, 1793) Breeding in captivity in Aragon, Spain. In: Second International seminar on the rearing of unionoid mussels. Clervaux, Luxembourg. DOI: 10.13140/RG.2.2.12440.49921

330 NATURESERVE. 2013. Petromyzon marinus. The IUCN Red List of Threatened Species 2013: e.T16781A18229984. http://dx.doi.org/10.2305/IUCN.UK.2013-1.RLTS. T16781A18229984.en. Downloaded 4 June 2018.

NEGISHI, J. N., KATSUKI, K., KUME, M., NAGAYAMA, S. & KAYABA, Y. 2014. Terrestrialization alters organic matter dynamics and habitat quality for freshwater mussels (Unionoida) in floodplain backwaters. Freshwater Biology, 59: 1026–1038.

NESEMANN, H., & NAGEL, K. O. 1989. Die Flussmuscheln (Bivalvia: Unionacea) im Einzugsgebiet der Loire (Zentralfrankreich) - eine erste Bestandserfassung. Mitteilungen der Deutschen Malakozoologischen Gesellschaft, 44-45: 1–15.

NEVES, R. J. 1999. Conservation and commerce: management of freshwater mussel (Bivalvia: Unionoidea) resources in the United States. Malacologia, 41: 461–474.

NEVES, R. J. & MOYER, S. N. 1988. Evaluation of techniques for age determination of freshwater mussels (Unionidae). American Malacological Bulletin, 6: 179–188.

NEZLIN, L. P., CUNJAK, R. A., ZOTIN, A. A. & ZIUGANOV, V. V. 1994. Glochidium morphology of the freshwater pearl mussel (Margaritifera margaritifera) and glochidiosis of Atlantic salmon (Salmo salar): a study by scanning electron microscopy. Canadian Journal of Zoology, 72: 15–21.

NICHOLS, S. J., ALLEN, J., WALKER, G., YOKOYAMA, M., & GARLING, D. 2001. Lack of surface-associated microorganisms in a mixed species community of freshwater Unionidae. Journal of Shellfish Research, 20: 329–335.

NICHOLS, S. J. & GARLING, D. 2000. Food-web dynamics and trophic-level interactions in a multispecies community of freshwater unionids. Canadian Journal of Zoology, 78: 871–882.

NICHOLS, S. J. & GARLING, D. 2002. Evaluation of substitute diets for live algae in the captive maintenance of adult and subadult unionidae. Journal of Shellfish Research, 21: 875– 881.

NICOL, D. 1984. Constraints to adaptive radiation in deposit-feeding pelecypods. The Nautilus, 98: 165–166.

331 NIENHUIS, J. A. J. H. 2003. The rediscovery of Spengler’s freshwater pearlmussel Pseudunio auricularius (Spengler, 1793) (Bivalvia, Unionoidea, Margaritiferidae) in two river systems in France, with an analysis of some factors causing its decline. Basteria, 67: 67–86.

NORDLIE, F. G. & MIRANDI, A. 1996. Salinity relationships in a freshwater population of eastern mosquitofish. Journal of Fish Biology, 49: 1226–1232.

O’BEIRN, F. X., NEVES, R. J. & STEG, M. B. 1998. Survival and growth of juvenile freshwater mussels (Unionidae) in a recirculating aquaculture system. American Malacological Bulletin, 14: 165–171.

O'DEE, S. H. & WATTERS, G. T. 1998. New or confirmed host identifications for ten freshwater mussels. In: Conservation, Captive Care, and Propagation of Freshwater Mussels Symposium. Ohio Biological Survey, Ohio.

O’FOIGHIL, D., KINGZETT, B., O’FOIGHIL, G. & BOURNE, N. 1990. Growth and survival of Japanese scallops, Patinopecten yessoensis, in nursery culture. Journal of Shellfish Research, 9: 135–144.

ORTMANN, A. E. 1909. The breeding season of the Unionidae in Pennsylvania. Nautilus, 22: 91–103.

ORTMANN, A. E. 1911. A monograph of the najads of Pennsylvania. Memoirs of the Carnegie Museum, 4: 279–347.

ORTMANN, A.E. 1912. Notes upon the families and genera of the najades. Annals of the Carnegie Museum, 8: 222–365.

ÖSTERLING, M. E., GREENBERG, L. A. & ARVIDSSON, B. A. 2008. Relationship of biotic and abiotic factors to recruitment patterns in Margaritifera margaritifera. Biological Conservation, 141: 1365–1370.

PATTERSON, M. A., JONES, J. W. & GATENBY, C. M. 2018. Why propagate freshwater mussels? In: Freshwater mussel propagation for restoration (M. A. Patterson, R. A. Mair, N. L. Eckert, C. M. Gatenby, T. Brady, J. W. Jones, B. R. Simmons & J. Devers, eds.), pp. 1–24. Cambridge University Press, Cambridge.

332 PECK, R.H. 1877. The minute structure of the gills of lamellibranch mollusca. Quarterly Journal of microscopical science, 17: 43-66.

PEKKARINEN, M. & VALOVIRTA, I. 1996. Anatomy of the glochidia of the freshwater pearl mussel, Margaritifera margaritifera (L.). Archiv für Hydrobiologie, 137: 411–423.

PHILIPPE, L., PATRY, N., MATHIEU, C. & WALTER, S. 2013. Suivi des populations de Grande Mulette déplacée dans le cadre de la LGV SEA. COSEA – Biotope.

PHILIPPE, L., PATRY, N. & MATHIEU, C. 2014. Suivi des populations de Grande Mulette déplacée dans le cadre de la LGV SEA. COSEA – Biotope.

PHILIPPE, L., PATRY, N. & MATHIEU, C. 2015. Suivi des populations de Grande Mulette déplacée dans le cadre de la LGV SEA. LISEA – Biotope

PHILIPPE, L., PATRY, N. & MATHIEU, C. 2016. Suivi des populations de Grande Mulette déplacée dans le cadre de la LGV SEA. LISEA – Biotope.

PHILIPPE, L., PRIÉ, V. & COCHET, G. 2009. Étude préliminaire au franchissement de la Vienne par la LGV SEA – Inventaire des mulettes protégées sur 59 cours d’eau entre Bordeaux et Tours et suivi de la population des stations de Grande Mulette de l’A10. Réseau Ferré de France – Biotope.

PHILIPPE, L., PRIÉ, V., CUCHERAT, X. & BOUSQUET, P. 2009. Inventaires des mollusques protégés sur l’Aisne et la Meuse dans le cadre du projet de réhabilitation de 29 barrages à aiguilles. Voies Navigables de France – Biotope.

PHILIPPE, L., PRIÉ, V., CUCHERAT, X., PATRY, N., FROMENT, D., FAHRNER, G., WALTER, S., MOLINA, Q. & TAPKO, N. 2012. Étude d’impact du franchissement de la Vienne par la LGV SEA – Déplacement de la population des stations de Grande Mulette et de Mulette épaisse impactées. COSEA – Biotope.

PHILIPPE, L., PRIÉ, V., CUCHERAT, X., PATRY, N., FROMENT, D., JAMIER, M., FAHRNER, G., FAURE, B., MATHIEU, C., WALTER, S., MOLINA, Q. & TAPKO, N. 2012. Étude d’impact du franchissement de la Vienne par la LGV SEA – Déplacement de la population des stations de Grande Mulette et de Mulette épaisse impactées. COSEA – Biotope.

333 PHILIPPE, L., PRIÉ, V., CUCHERAT, X., TAPKO, N. & BOUSQUET, P. 2010. Étude d’impact du franchissement de la Vienne par la LGV SEA - Suivi de la population des stations de Grande Mulette et de Mulette épaisse potentiellement impactées. COSEA – Biotope.

PHILIPPE, L., PRIÉ, V. & PATRY, N. 2011. Étude d’impact du franchissement de la Vienne par la LGV SEA – Test de marquage et de déplacement sur la population de Grande Mulette impactée. COSEA – Biotope.

PLAUT, I. 1998. Comparison of salinity tolerance and osmoregulation in two closely related species of blennies from different habitats. Fish Physiology and Biochemistry, 19: 181–188.

PORTAFAIX, P., SENECAL, A. & BAISEZ, A. 2015. Evaluation de la phase juvénile de la lamproie marine (Petromyzon marinus) sur le bassin de la Vienne. Rapport ammocètes bassin de la Vienne, Association LOGRAMI.

PREECE, R. C. 1988. A second British interglacial record of Margaritifera auricularia. Journal of Conchology, 33: 50–51.

PREECE, R. C., BURLEIGH, R., KERNEY, M. P. & JARZEMBOWSKI, E. A. 1983. Radiocarbon age determinations of fossil Margaritifera auricularia (Spengler) from the River Thames in west London. Journal of Archaeological Science, 10: 249–257.

PRESTON, S. J., KEYS, A. & ROBERTS, D. 2007. Culturing freshwater pearl mussel Margaritifera margaritifera: a breakthrough in the conservation of an endangered species. Aquatic Conservation: Marine and Freshwater Ecosystems, 17: 539–549.

PRIÉ, V. 2010. Inventaire de la Grande mulette Margaritifera auricularia dans le Sud-Ouest de la France (I). Agence de l’Eau Adour-Garonne – Biotope.

PRIÉ, V. 2010. La Grande Mulette Margaritifera auricularia sur la Charente, complément d’inventaire sur le site Natura 2000 FR 5400-472. Ligue pour la Protection des Oiseaux – Biotope.

PRIÉ, V. 2010. Margaritifera auricularia. The IUCN Red List of Threatened Species 2010: e.T12798A3381899. http://dx.doi.org/10.2305/IUCN.UK.2010-4.RLTS.T12798A3381899.en

PRIÉ, V. 2011. Inventaire de la Grande mulette Margaritifera auricularia dans le Sud-Ouest de la France (II). Agence de l’Eau Adour-Garonne – Biotope.

334 PRIÉ, V. 2012. Inventaire de la Grande mulette Margaritifera auricularia dans le Sud-Ouest de la France (III). Agence de l’Eau Adour-Garonne – Biotope.

PRIÉ, V. 2013. Recherche et caractérisation des populations de Grande Mulette Margaritifera auricularia dans la Dronne. EPIDOR – Biotope.

PRIÉ, V. 2017. Newly overlapping ranges: First records of Potomida littoralis (Cuvier, 1798) infestation by the European Bitterling Rhodeus amarus (Bloch, 1782). Journal of Conchology, 42: 381–382.

PRIÉ, V., ADAM, B., CUCHERAT, LEGRAND, N., PATRY, N. & PHILIPPE, L. 2015. Étude biogéographique de la Grande Mulette Margaritifera auricularia en France. Synthèse bibliographique et recherches muséographiques. Université François-Rabelais de Tours – Biotope.

PRIÉ, V., ADAM, B., CUCHERAT, X., LEGRAND, N., PATRY, N. & PHILIPPE, L. 2016. Étude biogéographique de la Grande Mulette Margaritifera auricularia en France. Prospections dans le Cher, la Saône, la Vézère, l’Arros et la Garonne. Université François-Rabelais de Tours – Biotope.

PRIÉ, V. & BOUSQUET, P. 2010. La Grande Mulette Margaritifera auricularia dans le Sud- Ouest de la France, rapport intermédiaire prospections 2010. Rapport d’expertise. Agence de l’eau Adour-Garonne.

PRIÉ, V., BOUSQUET, P., SERENA, A., TABACCHI, E., JOURDE, P., ADAM, B., DESCHAMPS, T., CHARNEAU, M., TICO, T., BRAMARD, M. & COCHET, G. 2010. Nouvelles populations de Grande Mulette Margaritifera auricularia (Spengler, 1793) (Bivalvia, Margaritiferidae) découvertes dans le Sud-ouest de la France. MalaCo, 6: 294–297.

PRIÉ, V., COCHET, G. & PHILIPPE, L. 2008. La Grande Mulette Margaritifera auricularia dans l'Oise - Chronique d'une mort annoncée. Le Courrier de la Nature, 239: 20–24.

PRIÉ, V., COCHET, G., PHILIPPE, L., RETHORET, H. & FILALI, R. 2008. Une population majeure de la très rare Grande Mulette Margaritifera auricularia (Spengler, 1793) (Bivalvia : Margaritiferidae) dans le fleuve Charente (France). MalaCo, 5: 230–239.

335 PRIÉ, V., MOLINA, Q. & GAMBOA, B. 2014. French Naiad (Bivalvia: Margaritiferidae, Unionidae) species Distribution Models: prediction maps as tools for conservation. Hydrobiologia, 735: 81–94.

PRIÉ, V. & MOUTON, N. 2016. Bathymétrie de la Charente – Caractérisation de l’habitat favorable à la Grande Mulette à partir de relevés bathymétriques. Biotope - INGEO.

PRIÉ, V. & NÉRI, F. 2016. Recherche de la Grande Mulette Margaritifera auricularia sur l’Arros (Gers). Conservatoire d’Espaces Naturels Midi-Pyrénées – Biotope.

PRIÉ, V., PHILIPPE, L. & COCHET, G. 2007. Étude d’impact du projet de curage de la Charente sur les bivalves et recherche de la Grande Mulette Margaritifera auricularia. Etablissement Public Territorial de Bassin du fleuve Charente – Biotope.

PRIÉ, V., PHILIPPE, L. & COCHET, G. 2007. Étude d’impact du projet de canal Seine-Nord- Europe sur les Naïades – rivière Oise. Voies Navigables de France – Biotope.

PRIÉ, V., PHILIPPE, L. & COCHET, G. 2007. Evaluation de l’impact d’un projet de canal sur les naïades de l’Oise (France) et découverte de valves récentes de Margaritifera auricularia (Spengler, 1793) (Bivalvia : Margaritiferidae). MalaCo, 4: 178–182.

PRIÉ, V., PHILIPPE, L. & COCHET, G. 2011. Plan National d’Actions en faveur de la Grande Mulette Margaritifera auricularia – 2012-2017. Ministère de l’écologie, du développement durable, des transports et du logement, Paris.

PRIÉ, V., PHILIPPE, L. & PATRY, N. 2011. Recherche de la présence d'espèces patrimoniales de mollusques aquatiques dans la rivière le Cher. DREAL Centre – Biotope.

PRIÉ, V. & PUILLANDRE, N. 2014. Molecular phylogeny, taxonomy and distribution of French Unio species (Bivalvia, Unionidae). Hydrobiologia, 735: 95-110.

PRIÉ, V., PUILLANDRE, N. & BOUCHET, P. 2012. Bad taxonomy can kill: molecular reevaluation of Unio mancus Lamarck, 1819 (Bivalvia: Unionidae) and its accepted subspecies. Knowledge and Management of Aquatic Ecosystems, 405: DOI http://dx.doi.org/10.1051/kmae/2012014

PRIÉ, V., SOLER, J., ARAUJO, R., CUCHERAT, X., PHILIPPE, L., PATRY, N., ADAM, B., LEGRAND, N., JUGÉ, PH., RICHARD, N. & WANTZEN, K. M. 2018. Margaritifera

336 auricularia (Bivalvia: Unionida), a critically endangered flagship species for challenging downstream ecosystems exploration: results of last 10 years’ dedicated surveys. Hydrobiologia, 810: 157–175.

RAIKOW, D. F. & HAMILTON, S. K. 2001. Bivalve diets in a midwestern U. S. stream: a stable isotope enrichment study. Limnology & Oceanography, 46: 514–522.

REICHARD, M., DOUDA, K., PRZYBYŁSKI, M., POPA, O. P., KARBANOVA, E., MATASOVA, K., RYLKOVA, K., POLACIK, M., BLAZEK, R. &. SMITH, C. 2015. Population-specific responses to an invasive species. Proceedings of the Royal Society B, 282: 20151063.

REICHARD, M., ONDRACKOVA, M., PRZYBYLSKI, M., LIU, H. &. SMITH, C. 2006. The costs and benefits in an unusual symbiosis: experimental evidence that bitterling fish (Rhodeus sericeus) are parasites of unionid mussels in Europe. Journal of Evolutionary Biology, 19: 788– 796.

REICHARD, M., POLACIK, M., TARKAN, A. S., SPENCE, R., GAYGUSUZ, O., ERCAN, E., ONDRACKOVA, M. & SMITH, C. 2010. The bitterling mussel coevolutionary relationship in areas of recent and ancient sympatry. Evolution, 64: 3047–3056.

REID, R. G. B., McMAHON, R. F., Ó FOIGHIL, D. & FINNIGAN, R. 1992. Anterior inhalant currents and pedal feeding in Bivalves. The Veliger, 35: 93–104.

REIS, J. 2006. Atlas dos bivalves de água doce de Portugal continental. Instituto de Conservação da Natureza, Lisboa.

REIS, J. & ARAUJO, R. 2009. Redescription of Unio tumidiformis Castro, 1885 (Bivalvia, Unionidae), an endemism from the south-western Iberian Peninsula. Journal of Natural History, 43: 1929–1945.

REYNOLDS, J. D., DEBUSE, V. J. & ALDRIDGE, D. C. 1997. Host specialization in an unusual symbiosis: European bitterlings spawning in freshwater mussels. Oikos, 78: 539–545.

RIBEIRO, F., COLLARES-PEREIRA, M. J. & MOYLE, P. B. 2009. Non-native fish in the fresh waters of Portugal, Azores and Madeira Islands: a growing threat to aquatic biodiversity. Fisheries Management and Ecology, 16: 255–264.

337 RICCIARDI, A., NEVES, R. J. & RASMUSSEN, J. B. 1998. Impending extinctions of North American freshwater mussels (Unionoida) following the zebra mussel (Dreissena polymorpha) invasion. Journal of Animal Ecology, 67: 613–619.

RICCIARDI, A. & RASMUSSEN, J. B. 1999. Extinction rates in North American freshwater fauna. Conservation Biology, 13: 1220–1222.

RICHARD, N., JUGÉ, PH., RODRIGUES, S. 2015. Impacts de l'arasement du barrage de Maisons-Rouges sur la dynamique sédimentaire ainsi que sur les habitats et espèces faunistiques et floristiques de la Vienne et de la Creuse. Campagne 2012-15. CETU Elmis Ingénieries de l’Université François-Rabelais, Tours.

RICHARD, P. E., DIETZ, T. H. & SILVERMAN, H., 1991. Structure of the gill during reproduction in the unionids Anodonta grandis, Ligumia subrostrata and Carunculina parva texasensis. Canadian Journal of Zoology, 69: 1744–1754.

RICHMAN, N. L., BÖHM, M., ADAMS, S. B. et al. 2015. Multiple drivers of decline in the global status of freshwater crayfish (Decapoda: Astacidea). Philosophical Transactions of the Royal Society B, 370: 1–11.

ROE, K. J., SIMONS, A. M. & HARTFIELD, P. 1997. Identification of a fish host of the inflated heelsplitter Potamilus inflatus (Bivalvia: Unionidae) with a description of its glochidium. American Midland Naturalist, 138: 48–54.

ROSE, K. E., ATKINSON, R. L., TURNBULL, L. A. & REES. M. 2009. The costs and benefits of fast living. Ecology Letters, 12: 1379–1384.

ROSS, H. C. G. 1992. The reproductive biology of the freshwater pearl mussel Margaritifera margaritifera (L.) in Co Donegal. Irish Naturalist Journal, 24: 43–50.

RUBIO MILLÁN, C., MUÑOZ-CAMARILLO, G., SANZ BAYÓN, I., ÁLVAREZ HALCÓN, R. M. & CALVO TOMÁS, A. 2016. Estado de conservación de las náyades en el Ebro medio. Naturaleza Aragonesa, 33.

SAENKO, E. M., SHEDKO, M. B. & KHOLIN, S. K. 2001. Morphology and feature of bionomics of glochidia mollusks of the genus Beringiana (Bivalvia, Unionidae) of Kamchatka and Northern Kuriles. Vestnik Zoologii, 35: 59–68. (in Russian).

338 SALONEN, J. K., MARJOMÄKI, T. J. & TASKINEN, J. 2016. An alien fish threatens an endangered parasitic bivalve: the relationship between brook trout (Salvelinus fontinalis) and freshwater pearl mussel (Margaritifera margaritifera) in northern Europe. Aquatic Conservation: Marine and Freshwater Ecosystems, 26: 1130–1144.

SAMBROOK, J., FRITSCH E. F. & MANIATIS, T. 1989. Molecular Cloning. A Laboratory Manual. Second Edition. Cold Spring Harbor, New York.

SAN MIGUEL, E., MONSERRAT, S., FERNÁNDEZ, C., AMARO, R., HERMIDA, M., ONDINA, P. & ALTABA, C. R. 2004. Growth models and longevity of freshwater pearl mussels (Margaritifera margaritifera) in Spain. Canadian Journal of Zoology, 82: 1370–1379.

SCHALIE, H. 1970. Hermaphroditism among North American freshwater mussels. Malacologia, 10: 93–112.

SCHARTUM, E., MORTENSEN, S., PITTMAN, K. & JAKOBSEN, P. J. 2016. From pedal to filter feeding: ctenidial organogenesis and implications for feeding in the postlarval freshwater pearl mussel Margaritifera margaritifera (Linnaeus, 1758). Journal of Molluscan Studies, 83, 36–42. https://doi.org/10.1093/mollus/eyw037

SCHEDER, C., GUMPINGER, C. & CSAR, D. 2011. Application of a five-stage field key for the larval development of the freshwater pearl mussel (Margaritifera margaritifera Linné, 1758) under different temperature conditions. A tool for the approximation of the optimum time for host fish infection in captive breeding. Ferrantia, 64: 13–22.

SCHEDER, C., LERCHEGGER, B., JUNG, M., CSAR, D. & GUMPINGER, C. 2014. Practical experience in the rearing of freshwater pearl mussels (Margaritifera margaritifera): advantages of a work’ saving infection approach, survival, and growth of early life stages. Hydrobiologia, 735: 203–212.

SCHLUMBERGER, O., SAGLIOCCO, M. & PROTEAU, J. P. 2001. Biogéographie du silure glane (Silurus glanis): Causes hydrographiques, climatiques et anthropiques. Bulletin Français de la Peche et de la Pisciculture, 357: 533–547.

SCHMIDT, C. & VANDRÉ, R. 2010. Ten years of experience in the rearing of young freshwater pearl mussels (Margaritifera margaritifera). Aquatic Conservation: Marine and Freshwater Ecosystems, 20: 735–747.

339 SCHÖNE, B. R., DUNCA, E., FIEBIG, J. & PFEIFFER, M. 2005. Mutvei´s solution: An ideal agent for resolving microgrowth structures of biogenic carbonates. Palaeogeography, Palaeoclimatology, Palaeoecology, 228: 149–166.

SCHÖNE, B. R., DUNCA, E., MUTVEI, H. & NORLUND, U. 2004. A 217-year record of summer air temperature reconstructed from freshwater pearl mussels (M. margaritifera, Sweden). Quaternary Science Reviews, 23: 1803–1816.

SCHWARTZ, M. L. & DIMOCK, R. V. 2001. Ultrastructural evidence for nutritional exchange between brooding unionid mussels and their glochidia larvae. lnvertebrate Biology, 120: 227– 236.

SELLMER, G. P. 1967. Functional morphology and ecological life history of the gem clam, Gemma gemma (Eulamellibranchia: Veneridae). Malacologia, 5: 137–223.

SIBLY, R. M. & CALOW, P. 1986. Physiological Ecology of Animals. Blackwell Publications, Oxford.

SIETMAN, B., DAVIS, M., HOVE, M., PLETTA, M., WAGNER, T., MARR, S., SECRIST, Z., FREEBURG, M., SCHEUNEMANN, A., KRUPP, K., HAGEMEYER, E., FRANZEN, A., SWANSON, C. & SAMPSON, A. 20017. Cumberlandia monodonta—host enigma resolved. Ellipsaria, 19: 18–20.

SILVERMAN, H., KAYS, W. T. & DIETZ, T. H. 1987. Maternal calcium contribution to glochidial shells in freshwater mussels (Eulamellibranchia: Unionidae). Journal of Experimental Zoology, 242: 137–146.

SILVERMAN, H., NICHOLS, S. J., CHERRY, J. S., ACHBERGER, E., LYNN, J. W. & DIETZ, T. H. 1997. Clearance of laboratory-cultured bacteria by freshwater bivalves: differences between lentic and lotic unionids. Canadian Journal of Zoology, 75: 1857–1866.

SIMBERLOFF, D. & VON HOLLE, B. 1999. Positive interactions of nonindigenous species: invasional meltdown? Biological Invasions, 1: 21–32.

SMITH, C., REICHARD, M. JURAJDA, P. & PRZYBYLSKI, M. 2004. The reproductive ecology of the European bitterling (Rhodeus sericeus). Journal of Zoology, 262: 107–124.

340 SMITH, C., REYNOLDS, J. D. & SUTHERLAND, W. J. 2000. The population consequences of reproductive decisions. Proceedings of the Royal Society of London B Biological Sciences, 267: 1327–1334.

SMITH, C., RIPPON, K., DOUGLAS, A. & JURAJDA, P. 2001. A proximate cue for oviposition site choice in the bitterling (Rhodeus sericeus). Freshwater Biology, 46: 903–911.

SMITH, D. G. 1976. Notes on the biology of Margaritifera margaritifera (Lin.) in central Massachusetts. American Midland Naturalist, 96: 252–256.

SMITH, D. G. 1979. Marsupial anatomy of the demibranch of Margaritifera margaritifera (Lin.) in northeastern North America (Pelecypoda: Unionacea). Journal of Molluscan Studies, 45: 39–44.

SMITH, D. G. 1980. Anatomical studies on Margaritifera margaritifera and Cumberlandia monodonta (Mollusca: Pelecypoda: Margaritiferidae). Zoological Journal of the Linnean Society, 69: 257–270.

SMITH, D. G. 1985. Recent range expansion of the freshwater mussel Anodonta implicata and its relationship to clupeid fish restoration in the Connecticut river system. Freshwater Invertebrate Biology, 4: 105–108.

SMITH, D. G. 1988. Notes on the biology and morphology of Margaritifera hembeli (Conrad, 1838) (Unionacea: Margaritiferidae). Nautilus, 102: 159–163.

SMITH, D. G. 2001. Systematics and distribution of Recent Margaritiferidae. In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 33–49. Springer-Verlag, Berlin Heidelberg.

SMITH, D. G. & HARTEL, K. E. 1999. Margaritiferidae (Mollusca: Unionoida): possible hosts for Rhodeus (Piscies: Cyprinidae). Polskie Archiwum Hydrobiologii, 46: 277–281.

SOLER, J., BOISNEAU, C., WANTZEN, K. M., & ARAUJO, R. 2018. Gasterosteus aculeatus Linnaeus, 1758, a new host fish for the endangered Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, DOI:10.1093/mollus/eyy038

341 SOLER J., WANTZEN, K. M., JUGÉ, PH. & ARAUJO, R. 2018. Brooding and glochidia release in Margaritifera auricularia (Spengler, 1793) (Unionoida, Margaritiferidae). Journal of Molluscan Studies, 84: 182–189.

SOUCHON, Y. & TISSOT, L. 2012. Synthesis of thermal tolerances of the common freshwater fish species in large Western Europe rivers. Knowledge and Management of Aquatic Ecosystems, 405: 23–70.

SOUSSA, R., NOVAIS, A., COSTA, R. & STRAYER, D. L. 2014. Invasive bivalves in fresh waters: impacts from individuals to ecosystems and possible control strategies. Hydrobiologia, 735: 233–255.

SOUSA, R., VARANDAS, S., TEIXEIRA, A., GHAMIZI, M., FROUFE, E. & LOPES-LIMA, M. 2016. Pearl mussels (Margaritifera marocana) in Morocco: Conservation status of the rarest bivalve in African fresh waters. Science of The Total Environment, 547:405–412.

SPENCE, R. & SMITH, C. 2013. Rose bitterling (Rhodeus ocellatus) embryos parasitize freshwater mussels by competing for nutrients and oxygen. Acta Zoologica, 94: 113–118.

SPENGLER, L. 1793. Om Slaegterne Chaena, Mya og Unio. Skrifter af Naturhistorie Selskabet, 3: 16–69.

SPOONER, D. E. & VAUGHN, C. C. 2006. Context-dependent effects of freshwater mussels on stream benthic communities. Freshwater Biology, 51: 1016–1024.

STASEK, C. R. 1963. Synopsis and discussion of the association of ctenidia and labial palps in the bivalved mollusca. The Veliger, 6: 91–97.

STERKI, V. 1903. Notes on the Unionidae and their classification. American Naturalist, 37: 103–1013.

STERN, E. M. & FELDER, D. L. 1978. Identification of host fishes for four species of freshwater mussels (Bivalvia: Unionidae). American Midland Naturalist, 100: 233–236.

STOECKL, K., TAEUBERT, J. E. & GEIST, J. 2015. Fish species composition and host fish density in streams of the thick-shelledriver mussel (Uniocrassus) – implications for conservation. Aquatic Conservation: Marine and Freshwater Ecosystems, 25: 276–287.

342 STOECKLE, B. C., ARAUJO, R., GEIST, J., KUEHN, R., TOLEDO, C. & MACHORDOM, A. 2016. Strong genetic differentiation and low genetic diversity of the freshwater pearl mussel (Margaritifera margaritifera L.) in the southwestern European distribution range. Conservation Genetics https://doi.org/ 10.1007/s10592-016-0889-3.

STRAYER, D. L. 2006. Challenges for freshwater invertebrate conservation. Journal of the North American Benthological Society, 25: 271–287.

STRAYER, D. L. 2008. Freshwater mussel ecology. A Multifactor Approach to Distribution and Abundance. Freshwater Ecology Series, University of California Press, Berkeley.

STRAYER, D. L., DOWNING, J. A., HAAG, W. R., KING, T. L., LAYZER, J. B., NEWTON, T. J. & NICHOLS, S. J. 2004. Changing perspectives on pearly mussels, North America’s most imperiled animals. BioScience, 54: 429–439.

STRAYER, D. L. & MALCOM, H. M. 2012. Causes of recruitment failure in freshwater mussel populations in southeastern New York. Ecological Applications, 22: 1780–1790.

SYVÄRANTA, J., CUCHEROUSSET, J., KOPP, D., MARTINO, A., CEREGHINO, R. & SANTOUL, F. 2009. Contribution of anadromous fish to the diet of European catfish in a large river system. Naturwissenschaften, 96: 631–635.

TAEUBERT, J. E., GUM, B. & GEIST, J. 2012. Host-specificity of the endangered thick- shelled river mussel (Unio crassus, Philipsson 1788) and implications for conservation. Aquatic Conservation: Marine and Freshwater Ecosystems, 22: 36–46.

TAEUBERT, J. E., GUM, B. & GEIST, J. 2013. Variable development and excystment of freshwater pearl mussel (Margaritifera margaritifera) at constant temperature. Limnologica – Ecology and Management of Inland Waters, 43, 319–322.

TAEUBERT, J. E., POSADA MARTINEZ, A. M., GUM, B. & GEIST, J. 2012. The relationship between endangered thick-shelled river mussel (Unio crassus) and its host fishes. Biological Conservation, 155: 94–103.

TANKERSLEY, R. A. 1996. Effect of larval brooding on the feeding physiology of freshwater unionid mussels. Invertebrate Biology, 115: 243–255.

343 TANKERSLEY, R. A. & DIMOCK, R. V. 1992. Quantitative analysis of the structure and function of the marsupial gills of the freshwater mussel Anodonta cataracta. Biological Bulletin, 182: 145–184.

TANKERSLEY, R. A. & DIMOCK, R. V. 1993. The effect of larval brooding on the respiratory physiology of the freshwater unionid mussel Pyganodon cataracta. American Midland Naturalist, 130: 146–163.

TATOJ, K., CMIEL, A. M., KWASNA, D., LIPINASKA, A. M., ZAJAK, K. &. ZAJAC, T. 2017. The endangered thick-shelled river mussel (Unio crassus): a new host species for the European bitterling (Rhodeus amarus). Biodiversity and Conservation, 26: 1217–1224.

TAVERNY, C. & ELIE, P. 2010. Les lamproies en Europe de l’Ouest. Écophases, especes et habitats. Edition Quae, Paris.

TAVERNY, C., URDACI, M., ELIE, A.-M., BEAULATON, L., ORTUSI, I., DAVERAT, F. & ELIE, P. 2005. Biologie, écologie et pêche des lamproies migratrices (Agnathes amphihalins). Rapport no. 99, Groupement de Bordeaux, CEMAGREF.

TEIXEIRA, A., DIAS, A. R., MIRANDA, F., LOPES-LIMA, M., VARANDAS, S., FROUFE, E., FILIPE, A. F., BEJA, P. & SOUSA, R. 2018. Can invasive alien fish species act as effective hosts of native freshwater mussels (Unionidae) in Iberia? In: 2018 Iberian Congress of Ichtiology. Faro, Portugal.

TEUTSCH, S. 1997. Populationsbiologische Untersuchungen an Anodonta: Fortpflanzungsstrategien und Charakterisierung von Teilpopulationen innerhalb eines Stillgewassers. Diplomarbeit, Lehrstuhl fur Okologie, Morphologie und Evolutionsbiologie, Universitat Freiburg.

THIELEN, F. 2011. Rearing of unionoid mussels (with special emphasis on the Freshwater Pearl Mussel Margaritifera margaritifera). Ferrantia, 64: 5–66.

THRELFALL, W. 1986. Seasonal occurrence of Anodonta cataracta Say 1817, glochidia on three-spined sticklebacks, Gasterosteus aculeatus (Linneaus). Veliger, 29: 231–234.

344 TRIANTAFYLLIDIS, A., KRIEG, F., COTTIN, C., ABATZOPOULOS, T. J., TRIANTAPHYLLIDIS, C. & GUYOMARD, R. 2002. Genetic structure and phylogeography of European catfish (Silurus glanis) populations. Molecular Ecology, 11: 1039–1055.

TRUMP, M. P. 2010. Feeding morphology of juvenile Utterbackia imbecillis (Bivalvia: Unionidae) MD, Department of Biology, Wake Forest University.

TUCKER, M. E. 1927. Morphology of the glochidium and juvenile of the mussel Anodonta imbecillis. Transactions of the American Microscopical Society, 46: 286–293.

UTHAIWAN, K., CHATCHAVALVANICH, K., NOPARATNARAPORN, N. & MACHADO, J. 2001. Scanning electron microscopy of glochidia and juveniles of the freshwater mussel, Hyriopsis myersiana. Invertebrate Reproduction and Development, 40: 143–151.

VALADOU, B. 2007. Le silure glane (Silurus glanis, L.) en France. Evolution de son aire de répartition et prédiction de son extension. Institut de recherche et de Développement, Conseil Supérieur de la Pêche Protection des Milieux, CEMAGREF.

VALENCIENNES, A., 1829-1848. Histoire naturelle des Poissons. Bertrand ed., Paris.

VALLEDOR, A. & ARAUJO, R. 2006. The historical misidentification of Margaritifera auricularia for M. margaritifera (Bivalvia, Unionoidea) explained by their iconography. Malacologia, 48: 285−294.

VAN DAMME, D., BOGUTSKAYA, N., HOFFMANN, R. C. &. SMITH, C. 2007. The introduction of the European bitterling Rhodeus amarus to west and central Europe. Fish and Fisheries, 8: 79–106.

VAUGHN, C. C. & HAKENKAMP, C. C. 2001. The functional role of burrowing bivalves in freshwater ecosystems. Freshwater Biology, 46: 1431–1446.

VAUGHN, C. C., NICHOLS, S. J. & SPOONER, D. E. 2008. Community and foodweb ecology of freshwater mussels. Journal of the North American Benthological Society, 27: 409– 423.

VAUGHN, C. C. & SPOONER, D. E. 2006. Scale-dependent associations between native freshwater mussels and invasive Corbicula. Hydrobiologia, 568: 331–339.

345 VAUGHN, C. C., SPOONER, D. E. & GALBRAITH, H. S. 2007. Context- dependent species identity effects within a functional group of filter-feeding bivalves. Ecology, 88: 1654–1662.

VAUGHN, C. C. & TAYLOR, C. M. 1999. Impoundements and the decline of freshwater mussels: a case study of an extinction gradient. Conservation Biology, 13: 912–920.

VENIOT, A., BRICELJ, V. M. & BENINGER, P. G. 2003. Ontogenetic changes in gill morphology and potential significance for food acquisition in the scallop Placopecten magellanicus. Marine Biology, 142: 123–131.

VICENTINI, H. 2005. Unusual spurting behaviour of the freshwater mussel Unio crassus. Journal of Molluscan Studies, 71: 409–410.

VILLASANTE, F., ABAD, J., ARAUJO, R. & BALSET, J. 2016. Aportación al conocimiento de la presencia de Margaritifera auricularia (Spengler, 1793) en el cauce del río Tajo (España). Iberus, 34: 79–82.

WÄCHTLER, K., DREHER-MANSUR, M. C. & RICHTER, T. 2001. Larval Types and Early Postlarval Biology in Naiads (Unionoida). In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer, G. & K. Wächtler, eds), pp. 93–123. Springer-Verlag, Berlin Heidelberg.

WALKER, K. F., BYRNE, M., HICKEY, C. W. & ROPER, D. S. 2001. Freshwater mussels (Hyriidae) of Australasia. In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 5–31. Springer-Verlag, Berlin Heidelberg.

WATTERS, G. T. 1992. Unionids, fishes, and the species–area curve. Journal of Biogeography, 19: 481–490.

WATTERS, G. T. 1995. Small dams as barriers to freshwater mussels (Bivalvia, Unionoida) and their hosts. Biological Conservation, 75: 79–85.

WATTERS, G. T. 2001. The evolution of the Unionacea in North America, and its implications for the worldwide fauna. In: Ecology and evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 281–307. Springer-Verlag, Berlin Heidelberg.

WELKER, M. & WALZ, N. 1998. Can mussels control the plancton in rivers? A planktological approach applying a Lagrangian sampling strategy. Limnology & Oceanography, 43: 753–762.

346 WELLMANN, G. 1943. Fischinfektionen mit glochidien der Margaritana margaritifera. Zeitschrift fuer Fischerei Band, 41: 385–390.

WHEELER, A. & MAINTLAND, P. S. 1973. The scarcer freshwater fishes of the British Isles I. Introduced species. Journal of Fish Biology, 5: 49–68.

WHITE, G. C. & BURNHAM, K. P. 1999. Program MARK: Survival estimation from populations of marked animals. Bird Study, 46: 120–138.

WIEPKEMA, P. R. 1961. An ethological analysis of the reproductive behaviour of the bitterling (Rhodeus amarus Bloch). Archives Néerlandaises de Zoologie, 14: 103–199.

WILES, M. 1974. The glochidia of certain Unionidae (Mollusca) in Nova Scotia and their fish hosts. Canadian Journal of Zoology, 53: 33–41.

WILLIAMS, J. D., WARREN, M. L., CUMMINGS, K. S., HARRIS, J. L. & NEVES, R. J. 1993. Conservation status of the freshwater mussels of the United States and Canada. Fisheries, 18: 6–22.

WOOD, E. M. 1974. Development and morphology of the glochidium larva of Anodonta cygnea (Mollusca: Bivalvia). Journal of Zoology, 173: 1–13.

WRIGHT, A. H. 1995. Particle selection and feeding morphology of in vitro and in vivo transformed juvenile Uterbackia imbecillis. Master thesis, Wake Forest University, Winston- Salem, North Carolina.

YASUNO, N., SHINDO, K., TAKAGI, Y., KANAYA, G., SHIKANO, S., FUJIMOTO, Y., SHIMADA, T., KIKUCHI, E. 2014. Ontogenetic changes in the trophic position of a freshwater Unionidae mussel. Fundamental & Applied Limnology, 184: 341–349.

YEAGER, M. M., CHERRY, D. S. & NEVES, R. J. 1994. Feeding and burrowing behavior of juvenile rainbow mussels, Villosa iris (Bivalvia: Unionidae). Journal of the North American Benthological Society, 13: 217–222.

YEAGER, M. M., NEVES, R. J. & CHERRY, D. S. 1999. Competitive interactions between early life stages of Villosa iris (Bivalvia: Unionidae) and adult Asian clams (Corcicula fluminea). In: First Freshwater Mollusk Conservation Society Symposium. Ohio Biological Survey, Ohio.

347 YOUNG, M., COSGROVE, P. & HASTIE, L. 2001. The extent of, and causes for, the decline of a highly threatened naiad: Margaritifera margaritifera. In: Ecology and Evolutionary Biology of the freshwater mussels Unionoidea (G. Bauer & K. Wächtler, eds.), pp. 337–357. Springer- Verlag, Berlin Heidelberg.

YOUNG, M. R. & WILLIAMS, J. C. 1984. The reproductive biology of the freshwater pearl mussel Margaritifera margaritifera (Linn.) in Scotland I. Field studies. Archiv für Hydrobiologie, 99: 405–422.

YOUNG, M. & WILLIAMS, J. 1984. The reproductive biology of the freshwater pearl mussel Margaritifera margaritifera (Linn.) in Scotland II. Laboratory studies. Archiv für Hydrobiologie, 100: 29–43.

YOUSON, J. H. 1980. Morphology and physiology of lamprey metamorphosis. Canadian Journal of Fisheries and Aquatic Sciences, 37: 1687−1710.

ZAJAC, K., ZAJAC, T. & CMIEL, A. 2017. What can we infer from the shell dimensions of the thick-shelled river mussel Unio crassus? Hydrobiologia, DOI 10.1007/s10750-017-3098-2

ZALE, A. V. & NEVES, R. J. 1982: Fish host of four species of lampsiline mussels (Mollusca: Unionidae) in Big Moccasin Creek. Canadian Journal of Zoology, 60: 2535–2542.

ZAPATER, M., ARAUJO, R., ÁLVAREZ HALCÓN, R., NAKAMURA, K. & ALCÁNTARA DE LA FUENTE, M. (coord.). 2006. Las almejas de agua dulce en Aragón: Margaritifera auricularia y otros bivalvos. Serie Especie. Consejo de Protección de la Naturaleza de Aragón, D.L.

ZHULKOV, A. I. & NIKIFOROV, S. N. 1998. Some data on the morphology and biology of the bitterling Rhodeus sericeus of the Tym River, Sakhalin. Journal Voprosy Ikhtiologii, 28:149–153.

ZIERITZ, A., HOFFMAN, J. L., AMOS, W. & ALDRIDGE, D. C. 2010. Phenotypic plasticity and genetic isolation-by-distance in the freshwater mussel Unio pictorum (Mollusca: Unionoida). Evolutionary Ecology, 24: 923–938.

ZIUGANOV, V., KALUZHIN, S., BELETSKY, V. & POPKOVITCH, E. 2001. The pearl mussel-salmon community in the Varzuga River, Northwest Russia: problems of environmental

348 impacts. In: Ecology and Evolution of the freshwater mussels Unionoida (G. Bauer & K. Wächtler, eds.), pp. 359–366. Springer-Verlag, Berlin Heidelberg.

ZIUGANOV, V., ZOTIN, A., NEZLIN, L. & TRETIAKOV, V. 1994. The freshwater pearly mussels and their relationships with salmonid fish. Vniro Publishing House, Moscow.

349

350 Appendices

351

352

Appendix 1 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 1

353

354 181010_M aur Vol 1 Synthesis on existing literature_final.docx

The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook

Volume 1 – Synopsis on the current scientific literature by Joaquin Soler, Rafael Araujo, and Karl M. Wantzen

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 1

355

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 2

356 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook

Volume 1 – Synopsis on the current literature by Joaquin Soler, Rafael Araujo, and Karl M. Wantzen

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 3

357 This publication should be cited as follows:

Soler, J., Araujo R. & Wantzen K. M. (2018): The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 1 – Synopsis on the current literature. University of Tours, France, CNRS UMR CITERES, LIFE+ project 13BIO/FR/001162 „Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe“. Tours (France) 68 pp.

Notice: This text contains graphs that are protected by copyrights. It must not be reproduced without having permits for reproduction.

Frontispiece: Margaritifera auricularia, photographed (c) by Philippe Jugé

This publication was conducted within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’ with the contribution of the financial instrument LIFE of the European Union. This paper has been produced under the auspices of the UNESCO Chair “Fleuves et Patrimoine / River Culture”.

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 4

358 Outline:

The Giant Freshwater Pearl Mussel (Margaritifera auricularia) is one of the rarest invertebrate species worldwide. This two-volume book aims to bring together all the so far available information on the species. Both volumes are independent books, although they can be seen as complimentary, giving scientific and technical information. They result from work by the authors in the context of the LIFE+ project 13BIO/FR/001162 „Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe“ and include additional work by the contributors from other conservation and research projects.

Volume 1 is dedicated to a synopsis of the current knowledge about the species, their biology and ecology, environmental impacts that have led to the reduction of the populations, as well suggestions to save this and other unionoid species in the Anthropocene. For a review on conservation and rearing techniques, please refer to Volume 2.

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 5

359 Contributors and their affiliations:

Joaquin Soler CNRS UMR CITERES, Université de Tours, 33, Av. F. de Lesseps, 37200 Tours, France, and Museo Nacional de Ciencias Naturales-CSIC, c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain, E-mail: [email protected]

Rafael Araujo Museo Nacional de Ciencias Naturales-CSIC, c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain, E-mail: [email protected]

Karl M. Wantzen UNESCO Chair for River Culture/Fleuves et Patrimoine CNRS UMR CITERES, Université de Tours, 33, Av. F. de Lesseps, 37200 Tours, France, E-mail: [email protected]

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 6

360

Table of contents:

Preface: Why study freshwater mussels, specifically M. auricularia? ...... 9 1. History of studies on M. auricularia ...... 11 2. Taxonomic position of M. auricularia in the context of evolution of Unionoida...... 13 2.1 Evolution and global diversity of Unionoida ...... 13 2.1.1 Systematics of M. auricularia ...... 15 3. Morphology of Unionoida, specifically M. auricularia ...... 17 3.1 Unionoida terminology and basic anatomy ...... 17 3.2 Morphology of M. auricularia ...... 20 4. Life cycle of Unionoida, specifically M. auricularia ...... 21 4.1 Growth and longevity ...... 21 4.2 Reproductive Strategies ...... 23 4.3 Parasitic Stage ...... 26 4.4 Post parasitic stage (with some details about feeding ecology) ...... 29 5. Biogeography and current distribution of M. auricularia ...... 34 5.1 Prehistorical, historical, and extant distributions of M. auricularia ...... 34 1.2.5. Population structure ...... 36 6. Habitat structure of sites where M. auricularia occurs ...... 38 7. Known and potential host fish of M. auricularia ...... 41 8. Human impacts on and conservation of M. auricularia ...... 47 8.1 Conservation status of M. auricularia ...... 47 8.2 Human impacts conducive to the population decrease of M. auricularia ...... 48 8.3 Conservation activities in favour of M. auricularia ...... 52 9.Conclusions – (How) can Margaritifera auricularia be saved in the Anthropocene?...... 54 10. References ...... 56

Margaritifera auricularia handbook vol. 1: Synopsis of current literature 7

361

362

Appendix 2 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 2

363

364 181010_M auricularia handbook vol 2 Technical manual_final version.docx The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook

Volume 2 – Technical Manual: Monitoring, artificial reproduction, rearing techniques, and suggestions for habitat conservation coordinated by Karl M. Wantzen and Rafael Araujo with contributions by Joaquin Soler, Catherine Boisneau, Nina Richard, Philippe Jugé, Yann Guerez, Laure Morisseau, Michèle De Monte, Keiko Nakamura and Vincent Prié

Margaritifera auricularia handbook vol. 2: Technical manual 1 365

Margaritifera auricularia handbook vol. 2: Technical manual 2 366 The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook

Volume 2 – Technical Manual: Monitoring, artificial reproduction, rearing techniques, and suggestions for habitat conservation coordinated by Karl M. Wantzen and Rafael Araujo, with contributions by Joaquin Soler, Catherine Boisneau, Nina Richard, Philippe Jugé, Yann Guerez, Laure Morisseau, Michèle De Monte, Keiko Nakamura and Vincent Prié

Margaritifera auricularia handbook vol. 2: Technical manual 3 367 This publication should be cited as follows:

Wantzen K. M. and Araujo R. (eds, 2018): The Giant Freshwater Pearl Mussel (Margaritifera auricularia) Handbook Volume 2 – Technical Manual: Monitoring, artificial reproduction, rearing techniques, and suggestions for habitat conservation with contributions by Karl M. Wantzen, Rafael Araujo, Joaquin Soler, Catherine Boisneau, Nina Richard, Philippe Jugé, Yann Guerez, Laure Morisseau, Michèle De Monte, Keiko Nakamura and Vincent Prié University of Tours, France, CNRS UMR CITERES, LIFE+ project 13BIO/FR/001162 „Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe“. Tours (France) 109 pp. Frontispiece: Margaritifera auricularia, photographed (c) by Philippe Jugé

This publication was conducted within the scope of the LIFE project ‘LIFE13 BIO/FR/001162 Conservation of the Giant Pearl Mussel in Europe’ with the contribution of the financial instrument LIFE of the European Union. This paper has been produced under the auspices of the UNESCO Chair “Fleuves et Patrimoine / River Culture”.

Margaritifera auricularia handbook vol. 2: Technical manual 4 368

Outline:

The Giant Freshwater Pearl Mussel (Margaritifera auricularia) is one of the rarest invertebrate species worldwide. This two-volume book aims to bring together all the so far available information on the species. Both volumes are independent books, although they can be seen as complimentary, giving scientific and technical information. They result from work by the authors in the context of the LIFE+ project 13BIO/FR/001162 „Conservation of the Giant Freshwater Pearl Mussel (Margaritifera auricularia) in Europe“ and include additional work by the contributors from other conservation and research projects.

Volume 2 is a manual that focuses on the practical aspects. It delivers hands-on information on how to find then animals in the field, how to identify them, how to study their habitats, how to prepare and run a laboratory for artificial reproduction and rearing, informs about release techniques, as well as descriptions of methods how to reinforce populations by infesting and releasing alternative host fish, or how to preserve mussel habitats. For a review of the literature on the species, please refer to Volume 1.

Margaritifera auricularia handbook vol. 2: Technical manual 5 369 Contributors and their affiliations:

Karl M. Wantzen UNESCO Chair for River Culture/Fleuves et Patrimoine CNRS UMR CITERES, Université de Tours, 33, Av. F. de Lesseps, 37200 Tours, France, E-mail: [email protected]

Rafael Araujo Museo Nacional de Ciencias Naturales-CSIC, c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain. E-mail: [email protected]

Joaquin Soler CNRS UMR CITERES, Université de Tours, 33, Av. F. de Lesseps, 37200 Tours, France, and Museo Nacional de Ciencias Naturales-CSIC, c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain. E-mail: [email protected]

Catherine Boisneau CNRS UMR CITERES, Université de Tours, 33, Av. F. de Lesseps, 37200 Tours, France, E-mail: [email protected]

Michèle de Monte PST Animaleries de l'Université de Tours, 60 rue du Plat d'Étain, 37000 Tours, E-mail: [email protected]

Laure Morisseau Projet LIFE "Grande Mulette", Université de Tours, UMR 7324 CITERES 11 quai Danton, 37500 Chinon, France, E-mail:[email protected]

Nina Richard, Philippe Jugé, Yann Guerez Projet LIFE "Grande Mulette", et CETU Elmis Université de Tours, 11 quai Danton, 37500 Chinon, France, E-mails [email protected], [email protected], [email protected]

Keiko Nakamura Environmental Service Department, Sociedad Aragonesa de Gestión Agroambiental (SARGA), Ranillas Avenue, 5-A, 3th floor, 50018 Zaragoza, Spain, e-mail: [email protected];

Vincent Prié Biotope, Service Recherche et Développement, 22 Bd Maréchal Foch 34 530 Mèze, France, Email [email protected]

Margaritifera auricularia handbook vol. 2: Technical manual 6 370 Table of contents:

Preface 8 1. Conservation status of Margaritifera auricularia and legal restrictions 10 2. Conservation status of host fish of Margaritifera auricularia and legal restrictions 12 3. Monitoring techniques to find Margaritifera auricularia in the field 16 3.1 Field observation using the Aquascope (valid for shallow waters) 17 3.2 Field observation using SCUBA devices (valid for deeper waters) 18 3.3 e-DNA sampling 20 3.4 Marking techniques 20 4. Preliminary considerations about artificial reproduction and rearing of M. auricularia 22 5. 5. Technical advices for building up a laboratory for artificial rearing of M. auricularia 30 6. Breeding period, sampling of gravid adults, transport and sampling of glochidia in the laboratory 38 6.1. Assessment of breeding period and glochidial release 38 6.2. Sampling gravid adults and transport to the laboratory 42 6.3. Sampling of glochidia from gravid adults in the laboratory 43 7. Maintenance and infestation of host fish in aquaria 46 7.1 Preparing aquaria and filters for keeping host fish 47 7.2 Maintenance of fish prior to infestation 48 7.3 Infestation of host fish 51 7.4 Maintenance and water quality control of host fish in cylindroconic tanks 55 8. Sampling and cleaning of Margaritifera auricularia juveniles when excysting from host fish gills 58 9. Different rearing techniques for juveniles of Margaritifera auricularia 63 9.1. “Detritus boxes” and Multiple beakers (drop by drop systems) 64 9.2. Artificial flumes 67 9.3. Types of diets 72 9.4. Monitoring of captive juveniles 75 10. Identification of reintroduction sites 77 11. Different reintroduction techniques for juveniles of Margaritifera auricularia 86 12. Monitoring of released juveniles of Margaritifera auricularia in the field 91 13. Search for, infestation and release of alternative host 93 13.1 Test for natural infestation 93 13.2 Identification of the glochidia on the gills 94 13.3. Onsite infestation and release of host fish 96 14. List of known and potential stressors, and protective measures to preserve M. auricularia 97 15. Appendix: Questionnaire for unionoid mussel raising 101 16. References 106

Margaritifera auricularia handbook vol. 2: Technical manual 7 371

372 Joaquín SOLER

Conservation ecology of Margaritifera auricularia (Spengler, 1793) in France Résumé

L'objectif général de cette thèse était d'améliorer les connaissances sur l'état de conservation des populations françaises de Margaritifera auricularia, la biologie de la reproduction et les premiers stades de vie de l'espèce, permettant ainsi de mieux comprendre le degré et les causes de son déclin et de proposer des mesures de conservation. Pour ce faire, quatre aspects ont été ciblés: l'évolution de l'état de conservation de l'espèce, déduit des populations actuellement connues, les références bibliographiques et les collections de musée; le processus d’incubation, y la fécondité et le comportement et le moment de la libération des glochidies; le stade parasite des moules et l'identification de nouvelles espèces hôtes, ainsi que les premiers stades de vie des moules d'eau douce et l'élevage des juvéniles de M. auricularia.

Mots clés

Margaritifera auricularia; moules d'eau douce; incubation; libération de glochidies; fécondité; statut de conservation; poisson hôte; juvéniles; élevage en captivité.

Résumé en anglais

The general goal of this thesis was to improve the knowledge on the conservation status of the Margaritifera auricularia French populations, the reproductive biology and early life stages of the species, allowing to better understand the degree and causes of its decline and to propose conservation measures. To do so, four aspects were focused: the evolution of the conservation status of the species, as inferred by its currently known populations, bibliographic references and museum collections; the brooding process, including the brood size and the glochidia release behaviour and timing; the parasitic stage of the mussels and the identification of new host species, and the early stages of freshwater mussels and the artificial rearing of M. auricularia.

Keywords:

Margaritifera auricularia; freshwater mussel; brooding; glochidia release; fecundity; conservation status; host fish; juveniles; captive rearing.

373