UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO POSGRADO EN CIENCIAS BIOLÓGICAS INSTITUTO DE ECOLOGÍA BIOLOGIA EVOLUTIVA Estructura genética y filogeografía del mangle negro, Avicennia germinans (L.) L. (Avicenniaceae), en México TESIS QUE PARA OPTAR POR EL GRADO DE: DOCTORA EN CIENCIAS PRESENTA: MARIED OCHOA ZAVALA TUTOR PRINCIPAL DE TESIS: Dr. Juan Núñez Farfán Instituto de Ecología, UNAM COMITÉ TUTOR: Dr. Daniel Piñero Dalmau Instituto de Ecología, UNAM Dr. Alejandro Nettel Hernanz Instituto de Ciencias Biológicas, UNICACH CD. MX. NOVIEMBRE 2019 UNAM – Dirección General de Bibliotecas Tesis Digitales Restricciones de uso DERECHOS RESERVADOS © PROHIBIDA SU REPRODUCCIÓN TOTAL O PARCIAL Todo el material contenido en esta tesis esta protegido por la Ley Federal del Derecho de Autor (LFDA) de los Estados Unidos Mexicanos (México). El uso de imágenes, fragmentos de videos, y demás material que sea objeto de protección de los derechos de autor, será exclusivamente para fines educativos e informativos y deberá citar la fuente donde la obtuvo mencionando el autor o autores. Cualquier uso distinto como el lucro, reproducción, edición o modificación, será perseguido y sancionado por el respectivo titular de los Derechos de Autor. UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO POSGRADO EN CIENCIAS BIOLÓGICAS INSTITUTO DE ECOLOGÍA BIOLOGÍA EVOLUTIVA Estructura genética y filogeografía del mangle negro, Avicennia germinans (L.) L. (Avicenniaceae), en México TESIS QUE PARA OPTAR POR EL GRADO DE: DOCTORA EN CIENCIAS PRESENTA: MARIED OCHOA ZAVALA TUTOR PRINCIPAL DE TESIS: Dr. Juan Núñez Farfán Instituto de Ecología, UNAM COMITÉ TUTOR: Dr. Daniel Piñero Dalmau Instituto de Ecología, UNAM Dr. Alejandro Nettel Hernanz Instituto de Ciencias Biológicas, UNICACH MÉXICO, CD. MX. NOVIEMBRE 2019 POSGRADO POLO , TENCIJIAS * BIOLÓGICAS? COORDINACIÓN DEL POSGRADO EN CIENCIAS BIOLÓGICAS ENTIDAD INSTITUTO DE ECOLOGÍA OFICIO CPCB/1154/2019 ASUNTO: Oficio de Jurado M. en C. Ivonne Ramírez Wence Directora General de Administración Escolar, UNAM Presente Me permito informar a usted que en la reunión ordinaria del Subcomité de Ecología y Biología Evolutiva, en su sesión ordinaria del día 26 de agosto de 2019 se aprobó el siguiente jurado para el examen de grado de DOCTORA EN CIENCIAS del estudiante OCHOA ZAVALA MARIED con número de cuenta 514012300 con la tesis titulada "ESTRUCTURA GENÉTICA Y FILOGEOGRAFÍA DEL MANGLE NEGRO (Avicennia germinans (L) L. (Avicenniaceae)”, realizada bajo la dirección del DR. JUAN SERVANDO NÚÑEZ FARFÁN Presidente: DR. JUAN PABLO JARAMILLO CORREA Vocal: DR. PÍNDARO DÍAZ JAIMES Secretario: DR. DANIEL IGNACIO PIÑERO DALMAU Suplente: DR. MIGUEL ÁNGEL PÉREZ FARRERA Suplente DR. CRISTIAN TOVILLA HERNÁNDEZ Sin otro particular, me es grato enviarle un cordial saludo. ATENTAMENTE ' “POR MI RAZA HABLARA EL ESPÍRITU” Ciudad Universitaria, Cd. Mx., a 30 de octubre de 2019 COORDINADOR DEL PROGRAMA DR. ADOLFO GERARDO NAVARRO SIGÚENZA c. Cc. p. Expediente del alumno COORDINACIÓN DEL POSGRADO EN CIENCIAS BIOLOGICAS UNIDAD DE POSGRADO Edificio D, 1? Piso. Circuito de Posgrados, Ciudad Universitaria Alcaldía Coyoacán. C. P. 04510 CDMX Tel. (+5255)5623 7002 http://pcbiol.posgrado.unam.mx/ Agradecimientos institucionales Al Posgrado en Ciencias Biológicas de la Universidad Nacional Autónoma de México (UNAM). Al Consejo Nacional de Ciencia y Tecnología (CONACyT) por la beca otorgada y a la Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO) por el financiamiento otorgado al proyecto de investigación KE008 con los que se realizó esta tesis. Finalmente, al tutor principal Dr. Juan Núñez Farfán, por la oportunidad brindada, por su apoyo, motivación y paciencia en el desarrollo de esta tesis. Gracias por guiarme en mi desarrollo profesional y ser un miembro activo y comprometido en este proyecto de investigación. A los miembros del comité tutoral, Dr. Juan Núñez Farfán, Dr. Daniel Piñero Dalmau y Dr. Alejandro Nettel Hernanz por su tiempo, orientación y dedicación para desarrollar y mejorar el proyecto de investigación. Agradecimientos a título personal A Juan Núñez y Daniel Piñero por el apoyo económico otorgado durante los meses posteriores del programa de estudios para terminar esta tesis. A Rosalinda Tapia López por su apoyo en los experimentos de laboratorio y ayuda logística en las salidas de campo ¡Infinitas gracias por tus consejos y apoyo incondicional! A todos los integrantes del Laboratorio de Genética Ecológica y Evolución por su apoyo en campo y laboratorio. A Luis Osorio Olvera, Juan Pablo Jaramillo Correa y Alejandro Nettel Hernanz, por su valiosa contribución académica en esta tesis. A todos mis colegas y amigos por su valioso apoyo académico, pero sobretodo por su amistad y momentos compartidos. A Jorge Juárez, Alejandra Gutiérrez, Marisol de la Mora, Rosalinda Tapia, Laura Giraldo, Laura Lorena, Pilar Suárez, Eunice, Mijail, Diana López y Juan Núñez gracias además por su apoyo en los momentos difíciles. A los técnicos Adriana y Rafael por su ayuda en laboratorio. A Miguel Ángel Vargas y Juan Manríquez por su amistad, consejos y apoyo incondicional para terminar esta tesis. Finalmente, pero no menos importante, a mi familia y en especial a mi madre, Maribel Zavala Sánchez, sin ti no hubiese podido terminar. A “mi gordito” Mathías Chacón, por toda la alegría que me brindaste cada día. En memoria de mi papá, Edgardo Ochoa Aguilar Índice General Resumen ................................................................................................................... 9 Abstract .................................................................................................................... 11 Introducción general ............................................................................................... 13 Distribución de la variación genética .................................................................... 13 Eventos climáticos y geológicos en el pasado reciente .................................... 14 La heterogeneidad del paisaje .......................................................................... 15 Nicho fundamental de las especies .................................................................. 15 Los manglares ....................................................................................................... 16 Distribución y extensión territorial .................................................................. 16 Adaptaciones .................................................................................................... 18 Avicennia germinans (L.) L. ............................................................................ 19 Descripción botánica ................................................................................... 21 Genética de manglares en México ........................................................................ 21 Sitio de estudio ...................................................................................................... 23 Objetivos y estructura de la tesis ......................................................................... 23 Referencias ............................................................................................................ 25 Capítulo 1. Patrones de colonización contratantes de los acervos genéticos del mangle negro (Avicennia germinans (L.) L.) a lo largo de las costas mexicanas .................. 33 Apéndices .............................................................................................................. 48 Capitulo 2. Infiriendo barreras potenciales para el flujo genético en poblaciones tropicales de Avicennia germinans ............................................................................................. 78 Apéndices .............................................................................................................. 111 Capítulo 3. Distancia al centro del nicho como predictor de la diversidad genética de las poblaciones del mangle negro ................................................................................... 125 Apéndices .............................................................................................................. 149 Discusión general ..................................................................................................... 153 Perspectivas .......................................................................................................... 160 Conclusiones generales ......................................................................................... 162 Referencias ............................................................................................................ 163 9 Resumen La diversidad genética y su distribución entre poblaciones de una misma especie está determinada por varios factores, muchos de los cuales varían significativamente en el espacio y tiempo. En esta tesis evaluamos los patrones geográficos de la variación genética de Avicennia germinans con microsatélites nucleares dentro y entre poblaciones muestreadas a lo largo de las costas del Pacífico y Atlántico de México, con el objetivo de dilucidar los efectos de factores históricos y contemporáneos en su distribución. Integramos la genética de poblaciones, filogeografía, genética del paisaje, y las técnicas de modelado de nicho que, en conjunto, aportaron información relevante acerca de los factores que han influenciado la estructura y diversidad genética de una de las especies de mangle más representativas de México. A lo largo de los capítulos en los que está organizada esta tesis, estimamos el tiempo de divergencia de las poblaciones de la costa Atlántico y Pacífico de México y comparamos la dinámica de colonización post-glacial de ambas costas. Evaluamos algunos aspectos del paisaje que influencian los patrones de flujo genético entre poblaciones de A. germinans y estudiamos los niveles de diversidad genética dentro y entre poblaciones. Finalmente, evaluamos la relación entre la calidad del hábitat y la heterocigosidad, así como, las variables ambientales que mejor explican su variación en el espacio geográfico y cómo éstas tienen un impacto en las dinámicas poblacionales. Nuestros resultados indicaron que 1) la estructura genética de A. germinans está organizada en diferentes escalas geográficas con relativamente poco flujo genético entre poblaciones. 2) Los linajes Pacífico y Atlántico de México divergieron hace 0.75 Ma aproximadamente; aunque las poblaciones en ambas costas se distribuyen en latitudes similares, exhiben firmas genéticas distintas. 3) Los estimados de migración sugieren que las poblaciones en ambas costas están pobremente conectadas por flujo genético, especialmente en el Pacífico, las cuales se caracterizaron por tener menor diversidad genética y mayor estructura. 4) Las poblaciones del norte y sur de México están evolucionando de manera independiente. 5) Cuatro barreras potencialmente influyen en la tasa de flujo genético de A. germinans: La Península de Baja California, el patrón de circulación oceánica en la boca del Golfo de California y Golfo de Tehuantepec, así como el patrón de corrientes oceánicas en la Península de Yucatán. 6) Las variables que definieron el 10 nicho fundamental de A. germinans están relacionadas a la textura del suelo. Como lo predice la hipótesis del centro abundante medioambiental, encontramos una fuerte correlación con la heterocigosidad y las distancias al centroide del nicho. Estas relaciones sugirieron la fuerte influencia que la textura del suelo puede tener en el crecimiento y establecimiento de A. germinans, el cual impacta en la variación genética de las poblaciones. Las poblaciones de ambas costas muestran diferencias notables tanto en diversidad genética como en patrones filogeográficos, por lo que es fundamental la inclusión de esta información en cualquier plan de manejo y conservación. 11 Abstract Genetic diversity and its distribution between populations of one species is determined by several factors, many of which vary significantly in space and time. In this thesis we assessed the geographic patterns of Avicennia germinans genetic variation with nuclear microsatellites within and between populations sampled along the Mexico's Pacific and Atlantic coasts, with the goal of elucidating the effects of historic and contemporary factors in its distribution. We integrated population genetics, phylogeography, landscape genetics and niche modeling techniques that altogether contributed with relevant information on the factors that have influenced the genetic structure and diversity of one of the most representative species of mangrove in Mexico. Throughout the chapters of this thesis, we estimate the time of divergence between the populations from the Atlantic and Pacific coast of Mexico and compared the post-glacial colonization dynamics of both coasts. We assessed some aspects of the landscape that influence the patterns of gene flow between populations of A. germinans and studied the amount of genetic diversity within and between populations. Finally, we assessed the relationship between habitat quality and heterozygosity, as well as the environment variables that better explain its variation in the geographic space, and how these have an impact in population dynamics. Our results indicate that 1) the genetic structure of A. germinans is organized in different geographic scales with relatively low gene flow between populations. 2) The Pacific and Atlantic lineages diverged 0.75 Ma ago approximately; even if populations in both coasts are distributed at similar latitude, they show distinct genetic signatures. 3) Estimates of migration suggest that populations in both coasts are poorly connected by gene flow, especially in the Pacific, where populations exhibited less genetic diversity and greater structure. 4) Populations from northern and southern Mexico are evolving independently. 5) Four barriers are likely influencing the rate of gene flow in A. germinans: The Baja California Peninsula, the pattern of oceanic circulation at the mouth of the Gulf of California and Gulf of Tehuantepec, and also the pattern of ocean currents in the Yucatan Peninsula. 6) The variables that defined the ecological niche of A. germinans are related to the soil texture. As predicted by the niche-centroid hypothesis hypothesis, we found a strong correlation between heterozygosity and distance to the niche 12 centroid. These relationships suggest the strong influence that soil texture can exert on the growth and establishment of A. germinans, that impacts on the genetic variation of the populations. Populations from both coasts have noticeable differences both in their genetic diversity and phylogeographic patterns, showcasing the great importance of the inclusion of this information in any conservation and management plan. 13 Introducción General Distribución de la variación genética La diversidad y la estructura genética de las poblaciones reflejan la interacción entre los procesos históricos de una especie y las fuerzas evolutivas. En una población, las frecuencias alélicas pueden cambiar debido a la acción de cuatro fuerzas fundamentales: selección natural, deriva genética, mutación y flujo genético. La estructura genética, es principalmente el resultado de la deriva genética y el flujo genético, y la variación espacial en la selección natural. Por lo tanto, cuando el flujo genético se reduce entre dos poblaciones, la oportunidad de divergencia en las frecuencias alélicas por medio de deriva genética se incrementa, y a su vez, la estructuración genética. En poblaciones naturales, las fuerzas evolutivas no actúan de forma aislada. Desde Avise (2000), se ha considerado al aislamiento espacial, la capacidad de dispersión y las barreras geográficas como los mecanismos más importantes de divergencia poblacional a escala evolutiva. Sin embargo, la cantidad de diversidad genética y su distribución están regidas, por otros factores que operan a distintas escalas temporales (Manel et al., 2003; Ellegren y Galtier, 2016) además de las características intrínsecas de las especies (e. g. sistemas de apareamiento, formas de vida y habilidad de dispersión). Uno de los principales objetivos de la genética evolutiva y del paisaje es identificar qué factores son los responsables de los patrones observados. Una gran cantidad de estudios han interpretado la estructuración espacial de la diversidad genética en términos de eventos climáticos y geológicos, la configuración del paisaje (Hewitt, 2000; Manel et al., 2003; Caplins et al., 2014) y, más recientemente, a variables ambientales que definen el nicho fundamental de las especies (Martínez-Meyer et al., 2013; Lira-Noriega y Manthey, 2014; Micheletti y Storfer, 2015). La teoría neutral de evolución molecular predice que la diversidad genética en una población de tamaño constante debe ser proporcional al tamaño poblacional efectivo (Kimura, 1983). Sin embargo, el tamaño efectivo poblacional no es constante a través del tiempo, por lo que el conocimiento de diversos procesos tanto históricos como contemporáneos, son clave para entender los niveles de diversidad genética 14 actuales y su distribución (Ellegren y Galtier, 2016). La estructuración genética constituye entonces el vínculo directo a la historia evolutiva de las especies. Nos permite hacer inferencias acerca del potencial evolutivo de las poblaciones y su historia demográfica. Consecuentemente, entender en qué medida distintos factores influencian la diversidad genética de las poblaciones es de vital importancia para asegurar el éxito de los programas de conservación y manejo de las mismas. Eventos climáticos y geológicos en el pasado reciente La tierra ha estado bajo importantes cambios geológicos y climáticos desde finales del Mioceno (~10 Ma), iniciando con la colisión del Arco de Panamá con Sudamérica (O’Dea et al., 2007), seguido por una serie de oscilaciones climáticas (e. g. eras de hielo) que caracterizaron el Cuaternario (Webb y Bartlein, 1992). El levantamiento del Istmo Centroamericano (CAI), el cual estableció un puente de tierra entre el norte y sur de América, tuvo un profundo impacto en el patrón de circulación oceánica y el clima global (Bacon et al., 2013), dando lugar a una barrera para la dispersión de especies marinas y costeras (Lessios, 2008; Nettel y Dodd, 2007; Cerón-Souza et al., 2015). Los periodos glaciares del Pleistoceno (~2.58 Ma) se caracterizaron por la extensión de las capas de hielo en latitudes norteñas (Lambeck et al., 2002) y la disminución global del nivel del mar (Pillans et al., 1998), el cual estuvo cerca de 130 m por debajo de los niveles actuales durante el último máximo glacial (LGM; Alongi, 2015). Al final de este periodo glacial, mientras la Tierra se calentaba, la línea costera se reconfiguró a su forma actual (Lambeck et al., 2002). Las condiciones que prevalecieron durante el LGM – 22,000 a 19,000 ± 250 años (Yokoyama et al., 2000) – restringió los rangos de distribución de muchas especies subtropicales en poblaciones refugios, y se expandieron nuevamente mientras el clima se calentó (Hewitt, 2000). Tal recolonización post-glacial resultó en una firma genética característica que puede variar entre taxones, debido a las diferencias en los rasgos de historias de vida, la geografía local (Hewitt, 2004) y la tolerancia a las condiciones ambientales (Davis y Shaw, 2001). Sin embargo, cuando la expansión poblacional sigue un frente continuo, se espera una clina en las frecuencias alélicas entre los refugios y las nuevas poblaciones fundadas (Excoffier y Ray, 2008 y referencias en el mismo) caracterizadas por 15 un incremento en la homocigosidad debido a los cuellos de botella y eventos fundador (Hewitt, 1996, 2000; Hallatschek y Nelson, 2008; Excoffier y Ray, 2008). La heterogeneidad del paisaje En plantas, la fase reproductiva es de particular importancia ya que representa una de las únicas oportunidades de los genes para dispersarse entre las poblaciones a través del movimiento del polen y las semillas. Debido a su inmovilidad, las plantas requieren de vectores para transferir sus genes, dando lugar a diversas adaptaciones asociadas al agente particular responsable de la dispersión (animales, viento o agua) (Barret, 1998; Barrett y Harder, 1995). La dispersión involucra todos aquellos mecanismos de los propágulos para llegar a un hábitat adecuado para la germinación, crecimiento y reproducción (Harper, 1977). Sin embargo, la dispersión de los individuos o patrones de flujo genético, pueden estar influenciados por las características del paisaje (Manel et al., 2003). La incorporación de la heterogeneidad del hábitat en la conectividad de las poblaciones ha tenido un impacto importante para identificar características del paisaje que puedan tener un rol significativo en el movimiento de los genes (Manel et al., 2003; Holderegger et al., 2010) y, por lo tanto, en la distribución de la variación genética. Explicar las discontinuidades genéticas en términos de heterogeneidad del paisaje (barreras) es un tema central en biología evolutiva y conservación, especialmente para aquellas especies donde las barreras no son explícitas. Además de la vicarianza, muchos estudios han enfatizado en otros factores bióticos (e. g. tiempo de viabilidad de propágulos) y abióticos (e. g. distancias geográficas, corrientes oceánicas, etc.) que pueden potencialmente facilitar o impedir la dispersión y consecuentemente, el flujo genético entre poblaciones (revisado por Storfer et al., 2010). Nicho fundamental de las especies El nicho ecológico de las especies es un hipervolumen de n-dimensiones en donde se reúnen las condiciones ambientales favorables para que una especie pueda sobrevivir (Hutchinson 1957). El nicho ecológico tiene una estructura interna determinada por las condiciones ambientales que influyen en la adecuación, existiendo condiciones óptimas, subóptimas y marginales dentro del hipervolumen. Las condiciones óptimas se encontrarían hacia el centroide del nicho en donde la tasa de natalidad sería máxima y la de mortalidad mínima; 16 por lo tanto, los tamaños poblacionales serían mayores (Maguire, 1973). Los modelos correlativos de nicho requieren de información sobre la presencia de las especies y una serie de parámetros climáticos actuales para generar un modelo de las condiciones que favorezcan la presencia de una especie. El modelo es proyectado al espacio geográfico para generar un mapa que representa la distribución de las condiciones favorables para la especie o su distribución potencial (Peterson et al., 2011). Recientemente, se ha explorado la relación que pueden tener los modelos de nicho con aspectos más relacionados con la biología y adecuación de las especies, así como los patrones geográficos de abundancia (VanDerWall et al., 2009; Tôrres et al., 2012; Yañez-Arenas et al., 2012). La hipótesis del centro-abundante establece que la abundancia de una especie es mayor en el centro de su rango de distribución geográfica y disminuye hacia el límite de esta (Sagarin y Gaines, 2002). Consecuentemente, debido al incremento de la deriva génica se espera que las poblaciones en el límite de la distribución geográfica (periféricas) pierdan variación genética, contrario a aquellas que se encuentren en el centro de su distribución (Diniz-Filho et al., 2009). Estudios recientes sugieren que los procesos demográficos pueden estar más directamente relacionados con la calidad de las condiciones locales (Martínez- Mayer et al., 2013), en términos del nicho ecológico fundamental de las especies (Hutchinson 1957, 1978). Bajo este argumento, se espera que la mayor abundancia y variación génica se encuentren en localidades donde la idoneidad del hábitat es mayor, coincidiendo con el centro del nicho de las especies (Martínez-Meyer et al., 2013; Lira-Noriega y Manthey, 2014), en lugar del centro de su rango de distribución geográfica. Por lo tanto, las relaciones existentes entre la idoneidad del nicho y la diversidad genética deben reflejar los efectos de la variación en el tamaño efectivo poblacional (Martínez-Meyer et al., 2013; Diniz-Filho et al., 2015). Los manglares Distribución y extensión territorial El ecosistema manglar incluye a los animales y plantas asociadas con un elemento forestal dominante, el mangle. En sentido más estricto, los manglares son árboles tropicales tolerantes a la salinidad que se caracterizan por su alta fidelidad al ecotono influenciado por las mareas 17 (Tomlinson, 2016). Habitan las zonas costeras del mundo entre los 30º norte y sur del Ecuador, con variaciones en ciertas zonas como Bermudas (32º N), Japón (31º N), Australia, Nueva Zelanda (38º S) y el Sureste de África (32º S) (Tomlinson, 1986; Spalding et al., 2010). Se conocen alrededor de 70 especies distribuidas en dos hemisferios principales; la región del Indo-Oeste del Pacífico (IWP), que incluye África Oriental, Indo-Malasia y Australia, y la región del Atlántico-Este del Pacífico (AEP), que incluye el Pacífico y Atlántico de América y África Occidental (Fig. 1). La mayor riqueza de especies se encuentra en el Archipiélago Malayo en la región IWP, con aproximadamente 36 a 46 especies de manglares (Polidoro et al., 2010). El IWP es considerado como el centro de origen de los manglares, por lo que además se suele distinguir entre los manglares del viejo y del nuevo mundo; siendo éstos últimos los que se encuentran en la región AEP (Kathiresan, 2009), con una riqueza aproximada de 4 a 13 especies (Polidoro et al., 2010). Figura 1. Distribución global de los manglares (rojo y verde) en ambas regiones (Indo-Oeste del Pacífico y Atlántico-Este del Pacífico) y la distribución actual de Avicennia germinans (L.) L. (verde). En el continente americano, los manglares cubren más de 40, 000 km2 y se distribuyen en el litoral de la mayoría de los países de América Latina y el Caribe (Yánez-Arancibia y 18 Lara-Domínguez, 1999). En México, las especies de mangle más representativas son Rhizophora mangle (mangle rojo, Rhizophoraceae), Avicennia germinans L. (mangle negro, madre de sal, Avicenniaceae), Laguncularia racemosa (L.) Gaerth. (mangle blanco, Combretaceae) y Conocarpus erectus L. (mangle botoncillo, Combretaceae) (Basañez et al., 2008). Únicamente para la costa de Chiapas, se han registrado Rhizophora x harrisonnii [un morfotipo resultado de fenómenos repetidos de hibridación y de cruzas con individuos de R. mangle y R. racemosa (Cerón-Souza et al., 2010)] y Avicennia bicolor Standl. (Rico 1981; Tovilla-Hernández, 2006; Tovilla-Hernández et al., 2007). En México, los manglares cubren una superficie estimada de 764, 486 ha, haciéndolo uno de los países con mayor cobertura de este tipo de vegetación a nivel mundial (Spalding et al., 2010). En la vertiente del Golfo de México, solo los estados de Campeche, Yucatán y Quintana Roo reúnen más de la mitad de la cobertura total de manglares (417, 025 ha), mientras que, en el Pacífico mexicano, los estados con mayor cobertura son Sinaloa, Nayarit y Chiapas que sumados abarcan 24.9% de la cobertura nacional (CONABIO, 2010). Adaptaciones Los manglares habitan sitios altamente dinámicos y fisiológicamente estresantes. Además de las mareas, experimentan fluctuaciones en salinidad, disponibilidad de nutrientes, temperatura, precipitación e hipoxia. Para persistir en estas condiciones ambientales, los manglares desarrollaron una serie de adaptaciones al estrés fisiológico, aunque no todas esas características se encuentran necesariamente reunidas en cada especie. Algunas especies poseen un sistema de raíces que les proporciona mayor estabilidad en suelos blandos, otras tienen estructuras especializadas para el intercambio de gases a través de una serie de poros llamados lenticelas y neumatóforos, estructuras modificadas de raíces con geotropismo negativo (Pizarro, et al., 2004). La particularidad de los manglares para sobrevivir en ambientes salinos se debe a los mecanismos y sistemas de filtración y excreción, que les permite eliminar pequeñas cantidades de sal que logran penetrar la planta. Las adaptaciones anatómicas incluyen glándulas especializadas para excretar sal por la base del pecíolo y tejidos foliares. Otro mecanismo es la exclusión selectiva que permite regular la cantidad de sales que ingresan al sistema radicular de la planta (Tomlinson, 1986; Jiménez, 1994). La forma y mecanismos de propagación de las semillas de los manglares es otra 19 adaptación a las condiciones de inundación a las cuales están sometidos estos ecosistemas. Las semillas se desarrollan y maduran hasta formar plántulas en estado embrionario sujetas al árbol madre, se les conoce como propágulos y a esta forma de reproducción se le da el nombre de viviparismo; cuando la radícula no logra romper el fruto se le conoce como viviparidad incompleta o criptoviviparidad (Jiménez, 1994). Otra característica de las semillas es su capacidad de flotar y ser arrastradas por las corrientes hasta que logran retenerse en el sustrato (Pizarro et al., 2004). En un ambiente intermareal, la oportunidad de establecimiento es limitado, por lo que la viviparidad o criptoviparidad han sido desarrolladas para acumular reservas de carbohidratos y permitir el rápido enraizamiento (Friess, 2016). Avicennia germinans (L.) L. Es una de las ocho especies que pertenecen al único género de la familia Avicenneaceae Endl. (Duke, 1991; Schwarzbach y Mcdade, 2002; Tomlinson, 2016). El género Avicennia es el más tolerante a las bajas temperaturas y es uno de los únicos dos géneros de manglares verdaderos que se distribuye a lo largo de las costas del viejo y nuevo mundo (Schwarzbach y McDade, 2002). El mangle negro o madresal, Avicennia germinans (Fig. 2a), ha estado presente en el Neotrópico desde hace 16 ma (Plaziat et al., 2001) y se distribuye actualmente en la costa oeste de África, desde Mauritania hasta Angola, en la Costa Atlántica de América, desde el sur de Brasil hasta las Bermudas y a través de la costa Pacífico de América, desde México hasta Perú (Fig. 1; Nettel y Dodd, 2007). El sistema de apareamiento de la especie es mixto, predominantemente por fertilización cruzada, con niveles moderados de auto- fertilización y signos de endogamia biparental (Nettel-Hernanz et al., 2013; Mori et al., 2015). Sus flores perfectas (Fig. 2b) son las más grandes en el género; son blancas y zigomorfas, pequeñas y sésiles; miden de 10 a 13 mm de ancho, poseen órganos femeninos y masculinos (especie monoica) y además son protandras (Tomlinson, 2016). La especie es polinizada por abejas (Tomlinson, 1986; Nettel-Hernanz et al., 2013), aunque también recibe la visita de avispas, moscas de la familia Syrphidae (Lemus-Jiménez y Ramírez, 2003) y mariposas. Tiene un tiempo generacional de cinco años aproximadamente, produce propágulos ovoides criptovivíparos con una longevidad de cuatro meses y es capaz de dispersarse a largas distancias por medio de las corrientes oceánicas (Nettel y Dodd, 2007). Tolera un gran espectro de condiciones climáticas y edáficas que le permiten ser dominante 20 o exclusiva de ambientes marginales en los límites latitudinales o en áreas donde los suelos tienen altas concentraciones de sal (Cintrón y Schaeffer-Novelli, 1983; Rodríguez-Ramírez et al., 2004). Figura 2. Características botánicas de Avicennia germinans. a) Muestra la apariencia de los individuos adultos en Bahía de Altata, Sinaloa; b) flor de A. germinans mostrando órganos masculinos y femeninos (tomada de Peterson, 2012.); c) Corteza en placas irregulares (Sontecomapan, Veracruz); d) pneumatóforos de A. germinans rodeando plántulas de R. mangle; e) Hojas opuestas; f y g) propágulos ovoides criptovivíparos de A. germinans (Lázaro Cárdenas, Michoacán); h) cristales de sal excretados por glándulas especializadas (Bahía Concepción, Baja California). 21 Descripción botánica. – Es un árbol pequeño o arbusto de gran talla, perenne, generalmente de 2 a 8 m de altura (Fig. 2a); en algunos casos llega alcanzar hasta los 30 m (e. g. costas de Chiapas). Su tronco llega a medir entre 20 a 60 cm de diámetro; su corteza es gruesa, de color marrón oscuro fragmentado en ásperas placas irregulares (Fi. 2c) con un tono rojizo en su interior. Sus raíces son superficiales con neumatóforos (Fig. 2d) que emergen alrededor del tronco principal (Tomlinson, 2016). Sus hojas son opuestas, lanceoladas y de tamaño variable (Fig. 2e); tienen entre 3 y 12 cm de largo por 1 a 4 cm de ancho (Jiménez, 1999); el envés de las hojas y ramas son de color glauco en contraste con el verde más oscuro de su superficie adaxial (Tomlinson, 2016). Los pétalos son de color crema o blanco, y la parte interna de la corola posee una coloración amarilla (Fig. 2b) (Jiménez, 1999). Tiene cuatro estambres (órganos masculinos) de 2 a 3 mm de largo, alternos a los pétalos (Fig. 2b). El estilo (órgano femenino) mide de 1 a 3 mm. Su fruto es verde pálido, redondeado y comprimido lateralmente; mide de 2 a 4 cm de longitud, con pequeños pelos que le dan la apariencia de polvo (Fig. 1f, g) (Tomlinson, 2016). Pertenece al grupo de especies secretoras, lo que implica que filtra entre el 85 y 95 % de sal en las raíces y el exceso se secreta por medio de glándulas especializadas presentes en las hojas (Fig. 2h) (Shimony et al., 1973; Drennan y Pammenter 1982; Sobrado 2001). Genética de manglares en México Desde el origen y diversificación de los manglares, hace aproximadamente 66 Ma (Ellison et al., 1999), así como la dispersión de los géneros Rhizophora y Avicennia al nuevo mundo, estas especies han estado bajo constantes cambios demográficos, impactando directamente en su diversidad genética (Xu et al., 2017; Guo et al., 2018). De las cuatro especies de mangles presentes en México, R. mangle y A. germinans han recibido mayor atención en estudios filogeográficos y de genética de poblaciones, probablemente debido a su abundancia y amplia distribución. Uno de los primeros estudios, que involucró linajes mexicanos, en sugerir una dinámica continua en la extinción y recolonización de A. germinans fue el realizado por Sherrod y McMillan (1985), quienes propusieron dos rutas de recolonización del Golfo de 22 México y Florida, desde poblaciones que se encuentran más al sur en la Península de Yucatán y Centroamérica (McMillan, 1986). En el caso específico del linaje de A. germinans en el Pacífico mexicano, se han encontrado patrones consistentes con un escenario de recolonización reciente a finales del Pleistoceno e inicios del Holoceno (Nettel y Dodd, 2007; Sandoval-Castro et al., 2014). Por otro lado, un estudio pionero en México, sugirió la importancia del levantamiento del Istmo Centroamericano (CAI) en la estructuración genética de R. mangle; sin embargo, contrario a lo esperado, la diferenciación genética entre los linajes del Pacífico y Atlántico de México no fue significativa (Núñez-Farfán et al., 2002). Posteriormente, para poblaciones del Pacífico de Panamá y Costa Rica se encontraron claras diferencias genéticas entre ambos linajes (Cerón-Souza et al., 2010). Resultados similares se han encontrado en otras especies incluyendo R. mangle (Takayama et al., 2013; Sandoval- Castro et al., 2014) y A. germinans (Nettel y Dodd, 2007; Sandoval-Castro et al., 2014), dando soporte a la hipótesis del cierre definitivo del CAI como barrera para el flujo genético. A pesar de que las especies de mangle tienen un alto potencial de dispersión a larga distancia (Dodd et al. 2002, Nettel y Dodd 2007), diversos factores locales pueden afectar las frecuencias alélicas y genotípicas en un determinado espacio geográfico, moldeando la estructura local del ecosistema (Sousa et al. 2007). En el caso de R. mangle se han encontrado resultados contrastantes. Por un lado Núñez-Farfán et al. (2002) con isoenzimas, encontraron mayor diferenciación dentro de la costa Atlántico en comparación con el Pacífico, en contraparte con lo observado por Sandoval-Castro et al. (2014) con microsatélites. En el caso de A. germinans, de acuerdo con los resultados de Nettel y Dodd (2007) con ADN de cloroplasto y secuencias ITS, las poblaciones del Pacífico Mexicano son genéticamente muy similres entre sí. Sin embargo, con marcadores más variables (microstélites) el nivel de estructuración dentro de la misma cuenca resultó ser mayor (Sandoval-Castro et al., 2014), por lo que el escenario más plausible es una diferenciación reciente. Por lo anteriormente expuesto, se ha sugerido que los eventos ocurridos durante el Pleistoceno tuvieron un impacto significativo en la distribución de la diversidad genética de R. mangle y A. germinans en el Pacífico de México, por el cual se observa claramente una disminución de la diversidad genética a mayores latitudes (Sandoval-Castro et al., 2014). 23 Sitio de estudio México tiene una alta diversidad biológica debido en parte a su posición geográfica, entre la región Neártica y Neotrópica y a su alta complejidad topográfica; lo que ha dado lugar a una diversa combinación de condiciones ambientales que albergan numerosos ecosistemas (Rzedowski, 2006), entre ellos, el ecosistema manglar. Alberga una de las mayores extensiones de manglar a nivel mundial a lo largo de sus 11, 592.77 km de línea de costa (Spalding et al., 2010; de la Lanza et al., 2013); se caracteriza por la presencia de diversas geoformas (e. g. lagunas, bahías, estuarios) que exhiben diferencias regionales tanto en su origen, longitud y naturaleza (de la Lanza et al., 2013). Los ambientes costeros ofrecen condiciones particulares climáticas y edáficas, tales como exposición a la salinidad, inundación temporal, vientos fuertes, alta radicación UV y suelos anaeróbicos. Estas variables que caracterizan a los ecosistemas costeros fluctúan diariamente con el cambio de mareas y su intensidad varía entre sitios (Gunderson et al., 2016). Adicionalmente, la forma de los hábitats costeros es inherentemente linear por lo que condiciona la distribución de los taxones en un espacio geográfico estrecho. Objetivos y estructura de la tesis La variación genética no se distribuye al azar, por lo que determinar qué factores contribuyen a la distribución espacial de la variación genética es fundamental para la conservación de poblaciones silvestres. En este trabajo integramos la genética de poblaciones, filogeografía, genética del paisaje, y técnicas de modelado de nicho para aportar información relevante acerca de los factores que han influenciado la estructura y diversidad genética de una de las especies de mangle más representativas de México, el mangle negro, Avicennia germinans. Los objetivos de este trabajo fueron: i) determinar cómo está organizada la diversidad genética entre y dentro de poblaciones de A. germinans y ii) evaluar el rol de factores tanto históricos como contemporáneos que puedan tener un rol significativo en determinar la distribución espacial de la variación genética de A. germinans en México. En el capítulo uno estudiamos el tiempo de divergencia de las poblaciones de la costa Atlántico y Pacífico de México y comparamos la dinámica de colonización post-glacial de ambas costas usando 24 marcadores microsatélites. En el capítulo dos evaluamos algunos aspectos del paisaje que potencialmente pueden estar influenciando los patrones de flujo genético entre poblaciones de A. germinans. Finalmente, en el capítulo tres evaluamos la relación entre la idoneidad del hábitat y la diversidad genética, así como, las variables ambientales que mejor explican su variación en el espacio geográfico. 25 Referencias Alongi, D.M. (2015) The impact pf climate change on mangrove forest. Current Climate Change Reports, 1, 30-39. Avise, J. C. (2000) Phylogeography: The history and formation of species. Harvard University Press, Cambridge. 453 pp. Bacon, C.D., Mora, A., Wagner, L.W. & Jaramillo, C.A. (2013) Testing geological models of evolutions of the Isthmus of Panama in a phylogenetic framework. Botanical Journal of the Linnean Society, 171, 287-300. Barret, S. C. H. (1998) The genetics of plant migration and colonization. In: Brown, A. H. D. Clegg, M. T. Kahler, A. L. y Weir, B. S. Plant population genetics, breeding and genetic resources. Sinauer, Sunderland. 449 pp. Barret, S. C. H. y Harder, D. L. 1996. Ecology and Evolution of plant mating. Tree 11, 73- 79. Basañez, M. A. J., Cruz, L. M. A., Domínguez, B. C. González, G. C., Serrano, S. A. & Hernández, A. A. (2008) Estructura y producción de Conocarpus erectus L. en el Sitio Ramsar "Manglares y Humedales de Tuxpan", Veracruz, México. UDO Agrícola, 8, 78- 87. Caplins, S. A. et al. 2014. Landscape structure and the genetic effects of a population collapse. Proc. R. Soc. B, 281, 20141798. Cerón-Souza, I., Gonzalez, E.G., Schwarzbach, A.E., Salas-Leiva, D.E., Rivera-Ocasio, E., Toro-Perea, N., Bermingham, E. & McMillan, W.O. (2015) Contrasting demographic history and gene flow patterns of two mangrove species on either side of the Central American Isthmus. Ecology and Evolution, 5, 3486-3499. Cerón-Souza, I., Rivera-Ocasio, E., Medina, E., Jiménez J.A., McMillan, W.O. & Birmingham, E. (2010) Hybridization and introgression in New World red mangroves, Rhizophora (Rhizophorarceae). American Journal of Botany, 97, 945-957. Citrón, G. & Schaeffer-Novelli, Y. (1983) Introducción a la ecología del manglar. UNESCO, Montevideo. 109 p. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO). (2010). Extensión y distribución de los manglares en México. Consultado el 18 de Diciembre 26 2018 en https://www.biodiversidad.gob.mx/ecosistemas/manglares2013/extensionDist.html Davis, M.B. & Shaw R.G. (2001) Range shifts and adaptative responses to Quaternary climate change. Science, 292, 673-679. De la Lanza, G., Ortíz, M. A., & Carbajal, J. L. (2013). Diferenciación hidrogeomorfológica de los ambientes costeros del Pacífico, del Golfo de México y del Mar Caribe. Investigaciones Geográficas, Boletín del Diniz-Filho, J. A. F., Nabout, J. C., Bini, L. M., Soares, T. N., Pires de Campos Telles, M., de Marco, P. Jr., Collevatti, R. G. (2009). Niche modelling and landscape genetics of Caryocar Brasiliense (“Pequi” tree: Caryocaraceae) in Brazilian Cerrado: an integrative approach for evaluating central-peripheral population patterns. Tree Genetics & Genomes, 5, 617-627. Diniz-Filho, J. A., Rodrigues, H., Pires de Campos Telles, M., de Oliveira, G., Terribile, L. C., Soares, T. N. & Nabout, J. C. (2015). Correlation between genetic diversity and environmental suitability: taking uncertainty from ecological niche models in to account. Molecular Ecology Resources, 15, 1059-1066. Dodd, R.S., Afzal-Rafii Z., Kashani N. & Budrick J. (2002) Land barriers and open oceans: effects on gene diversity and population structure in Avicennia germinans L. (Avicenniaceae). Molecular Ecology, 11, 1327-1338. Drennan, P. & Pammenter N. W. (1982) Physiology of salt secretion in the mangrove Avicennia marina (Forsk.) Vierh. The New Phytologist, 91, 1000– 1005. Duke, N. C. (1991) A systematic revision of the mangrove genus Avicennia (Avicenniaceae) in Australasia. Australian Systematic Botany, 4, 299ñ324. Ellegren, H. & Galtier, N. (2016) Determinants of genetic diversity. Nature Reviews Genetics, 17, 422-433. Ellison, A.M., Farnsworth, E.J. & Merkt, R.E. (1999) Origins of mangrove ecosystems and mangrove biodiversity anomaly. Global Ecology and Biogeography, 8, 95-115. Excoffier, L. & Ray, N. (2008) Surfing during population expansions promotes genetic revolutions and structuration. Trends in Ecology and Evolution, 23, 347-351. Friess, D. A. (2016) Quick guide mangrove forest. Current Biology Magazine, 26, R746- R748. 27 Gunderson, A. R., Armstrong, E. J. y Stillman, J. H. (2016). Multiple Stressors in a Changing World: The Need for an Improved Perspective on Physiological Responses to the Dynamic Marine Environment. Annual Review of Marine Science, 8, 357-378. Guo, Z., Li, X., He, Z., Yang, Y., Wang, W., Zhong, C., … Duke, N.C. (2018) Extremely low genetic diversity across mangrove taxa reflects past sea level changes and hints at poor future responses. Global Change Biology, 24, 1741-1748. Hallatschek, O. & Nelson, D.R. (2008). Gene surfing in expanding populations. Theoretical Population Biology, 73, 158-170. Harper, J.L., 1977. Population Biology of Plants. Academic Press, New York. Hewitt, G.M. (1996) Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society 58, 247-276. Hewitt, G.M. (2000) The genetic legacy of the Quaternary ice ages. Nature, 405, 907-913. Hewitt, G.M. (2004) Genetic consequences of climatic oscillations in the Quaternary. Philosophical Transactions of the Royal Society B 359(1442): 183-195. Holderegger, R., Buehler, D., Gugerli, F. & Manel, S. (2010) Landscape genetics of plants, Trends in Plant Science, 15, 675-683. Hutchinson, G. E. (1957) Concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology, 22, 415-427. Hutchinson, G. E. (1978) An introduction to population ecology. Yale University Press, New Haven, Connecticut. 260 pp. Instituto de Geografía, UNAM, 2013, 33–50. Jiménez, J. A. (1994) Los manglares del Pacífico Centroamericano. Heredia. C.R. Editorial EFUNA. 352 pp. Jiménez, J. A. (1999) Ambiente, distribución y características estructurales en los manglares del Pacífico de Centro América: Contrastes climáticos, p. 51-70. En: Yáñez-Arancibia, A. & Lara-Domínguez, A. L. (eds.). Ecosistemas de manglar en América tropical. Instituto de Ecología A.C. México, UICN/ORMA, Costa Rica, NOAA/NMFS Silver Spring MD USA. 380 pp. Kathiresan, K. (2009) Mangrove ecosystems. Distribution of mangroves. En: Balasubramanian, T., Kathiresan, K. & Ajmal Khan, S. 2009. Training Course on Mangroves and Biodiversity. Pp 92-100. Kimura, M. (1983). The Neutral Theory of Molecular Evolution (Cambridge Univ. Press). 28 Lambeck, K., Esat, T.M. & Potter, E-K. (2002) Links between climate and sea levels for the past three million years. Nature, 419, 199-206. Lemus-Jiménez, L. J. & Ramírez, N. (2003). Polinización y polinazadores en la vegetación de la planicie costera de Paraguana, Estado Falcon, Venezuela. Acta Científica Venezolana. 54, 97-114. Lessios, H. A. (2008). The great american schism: divergence of marine organisms after the rise of the Central American Isthmus. The Annual Review of Ecology, Evolution, and Systematics, 39, 63-91. Lira-Noriega, A. & Manthey, J. D. (2014) Relationship of genetic diversity and niche centrality: A survey and analysis. Evolution, 68, 1082-1093. Maguire Jr., B. 1973. Niche response structure and the analytical potentials of its relationship to the habitat. The American Naturalist. 107: 213-246. Manel, S., Schwarts, M K., Luikart, G. & Taberlet, P. (2003) Landscape genetics: combining landscape ecology and population genetics, Trends in Ecology and Evolution, 18, 189- 197. Martínez-Meyer, E., Díaz-Porras, D., Peterson, A. T. & Yáñez-Arenas, C. (2013). Ecological niche structure and rangewide abundance patterns of species. Biology Letters, 9, 20120637. McMillan, C. (1986) Isozyme patterns among populations of black mangrove, Avicennia germinans, from the Gulf of Mexico-Caribbean and Pacific Panama. Contributions in Marine Science, 29, 17-25. Micheletti, S. J. & Storfer, A. (2015). A test of the central-marginal hypothesis using population genetics and ecological niche modelling in an endemic salamander (Ambystoma barbourin). Molecular Ecology, 24, 967-979. Nettel-Hernanz, A., Dodd, R.S., Ochoa-Zavala, M., Tovilla-Hernández, C. & Días-Gallegos, J.R. (2013) Mating system analyses of tropical populations of the black mangrove, Avicennia germinans (L.) L. (Avicenniaceae). Botanical Sciences, 91, 115-117. Nettel, A. & Dodd, R.S. (2007) Drifting propagules and receding swamps: Genetic footprints of mangrove recolonization and dispersal along tropical coasts. Evolution, 61, 958-971. Núñez-Farfán J., Domínguez, C.A., Eguiarte, L.E., Cornejo, A., Quijano, M., Vargas, J. & Dirzo, R. (2002) Genetic divergence among Mexican populations of red mangrove 29 (Rhizophora mangle): Geographic and historic effects. Evolutionary Ecology Research, 4, 1049-1064. O’Dea A., Rodríguez F., De Gracia C. & Coates A.G. (2007) Patrimonio paleontológico. Canto Rodado, 2, 149-179. Peterson, B. 2012. Black mangrove (Avicennia germinans). Download 08 Julio, 2019 de https://es.wikipedia.org/wiki/Archivo:Black_Mangrove_(Avicennia_germinans)_(72705 58948).jpg Peterson, A.T., Soberón, J., Pearson, R. G., Anderson, R. P., Martínez Meyer, E., Nakamura, M. & Araújo, M. B. 2011. Ecological Niches and Geographic Distributions. Monographs in Population Biology. Princeton University Press, Princeton, N.J. 314 p. ISBN: 978-1- 4008-4067-0. Pillans, B., Chappell, J. & Naish, T.R. (1998) A review of the Milankovitch climatic beat: template for Plio-Pleistocene sea-level changes and sequence stratigraphy. Sedimentary Geology, 122, 5–21. Pizarro, F., Piedra, L., Bravo, J., Asch, J. & Asch, C. (2004) Manual de procedimientos para el manejo de manglares en Costa Rica. Fundación UNA (EFUNA). Costa Rica. 132 pp. Plaziat, J-C., Cavagnetto, C., Koeniguer, J-C. & Baltzer, F. (2001) History and biogeography of the mangrove ecosystem, based on a critical reassessment of the paleontological record. Wetlands Ecology and Management, 9, 161-179. Polidoro, A. B., Carpenter, E. K., Collins, L., Duke, C. N., Ellison, M. A., Ellison, C. j… & Hong Yong, J. W. (2010) The loss of species: Mangrove extinction risk and geographic areas of global concern. Plos One 5, e10095. Rico, G. V. (1981) Rhizophora harrisonni (rhizophoraceae), un nuevo registro de las costas de México. Boletín de la Sociedad Botánica de México, 41, 163-166. Rodríguez-Ramírez, A., Nivia-Ruíz, J. & Garzón-Ferreira, J. (2004) Características estructurales y funcionales de Avicennia germinans en la Bahía de Chengue (Caribe Colombiano). Boletín de Investigaciones Marinas y Costeras, 33, 223-244. Rzedowski, J. (2006). Vegetación de México. México: Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO), 504 pp. Sagarin, R. D. & Gaines, S. (2002) The ‘abundant centre’ distribution: to what extent is it a biogeographical rule? Ecology Letters,5, 137-147. 30 Sandoval-Castro, E., Dodd, R.S., Riosmena-Rodríguez, R., Enríquez-Paredes, L. M., Tovilla-Hernández, C., López-Vivas, J. M., Aguilar-May. B. & Muñiz-Salazar, R. (2014) Post-Glacial expansion and population genetic divergence of mangroves species Avicennia germinans (L.) Stearn and Rhizophora mangle L. along the Mexican coast. Plos One, 9, e93358. Schwarzbach, E. A. & McDade, A. L. (2002) Phylogenetic relationship of the mangrove family Avicenniaeceae based on chloroplast and nuclear ribosomal DNA sequences, Systematic Botany 27, 84-98. Sherrod, C.L. & McMillan, C. (1985) The distributional history and ecology of mangrove vegetation along the northern Gulf of Mexico coastal region. Contributions in Marine Science, 28, 129-140. Shimony, C., Fahn A. & Rheinhold, L. (1973) Ultrastructure and ion gradients in the salt glands of Avicennia marina (Forskal) Vierh. The New Phytologist, 72, 27–36. Sobrado, M. A. (2001) Effect of high external NaCl concentration of xylem sap, leaf tissue and leaf glands secretion of the mangrove Avicennia germinans (L.) L. Flora, 196, 63– 70. Sousa, P.W., P. Kennedy, G. Mitchell, J. Betsy & L.B. Ordóñez. (2007) Supply-Side ecology in mangroves: Do propagule dispersal and seedling establishment explain forest structure? Ecological Monographs 77, 53-76. Spalding, M., Kainuma, M. & Collins, L. (2010) World atlas of mangroves. 2a edición. Earthscan. USA. 319 pp. Storfer, A., Murphy, M. A., Spear, S. F., Holderegger, R. & Waits, L. P. (2010) Landscape genetics: where are we now? Molecular Ecology, 19, 3496-3514. Takayama, K., Tamura, M., Tateishi, Y., Webb, E.L. & Kajita, T. (2013) Strong genetic structure over the American continents and transoceanic dispersal in the mangrove genus Rhizophora (Rhizophoraceae) revealed by broad-scale nuclear and chloroplast DNA analysis. American Journal of Botany 100, 1191-1201. Tomlinson, P.B. (1986) The botany of mangroves. Cambridge, UK. Cambridge University Press. 419 pp. --------------------- (2016) The botany of mangroves. 2nd Edition. Cambridge, UK. Cambridge University Press. 419 pp. 31 Tôrres, N. M., Júnior, P. De M., Santos, T., Silveira, L., de Almeida-Jácomo, A. T. & Diniz- Filho, J. A. 2012. Can species distribution modelling provide estimates of population densities? A case study with jaguars in the Neotropics. Diversity & Distributions. 18 (6): 615-627. Tovilla-Hernández, C. (2006) Propuesta para la conservación, manejo y restauración en los manglares de la costa de Chiapas. Laboratorio de Ecología de Manglares y Zona Costera. El Colegio de la Frontera Sur, Unidad Tapachula y Consejo de Ciencia y Tecnología del estado de Chiapas. México. 148 pp. Tovilla-Hernández, C., Salas, R. R.L., de la Presa, P. J. C., Romero, B. E., Ovalle, F. E. & Ortega, R. O. (2007) Inventario de los bosques de mamglar de la costa de Chiapas: Informe final COCYTECH. Laboratorio de Ecología de Manglares y Zona Costera, El Colegio de la Frontera Sur, Unidad Tapachula y Consejo de Ciencia y TEcnología del Estado de Chiapas. México, 98 pp. VanDerWal, J., Shoo, L. P., Johnson, C. N. & Williams, S. E. 2009. Abundance and the environmental niche: environmental suitability estimated from niche models predicts the upper limit of local abundance. The American Naturalist. 174 (2): 282-291. Webb, T. & Bartlein, P.J. (1992). Global changes during the last 3 million years: Climatic controls and biotic responses. Annual Review of Ecology and Systematics, 23, 141-173. Xu, S., He, Z., Zhang, Z., Guo, Z., Lyu, H., Li, J., … Shi, S. (2017) The origin, diversification and adaptation of a major mangrove clade (Rhizophoreae) revealed by whole-genome sequencing. Molecular Biology & Genetics, 4, 721-734. Yañez-Arancibia, A. & Lara-Domínguez, A. L. (1999) Los manglares de América Latina en La Encrucijada, p 9-16 en: Yáñez-Arancibia, A. & Lara-Domínguez, A.L. (eds.) Ecosistemas de manglar en América tropical. Instituto de Ecología, A.C. México, UICN/ORMA. Costa Rica. 380 pp. Yañez-Arenas, C, Martínez-Meyer, E., Mandujano, S. & Rojas-Soto, O. 2012. Modelling geographic patterns of population density of the white-tailed deer in central Mexico by implementing ecological niche theory. Oikos. 121 (12): 2081-2089. Yokoyama, Y., Lambeck, K., De Deckker, P., Johnston, P. & Fifield L.K. (2000). Timing of the Last Glacial Maximum from observed sea-level minima. Nature, 406, 713-716 Capítulo 1 Patrones de colonización contrastantes de los acervos genéticos del mangle negro (Avicennia germinans (L.) L.) a lo largo de las costas mexicanas 884 | wileyonlinelibrary.com/journal/jbi Journal of Biogeography. 2019;46:884–898.© 2019 John Wiley & Sons Ltd Accepted: 22 January 2019 DOI: 10.1111/jbi.13536 R E S E A R C H P A P E R Contrasting colonization patterns of black mangrove (Avicennia germinans (L.) L.) gene pools along the Mexican coasts Maried Ochoa-Zavala1,2 | Juan Pablo Jaramillo-Correa1 | Daniel Piñero1 | Alejandro Nettel-Hernanz3 | Juan Núñez-Farfán1 1Departamento de Ecología Evolutiva, Instituto de Ecología, Universidad Nacional Autónoma de México, Ciudad de México, Mexico 2Posgrado en Ciencias Biológicas, Unidad de Posgrado, Ciudad de México, Mexico 3Instituto de Ciencias Biológicas, Universidad de Ciencias y Artes de Chiapas, Tuxtla Gutiérrez, Mexico Correspondence Juan Núñez-Farfán, Departamento de Ecología Evolutiva, Instituto de Ecología, Universidad Nacional Autónoma de México, Ciudad de México, México. Email: farfan@unam.mx Funding information National Council of Science and Technology (CONACyT); UNAM; National Commission for the Knowledge and Use of Biodiversity, Grant/Award Number: KE008 Editor: Enrique Martínez-Meyer Abstract Aim: Historical and geological events can impact the genetic structure of species, pro- ducing signatures that vary among taxa and among gene pools within taxa. Such signa- tures can also be affected by local geography and tolerance to environmental conditions. However, disentangling the different drivers of population structure is often difficult. In an attempt to do so, we surveyed two independent gene pools of the same species that followed similar paths of post- glacial colonization across contrasting landscapes and environmental conditions. We aimed to determine how these differ- ences have affected the post- glacial population dynamics of each gene pool. Location: The Pacific and Atlantic coasts of Mexico. Taxon: Black mangrove (Avicennia germinans; Avicenniaceae). Methods: Using microsatellite variation, we estimated the divergence time of black mangrove populations through approximated Bayesian computation and imple- mented a comparative approach to evaluate different demographic hypotheses within and between coasts. Results: The Pacific and Atlantic gene pools diverged long after the rise of the Central American Isthmus (Mid- Pleistocene), although occasional transisthmian gene exchanges were also inferred. Both coasts showed the characteristic isolation by distance (IBD) pattern expected for expanding gene pools. However, populations from the Atlantic coast were more genetically diverse and admixed than those from the Pacific basin. Both our migration models and the climate data gathered suggested a more ancient es- tablishment and/or more stable conditions for black mangrove on the Atlantic coast. Main conclusions: The Atlantic basin likely bore more favourable climate conditions than the Pacific, allowing for the survival of A. germinans during the Last Glacial Maximum in situ. Populations from the northern Pacific coast became established after the Holocene warming, leading to contrasting genetic patterns between the two gene pools. Nevertheless, the action of environmental factors in determining the contemporary distribution of genetic variation in A. germinans cannot be discarded. K E Y W O R D S Black mangrove, contrasting landscape, isolation by distance, microsatellites, Pacific–Atlantic divergence, post-glacial colonization dynamics 33 |OCHOA- ZAVALA et AL. | Describing how genetic variability is distributed within and be- tween populations of a species is key to understanding the evolu- tionary forces that shape genetic structure. Geological and climatic events, such as the rise of the Central American Isthmus (CAI) or the Pleistocene glaciation cycles, strongly impact species’ genetic struc- ture, leaving ‘genetic signatures’ that can be used to reconstruct their evolutionary history (Cerón- Souza et al., 2015; Takayama, Tamura, Tateishi, Webb, & Kajita, 2013). It has been argued that the conditions that prevailed during the Last Glacial Maximum (LGM) restricted many temperate and subtropical species into southern refugial populations (i.e. those located in areas of relative climatic stability during glacial cycles, enabling the maintenance of genetic diversity and promoting the differentiation of populations Hewitt, 1996), which then expanded north as the climate warmed (Hewitt, 2000). Such post- glacial recolonization resulted in genetic signatures that vary among taxa due to life history differences, local geography (Hewitt, 2004), and tolerance to the environmental conditions in the colonized areas (Davis & Shaw, 2001). Classic comparative phylo- geographical studies address such differences by surveying patterns among co- distributed taxa, although they often struggle to disentan- gle individual drivers of population structure (Carstens, Brunsfeld, Demboski, Good, & Sullivan, 2005; Carstens & Richards, 2007; Cerón- Souza et al., 2015; Fouquet et al., 2012; Guo, Guo, et al., 2018; Maliouchenko, Palmé, Buonamici, Vendramin, & Lascoux, 2007; Marske, Leschen, & Buckley, 2012). One way to tease apart these factors is to survey different gene pools of the same species that follow similar colonization paths (e.g. from south to north) but across contrasting landscapes or environmental conditions. However, suitable model species to perform such studies are scarce, and the results are often difficult to interpret because refugial populations no longer exist and/or because secondary contact between gene pools blurs modern pat- terns (Bai, Wang, & Zhang, 2014; Liepelt, Bialozyt, & Ziegenhagen, 2002). Mangrove species growing along the subtropical Atlantic and Pacific coastlines of southern Mexico appear ideal to address such questions. The populations from the two coasts are separated by hundreds of kilometres and are both roughly linearly distributed, fol- lowing similar north–south clines (Figure 1). Furthermore, their sub- tropical location increases the likelihood that glacial populations still exist, in contrast to temperate or boreal taxa (Castellanos- Morales, Gámez, Castillo- Gámez, & Eguiarte, 2016; Roberts & Hamann, 2015; Scheinvar, Gámez, Castellanos- Morales, Aguirre- Planter, & Eguiarte, 2016). Moreover, they have been likely separated since the clo- sure of the CAI (Cerón- Souza et al., 2010, 2015; Dodd, Afzal- Rafii, Kashani, & Budrick, 2002; Nettel & Dodd, 2007) which decreases the likelihood that secondary contact will affect the results (but see Takayama et al., 2013). Mexico's complex evolutionary history and the fact that it spans both sides of the Tropic of Cancer has led to wide variation in climatic and orographic conditions and great biological diversity (Espinosa, Ocegueda, Aguilar, Flores, & Llorente- Bousquets, 2008; Rzedowski, 2006). The Atlantic and Pacific coasts show similar latitudinal changes in land and ocean surface temperatures (Figure 2a; Fernández- Eguiarte, Zavala- Hidalgo, & Romero- Centeno, 2018a), but differ in several im- portant ways, like in salinity gradients, that are almost absent in the Atlantic (Figure 2b; Antonov et al., 2010). Moreover, the Atlantic coast is less than half as long as the Pacific (De la Lanza, Ortíz, & Carbajal, 2013), and the dominance of trade winds makes the oceanic effect Geographical location of 29 populations of Avicennia germinans analysed for microsatellite loci along the coasts of Mexico. PN = Pacific North; PC = Pacific Central; PS = Pacific South; GM = Gulf of Mexico; PY = Yucatan Peninsula. Names of localities in Table 1. Lower panel shows Structure clustering considering all populations (K = 2; Pacific vs. Atlantic) 34 | OCHOA- ZAVALA et AL. G e o g ra p h ic l o c a ti o n , sa m p le s iz e , a n d d e sc ri p ti v e g e n e ti c s st a ti st ic s o f 2 9 p o p u la ti o n s o f A v ic e n n ia g e rm in a n s (c f. F ig . 1 ) a lo n g t h e P a c if ic a n d A tl a n ti c c o a st s o f M e x ic o P o p u la ti o n n a m e s G e o g ra p h ic a l lo c a ti o n n A P A A R A T H O H E F IS L a ti tu d e ( N ) L o n g it u d e ( W ) P a c if ic c o a st E l S a rg e n to P N 0 4 2 9 .3 2 8 5 1 7 4 0 2 ( 0 ) 0 1 .1 2 3 9 0 .0 5 ( 0 ) 0 .1 4 ( 0 ) 0 .6 4 7 ** B a h ía C o n c e p c ió n P N 0 2 2 6 .7 6 2 1 3 3 4 0 2 .2 ( 0 .4 4 7 ) 0 1 .4 7 1 4 0 .0 3 2 ( 0 .0 1 1 ) 0 .0 4 7 ( 0 .0 2 4 ) 0 .3 6 5 * T o p o lo b a m p o P N 0 6 2 5 .5 8 2 1 9 1 4 0 2 .6 6 7 ( 1 .2 1 1 ) 0 2 .0 1 1 8 0 .1 4 1 ( 0 .1 1 4 ) 0 .1 4 2 ( 0 .1 ) 0 .0 0 6 B a h ía M a g d a le n a P N 0 3 2 4 .7 9 4 1 8 3 4 0 2 ( 0 ) 1 1 .4 8 7 1 3 0 .1 1 7 ( 0 .1 9 4 ) 0 .1 5 9 ( 0 .1 9 2 ) 0 .0 9 4 L a P a z P N 0 5 2 4 .1 8 1 5 9 3 4 0 3 ( 0 ) 0 1 .6 2 7 1 4 0 .2 4 1 ( 0 .1 8 ) 0 .2 8 7 ( 0 .1 9 5 ) 0 .1 6 1 B a h ía d e A lt a ta P N 1 1 2 4 .6 1 5 6 6 7 4 0 3 .3 7 5 ( 1 .0 6 1 ) 2 2 .8 6 1 2 7 0 .2 0 2 ( 0 .1 6 1 ) 0 .2 9 1 ( 0 .2 2 1 ) 0 .2 9 1 ** * E l C a im a n e ro P N 1 2 2 2 .8 8 2 2 0 8 4 0 3 .1 4 3 ( 1 .4 6 4 ) 2 2 .3 8 9 2 3 0 .1 6 3 ( 0 .1 4 3 ) 0 .1 7 7 ( 0 .1 3 8 ) 0 .0 9 5 M a ri sm a s N a c io n a le s P N 1 0 2 2 .3 7 4 4 6 7 4 0 4 ( 1 .7 3 2 ) 3 2 .9 0 6 2 9 0 .2 1 4 ( 0 .1 7 6 ) 0 .2 7 6 ( 0 .2 3 3 ) 0 .1 9 7 ** P u n ta P é ru la P C 1 5 1 9 .5 9 2 0 2 7 4 0 3 ( 1 .1 5 5 ) 0 1 .8 8 1 6 0 .3 0 7 ( 0 .2 2 1 ) 0 .3 3 1 ( 0 .2 ) 0 .0 5 B a rr a d e N a v id a d P C 1 1 1 9 .1 9 5 5 3 2 4 0 2 .5 ( 0 .8 3 7 ) 2 1 .7 8 1 7 0 .1 3 2 ( 0 .2 3 4 ) 0 .1 4 5 ( 0 .1 7 1 ) 0 .0 7 8 T é c p a n d e G a le a n a P C 1 9 1 7 .2 1 3 9 3 5 4 0 4 ( 1 .8 2 6 ) 0 2 .3 3 9 2 0 0 .4 3 7 ( 0 .1 8 ) 0 .4 4 3 ( 0 .1 6 ) 0 .0 1 2 B a rr a d e T e c o a n a p a P S 1 7 1 6 .4 9 7 8 4 0 3 .8 3 3 ( 2 .2 2 9 ) 0 2 .5 2 5 2 5 0 .2 7 ( 0 .2 5 5 ) 0 .2 6 9 ( 0 .2 3 ) L a g u n a s d e C h a c a h u a P S 2 0 1 6 .0 1 1 4 4 0 3 .3 3 3 ( 1 .0 3 3 ) 0 2 .5 9 2 2 2 0 .3 5 8 ( 0 .1 4 8 ) 0 .4 0 5 ( 0 .1 9 7 ) 0 .1 2 4 * M a r M u e rt o P S 2 9 1 6 .0 1 0 4 4 0 4 .5 ( 2 .6 1 9 ) 2 3 .8 5 8 3 6 0 .3 4 9 ( 0 .2 9 3 ) 0 .4 0 5 ( 0 .2 9 5 ) 0 .1 2 8 ** L a E n c ru c ij a d a P S 2 4 1 5 .1 6 8 3 6 7 4 0 4 ( 2 ) 2 3 .0 5 2 2 6 0 .4 7 6 ( 0 .2 8 6 ) 0 .4 9 8 ( 0 .2 7 2 ) 0 .0 4 6 M e a n 0 .1 6 5 0 .1 9 0 0 .1 2 9 A tl a n ti c c o a st R ío B ra v o G M 3 7 2 5 .9 4 6 1 3 3 4 0 2 .4 ( 0 .8 9 4 ) 0 1 .8 0 3 1 5 0 .2 3 ( 0 .1 7 8 ) 0 .2 9 6 ( 0 .2 1 7 ) 0 .2 2 6 ** L a P e sc a G M 4 1 2 3 .7 7 6 0 8 3 4 0 2 .8 3 3 ( 1 .1 6 9 ) 1 2 .2 4 2 1 9 0 .2 1 2 ( 0 .1 8 2 ) 0 .2 4 5 ( 0 .1 9 9 ) 0 .1 3 7 T u x p a n G M 5 6 2 0 .9 8 0 7 4 0 4 ( 1 .6 7 3 ) 2 3 .0 3 5 2 6 0 .4 3 6 ( 0 .1 5 7 ) 0 .4 1 3 ( 0 .1 6 4 ) L a M a n c h a G M 4 0 1 9 .5 8 6 1 6 7 3 4 3 .4 2 9 ( 1 .1 3 4 ) 2 2 .9 8 2 2 5 0 .2 7 4 ( 0 .2 0 1 ) 0 .3 0 9 ( 0 .2 4 ) 0 .1 A lv a ra d o G M 5 3 1 8 .7 3 2 8 5 4 1 3 .7 5 ( 1 .8 3 2 ) 3 3 .3 3 6 3 0 0 .2 3 1 ( 0 .2 0 9 ) 0 .2 7 1 ( 0 .2 ) 0 .1 0 4 S o n te c o m a p a n G M 5 4 1 8 .5 0 8 4 1 7 3 4 3 ( 1 .2 6 5 ) 1 2 .4 5 7 2 0 0 .2 9 5 ( 0 .2 5 2 ) 0 .2 8 ( 0 .2 2 1 ) 0 .0 6 4 C o a tz a c o a lc o s G M 3 6 1 8 .0 9 7 1 – 9 4 .4 3 1 3 8 3 3 9 3 .2 5 ( 1 .7 5 3 ) 1 3 .0 5 2 6 0 .2 8 7 ( 0 .2 5 8 ) 0 .3 0 7 ( 0 .2 2 8 ) 0 .1 0 8 M e c o a c á n G M 4 6 1 8 .4 1 9 6 2 6 3 5 3 .7 1 4 ( 1 .3 8 ) 3 3 .2 3 2 2 7 0 .4 1 2 ( 0 .1 9 ) 0 .4 2 5 ( 0 .1 9 6 ) P o m - A ta st a P Y 6 7 1 8 .5 7 4 5 4 0 4 .1 6 7 ( 1 .4 7 2 ) 1 3 .2 2 2 7 0 .5 1 4 ( 0 .1 6 2 ) 0 .4 6 3 ( 0 .1 2 1 ) L o s P e te n e s P Y 6 6 2 0 .1 5 8 9 1 7 4 1 3 .5 7 1 ( 1 .2 7 2 ) 3 3 .0 9 8 2 6 0 .4 3 5 ( 0 .2 5 6 ) 0 .4 5 2 ( 0 .2 4 ) R ío L a g a rt o s P Y 7 0 2 1 .6 1 0 2 4 0 3 .6 6 7 ( 0 .8 1 6 ) 1 2 .9 0 2 2 4 0 .3 3 3 ( 0 .2 7 3 ) 0 .3 3 6 ( 0 .2 1 4 ) 0 .0 1 2 Y u m b a la m P Y 8 1 2 1 .4 4 3 5 8 3 4 0 3 .5 7 1 ( 1 .9 0 2 ) 1 2 .9 8 3 2 6 0 .3 5 8 ( 0 .2 7 5 ) 0 .3 1 6 ( 0 .2 1 7 ) C o zu m e l P Y 6 5 2 0 .5 5 5 6 3 3 4 0 4 .1 6 7 ( 1 .3 2 9 ) 0 3 .2 4 1 2 7 0 .4 9 1 ( 0 .3 ) 0 .4 6 9 ( 0 .2 0 2 ) S ia n K a ’a n P Y 7 7 2 0 .1 1 0 9 6 7 4 0 4 .1 6 7 ( 1 .3 2 9 ) 0 3 .2 6 2 2 7 0 .4 9 3 ( 0 .1 6 9 ) 0 .5 ( 0 .1 5 3 ) M e a n 0 .2 8 7 0 .2 9 2 0 .0 1 8 S ta n d a rd d e v ia ti o n s a re s h o w n i n p a re n th e se s. n , sa m p le s iz e ; A , a v e ra g e n u m b e r o f a lle le s p e r lo cu s; P A , p ri v a te a lle le s; A R , a lle lic r ic h n e ss ; A T , to ta l n u m b e r o f a lle le s; H O , H E , o b se rv e d a n d e x p e ct e d h e te ro zy - g o si ty , re sp e ct iv e ly ; F IS , in b re e d in g c o e ff ic ie n t. *P < 0 .0 5 ; ** P < 0 .0 1 ; ** *P < 0 .0 0 1 . 35 |OCHOA- ZAVALA et AL. more intense on the Atlantic coast (Rzedowski, 2006). As a conse- quence, the Atlantic slope is much wetter than its Pacific counterpart (Figure 2c; Espinosa et al., 2008; Rzedowski, 2006) and receives less hours of insolation per year (Figure 2d; Pérez- Villegas, 1990). Ocean currents also play an important role in the wet conditions of each coast. On the Atlantic side, the Gulf of Mexico is dominated by a northward- moving warm marine current that constitutes a rich and constant con- tribution of water vapour, while in the Pacific, the influence of the cold southward- moving California current leads to dry conditions on the Baja California Peninsula and Sonora (Figure 2b; Espinosa et al., 2008). Mexico possesses one of the largest areas covered by mangrove forests worldwide (Giri et al., 2011; Spalding, Kainuma, & Collins, 2010). These forests are tropical and subtropical salt- tolerant eco- systems that grow along coastlines around the world (Tomlinson, 1986). They are one of the most productive and economically import- ant ecosystems on the planet; they house a wide variety of marine and terrestrial species (Alongi, 2002) and provide numerous envi- ronmental services, including capture and storage of CO2 (Donato et al., 2011), accretion of sediments, protection of seashores (Alongi, 2008) and biofiltering of heavy metals (Analuddin et al., 2017; Keller, Wilson, Reeve, & Platenberg, 2017; Marchand, Fernández, & Moreton, 2016; Yan, Sun, Zhang, & Li, 2017). Furthermore, these forests have been subjected to past climate events and sea level changes that have impacted the genome- wide nucleotide diversity of mangrove species (Guo, Li, et al., 2018; Xu et al., 2017), allowing demographic variations to be traced back and compared between gene pools within the same species. Mexican mangrove ecosystems are frequently dominated by the black mangrove (Avicennia germinans (L.) L). This species is distributed from the west coast of Africa to the Atlantic and Pacific coastlines of tropical and subtropical America. It produces small cryptoviviparous propagules with a longevity of nearly 4 months (Alleman & Hester, 2011), which are dispersed through ocean currents. Avicennia germi- nans has a predominantly outcrossing mating system, although it can also endure moderate levels of inbreeding (Mori, Zucchi, & Souza, 2015; Nettel- Hernanz, Dodd, Ochoa- Zavala, Tovilla- Hernández, & Días- Gallegos, 2013). Five- year- old treelets can already produce vi- able propagules (Nettel & Dodd, 2007). Here, after extensively sampling A. germinans in the Pacific and Atlantic (the Gulf of Mexico and the Caribbean sea) coasts of Mexico, we assess the divergence time between these gene pools using approximated Bayesian computation (ABC) and infer and compare their likely post- glacial colonization patterns using microsatellite data. Implementation of different approaches, such as model comparisons, can improve our knowledge on how species respond to glacial periods (Barrow, Bigelow, Phillips, & Lemmon, 2015; Carstens et al., 2013), particularly when ‘du- plicates’ of the same process are available for independent gene pools. We aimed to (a) evaluate whether post- glacial population dynamics could be related to the landscape and environmental differences between the coasts of Mexico, and (b) determine if the divergence of the two gene pools is consistent with vicariance following the final closure of the CAI and, within each coast, an isolation by distance (IBD) model, resulting from a slow- moving front from south to north. The implications for the conservation of this keystone species are also discussed. Some environmental features of the Mexican coasts. (a) Terrestrial and ocean surface temperature during February (Fernández- Eguiarte et al., 2018a). (b) Direction of surface water currents and ocean salinity on December 2018. The magnitude of the current is indicated by the length and width of the streaklet. For salinity, a colour scale from light to dark red indicate values from 32 to 36 practical salinity scales (PSS; nowCOAST, 2018). (c) Terrestrial humidity ranges, including the Köppen classification adjusted for Mexico by García (1990). (d) Terrestrial insolation ranges (hours of insolation per year) (Pérez- Villegas, 1990) (a) (b) (c) (d) 36 | OCHOA- ZAVALA et AL. | | We sampled 1144 individuals of A. germinans from 29 populations along the Pacific and Atlantic coasts of Mexico (600 and 544 indi- viduals respectively; Figure 1, Table 1). Foliage was collected from georeferenced adult trees separated by at least 50 m to avoid resa- mpling the same or closely related genotypes. Leaves were stored in liquid nitrogen after collection. Genomic DNA was isolated from ~200 mg of tissue using the DNeasy Plant Mini Kit (Qiagen) follow- ing the Quick- Start Protocol with some modifications (the incuba- tion time was adjusted to 40 and 30 min at steps two and three respectively). DNA quality and concentration were determined with a NanoDrop Lite Spectrophotometer (Thermo Scientific). | We used 10 previously developed nuclear microsatellite loci (Cerón- Souza, Bermingham, McMillan, & Jones, 2012; Cerón- Souza, Rivera- Ocasio, Funk, & McMillan, 2006; Mori, Zucchi, Sampaio, & Souza, 2010; Nettel, Rafii, & Dodd, 2005) to assess genetic diversity and population structure. Details are provided in Appendix S1: Table S1.1. PCR reactions were performed using a Veriti 96- well thermal cycler (Applied Biosystems). Amplification success was verified on 2% agarose gel and PCR products were sent to the Core Sequencing Facility at University of Illinois at Urbana- Champaign (UIUC) for analysis. Allele size and genotypes were scored using GeneMarker 2.4.0 (SoftGenetics). | Micro- checker 2.2.3 (Van Oosterhout, Hutchinson, Wills, & Shipley, 2004) was used to infer null alleles and scoring errors caused by stut- tering or large allele dropout. After correction, the final data set was tested, locus by locus and in each population, for Hardy–Weinberg equilibrium (HWE) and linkage disequilibrium (LD) between pairs of loci using the Expectation–Maximization (EM) algorithm. Genetic di- versity was assessed as the total number of alleles (AT), the average number of alleles per locus (A), the number of private alleles (PA), allelic richness (AR), and the observed and expected heterozygosi- ties (HO and HE, respectively). HWE, LD, A, HO, HE and inbreeding coefficients (FIS) were calculated using arlequin 3.5 (Excoffier & Lischer, 2010), and AT, PA and AR were determined with fstat 2.9.3.2 (Goudet, 2002). | To assess genetic structure within each coast, we first evaluated the spatial decrease in heterozygosity along each coastline with Spearman's rank correlations (rs) between latitude and HE, as imple- mented in R 3.2.5 (R Core Team, 2016). For the Atlantic coast, we re- peated this analysis removing the following populations: PY67, PY66, PY70, PY81, PY65 and PY77 to avoid some bias due to the concave shape of the coast. We then performed a Mantel test (Mantel, 1967) to test whether populations adjusted to an IBD model. We used paired matrices of genetic (Nei, Tajima, & Tateno, 1983) and geographic distances using the ‘ade4’ package for R (Dray & Dufour, 2007). Geographic distances (km) among populations were deter- mined using the Google Earth 7.1.7.2600 (Google, 2016) rule tool along the coastline at an elevation of c. 500 m, except at those sites, where the coastline is highly dissected; at these sites, we used the tool rule at an altitude of 3 km (v.gr., only few localities in the Pacific coast). This procedure provided a more accurate estimate of the distance between localities than Euclidean distances. In addition, when possible, we incorporated the direction and patterns of ocean currents (Fernández- Eguiarte, Zavala- Hidalgo, & Romero- Centeno, 2018b). We inferred the number of genetic clusters (K) at two lev- els—within coasts and between coasts (Pacific vs. Atlantic pop- ulations)—with the Bayesian method, implemented in structure 2.3.4 (Pritchard, Stephens, & Donnelly, 2000). We used the admix- ture model with correlated allele frequencies among populations without information on sampling location, and assuming K values ranging from 1 to 10, with 10 replicates for each K value. All runs consisted of 2,000,000 Markov chain Monte Carlo simulations, 100,000 of which were discarded as burn- in. We used structure harvester 0.6.94 (Earl & vonHoldt, 2012) to determine the optimum K value based on (Evanno, Regnaut, & Goudet, 2005) and to evaluate the probability of the data (ln P(D)) for each K value. To avoid only detecting the uppermost level of a putative hierarchical genetic structure (see Janes et al., 2017), we searched for K values where secondary peaks and a plateau were observed in the distri- bution of and ln P(D), respectively. Then, we complemented this analysis by a discriminant analysis of principal components (DAPC; Jombart, Devillard, & Balloux, 2010) implemented in ‘adegenet’ (Jombart, 2008) in R. | To assess the divergence time between Pacific and Atlantic popu- lations, we used ABC (diyabc 2.1.0, Cornuet et al., 2014) to simu- late 1,000,000 data sets under a scenario of two populations of size N1 (Atlantic) and N2 (Pacific) that diverged t generations ago from an ancestral population of size NA (Appendix S2: Figure S2.1). After fine- tuning (Appendix S2: Figures S2.2 and S2.3), prior dis- tributions of demographic parameters were set as follows: Normal (min = 0, max = 1,000, M = 500 and SD = 50) for N1 and N2, nor- mal (min = 0, max = 5000, M = 2560 and SD = 100) for NA and for t, uniform (min = 5, max = 400000) with the condition N2 < N1. We assumed a stepwise mutation model with a mean mutation rate of 1 × 10 –1 × 10 per generation and a gamma distribution. The summary statistics of each simulation included the mean gene di- versity across loci (Nei, 1987), the mean genetic diversity, the mean allele size variance between pairs of populations, and FST (sensu Weir & Cockerham, 1984). 37 |OCHOA- ZAVALA et AL. Given that putative refugial populations (i.e. the southernmost populations) may contribute to recolonized areas, estimating gene flow is critical for inferences of post- glacial migration dynamics (Carstens et al., 2013). We implemented the Bayesian inference strategy from Migrate- n 3.2.15 (Beerli, 2006; Beerli & Felsenstein, 2001; Beerli & Palczewski, 2010) to compare different migration models that describe recolonization from different glacial refugia on each coast (Table 2). Here, we constructed 10 models (6 and 4 for the Pacific and Atlantic coasts, respectively) based on the pre- viously identified population clusters within each basin (Figures 3 and 4, see Section 3), and evaluated several alternative hypothe- ses of gene flow patterns, including, for example, the patterns ex- pected under the assumption that ocean currents drive gene flow. These alternative models are shown in Table 2, with further detail in Appendix S3. In addition, we used the reconstruction of air and sea surface temperatures during the LGM (Appendix S3: Tables S3.2 and S3.3) by Annan and Hargreaves (2013) to determine clus- ters that were most likely refugia at sites with warmer conditions. We then compared the marginal likelihoods and Bayes factors (BF) of the different hypotheses (Table 2), which differed in the number of refugia and directionality of gene flow, to test which best fit the data (Carstens et al., 2013). We used the marginal likelihoods by thermodynamic integration with a Bezier approximation to calcu- late the BF and posterior model probabilities (Beerli & Palczewski, 2010). | | Of the 10 microsatellite loci, 8 were in HWE and did not show spe- cific patterns of LD (Appendix S4: Tables S4.4–S4.7). The two re- maining loci (Agerm_CT_003, Agerm_GA_003) had high null allele frequencies (>0.20) in more than half of the populations and were removed from the analyses. Among all the surveyed black mangrove stands, there was a total of 81 different alleles with an average of 10.12 alleles per locus, with no apparent difference in total al- leles between coasts (57 and 59 alleles for the Atlantic and Pacific coast respectively). Overall, the Atlantic populations had higher av- erage gene diversity (HO = 0.287, HE = 0.292) and less inbreeding (FIS = 0.018) than the Pacific populations (HO = 0.165, HE = 0.190 and FIS = 0.129; P < 0.05, two- sided test with 10,000 permutations performed in fstat), suggesting contrasting demographic processes between coasts. Within the Pacific coast, northern populations (Pop PN04, PN02 and PN06) had less genetic diversity than the southern ones (Pop PS29 and PS24). Private alleles were found in less than half of the populations, mostly those located in the northern part of the Pacific coast, 6 were significantly inbred, with FIS values as high as 0.65 (Pop PN04; Table 1). On the Atlantic coast, population gene diver- sity varied much less widely than on the Pacific coast. Private alleles were present in all but three populations, and also at low frequencies FIS values were low and only the northernmost population (Pop GM37) was significantly inbred (Table 1). | On both coasts, southern populations (e.g. PS24 and PY77) had about twice the genetic diversity of the northern ones (e.g. PN02 and GM41). The relationship between geographic and genetic distances between populations (IBD) was linear and significant on both coasts, whereas the relationship between latitude and genetic diversity was significant only for the Pacific coast (Figure 5). On the Atlantic, this correlation remained non- significant even after accounting for the effect of coast shape (rs = 0.309, P = 0.538). structure analyses indicated strong genetic differentiation between the Pacific and Atlantic populations, with nearly no connectivity (Figure 1). Within each coast, we detected an opti- mum of five and three genetic clusters for the Pacific (Figure 3) and Atlantic (Figure 4) coasts, respectively. Individuals from clusters in the extremes of the Pacific (C1P and C5P) and Atlantic (C1A and C3A) gradients all had high membership co- efficients (Q > 60), while individuals from intermediate clusters were more admixed (Figures 3 and 4). Please refer to Appendix S5 for details of the optimum K choice, structure and Evanno results (see Appendix S5: Figures S5.5–S5.9). DAPC's scatter- plots revealed essentially the same patterns on both coasts, but illustrated more clearly the stepping- stone outline previ- ously inferred from the genetic diversity analyses (Appendix S6: Figures S6.10 and S6.11; see also Figure S6.12 for DAPC results between coasts). | The ABC analyses suggested that the populations from the two coasts diverged relatively recently, during the Mid- Pleistocene (0.75 Myr; 95% CI 0.18–1.75 Myr, Table 3). The comparison of migra- tion models showed strong support for the multiple glacial refugia hypothesis along both coasts. All of the most supported models in- cluded glacial refugia at sites where conditions (Annan & Hargreaves, 2013) were favourable for mangrove persistence at the southern Pacific and most of the Atlantic coast. The best- fitting model for the Pacific coast (PP = 0.9999; Table 2a) included three refugial popula- tions (C3P–C5P) with a northward stepping stone colonization pat- tern starting from C3P. These refugial stands were likely linked by gene flow, either bi- (i.e. between C3P and C4P) or unidirectionally (from C5P to C4P). Within this coast, estimated average Nm values ranged between 0.32 and 8.87 individuals per generation, and the highest migration exchange occurred between C4P and C3P (see Nm values on Table 2a). For the Atlantic coast, the best model (PP = 0.9999; Table 2b) included two refugia (C2A and C3A), from which northern col- onization of the Gulf of Mexico proceeded. As for the Pacific coast, the refugial stands were likely interconnected via bidirec- tional gene flow (C2A and C3A). The average Nm in these Atlantic 38 | OCHOA- ZAVALA et AL. populations ranged between 0.38 and 20.38 individuals per gen- eration, suggesting a founder effect in the north and virtual pan- mixia between refugial populations in the south (Table 2b). Theta parameters did not converge (Appendix S3: Figure S3.4), so Ne estimates are not shown. | Here, we explored the parallel recolonization pattern of two inde- pendent gene pools of black mangrove along the Atlantic and Pacific coasts of Mexico. Our results revealed multiple glacial refugia on each coast, from which northern recolonization took place, likely after the LGM. Although concordant genetic patterns were found on both coasts, some differences were inferred in the post- glacial dynamics between coasts. | Our data suggest that the Pacific and Atlantic populations of A. germinans have evolved independently since the Calabrian or the Middle Pleistocene, which is more recent than the final closure of the CAI (3 Ma). While uncertainty in the generation time of tree spe- cies with overlapping generations could lead to the underestimation of the time since divergence (Petit & Hampe, 2006; Tsuda, Nakao, Model comparison performed with Migrate- n within each coast of Mexico. All hypotheses describe a stepping stone migration pattern No. Migration modela Description Bezier l Lm LBF PP Rank (a) Pacific coast 1 Bidirectional migration between adjacent clusters <0.00001 4 2 Migration based on ocean currents pattern 0 6 3 C3P C4P C5P Three refugial populations with bidirectional migration among them, northern migration from C3P <0.00001 5 4 C4P C5P Two refugial populations with bidirectional migration and northern migration from C4P <0.00001 2 5 C5P Single refugial population with northern migration <0.00001 3 6 C1P ← 1.15 C2P ← 0.38 C3P ↔ 8.87 C4P ← 0.32 C5P Based on the pattern observed in STRUCTURE and experimental runs with Migrate- n. Three refugial populations, with bidirectional migration between C3P–C4P and unidirectional migration from C5P to C4P, northern migration from C3P 0 0.99999 1 (b) Atlantic coast 7 Based on the pattern observed in STRUCTURE and experimental runs with Migrate- n. South migration from C1A to C2A and bidirectional among C2A–C3A <0.00001 4 8 Migration pattern based on ocean currents <0.00001 3 9 C1A ← 0.38 C2A ↔ 20.38 C3A Two refugial populations with bidirectional migration among them and northern migration from C2A 0 0.99999 1 10 C3A Single refugia with northern migration 675.9 <0.00001 2 Bold letters indicate the refugial populations for Avicennia germinans. Nm values are shown for the best- fitted model. Bezier approximation scores of the log marginal likelihoods (l ml), log Bayes factors (LBF) and posterior probability (PP) of each model. aMigrations models were based on population clusters identified (cf. Figures 3 and 4). 39 |OCHOA- ZAVALA et AL. Ide, & Tsumura, 2015), there are several biological explanations of the apparent lag between the closure of the CAI and divergence. For instance, for divergence to take place in the absence of natural se- lection, gene flow must be reduced sufficiently after the appearance of a major geographical barrier (Slatkin, 1987). Although previous studies have found a general coincidence between the divergence time of various taxa on both sides of the CAI (O'Dea et al., 2016), the time elapsed until complete isolation depends on each species’ vagility (Groeneveld et al., 2014), the features of the barrier and the joint action of different evolutionary processes. Long- lived trees are reputed for having large amounts of gene flow and long generation times, which implies that environmental changes may take longer to impact their genetic structure (Petit & Hampe, 2006). Occasional transisthmian gene exchanges may also have occurred and delayed gene pool divergence. Recent studies in Rhizophora mangle have in- deed found admixture between populations on both sides of the CAI (Takayama et al., 2013). Possible explanations for this pattern may be temporary interconnections between oceans due to interglacial sea level rises (O'Dea et al., 2016; Takayama et al., 2013 and references there in). Our data for A. germinans also show some shared ancestry between populations (Figure 1), indicating that the Pacific–Atlantic isolation is not absolute. While some small amount of gene flow, perhaps due to hurricanes or other meteorological extreme events (Nathan et al., 2008) or by long- distance pollen transport by insects could occur, allele homoplasy should also be considered to account for this pattern (Ortega- Del Vecchyo, Piñero, Jardón- Barbolla, & van Heerwaarden, 2017). | between gene pools One of our assumptions was that the refugial populations of this spe- cies still exist. Mangrove forest cover decreases as the number of Reconstructions of surface air and sea temperatures during the LGM (Annan & Hargreaves, 2013) show that the temperature conditions in the central and southern portions of both Mexican coasts have been adequate for black mangrove survival for the last 22–19 ka. Thus, Geographical distribution of genetic clusters of Avicennia germinans identified along the Pacific coast, considering K = 5. Population names are defined in Table 1 40 | OCHOA- ZAVALA et AL. populations from clusters C3P, C4P and C5P on the Pacific coast and C2A and C3A on the Atlantic coast were the most likely source for the colonization of the species’ northernmost limits, at least in this part of its distribution. The LGM data available for Mexico is far from complete; however, it suggests drier conditions than at present in the Yucatan Peninsula, a cold dry climate in central Mexico, and a wetter environment than currently observed in northern Mexico (Lozano- García, Torres- Rodríguez, Ortega, Vázquez, & Caballero, 2013; Metcalfe, O'Hara, Caballero, & Davies, 2000). Nonetheless, it is likely that climate conditions along the Atlantic Mexican coast allowed for the survival of A. germinans during the LGM, while pop- ulations in the northern Pacific coast could only have established during the Holocene warming. Consequently, we hypothesize that environmental conditions were an important driver of population structure, producing contrasting genetic patterns between these two gene pools of the same species migrating along similar paths. Along the Pacific coast, we found evidence for subsequent mi- gration events that allowed the recolonization of the northern local- ities (C1P and C2P) following a stepping- stone pattern. Taking the reconstruction above into account, this recolonization likely took place during the Holocene. The stepping stone and IBD pattern in this basin are supported by the Bayesian clustering analysis, which shows a decrease in shared ancestry at higher latitudes (Figure 3). As expected under this hypothesis, populations within the C1P pop- ulation cluster were most recent and least genetically diverse, fol- lowed by those within C2P (Excoffier & Ray, 2008; Hallatschek & Nelson, 2008). On the Atlantic coast, our results suggest long- term per- sistence of two glacial populations, one on the Gulf of Mexico and one on the Yucatan Peninsula (C2A and C3A, respectively), which in sum, cover almost the entire Atlantic coast of Mexico. This was probably facilitated by the more favourable water and air tem- peratures in this region than in the Pacific (Annan & Hargreaves, 2013). We hypothesize that northern populations (from C1A) likely originated from the northernmost populations within the C2A cluster (i.e. GM56), probably following the Gulf current during the Holocene. Previous studies proposed two recoloni- zation routes for the northern parts of the Gulf of Mexico and Florida (Sherrod & McMillan, 1985), from sources in the Yucatan Peninsula and the Caribbean coasts in Central America (McMillan, 1986). However, our migration models and the previously inferred climatic conditions do not support a large population extinction Spatial location of the genetic clusters of Avicennia germinans identified along the Atlantic coast (K = 3). Population names are defined in Table 1 Approximate posterior distribution of parameters obtained using DIYABC Parameter Mean Median Mode Q0.025 Q0.050 Q0.950 Q0.975 N1 531 530 529 454 466 599 614 N2 471 472 479 388 401 537 549 NA 2,560 2,560 2,550 2,360 2,390 2,720 2,760 t 0.84 0.75 0.4385 0.1805 0.2315 1.755 1.865 μ 3.79 × 10 3.58 × 10 3.16 × 10 1.76 × 10 1.95 × 10 6.35 × 10 7.00 × 10 N1 Atlantic, N2 Pacific, NA ancestral population, t divergence time (million years) and mutation rate (μ). We consider a generation time of approximately 5 years for Avicennia germinans. 41 |OCHOA- ZAVALA et AL. along the Gulf of Mexico, which is further backed by the lack of correlation between latitude and heterozygosity along this coast (which remained non- significant even after taking the concave shape of this basin into account; rs = 0.309, P = 0.538; see also Figure 5). McMillan (1986) found similarities between populations of A. germinans from the northern parts of the Gulf of Mexico and Texas, which in turn differed from those from Florida, while the Central American populations (Panama and Belize) showed ad- mixed patterns. Our data suggest that populations in and around Texas likely originated from recent founder events from stands in the central Gulf of Mexico (Tamaulipas), while those in Florida could be the result of migration from populations in the Caribbean via the Loop Current, as reported by Kennedy et al. (2016) for R. mangle. Based on the information available for A. germinans, we further hypothesize that populations from the Yucatan Peninsula (PY70, PY81, PY65 and PY77) maintained a substantial gene flow with other stands in the Caribbean (e.g. in Panama and Belize; McMillan, 1986), and are more related to the populations from Florida (Cerón- Souza et al., 2015). | Isolation by distance seem typical of mangroves (Cerón- Souza et al., 2012; Kennedy et al., 2016; Sandoval- Castro et al., 2014). However, it has been suggested that factors other than absolute geographic distance should be also considered, including water velocity (Nathan et al., 2008), the retention of propagules by roots (Van der Stocken et al., 2015), tides (Ngeve, van der Stocken, Sierens, Koedam, & Triest, 2017), the direction of wind and local currents, and coastal to- pography (Yan, Duke, & Sun, 2016). We hypothesized that the level of genetic structure found on each coast would be related to differ- ences in each coast's features. First, the Atlantic coast of Mexico is a marginal semi- enclosed region with a shorter shore extension than the Pacific coast, which is more linear and topographically complex, with various geomorphic types (De la Lanza et al., 2013). In addi- tion, the pattern of ocean currents along the Pacific coast appears to be more dynamic than in the Atlantic (Figure 2b). The Pacific is dominated by intense cyclonic and anticyclonic eddies (Fernández- Eguiarte et al., 2018b), with extensive mixing during El Niño events (Muss, Robertson, Stepien, Wirtz, & Bowen, 2001). The current sys- tems in the Gulf of Mexico, in contrast, are relatively stable; they are dominated by the Loop Current and a westward drift anticyclonic gyre (Behringer, Molinari, & Festa, 1977). Genetic patterns similar to those described in this work (stronger structure along the Pacific coast than the Atlantic) have been doc- umented for other species, particularly in marine fishes with both demersal and pelagic habits (demersal: Bernardi, Findley, & Rocha- Olivares, 2003; Garber, Tringali, & Stuck, 2004; Heist & Gold, 2000; Landínez- García, Ospina- Guerrero, Rodríguez- Castro, Arango, & Relationship between populations’ expected heterozygosity and latitude, and between genetic and geographic distance between pairs of populations of Avicennia germinans in the Pacific (upper) and Atlantic (lower) coasts of Mexico 42 | OCHOA- ZAVALA et AL. Márquez, 2009; Munguia- Vega et al., 2018; Pruett, Saillant, & Gold, 2005, pelagic: Bernardi et al., 2003; Broughton, Stewart, & Gold, 2002; López, Alcocer, & Jaimes, 2010; Pacheco- Almanzar, Ramírez- Saad, Velázquez- Aragón, Serrato, & Ibáñez, 2017). This suggests that marine currents are important drivers of genetic structure, indepen- dent of the species’ biological traits. Though they are very different organisms, there are interesting parallels between mangroves and fishes; the dispersal of fish eggs and larvae can be influenced by the strength, direction and variability of the same ocean currents that would affect mangrove propagules, and juvenile fish inhabit shallow waters, including estuaries, littoral and/or coastal lagoons domi- nated by mangrove forests. Second, when populations occupy heterogeneous environments, the establishment of immigrants is expected to decrease (Orsini, Vanoverbeke, Swillen, Mergeay, & de Meester, 2013) because of nat- ural selection against immigrants (Sexton, Hangartner, & Hoffmann, 2014). For instance, along the coasts of Mexico, there is a clear lati- tudinal change in both terrestrial and sea surface maximum and min- imum temperatures (Figure 2a; Fernández- Eguiarte et al., 2018a). The northernmost Pacific populations experience a relatively high whereas in the southernmost stands (e.g. PS29, PS24) temperatures - - tions. Some studies in A. germinans suggest that temperature may strongly influence seed establishment due to the inhibition of the have shown that propagules mortality increased with decreasing lat- itude of source material (Krauss et al., 2008). If this scenario is cor- rect, it may have created contrasting selection pressures for migrant establishment between north and south populations within each coast. To further assess whether environmental factors contribute to genetic structure, two approaches could be used to explore the potential role of local adaptation; first, looking for signatures of nat- ural selection at the genomic level (Savolainen, Lascoux, & Merilä, 2013), and second, performing reciprocal transplant experiments to assess seedling survival and establishment success (Blanquart, Kaltz, Nuismer, & Gandon, 2013). Finally, there is also a north–south salinity gradient (Figure 2b), of c. 35.7–33 PSS along the Mexican Pacific coast, which is almost absent in the Atlantic basin (Antonov et al., 2010). Although salin- ity may constitute an ecological stressor for other mangroves, A. germinans seedlings are able to develop roots under twice the salt concentration of sea water (McMillan, 1971). However, there is no evidence regarding the effect of salt concentration on the root ex- tension rate over time during the initial anchoring phase of propa- gules (e.g. delayed root elongation decreases the likelihood that the seed will anchor), which constitutes a crucial trait for the success- ful establishment of Avicennia (Balke et al., 2011). Therefore, it is possible that even if there is space available for colonization along the Pacific coast, migrants would have little chance of surviving and contributing genes to the established population due to envi- ronmental stress and/or by an ongoing ‘founder takes all’ principle (Waters, Fraser, & Hewitt, 2013). On the other hand, salinity should not be a factor affecting populations on the Atlantic coast, which could be readily tested with parallel transplant experiments on the two coasts. Thus, the IBD pattern found herein could be the result of the interaction between contrasting coastal landscape and envi- ronmental stress and/or density- dependent processes. However, if reduced effective gene flow is due to a reduction in establishment success among migrants (selection against foreign genotypes), local genetic adaptation might be taking place (Orsini et al., 2013) and further tests would be needed to identify its role. Understanding the underlying mechanisms that promote genetic structure is critical to conservation efforts of keystone species like mangroves. Our results suggest that localities bearing long- term refugia have maintained significant levels of genetic variation and should thus be a priority for conservation and management (along both Mexican coasts). Furthermore, since contrasting environmental conditions may have led to differences in the post- glacial population dynamics, putative local adaptation should be tested for and inte- grated into such efforts. ACKNOWLEDG EMENTS The authors thank R. Tapia- López for laboratory assistance, and all members of Laboratorio de Genética Ecológica y Evolución for logis- tical and field support. This paper constitutes a partial fulfilment of the PhD Thesis of the Graduate Program in Biological Sciences, National Autonomous University of Mexico (UNAM) of M.O.- Z., who also acknowledges the scholarship granted for graduate stud- ies by the National Council of Science and Technology (CONACyT). J.P.J.- C. was supported by an international fellowship (PASPA) from the Dirección General de Asuntos del Personal Académico at UNAM. This study was funded by National Commission for the Knowledge and Use of Biodiversity (CONABIO; grant KE008 granted to J.N.F.). We thank the Dirección General de Vida Silvestre (SEMARNAT) for permits to collect mangroves on the coasts of Mexico (no. SGPA/ DGVS/08820/17). All microsatellite genotypes used for this study will be available as.csv format from the Comisión Nacional para Conocimiento y Uso de la Biodiversidad (CONABIO) Project KE008, and from the Dryad Digital Repository: Ochoa-Zavala, Jaramillo-Correa, Piñero, Nettel- Hernanz, and Núñez-Farfán (2019). Title: Data from: Contrasting colonization patterns of black man- grove (Avicennia germinans (L.) L.) gene pools along the Mexican coasts DOI: doi:10.5061/dryad.4r0q7m5 Journal: Journal of Biogeography Journal manuscript number: JBI-18-0269 43 |OCHOA- ZAVALA et AL. Juan Núñez-Farfán https://orcid.org/0000-0001-5829-8338 R E FE R E N CE S Alleman, L. K., & Hester, W. H. (2011). Reproductive ecology of black man- grove (Avicennia germinans) along the Louisiana coast: Propagule produc- tion cycles, dispersal limitations, and establishment elevations. Estuaries and Coasts, 34, 1068–1077. https://doi.org/10.1007/s12237-011-9404-8 Alongi, D. M. (2002). Present state and future of the world's mangrove forests. Environmental Conservation, 29, 331–349. Alongi, D. M. (2008). Mangrove forests: Resilience, protection from tsu- namis, and responses to global climate change. Estuarine Coastal and Shelf Science, 76, 1–13. https://doi.org/10.1016/j.ecss.2007.08.024 Analuddin, K., Sharma, S., Jamili, Septiana, A., Sahidin, I., Rianse, U., & Nadaoka, K. (2017). Heavy metal bioaccumulation in mangrove eco- system at the coral triangle ecoregion, Southeast Sulawesi, Indonesia. Marine Pollution Bulletin, 125, 472–480. https://doi.org/10.1016/j. marpolbul.2017.07.065 Annan, J. D., & Hargreaves, J. C. (2013). A new global reconstruction of temperature changes at the Last Glacial Maximum. Climate of the Past, 9, 367–376. https://doi.org/10.5194/cp-9-367-2013 Antonov, J. I., Seidov, D., Boyer, T. P., Locarnini, R. A., Mishonov, A. V., Garcia, H. E, … Johnson, D. R. (2010). World Ocean Atlas 2009, Volume 2: Salinity. In S. Levitus (Ed.) NOAA Atlas NESDIS 69. Washington, DC: U.S. Government Printing Office, 184 pp. Bai, W.-N., Wang, W.-T., & Zhang, D.-Y. (2014). Contrasts between the phylogeographic patterns of chloroplast and nuclear DNA highlight a role for pollen- mediated gene flow in preventing population diver- gence in an East Asian temperate tree. Molecular Phylogenetics and Evolution, 81, 37–48. https://doi.org/10.1016/j.ympev.2014.08.024 Balke, T., Bouma, T. J., Horstman, E. M., Webb, E. L., Erftemeijer, P. L. A., & Herman, P. M. (2011). Windows of opportunity: Thresholds to mangrove seedling establishment on tidal flats. Marine Ecology Progress Series, 440, 1–9. https://doi.org/10.3354/meps09364 Barrow, L. N., Bigelow, A. T., Phillips, C. A., & Lemmon, E. M. (2015). Phylogeographic inference using Bayesian model comparison across a fragmented chorus frog species complex. Molecular Ecology, 24, 4739–4758. https://doi.org/10.1111/mec.13343 Beerli, P. (2006). Comparison of Bayesian and maximum- likelihood infer- ence of population genetic parameters. Bioinformatics, 22, 341–345. https://doi.org/10.1093/bioinformatics/bti803 Beerli, P., & Felsenstein, J. (2001). Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences of the United States of America, 98, 4563–4568. https:// doi.org/10.1073/pnas.081068098 Beerli, P., & Palczewski, M. (2010). Unified framework to evaluate panmixia and migration direction among multiple sampling locations. Genetics, 185, 313–326. https://doi.org/10.1534/genetics.109.112532 Behringer, D. W., Molinari, R. L., & Festa, J. F. (1977). The variability of anticyclonic current patterns in the Gulf of Mexico. Journal of Geophysical Research, 82, 5469–5476. https://doi.org/10.1029/ JC082i034p05469 Bernardi, G., Findley, L., & Rocha-Olivares, A. (2003). Vicariance and dispersal across Baja California in disjunct marine fish populations. Evolution, 57, 1599–1609. https://doi.org/10.1111/j.0014-3820.2003. tb00367.x Blanquart, F., Kaltz, O., Nuismer, S. L., & Gandon, S. (2013). A practical guide to measuring local adaptation. Ecology Letters, 16, 1195–1205. https://doi.org/10.1111/ele.12150 Broughton, R. E., Stewart, L. B., & Gold, J. R. (2002). Microsatellite variation suggest substantial gene flow between king mackerel (Scomberomorus cavalla) in the western Atlantic Ocean and Gulf of Mexico. Fisheries Research, 54, 305–316. https://doi.org/10.1016/ S0165-7836(01)00275-2 Carstens, B. C., Brennan, R. S., Chua, V., Duffie, C. V., Harvey, M. G., Koch, R. A., … Sullivan, J. (2013). Model selection as a tool for phy- logeographic inference: An example from the willow Salix melan- opsis. Molecular Ecology, 22, 4014–4028. https://doi.org/10.1111/ mec.12347 Carstens, B. C., Brunsfeld, S. J., Demboski, J. R., Good, J. M., & Sullivan, J. (2005). Investigating the evolutionary history of the Pacific Northwest mesic forest ecosystem: Hypothesis testing within a comparative phytogeography framework. Evolution, 59, 1639–1652. https://doi.org/10.1554/04-661.1 Carstens, B. C., & Richards, C. L. (2007). Integrating coalescent and eco- logical niche modeling in comparative phytogeography. Evolution, 61, 1439–1454. https://doi.org/10.1111/j.1558-5646.2007.00117.x Castellanos-Morales, G., Gámez, N., Castillo-Gámez, R. A., & Eguiarte, L. E. (2016). Peripatric speciation of an endemic species driven by Pleistocene climate change: The case of the Mexican prairie dog (Cynomys mexicanus). Molecular Phylogenetics and Evolution, 94, 171– 181. https://doi.org/10.1016/j.ympev.2015.08.027 Cavanaugh, K. C., Kellner, J. R., Forde, A. J., Gruner, D. S., Parker, J. D., Rodriguez, W., & Feller, I. (2014). Poleward expansion of mangroves is a threshold response to decreased frequency of extreme cold events. Proceedings of the National Academy of Sciences of the United States of America, 111, 723–727. https://doi.org/10.1073/pnas.1315800111 Cerón-Souza, I., Bermingham, E., McMillan, W. O., & Jones, F. A. (2012). Comparative genetic structure of two mangrove species in Caribbean and Pacific estuaries of Panama. BMC Evolutionary Biology, 12, 205. https://doi.org/10.1186/1471-2148-12-205 Cerón-Souza, I., Gonzalez, E. G., Schwarzbach, A. E., Salas-Leiva, D. E., Rivera-Ocasio, E., Toro-Perea, N., … McMillan, W. O. (2015). Contrasting demographic history and gene flow patterns of two man- grove species on either side of the Central American Isthmus. Ecology and Evolution, 5, 3486–3499. https://doi.org/10.1002/ece3.1569 Cerón-Souza, I., Rivera-Ocasio, E., Funk, S. M., & McMillan, W. O. (2006). Development of six microsatellite loci for black mangrove (Avicennia germinans). Molecular Ecology Notes, 6, 692–694. https:// doi.org/10.1111/j.1471-8286.2006.01312.x Cerón-Souza, I., Rivera-Ocasio, E., Medina, E., Jiménez, J. A., McMillan, W. O., & Birmingham, E. (2010). Hybridization and introgression in New World red mangroves, Rhizophora (Rhizophorarceae). American Journal of Botany, 97, 945–957. https://doi.org/10.3732/ajb.0900172 nowCOAST. (2018). Global water currents, water temperature and sa- linity (NOAA). Layer: Surface water currents w/Salinity. Retrieved from https://nowcoast.noaa.gov/arcgis/rest/services/nowcoast/ guidance_model_coastalocean_gomofs_time/MapServer Cornuet, J.-M., Pudlo, P., Veyssier, J., Dehne-García, A., Gautier, M., Leblois, R., … Estoup, A. (2014). DIYABC v2.0: A software to make approximate Bayesian computation inferences about population history using single nucleotide polymorphism, DNA sequence and microsatellite data. Bioinformatics, 30, 1187–1189. https://doi.org/10.1093/bioinformatics/btt763 Davis, M. B., & Shaw, R. G. (2001). Range shifts and adaptive responses to Quaternary climate change. Science, 292, 673–679. https://doi. org/10.1126/science.292.5517.673 De la Lanza, G., Ortíz, M. A., & Carbajal, J. L. (2013). Diferenciación hi- drogeomorfológica de los ambientes costeros del Pacífico, del Golfo de México y del Mar Caribe. Investigaciones Geográficas, Boletín del Instituto de Geografía, UNAM, 2013, 33–50. Dodd, R. S., Afzal-Rafii, Z., Kashani, N., & Budrick, J. (2002). Land barri- ers and open oceans: Effects on gene diversity and population struc- ture in Avicennia germinans L. (Avicenniaceae). Molecular Ecology, 11, 1327–1338. https://doi.org/10.1046/j.1365-294X.2002.01525.x 44 | OCHOA- ZAVALA et AL. Donato, D. C., Kauffman, J. B., Murdiyarso, D., Kurnianto, S., Stidham, M., & Kanninen, M. (2011). Mangroves among the most carbon- rich forest in the tropics. Nature, 4, 293–297. Dray, S., & Dufour, A. B. (2007). The ade4 package: Implementing the duality diagram for ecologists. Journal of Statistical Software, 22(4), 1–20. Earl, D. A., & vonHoldt, B. M. (2012). STRUCTURE HARVESTER: A web- site and program for visualizing STRUCTURE output and implement- ing the Evanno method. Conservation Genetics Resources, 4, 359–361. https://doi.org/10.1007/s12686-011-9548-7 Espinosa, O. D., Ocegueda, C. S., Aguilar, Z. C., Flores, V. O., & Llorente- Bousquets, J. (2008). El conocimiento biogeográfico de las especies y su regionalización natural. In J. Soberón, G. Halfter & J. Llorente- Bousquets (Comp.), Capital natural de México: Conocimiento actual de la biodiversidad (vol. 1, pp. 33–65). México: Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO). Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the num- ber of clusters of individuals using the software STRUCTURE: A simulation study. Molecular Ecology, 14, 2611–2620. https://doi. org/10.1111/j.1365-294X.2005.02553.x Excoffier, L., & Lischer, H. E. L. (2010). Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10, 564–567. https://doi. org/10.1111/j.1755-0998.2010.02847.x Excoffier, L., & Ray, N. (2008). Surfing during population expansions promotes genetic revolutions and structuration. Trends in Ecology and Evolution, 23, 347–351. https://doi.org/10.1016/j.tree.2008. 04.004 Fernández-Eguiarte, A., Zavala-Hidalgo, J., & Romero-Centeno, R. (2018a). Climatología mensual promedio de temperatura superfi- cial del mar (1992-2012). Atlas Climático Digital de México. Centro de Ciencias de la Atmósfera. Universidad Nacional Autónoma de México. Retrieved from http://uniatmos.atmosfera.unam.mx/ Fernández-Eguiarte, A., Zavala-Hidalgo, J., & Romero-Centeno, R. (2018b). Climatología mensual de velocidad geostrófica absoluta (1992-2012). Atlas Climático Digital de México. Centro de Ciencias de la Atmósfera. Universidad Nacional Autónoma de México. Retrieved from http://uniatmos.atmosfera.unam.mx/ Fouquet, A., Noonan, B. P., Rodrigues, M. T., Pech, N., Gilles, A., & Gemmell, N. J. (2012). Multiple quaternary refugia in the Eastern Guiana shield revealed by comparative phylogeography of 12 frog species. Systematic Biology, 61, 461–489. https://doi.org/10.1093/ sysbio/syr130 Garber, A. F., Tringali, M. D., & Stuck, K. C. (2004). Population structure and variation in red snapper (Lutjanus campechanus) from the Gulf of Mexico and Atlantic coast of Florida as determined from mitochon- drial DNA control region sequence. Marine Biotechnology, 6, 175–185. García, E. (1990). Rangos de humedad. Extraído de Climas. IV.4.10. Atlas Nacional de México. Vol II. Scale 1: 4000000. Instituto de Geografía UNAM. Mexico. Retrieved from http://www.conabio.gob.mx/ informacion/gis/ Giri, C., Ochieng, E., Tieszen, L. L., Zhu, Z., Singh, A., Loveland, T., … Duke, N. (2011). Status and distribution of mangrove forests of the world using earth observation satellite data. Global Ecology and Biogeography, 20, 154–159. https://doi.org/10.1111/j.1466-8238.2010.00584.x Google. (2016). Google Earth (version 7.1.7.2600). Mountain View, CA: Google. Goudet, J. (2002). FSTAT software (version 2.9.3.2): A program to esti- mate and test gene diversities and fixation indices. Retrieved from http://www.unil.ch/popgen/softwares/fstat.htm Groeneveld, J., Hathorne, E. C., Steinke, S., DeBey, H., Mackensen, A., & Tiedemann, R. (2014). Glacial induced closure of the Panamanian gateway during Marine Isotope Stages (MIS) 95- 100 (~2.5 Ma). Earth and Planetary Science Letters, 404, 296–306. https://doi. org/10.1016/j.epsl.2014.08.007 Guo, Z., Guo, W., Wu, H., Fang, X., Ng, W. L., Shi, X., … Huang, Y. (2018). Differing phylogeographic patterns within the Indo- West Pacific mangrove genus Xylocarpus (Meliaceae). Journal of Biogeography, 45, 676–689. https://doi.org/10.1111/jbi.13151 Guo, Z., Li, X., He, Z., Yang, Y., Wang, W., Zhong, C., & Duke, N. C. (2018). Extremely low genetic diversity across mangrove taxa reflects past sea level changes and hints at poor future responses. Global Change Biology, 24, 1741–1748. https://doi.org/10.1111/gcb.13968 Hallatschek, O., & Nelson, D. R. (2008). Gene surfing in expanding pop- ulations. Theoretical Population Biology, 73, 158–170. https://doi. org/10.1016/j.tpb.2007.08.008 Heist, E. J., & Gold, J. R. (2000). DNA microsatellite loci and genetic structure of red snapper in the Gulf of Mexico. Transactions of the American Fisheries Society, 129, 469–475. https://doi.org/10.1577/15 48-8659(2000)129<0469:DMLAGS>2.0.CO;2 Hewitt, G. M. (1996). Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society, 58, 247–276. https://doi.org/10.1111/j.1095-8312.1996.tb01434.x Hewitt, G. M. (2000). The genetic legacy of the Quaternary ice ages. Nature, 405, 907–913. https://doi.org/10.1038/35016000 Hewitt, G. M. (2004). Genetic consequences of climatic oscillations in the Quaternary. Philosophical Transactions of the Royal Society B, 359, 183–195. https://doi.org/10.1098/rstb.2003.1388 Janes, J. K., Miller, J. M., Dupuis, J. R., Malenfant, R. M., Gorrel, J. C., Cullingham, C. I., & Andrew, R. M. (2017). The K = 2 conundrum. Molecular Ecology, 26, 3594–3602. https://doi.org/10.1111/mec.14187 Jombart, T. (2008). adegenet: A R package for the multivariate analy- sis of genetic markers. Bioinformatics, 24, 1403–1405. https://doi. org/10.1093/bioinformatics/btn129 Jombart, T., Devillard, S., & Balloux, F. (2010). Discriminant analysis of principal components: A new method for the analysis of genet- ically structured populations. BMC Genetics, 11, 94. https://doi. org/10.1186/1471-2156-11-94 Keller, J. A., Wilson, G. K., Reeve, A. S., & Platenberg, R. (2017). Mangroves buffer marine protected area from impacts of Bovoni Landfill, St. Thomas, United States Virgin Islands. Wetlands Ecology and Management, 25, 563–582. https://doi.org/10.1007/s11273-017-9536-0 Kennedy, J. P., Pil, M. W., Proffitt, E., Boeger, W. A., Stanford, A. M., & Devlin, D. (2016). Postglacial expansion pathways of red mangrove, Rhizophora mangle, in the Caribbean Basin and Florida. American Journal of Botany, 103, 260–276. https://doi.org/10.3732/ajb.1500183 Krauss, K. W., Lovelock, C. E., Mckee, K. L., López-Hoffman, L., Ewe, S. M. L., & Sousa, W. P. (2008). Environmental drivers in mangrove establishment and early development: A review. Aquatic Botany, 89, 105–127. https://doi.org/10.1016/j.aquabot.2007.12.014 Landínez-García, R. M., Ospina-Guerrero, S. P., Rodríguez-Castro, D.J., Arango, R., & Márquez, E. (2009). Genetic analysis of Lutjanus synagris populations in the Colombian Caribbean. Ciencias Marinas, 35, 321–331. Retrieved from http://www.redalyc.org/articulo.oa?id=48013184001> ISSN 0185-3880 https://doi.org/10.7773/cm Liepelt, S., Bialozyt, R., & Ziegenhagen, B. (2002). Wind- dispersed pol- len mediates postglacial gene flow among refugia. Proceedings of the National Academy of Sciences of the United States of America, 99, 14590–14594. https://doi.org/10.1073/pnas.212285399 López, M. D., Alcocer, M. U., & Jaimes, P. D. (2010). Phylogeography and historical demography of the Pacific Sierra mackerel (Scomberomorus sierra) in the Eastern Pacific. BMC Genetics, 11, 34. https://doi. org/10.1186/1471-2156-11-34 Lozano-García, S., Torres-Rodríguez, E., Ortega, B., Vázquez, G., & Caballero, M. (2013). Ecosystem responses to climate and dis- turbances in western central Mexico during late Pleistocene and Holocene. Palaeography, Paleoclimatology, Palaeoecology, 370, 184– 195. https://doi.org/10.1016/j.palaeo.2012.12.006 Maliouchenko, O., Palmé, A. E., Buonamici, A., Vendramin, G. G., & Lascoux, M. (2007). Comparative phytogeography and population 45 |OCHOA- ZAVALA et AL. structure of European Betula species, with particular focus on B. pendular and B. pubescens. Journal of Biogeography, 34, 1601–1610. https://doi.org/10.1111/j.1365-2699.2007.01729.x Mantel, N. A. (1967). The detection of disease clustering and a general- ized regression approach. Cancer Research, 27, 209–220. Marchand, C., Fernández, J.-M., & Moreton, B. (2016). Trace metal geo- chemistry in mangrove sediments and their transfer to mangrove plants (New Caledonia). Science of the Total Environment, 562, 216– 227. https://doi.org/10.1016/j.scitotenv.2016.03.206 Marske, K. A., Leschen, R. A. B., & Buckley, T. R. (2012). Concerted versus independent evolution and the search for multiple refugia: Comparative phylogeography of four forest beetles. Evolution, 66, 1862–1877. https://doi.org/10.1111/j.1558-5646.2011.01538.x McMillan, C. (1971). Environmental factors affecting seedling establish- ment of the black mangrove on the Central Texas coast. Ecology, 52, 927–930. https://doi.org/10.2307/1936046 McMillan, C. (1986). Isozyme patterns among populations of black man- grove, Avicennia germinans, from the Gulf of Mexico- Caribbean and Pacific Panama. Contributions in Marine Science, 29, 17–25. Metcalfe, S. E., O'Hara, S. L., Caballero, M., & Davies, S. J. (2000). Records of Late Pleistocene- Holocene climatic change in Mexico – a review. Quaternary Science Reviews, 19, 699–721. https://doi.org/10.1016/ S0277-3791(99)00022-0 Mori, G. M., Zucchi, M. I., Sampaio, I., & Souza, A. P. (2010). Microsatellites for the mangrove tree Avicennia germinans (Acanthaceae): Tools for hybridization and mating system studies. American Journal of Botany, 97, e79–e81. https://doi.org/10.3732/ ajb.1000219 Mori, G. M., Zucchi, M. I., & Souza, A. P. (2015). Multiple- geographic- scale genetic structure of two mangrove tree species: The roles of mating system, hybridization, limited dispersal and extrinsic factors. PLoS ONE, 10, e0118710. https://doi.org/10.1371/journal. pone.0118710 Munguia-Vega, A., Marinone, S. G., Paz-Garcia, D., Giron-Nava, A., Plomozo-Lugo, T., Gonzalez-Cuellar, O., … Reyes-Bonilla, H. (2018). Anisotropic larval connectivity and metapopulation structure driven by directional ocean currents in a marine fish targeted by small- scale fisheries. Marine Biology, 165, 16. https://doi.org/10.1007/ s00227-017-3267-x Muss, A., Robertson, D. R., Stepien, C. A., Wirtz, P., & Bowen, B. W. (2001). Phylogeography of Ophioblennius: The role of ocean currents and geography in reef fish evolution. Evolution, 55, 561–572. https://doi.org/10.1554/0014-3820(2001)055[0561: POOTRO]2.0.CO;2 Nathan, R., Schurr, F. M., Spiegel, O., Steinitz, O., Trakhtenbrot, A., & Tsoar, A. (2008). Mechanisms of long- distance seed dispersal. Trends in Ecology and Evolution, 23, 638–647. https://doi.org/10.1016/j. tree.2008.08.003 Nei, M. (1987). Molecular Evolutionary Genetics. New York, NY: Columbia University Press, 512 pp. Nei, M., Tajima, F., & Tateno, Y. (1983). Accuracy of estimated phyloge- netic trees from molecular data. Journal of Molecular Evolution, 19, 153–170. https://doi.org/10.1007/BF02300753 Nettel, A., & Dodd, R. S. (2007). Drifting propagules and receding swamps: Genetic footprints of mangrove recolonization and dis- persal along tropical coasts. Evolution, 61, 958–971. https://doi. org/10.1111/j.1558-5646.2007.00070.x Nettel, A., Rafii, F., & Dodd, R. S. (2005). Characterization of micro- satellite markers for the mangrove tree Avicennia germinans L. (Avicenniaceae). Molecular Ecology Notes, 5, 103–105. https://doi. org/10.1111/j.1471-8286.2004.00851.x Nettel-Hernanz, A., Dodd, R. S., Ochoa-Zavala, M., Tovilla-Hernández, C., & Días-Gallegos, J. R. (2013). Mating system analyses of tropi- cal populations of the black mangrove, Avicennia germinans (L.) L. (Avicenniaceae). Botanical Sciences, 91, 115–117. Ngeve, M. N., van der Stocken, T., Sierens, T., Koedam, N., & Triest, L. (2017). Bidirectional gene flow on a mangrove river land- scape and between- catchment dispersal of Rhizophora race- mosa (Rhizophoraceae). Hydrobiologia, 790, 93–108. https://doi. org/10.1007/s10750-016-3021-2 O'Dea, A., Lessios, H. A., Coates, A. G., Eytan, R. I., Restrepo-Moreno, S. A., Cione, A. L., & Jackson, J. B. C. (2016). Formation of the Isthmus of Panama. Science Advances, 2, e1600883. https://doi.org/10.1126/ sciadv.1600883 Ochoa-Zavala, M., Jaramillo-Correa, J. P., Piñero, D., Nettel-Hernanz, A. & Núñez-Farfán, J. (2019). Data from: Contrasting coloniza- tion patterns of black mangrove (Avicennia germinans (L.) L.) gene pools along the Mexican coasts. Dryad Digital Repository. https://doi. org/10.5061/dryad.4r0q7m5 Orsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J., & de Meester, L. (2013). Drivers of population genetic differentiation in the wild: Isolation by dispersal limitation, isolation by adaptation and isolation by coloni- zation. Molecular Ecology, 22, 5983–5999. https://doi.org/10.1111/ mec.12561 Ortega-Del Vecchyo, D., Piñero, D., Jardón-Barbolla, L., & van Heerwaarden, J. (2017). Appropriate homoplasy metrics in linked SSRs to predict an underestimamtion of demographic expansion times. BMC Evolutionary Biology, 17, 213. https://doi.org/10.1186/s12862-017-1046-4 Pacheco-Almanzar, E., Ramírez-Saad, H., Velázquez-Aragón, J. A., Serrato, A., & Ibáñez, A. L. (2017). Diversity and genetic structure of white mullet populations in the Gulf of Mexico analyzed by micro- satellite markers. Estuarine, Coastal and Shelf Science, 198, 249–256. https://doi.org/10.1016/j.ecss.2017.09.015 Pérez-Villegas, G. (1990). Insolación anual en “observatorios, estaciones meteorológicas e insolación”. IV.4.1 Atlas Nacional de México, Vol. II. Scale 1:8 000 000. Instituto de Geografía, UNAM. Mexico. Retrieved from http://www.conabio.gob.mx/informacion/gis/ Petit, R. J., & Hampe, A. (2006). Some evolutionary consequences of being a tree. Annual Review of Ecology, Evolution and Systematics, 37, 187– 214. https://doi.org/10.1146/annurev.ecolsys.37.091305.110215 Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of popu- lation structure using multilocus genotype data. Genetics Society of America, 155, 945–959. Pruett, C. L., Saillant, E., & Gold, J. R. (2005). Historical population de- mography of red snapper (Lutjanus campechanus) from the north- ern Gulf of Mexico based on analysis of sequences of mitochon- drial DNA. Marine Biology, 147, 593–602. https://doi.org/10.1007/ s00227-005-1615-8 R Core Team. (2016). R: A language and environment for statistical com- puting. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-project.org/ Roberts, D. R., & Hamann, A. (2015). Glacial refugia and modern genetic diversity of 22 western North American tree species. Proceedings of the Royal Society B, 282 (1804), 20142903. https://doi.org/10.1098/ rspb.2014.2903 Rzedowski, J. (2006). Vegetación de México. México: Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO), 504 pp. Sandoval-Castro, E., Dodd, R.S., Riosmena-Rodríguez, R., Enríquez- Paredes, L. M., Tovilla-Hernández, C., López-Vivas, J. M., … Muñiz- Salazar, R. (2014). Post- Glacial expansion and population genetic divergence of mangroves species Avicennia germinans (L.) Stearn and Rhizophora mangle L. along the Mexican coast. PLoS ONE, 9, e93358. https://doi.org/10.1371/journal.pone.0093358 Savolainen, O., Lascoux, M., & Merilä, J. (2013). Ecological genomics of local adaptation. Nature Reviews Genetics, 14, 807–820. https://doi. org/10.1038/nrg3522 Scheinvar, E., Gámez, N., Castellanos-Morales, G., Aguirre-Planter, E., & Eguiarte, L. E. (2016). Neogene and Pleistocene history of Agave lechuguilla in the Chihuahuan Desert. Journal of Biogeography, 44, 322–334. 46 | OCHOA- ZAVALA et AL. Sexton, J. P., Hangartner, S. B., & Hoffmann, A. A. (2014). Genetic isola- tion by environment or distance: Which pattern of gene flow is most common? Evolution, 68, 1–15. https://doi.org/10.1111/evo.12258 Sherrod, C. L., & McMillan, C. (1985). The distributional history and ecology of mangrove vegetation along the northern Gulf of Mexico coastal region. Contributions in Marine Science, 28, 129–140. Slatkin, M. (1987). Gene flow and the geographic structure of natu- ral populations. Science, 236, 787–792. https://doi.org/10.1126/ science.3576198 Spalding, M., Kainuma, M., & Collins, L. (2010). World atlas of mangroves. London: Earthscan, 336 pp. https://doi.org/10.4324/9781849776608 Takayama, K., Tamura, M., Tateishi, Y., Webb, E. L., & Kajita, T. (2013). Strong genetic structure over the American continents and transoceanic dispersal in the mangrove genus Rhizophora (Rhizophoraceae) revealed by broad- scale nuclear and chloroplast DNA analysis. American Journal of Botany, 100, 1191–1201. https:// doi.org/10.3732/ajb.1200567 Tomlinson, P. B. (1986). The botany of mangroves. Cambridge, UK: Cambridge University Press, 419 pp. Tsuda, Y., Nakao, K., Ide, Y., & Tsumura, Y. (2015). The population demogra- phy of Betula maximowicziana, a cool- temperate tree species in Japan, in relation to the last glacial periods: Its admixture- like genetic struc- ture is the result of simple population splitting not admixing. Molecular Ecology, 24, 1403–1418. https://doi.org/10.1111/mec.13123 Van der Stocken, T., De Ryck, D. J. R., Vanschoenwinkel, B., Deboelpaep, E., Bouma, T. J., Dahdouh-Guebas, F., & Koedam, N. (2015). Impact of landscape structure on propagule dispersal in mangrove forest. Marine Ecology Progress Series, 524, 95–106. https://doi.org/10.3354/ meps11206 Van Oosterhout, C., Hutchinson, W. F., Wills, D. P. M., & Shipley, P. (2004). MICRO- CHECKER: Software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes, 4, 535–538. https://doi.org/10.1111/j.1471-8286.2004.00684.x Waters, J. M., Fraser, C. I., & Hewitt, G. M. (2013). Founder takes all: Density- dependent processes structure biodiversity. Trends in Ecology & Evolution, 28, 78–85. https://doi.org/10.1016/j. tree.2012.08.024 Weir, B. S., & Cockerham, C. C. (1984). Estimating F- statistics for the analysis of population structure. Evolution, 38, 1358–1370. Xu, S., He, Z., Zhang, Z., Guo, Z., Lyu, H., Li, J., & Shi, S. (2017). The origin, diversification and adaptation of a major mangrove clade (Rhizophoreae) revealed by whole- genome sequencing. Molecular Biology & Genetics, 4, 721–734. Yan, Y.-B., Duke, N., & Sun, M. (2016). Comparative analysis of the pattern of population genetic diversity in three Indo- West Pacific Rhizophora mangrove species. Frontiers in Plant Science, 7, 1434. Yan, Z., Sun, X., Zhang, Q., & Li, X. (2017). Accumulation and tolerance of mangroves to heavy metals: A review. Current Pollution Reports, 3, 302–317. https://doi.org/10.1007/s40726-017-0066-4 Maried Ochoa-Zavala is broadly interested in the evolution- ary biology and phylogeography of tropical species. This paper constitutes a central goal of her PhD thesis at the Universidad Nacional Autónoma de México. The other authors collaborate on different aspects of evolutionary ecology and quantitative ge- netics in different species of plants and animals. Author contributions: J.N.F., D.P. and M.O.Z. conceived the ideas; M.O.Z. and J.N.F. collected the data; J.P.J.C., M.O.Z. ana- lysed the data; and M.O.Z. led the writing with active assistance from J.N.F., J.P.J.C., D.P. and A.N.H. Additional supporting information may be found online in the Supporting Information section at the end of the article. How to cite this article: Ochoa-Zavala M, Jaramillo-Correa JP, Piñero D, Nettel-Hernanz A, Núñez-Farfán J. Contrasting colonization patterns of black mangrove (Avicennia germinans (L.) L.) gene pools along the Mexican coasts. J Biogeogr. 2019;46:884–898. https://doi.org/10.1111/jbi.13536 47 48 Journal of Biogeography SUPPORTING INFORMATION Contrasting colonization patterns of black mangrove (Avicennia germinans (L.) L.) gene pools along the Mexican coasts Maried Ochoa-Zavala1,2. Juan Pablo Jaramillo-Correa1. Daniel Piñero1. Alejandro Nettel- Hernanz2. Juan Núñez-Farfán1. 49 Appendix S1. Primers used and amplification protocols We used 10 nuclear microsatellite loci (Table S1.1) developed by Cerón-Souza et al. (2006, 2012), Mori et al. (2010) and Nettel et al. (2005). The forward primer of each primer pair was labelled with one fluorescent label of 6-FAM, VIC, PET or NED (Applied Biosystems). Loci were amplified either individually using a GoTaq Flexi DNA Polymerase (Promega) or by three Qiagen multiplex PCR reactions. Each Multiplex reaction contained 1X of Multiplex PCR master mix buffer (including HotStarTaq DNA polymerase and synthetic factor MP), 0.2M of each primer and 40 ng/µL of genomic DNA. Each of the GoTaq Flexi DNA Polymerase reactions contained 1X colorless buffer, 1.5 mM of MgCl2 solution, 0.2 mM of dNTPs, 0.1 µM of each forward and reverse primers, 1U of DNA polymerase and 40 ng/µL of template DNA. Table S1.1. Characteristics of 10 nuclear microsatellites used in this study. Primer Tannealing (° C) Repeat sequence Label Reference Agerm_CT_003 TD 65.5 – 55.5 (CT)14 VIC Cerón-Souza et al., 2006 Agerm_CT_004 TD 65 – 55 (CT)17 PET Cerón-Souza et al., 2012 Agerm_GA_003 (GA)16 NED Agerm_GT_006 (GT)16 6-FAM Agerm-25 TD 65 – 60 (TG)10…(TG)4 6-FAM Mori et al., 2010 Agerm-22 64.5 (TTTCTT)4 PET Agerm-18 (AG)16 VIC AgD6 52 (ATT)4N7(GT)15 PET Nettel et al., 2005 AgT9 (CA)8(GA)2(CAGA)9 NED AgT4 (CATA)5CATG(CATA)9 6-FAM TD indicates touchdown PCR We formed three mixtures of primers for amplification with Qiagen Multiplex PCR Kit with the following PCR cycling protocol for each: 50 1) Primers Agerm_CT_004, Agerm_GA_003 and Agerm_GT_006. Touchdown-PCR with initial activation step for 15 min at 95° C followed by 12 cycles of 94° C for 40 s, 65° C per 1 min, 72° C during 40 s, a second round of 25 cycles of 40 s at 94° C, annealing temperature of 55° C for 1 min, 72° C per 40 s, and final extension of 4 min at 72° C. 2) Primers Agerm-22 and Agerm-18. Initial activation step for 15 min at 95° C, followed by 30 cycles of 94° C for 1 min, annealing temperature of 64.5° C for 40 s, 72° C for 2 min and final extension step at 72° C for 5 min. 3) Primers AgD6, AgT9 and AgT4. An activation step of 95° for 15 min followed by 35 cycles of 95° C for 30 s, 52° C for 30 s, 72° C for 45 s, and final extension at 72° C for 30 min. The remaining loci were amplified using GoTaq Flexi DNA Polymerase (Promega) under Touchdown-PCR with the following conditions: 1) Agerm_CT_003, 94° C for 3 min, followed by 12 cycles of 94° C for 40 s, 65.5° C for 1 min, 72° C for 40 s, a second round of 25 cycles of 94° C for 40 s, 55.5° C for 1 min, 72° C for 40s, and final extension of 72° C for 4 min. 2) Agerm-25, 94° C for 2 min, followed by 12 cycles of 94° C for 1 min, 65° C for 1 min, 72° C for 2 min, a second round of 20 cycles of 94° C for 1 min, 60° C for 1 min, 72° C for 2 min, and final extension step of 72° C for 5 min. 51 Appendix S2. ABC analyses By means of DIYABC 2.1.0 we assessed the divergence time between Atlantic (N1) and Pacific (N2) populations from ancestral population of size NA. We simulated 1,000,000 data sets under the scenario of Fig S2.1 Fig. S2.1 Scenario coded on the left side and the graphic representation given by DIYABC Before running the final analysis, we spent a lot of time in the fine-tuning: testing several combinations of priors of demographic parameters, including different distribution types, mutations rates (including the default values provided by the software) assumed for microsatellite loci in plants (i.e., Udupa & Baum, 2001; Vigouroux et al., 2002), and adding or removing summary statistics provided by DIYABC. Finally, we identified the prior combinations of demographic parameters and summary statistics that better adjusted to the observed data. Prior distribution of demographic parameters were as follow: Normal (min. = 0, max. = 1000, mean = 500 and SD = 50) for N1 and N2, normal (min.= 0, max. = 5000, mean = 2560 and SD = 100) for NA and for t, uniform (min. = 5, max. = 400000). We set a condition N2 < N1 based on our genetic diversity estimates (HO, HE) that points the Atlantic populations have higher effective population size than the Pacific ones. Besides, previous studies indicated that Avicennia could arrive at the Caribbean some 16 Mya N1 N2 0 sample 1 0 sample 2 t merge 1 2 t varNe 1 NA 52 (Plaziat et al., 2001) and probably colonized the Pacific coast before the final closure of the Central American Isthmus (Dodd et al., 1998), which should coincide with a larger population size on the Atlantic. In addition, we assumed a step-wise mutation model with a mean mutation rate of 1 ´ 10-5 – 1 ´ 10-3 per generation, as minimum and maximum, with gamma distribution. The summary statistics include mean gene diversity across loci, FST values, mean genetic diversity and mean allele size variance between pairs of populations. We obtained visual information (by running a PCA) on how the observed data sets under this scenario have a good fit of the posterior distribution (Fig S2.2). Fig S2.2 Fit of the posterior distribution for each PCA component. 53 We obtained the posterior distributions of the parameters. The red and green lines show the prior and posterior distribution, respectively. Above each panel, the mean of the posterior distribution is given (Fig S2.3). Fig S2.3 Posterior distribution of the parameters. Atlantic (N1) and Pacific (N2) population sizes, NA = Ancestral population size, t = divergence time and µmic = mutation rate 54 Appendix S3. Migration hypotheses Given that putative refugial populations (i.e the southernmost ones) may contribute to recolonized areas, estimated gene flow is critical to make inferences of post-glacial migration dynamics (Carstens et al., 2013). Because Migrate-n can estimate some population genetics parameters such as historical effective population size (Q = 4Neµ, where Ne is the effective population size and µ is the mutation rate) and migration rates (M = m/µ, where m is the migration rate), based on a coalescent approach, we implemented the Bayesian inference strategy from Migrate-n 3.2.15 (Beerli & Felsenstein, 2001; Beerli, 2006; Beerli & Palczewski, 2010) to compare different migration models that describe recolonization from different glacial refugia within each basin (Pacific or Atlantic). This method can be used to compare marginal likelihoods and Bayes factor (BF) of different hypotheses (models) on the directionality of gene flow and populations’ connectivity to test the best fit to the data (Carstens et al., 2013). We constructed 10 models (six and four for the Pacific and Atlantic coasts, respectively) based on the population clusters previously identified within each basin, ordered from north to south (see Figs. 3 and 4). We first define putative glacial populations during the Last Glacial Maximum (LGM) that could serve as refugia, for that, we relied on the fact that mangrove forests cover decreases as the number of days colder than -4° C in a year increases (Cavanaugh et al. 2014). Thus, by using the reconstruction of surface air and sea temperatures during the LGM (Table S3.2, S3.3) by Annan & Hargreaves (2013), we determined the clusters that most likely persisted during the LGM: C3P, C4P and C5P for the Pacific basin, and C2A and C3A for the Atlantic coast. Then, 10 migration hypotheses differing in the number of refugia and directionality of gene flow were built under three main assumptions: 1) A stepping-stone model; 2) shared ancestry pattern found by STRUCTURE (Pritchard et al., 2000) and the pattern of gene flow observed during experimental runs with Migrate-n (see below); and 3) the ocean currents pattern modeled by Fernández-Eguiarte et al. (2018), which may have promoted gene flow and colonization (see Table 2 for models description). 55 Table S3.2. Air and sea temperature during the Last Glacial Maximum along the Pacific coast. Approximately among the following latitudes Air temperature Sea temperature 29º N (Northern Pacific) –8º C –4º C 24º N to –2º C to – 2º C . . ~16º N (Southern Pacific) –2º C to –1ºC –1º C Note that this temperature estimation could differ with other studies (e.g., Bush & Philander, 1999). Here we use the latest information available because we consider it provides the best model-data synthesis. Table S3.3 Air and sea temperature during the Last Glacial Maximum along the Atlantic coast. Climatic conditions within this basin were slightly different. Approximately among the following latitudes Air temperature Sea temperature ~ 25º N (Northern limit of the Gulf of Mexico, south of Texas, USA) –8º C to –4º C –4º C to –2º C ~ 24º N - 20º N (almost all the basin) –4º C to –2º C ~ 21º N - 20º N (Yucatan Peninsula) –2º C to –1ºC Note that this temperature estimation could differ with other studies (e.g., Bush & Philander, 1999). Here we use the latest information available because we consider it provides the best model-data synthesis. All models were run in parallel on Four Intel® Xeon® Processors E7-4809 V3 (20M Cache, 2.00GHz). We first experimented with the full model to adjust the prior distribution 56 for each coast. We chose the following uniform prior distribution for the Pacific: Q min. = 0, max. = 2, delta = 0.2 and M: min. = 0, max. = 150, delta = 15; Atlantic basin: Q min. = 0, max. = 2, delta = 0.2 and M: min. = 0, max. = 100, delta = 10. Each run implemented the infinite allele model, initial values were computed using FST, and mutation rate was calculated from the data. We performed two replicates with a static heading scheme using four short chains and a single long chain. We discarded 10,000 trees as burn-in and recorded 50,000 steps with a sampling increment of 100. We calculated the historical number of migrants (Nm) as (QM)/4. To evaluate chains convergence, we examined the posterior distributions of histograms of each parameter (Fig S3.4) looking for a smooth curve with a single peak. We used the marginal likelihoods by thermodynamic integration with Bezier approximation to calculate BF and posterior model probabilities (Beerli & Palczewski, 2010). Finally, the best-fit models given by the data within each coast were run under a Brownian motion model, following the strategy described above, to obtain Nm estimations. 5 7 F ig S 3 .4 P o sterio r d istrib u tio n o f th e p aram eters q an d N m o b tain ed w ith M ig rate-n . T h e b est-fitted m o d el tested is d ep icted fo r th e A tlan tic an d P acific co asts. 5 8 A p p en d ix S 4 . M icro -ch eck er a n a ly ses, H a rd y – W ein b erg eq u ilib riu m (H W E ) a n d lin k a g e d iseq u ilib riu m (L D ) In d iv id u al trees o f A vicen n ia g erm in a n s w ere g en o ty p ed fo r ten n u clear m icro satellite m ark ers. T w o o f th em (A g erm _ C T _ 0 0 3 , A g erm _ G A _ 0 0 3 ) w ere ex clu d ed fro m fu rth er an aly ses after d etectin g h ig h freq u en cies ( > 0 .2 0 ) o f n u ll alleles in m o re th an h alf o f th e p o p u latio n s in b o th co asts, in o rd er to av o id sig n ifican t d ev iatio n s fro m H W E an d o v erestim atio n o f g en etic stru ctu re. In 1 5 p o p u latio n s in th e P acific an d 1 4 p o p u latio n s in th e A tlan tic b asin , w e fo u n d a freq u en cy o f 0 .2 1 an d 0 .1 7 , resp ectiv ely , o f lo ci sh o w in g ev id en ce o f H W E d ev iatio n . L ik ew ise, a lo w freq u en cy o f lo ci (0 .1 1 ) in L D w as fo u n d fo r b o th co asts. In g en eral, w e d id n o t fin d an y sp ecific p attern o f g en o ty p in g erro rs (T ab le S 4 .4 , S 4 .5 ), d ev iatio n o f H W E (T ab le S 4 .6 ) o r L D (T ab le S 4 .7 ) fo r th e rem ain in g eig h t lo ci. 5 9 T a b le S 4 .4 M icro -ch eck er an aly ses. R esu lts fo r stu tters (S t), allele d ro p o u t (A D ) an d n u ll alleles (N A ). A g erm _ C T _ 0 0 4 A g erm _ G T _ 0 0 6 A g erm -2 5 A g erm -2 2 A g erm -1 8 A g T 9 A g T 4 A g D 6 S t A D N A S t A D N A S t A D N A S t A D N A S t A D N A S t A D N A S t A D N A S t A D N A P N 0 4 × × P N 0 2 P N 0 6 P N 0 3 × × P N 0 5 P N 1 1 × × × × × × × P N 1 2 × P N 1 0 × × × × P C 1 5 × P C 1 1 × × P C 1 9 P S 1 7 × P S 2 0 × × P S 2 9 × × P S 2 4 × G M 3 7 G M 4 1 × G M 5 6 G M 4 0 G M 5 3 × × G M 5 4 G M 3 6 × × × 6 0 G M 4 6 P Y 6 7 P Y 6 6 P Y 7 0 × P Y 8 1 P Y 6 5 P Y 7 7 6 1 T a b le S 4 .5 N u ll allele freq u en cies estim ated b y O o sterh o u d ’s m eth o d w ith M icro -ch eck er, b y lo cu s b y p o p u latio n . V alu es in b o ld ty p e = m o n o m o rp h ic P o p ID A g erm C T 0 0 4 A g erm G T 0 0 6 A g erm -2 5 A g erm -2 2 A g erm -1 8 A g T 9 A g T 4 A g D 6 M ea n PACIFIC COAST P N 0 4 0 0 0 0 0 0 0 0 .1 7 2 3 0 .1 7 2 3 P N 0 2 -0 .0 1 9 4 0 0 0 -0 .0 1 2 6 0 .1 3 2 8 0 .1 2 0 2 -0 .0 1 2 9 0 .0 4 1 6 P N 0 6 0 -0 .0 2 5 3 -0 .1 0 5 6 0 0 .0 2 8 2 0 .1 1 9 9 -0 .0 1 2 6 -0 .1 5 3 8 -0 .0 2 4 9 P N 0 3 0 0 .2 3 5 6 0 .0 2 2 8 0 0 -0 .0 1 2 6 -0 .0 1 2 6 0 .1 0 6 8 0 .0 6 8 0 P N 0 5 0 0 0 .0 6 0 9 0 0 .0 8 1 5 0 0 0 .0 6 7 5 0 .0 7 0 0 P N 1 1 0 .1 3 7 2 0 .1 6 0 5 0 .1 7 2 8 -0 .0 1 2 9 0 .2 0 5 4 0 .1 4 7 2 0 .0 0 0 9 -0 .0 1 9 3 0 .0 9 9 0 P N 1 2 0 -0 .0 6 0 7 -0 .1 0 4 3 -0 .0 1 2 6 0 .1 4 1 6 -0 .0 1 2 6 0 .1 2 0 4 -0 .0 8 8 -0 .0 0 2 0 P N 1 0 0 .1 5 2 7 0 .1 4 7 4 0 .0 6 1 8 0 0 .1 9 9 5 -0 .0 3 7 9 -0 .0 3 7 7 -0 .1 0 2 6 0 .0 5 4 7 P C 1 5 0 0 0 0 0 .0 4 5 2 0 .1 4 5 5 -0 .0 1 6 0 .0 4 0 8 0 .0 5 3 9 P C 1 1 -0 .0 1 5 3 0 0 .1 3 4 1 0 -0 .1 7 7 4 0 .1 7 2 3 -0 .0 1 2 6 0 .1 7 3 1 0 .0 4 5 7 P C 1 9 0 0 0 0 -0 .0 2 6 2 0 .0 2 4 1 0 .0 4 5 9 -0 .0 1 3 6 0 .0 0 7 6 P S 1 7 0 0 .1 4 5 5 0 -0 .0 1 2 6 -0 .0 2 3 1 0 .0 4 1 8 -0 .0 9 2 4 -0 .0 0 6 6 0 .0 0 8 8 P S 2 0 0 -0 .0 5 1 3 0 0 .1 2 5 4 0 .0 9 6 4 -0 .0 3 8 1 0 .0 1 2 -0 .0 1 8 7 0 .0 2 1 0 P S 2 9 0 .1 5 0 8 0 .0 3 6 8 0 .1 3 3 0 .1 3 2 8 0 .1 6 9 -0 .0 6 2 5 0 .0 7 7 4 -0 .0 2 6 1 0 .0 7 6 4 P S 2 4 0 0 .0 0 7 4 -0 .0 5 1 3 0 -0 .0 2 7 5 0 .1 3 2 3 0 .0 4 6 1 -0 .0 4 3 3 0 .0 1 0 6 6 2 T a b le S 4 .5 . C o n tin u ed ATLANTIC COAST G M 3 7 0 .0 4 0 .1 1 1 9 0 0 0 .1 1 1 9 0 0 .1 1 8 -0 .0 1 2 6 0 .0 7 3 8 G M 4 1 0 .1 7 4 8 -0 .0 6 3 6 0 0 .0 5 8 1 -0 .0 1 2 6 0 -0 .0 0 0 3 -0 .0 3 7 9 0 .0 1 9 8 G M 5 6 -0 .0 2 2 7 -0 .1 6 9 5 0 -0 .1 1 7 3 -0 .0 8 9 7 0 0 .0 2 5 5 -0 .0 5 5 7 -0 .0 7 1 6 G M 4 0 0 .0 8 8 5 0 .0 2 9 2 -0 .0 1 4 8 -0 .0 7 6 5 -0 .0 2 9 6 0 0 .0 5 5 0 .0 1 6 3 0 .0 0 9 7 G M 5 3 -0 .0 1 6 1 -0 .1 9 6 8 -0 .0 2 5 2 0 .1 2 2 5 0 .1 9 6 3 0 .1 4 3 9 0 .0 8 5 9 0 .0 5 7 4 0 .0 4 6 0 G M 5 4 0 .0 1 9 3 -0 .0 5 9 7 0 -0 .2 1 9 7 -0 .0 1 5 7 0 0 .0 7 4 8 -0 .0 6 0 7 -0 .0 4 3 6 G M 3 6 -0 .0 4 9 3 0 .1 6 7 1 0 .1 4 9 -0 .0 7 -0 .0 3 9 2 0 .1 4 7 2 -0 .0 3 1 6 0 .0 3 0 2 0 .0 3 7 9 G M 4 6 0 .0 7 1 6 -0 .0 7 2 6 -0 .0 2 8 8 0 .0 8 8 6 -0 .1 5 4 8 0 -0 .0 0 4 7 -0 .0 2 2 4 -0 .0 1 7 6 P Y 6 7 -0 .1 3 7 2 0 .0 3 9 4 0 -0 .0 1 8 1 -0 .1 5 7 9 0 -0 .1 2 3 6 -0 .1 1 8 5 -0 .0 8 6 0 P Y 6 6 0 .1 2 6 3 -0 .0 2 6 2 -0 .0 2 4 5 0 .0 9 7 9 -0 .1 9 8 6 0 -0 .1 2 4 0 .0 7 7 6 -0 .0 1 0 2 P Y 7 0 0 .0 8 2 0 .1 4 4 4 0 -0 .0 1 4 -0 .3 3 7 5 0 -0 .1 5 4 0 .1 1 3 1 -0 .0 2 7 7 P Y 8 1 0 .0 1 5 8 -0 .1 6 3 7 -0 .0 1 2 6 0 -0 .5 5 2 8 -0 .0 1 2 6 -0 .0 1 1 6 0 .0 3 1 2 -0 .1 0 0 9 P Y 6 5 0 .0 4 6 1 0 .0 4 6 3 0 0 .1 3 2 8 -0 .3 6 9 7 0 -0 .1 0 8 1 0 .1 0 7 3 -0 .0 2 4 2 P Y 7 7 0 .0 8 3 4 0 .0 0 7 9 0 0 .0 9 4 9 -0 .2 4 4 3 0 -0 .0 0 4 5 0 .0 1 6 1 -0 .0 0 7 8 6 3 T a b le S 4 .6 H W E b y lo cu s, b y p o p u latio n . W e sh o w v alu es P < 0 .0 5 ; M , m o n o m o rp h ic P o p ID A g erm C T 0 0 4 A g erm G T 0 0 6 A g erm -2 5 A g erm -2 2 A g erm -1 8 A g T 9 A g T 4 A g D 6 P a c ific C o a st P N 0 4 M M M M M M M 0 .0 0 7 0 8 P N 0 2 M M M 0 .0 3 8 2 1 0 .0 3 6 3 6 P N 0 6 M M P N 0 3 M 0 .0 0 0 0 4 M M P N 0 5 M M M M M P N 1 1 0 .0 1 3 8 3 0 .0 0 0 4 6 0 .0 0 9 4 3 0 0 .0 1 2 7 3 P N 1 2 M 0 .0 0 3 5 2 P N 1 0 0 .0 1 4 0 7 0 .0 0 5 6 7 M 0 P C 1 5 M M M M 0 .0 1 2 6 6 P C 1 1 M 0 .0 3 8 9 M 0 .0 0 7 3 4 0 .0 0 1 4 5 P C 1 9 M M M M P S 1 7 M 0 .0 1 2 6 M P S 2 0 M M 0 .0 0 1 4 3 P S 2 9 0 .0 1 3 6 6 0 .0 2 2 0 7 0 .0 1 2 5 4 0 .0 3 7 9 7 0 .0 0 1 8 9 P S 2 4 0 .0 2 2 4 8 A tla n tic C o a st G M 3 7 M M M 6 4 G M 4 1 0 .0 0 1 0 6 M M G M 5 6 M M G M 4 0 0 .0 4 6 6 8 M G M 5 3 0 .0 0 0 1 2 0 .0 1 1 9 4 G M 5 4 M M G M 3 6 0 .0 0 1 3 3 0 .0 1 3 3 8 0 .0 1 3 1 5 G M 4 6 0 .0 3 1 0 2 0 .0 4 3 4 2 M P Y 6 7 M M P Y 6 6 0 .0 0 3 9 9 M P Y 7 0 0 .0 4 0 7 6 0 .0 0 4 1 1 M 0 M 0 .0 4 2 7 P Y 8 1 M 0 .0 0 0 0 2 P Y 6 5 0 .0 0 8 4 8 M 0 .0 3 8 0 6 0 M P Y 7 7 0 .0 3 3 0 6 M 0 .0 2 4 1 7 M 6 5 T a b le S 4 .7 S ig n ifican t L D b y lo cu s b y p o p u latio n . 0 , A g erm _ C T _ 0 0 4 ; 1 , A g erm _ G T _ 0 0 6 ; 2 , A g erm -2 5 ; 3 , A g erm -2 2 ; 4 , A g erm -1 8 ; 5 , A g T 9 ; 6 , A g T 4 ; 7 , A g D 6 . * P < 0 .0 5 0 -1 0 -2 0 -3 0 -4 0 -5 0 -6 0 -7 1 -2 1 -3 1 -4 1 -5 1 -6 1 -7 2 -3 2 -4 2 -5 2 -6 2 -7 3 -4 3 -5 3 -6 3 -7 4 -5 4 -6 4 -7 5 -6 5 -7 6 -7 P N 0 4 P N 0 2 * * P N 0 6 * * P N 0 3 * P N 0 5 P N 1 1 * * * * * * * * * P N 1 2 * P N 1 0 * * P C 1 5 * * * * * * P C 1 1 * * * * * * P C 1 9 P S 1 7 P S 2 0 * * * P S 2 9 * * * * * * * * * * * * * P S 2 4 * G M 3 7 G M 4 1 * * G M 5 6 * G M 4 0 G M 5 3 * * * * * * * * * * * * * * G M 5 4 * * G M 3 6 * * * * * * * * * * * * * * * * * * G M 4 6 * P Y 6 7 * 6 6 P Y 6 6 P Y 7 0 P Y 8 1 * P Y 6 5 * P Y 7 7 * * 67 Appendix S5. Genetic structure We used STRUCTURE software to assess the genetic differentiation between (Pacific vs. Atlantic; Fig S5.5) and within each coast (Figs S5.6 and S5.7). STRUCTURE HARVESTER was employed to determine the optimum K value based on ∆K and were incorporated by searching for the K value in which a secondary peak was observed for ∆K and were a plateau was reached for the Ln P(D) values. Fig S5.5 Structure plots and mean posterior probability (Ln P(D)) values per cluster (K), based on 10 replicates per K, generated by the STRUCTURE program and ∆K analysis of Ln P(D), according to Evanno et al. (2005) between coasts. 68 Fig S5.6 Structure plots and mean posterior probability (Ln P(D)) values per cluster (K), based on 10 replicates per K, generated by the STRUCTURE program and, ∆K analysis of Ln P(D) according to Evanno et al. (2005). Results of Bayesian clustering including all individuals sampled for the Pacific coast. Fig S5.7 Structure plots and mean posterior probability (Ln P(D)) values per cluster (K), based on 10 replicates per K, generated by the STRUCTURE program and, ∆K analysis of Ln P(D) according to Evanno et al. (2005). Results of Bayesian clustering including all individuals sampled for the Atlantic coast. 69 However, STRUCTURE tends to perform poorly under isolation by distance frameworks, such as in our case (Rosenberg et al., 2005; Schwartz & McKelvey, 2009; Jombart et al. 2010; Pritchard et al. 2010). In such cases, subsequent analyses are needed to detect further nested clusters and/or to infer genetic clines and IBD. Evanno et al. (2005) themselves state quite clearly that their test is simply ad hoc and should be combined with other analyses in cases of hierarchical structuring (see also François & Durand, 2010; Meirmans, 2015). Accordingly, we ran extra Bayesian analyses in order to identify nested clusters under K = 2 within each coast, following the same strategy described in the ‘Materials and methods’ section. We verified the level of genetic differentiation among gene pools (FCT) by Analysis of Molecular Variance (AMOVA) using ARLEQUIN with 10,000 permutations based on a stepwise mutation model and a fixed significance level of 0.01. In the Pacific coast, the uppermost level of genetic structure was K = 2, belonging North and South Pacific (Fig S5.6 and S5.8). In the reanalysis of North Pacific cluster, we found a clear split into two new clusters, which we named C1P and C2P. In the reanalysis of the South Pacific, we identified three clusters (C3P, C4P and C5P) in which an evident difference of C3P populations was observed (Fig S5.8). In addition, AMOVA analyses detected higher genetic differentiation among five genetic clusters than assuming K = 2 (Fig S5.8). These five genetic clusters fitting to K = 5 of the first analysis we performed and match with a secondary peak in ∆K of the Evanno method (Fig S5.6). 70 Fig S5.8. Clustering of Pacific populations. K = 2 (upper panel; including all individuals). In the middle, sub-clusters identified into North and South Pacific, respectively. Lower, we show K = 5 as the optimum genetic clusters determined in the Pacific. Regarding the Atlantic populations, initially we observed an uppermost level of genetic structure of two clusters: the Gulf of Mexico and Yucatan Peninsula (Fig S5.7 and Fig S5.9a) that according to AMOVA analysis, the genetic structure was not significant. In the reanalysis of the Gulf of Mexico cluster, we identified that the northernmost populations (C1A) splitted from the rest (C2A; Fig S5.9b). Assuming K = 3 (C1A, C2A and Yucatan Peninsula as C3A) we detected low but significant differentiation (Fig S5.9c). In the reanalysis of Yucatan Peninsula cluster, we found that these populations can be divided into other two sub-clusters (Fig S5.9d), however, the genetic variation explained by K = 4 was not much higher than K = 3. Thus, we assumed K = 3 as the optimum genetic clusters in the Atlantic. Moreover, this three genetic cluster fitted to K = 3 of our first analysis and a secondary peak was observed in Evanno method. 71 Fig S5.9. Results of the Bayesian clustering analyses in the Atlantic. a) K = 2 including all individuals, b) sub-clusters identified into Gulf of Mexico, c) K = 3 is the optimum number of genetic clusters found in the Atlantic coast and, d) reanalysis of Yucatan Peninsula cluster. 72 Appendix S6. Discriminant Analysis of Principal Components As a complement of Bayesian structure analyses, we used the Discriminant Analysis of Principal Components (DAPC, Jombart et al., 2010) method implemented in adegenet (Jombart 2008). Also, at two levels, within basins (Fig S6.10, S6.11) and between coasts (Fig S6.12). Inspection of the Bayesian information criterion values indicates a subdivision into 15 and 14 clusters, for the Pacific and Atlantic respectively, due to the stepping stone model (that incorporates the IBD pattern detected), easily inferred from the clinal arrangement of the clusters along to the coastlines. 73 Fig S6.10 DAPCs plot obtained from Pacific populations. We show (a) 1-2 and (b) 1-7 discriminant analysis (DA) eigenvalues to observe the clusters. We show the names of clusters identified by STRUCTURE analysis, see Results Fig. 3. 74 Figure S6.11 DAPCs plot obtained from Atlantic populations. We show the names of clusters identified by STRUCTURE analysis, see Results Fig. 4. Figure S6.12 DAPC’s scatterplots obtained between coasts. We show 1-3 DA eigenvalues. 75 Supplementary References Annan, J.D., & Hargreaves, J.C. (2013). A new global reconstruction of temperature changes at the Last Glacial Maximum. Climate of the Past, 9, 367-376. Beerli, P. (2006) Comparison of Bayesian and maximum-likelihood inference of population genetic parameters. Bioinformatics, 22, 341-345. Beerli, P. & Felsenstein, J. (2001) Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences of the United States of America, 98, 4563-4568. Beerli, P. & Palczewski, M. (2010) Unified framework to evaluate panmixia and migration direction among multiple sampling locations. Genetics, 185, 313-326. Bush, A. B. & Philander, S. G. H. (1999). The climate of the Last Glacial Maximum: Results from a coupled atmosphere-ocean general circulation model. Journal of Geophysical Research, 104, 24509-24525. Carstens, B.C., Brennan, R.S., Chua, V., Duffie, C.V., Harvey, M. G., Koch, R.A., Mcmahan, C.D., Nelson, B.J., Newman C.E., Satler, J.D., Seeholzer, G., Posbic, K., Tank, D.C. & Sullivan, J. (2013) Model selection as a tool for phylogeographic inference: an example from the willow Salix melanopsis. Molecular Ecology, 22, 4014-4028. Cavanaugh, K.C., Kellner, J.R., Forde, A.J., Gruner, D.S., Parker, J.D., Rodriguez, W. & Feller, I. (2014). Poleward expansion of mangroves is a threshold response to decreased frequency of extreme cold events. Proceedings of the National Academy of Sciences of the United States of America, 111, 723-727. Cerón-Souza, I., Bermingham, E., McMillan, W. O. & Jones, F. A. (2012). Comparative genetic structure of two mangrove species in Caribbean and Pacific estuaries of Panama. BMC Evolutionary Biology, 12, 205. Cerón-Souza, I., Rivera-Ocasio, E., Funk, M. S. & McMillan, W. O. (2006). Development of six microsatellite loci for black mangrove (Avicennia germinans). Molecular Ecology Notes, 6, 692-694. Dodd, R.S., Rafii, Z.A., Fromard, F. & Blasco, F. (1998) Evolutionary diversity among Atlantic coast mangroves. Acta Oecologica, 19, 323-330. Evanno, G., Regnaut, S. & Goudet, J. (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology, 14, 2611-2620. Fernández-Eguiarte, A., Zavala-Hidalgo, J. & Romero-Centeno, R. (2018). Climatología mensual de velocidad geostrófica absoluta (1992-2012). Atlas Climático Digital de México. Centro de 76 Ciencias de la Atmósfera. Universidad Nacional Autónoma de México. http://uniatmos.atmosfera.unam.mx/ François, O. & Durand, E. (2010) Spatially explicit Bayesian clustering models in population genetics. Molecular Ecology Resources, 10, 773-784. Jombart, T. (2008) adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics, 24, 1403-1405. Jombart, T., Devillard, S. & Balloux, F. (2010) Discriminant analysis of principal components: a new method for the analysis of genetically structured populations. BMC Genetics, 11, 94. Mori, G. M., Zucchi, M. I., Sampaio, I., & Souza, A. P. (2010). Microsatellites for the mangrove tree Avicennia germinans (Acanthaceae): Tools for hybridization and mating system studies. American Journal of Botany, 97, e79-e81. Nettel, A., Afzal-Rafii, F. & Dodd, R. S. (2005). Characterization of microsatellite markers for the mangrove tree Avicennia germinans L. (Avicenniaceae). Molecular Ecology Notes 5, 103– 105. Plaziat, J-C., Cavagnetto, C., Koeniguer, J-C. & Baltzer, F. (2001) History and biogeography of the mangrove ecosystem, based on a critical reassessment of the paleontological record. Wetlands Ecology and Management, 9, 161-179. Pritchard J.K., Stephens, M. & Donnelly, P. (2000) Inference of population structure using multilocus genotype data. Genetics Society of America, 155, 945-959. Rosenberg, N.A., Mahajan, S., Ramachandran, S., Zhao, C., Pritchard, J.K. & Feldman, M.W. (2005) Clines, clusters, and the effect of study design on the inference of Human population structure. Plos Genetics, 1(6), e70. Schwartz, M.K. & McKelvey, K.S. (2009) Why sampling scheme matters: The effect of sampling scheme on landscape genetic results. Conservation Genetics, 10, 441-452. Udupa, S.M. & Baum, M. (2001) High mutation rate and mutational bias at (TAA)n microsatellite loci in chickpea (Cicer arietinum L.). Molecular Genetics and Genomics, 265, 1097-1103. Vigouroux, Y., Jaqueth, S.J., Matsuoka, Y., Smith, O.S., Beavis, W.D., Smith, S.C. & Doebley, J. (2002) Rate and pattern f mutation at microsatellite loci in Maize. Molecular Biology and Evolution, 19, 1251-1260. Meirmans, P. (2015) Seven common mistakes in population genetics and how to avoid them. Molecular Ecology, 24, 3223-3231. Pritchard, J. K., Wn, X. & Falush, D. (2010) Documentation for structure software: Version 2.3. University of Chicago, Chicago, IL. http://pritch.bsd.uchicago.edu/structure_software/release_ versions/v2.3.4/structure_doc.pdf Capítulo 2 Infiriendo barreras potenciales para el flujo genético de poblaciones tropicales de Avicennia germinans AQBOT_2019_112 78 Inferring potential barriers to gene flow in tropical populations of Avicennia 1 germinans 2 3 Ochoa-Zavala, M.a, Osorio-Olvera, L.b, c, Piñero, D.a, and Núñez-Farfán, J.a 4 5 a Departamento de Ecología Evolutiva, Instituto de Ecología, Universidad Nacional 6 Autónoma de México. Circuito Exterior S/N anexo Jardín Botánico, Ciudad Universitaria, 7 C.P. 04510 Ciudad de México, México 8 9 b Biodiversity Institute, University of Kansas, Lawrence, KS 66045 10 11 c Centro del Cambio Global y la Sustentabilidad en el Sureste A.C., C.P. 86080 12 Villahermosa, Tabasco, México 13 14 Correspondence 15 Núñez-Farfán J. farfan@unam.mx 16 17 18 19 20 AQBOT_2019_112 79 Abstract 21 Genetic variation within- and divergence among-populations is essential for conservation 22 and management efforts. In plants that are dispersed mainly by ocean currents, physical 23 features of the landscape may influence rates of gene flow among populations by 24 preventing or facilitating dispersal of buoyant propagules. The complex landscape and the 25 pattern of superficial currents acting along the wide latitudinal range of the coasts of 26 Mexico offer an ideal opportunity to evaluate the relevance of oceanographic factors to the 27 genetic diversity of a mangrove species. Here, we infer the role of some oceanic features as 28 potential barriers to gene flow among populations of Avicennia germinans. We used eight 29 nuclear microsatellite markers to estimate recent migration rates and geographic barriers 30 among populations of A. germinans along Mexico’s Atlantic and Pacific coasts. Along the 31 Pacific, we identified potential barriers to gene flow related to the circulation pattern at the 32 mouth of the Gulf of California and Gulf of Tehuantepec in the Pacific. On the Atlantic, 33 population genetic variation coincides with the trends of major ocean currents around the 34 Yucatan Peninsula. Our results suggest that ocean currents and a physical barrier (Baja 35 California Peninsula) have maintained and generated genetic discontinuities along the 36 coasts of Mexico. We find compelling evidence that suggests that northern and southern 37 populations evolve as independent units. Genetic drift has had an important role in 38 determining distribution of genetic diversity in local populations of A. germinans. We 39 strongly recommend the conservation of northern and southern mangrove communities of 40 Mexico; we also suggest use individuals from adjacent populations as genetic source for 41 restoration programs. 42 43 Keywords 44 45 Avicennia germinans, Gene diversity, genetic differentiation, genetic drift, Gulf of 46 California, Gulf of Tehuantepec, mangroves, Yucatan Peninsula 47 48 AQBOT_2019_112 80 1. Introduction 49 50 Information of genetic variation within- and divergence among-populations of species is 51 essential for conservation and management efforts (Frankham, 2012). A number of 52 contemporary and historical factors, acting at different temporal and spatial scales, play 53 important roles in determining the amount and distribution of genetic diversity. For tropical 54 coastal plant species, like mangroves that have dispersal of propagules by sea currents, 55 studies have proposed several different scenarios to explain the structuring of genetic 56 diversity (Takayama et al. 2006). Since their origin, diversification and arrival of 57 Rhizophora and Avicennia genera to the Americas, mangrove forests have been subject to 58 climate events and sea level changes that have impacted their genome-wide nucleotide 59 diversity (Xu, et al., 2017; Guo et al., 2018). In the Americas, two major historical events 60 are known to have reduced dispersal and generated spatial structuring of genetic diversity 61 in different mangrove species: the rise of the Central American Isthmus (CAI) and the 62 Pleistocene glaciations (Núñez-Farfán et al., 2002; Nettel and Dodd, 2007; Pil et al., 2011; 63 Sandoval-Castro et al., 2014; Cerón-Souza et al., 2015; Ochoa-Zavala et al., 2019a). The 64 closure of the CAI fragmented the mangrove communities into two evolutionary units, the 65 Pacific and Atlantic gene pools (Cerón-Souza et al., 2015), while differences in glacial 66 conditions produce different phylogeographic patterns between gene pools of the same 67 species (Ochoa-Zavala et al., 2019a). 68 69 Additionality, in plants that are dispersed mainly by ocean currents, landscape 70 physical features may influence rates of gene flow among populations, by preventing or 71 facilitating the dispersal of buoyant propagules, and contributing to the distribution of 72 genetic variation in mangroves species (Nettel and Dodd, 2007; Takayama et al., 2013). 73 The complex landscape and the pattern of superficial sea currents along the coasts of 74 Mexico offer an ideal opportunity to evaluate the relevance of oceanographic factors as 75 barriers to gene flow of mangrove species. In Mexico, on the Pacific side, the peninsula of 76 Baja California separates the open Pacific Ocean from the Gulf of California, which is a 77 semi-enclosed marginal sea with seasonal variation in ocean circulation (Castro et al., 78 2000). Surface water is exchanged between the cold, southward-moving California Current 79 AQBOT_2019_112 81 (which moves water into the Pacific Ocean) and the warm Gulf of California current via 80 cyclonic circulation at the mouth of the Gulf of California (Castro et al., 2000; Fernández-81 Eguiarte et al., 2018a). In the Gulf of Tehuantepec, oceanic eddies are generated by strong 82 winds that are associated with the movement of cold fronts across the Gulf of Mexico 83 (Müller-Karger and Fuentes-Yaco, 2000). On the Atlantic side of Mexico, current system is 84 dominated by the Loop Current, which is a clockwise flow that extends northward into the 85 Gulf of Mexico and merges with the Yucatan Current, which flows from the south of 86 Cozumel Island and across the western side of the Yucatan Channel into the Gulf of 87 Mexico (Athié et al., 2011; Fernandez-Eguiarte et al., 2018a). 88 89 Among the six mangroves species that occur in the coasts of Mexico, Rhizophora 90 mangle and Avicennia germinans are the most abundant and widely distributed (Núñez-91 Farfán et al., 1996). However, the two species exhibit ecological differences. First, R. 92 mangle behaves like the typical pioneer mangrove species in the community, colonizing the 93 open front of coastal lagoons and rivers. In contrast, A. germinans occupies semi-inundated 94 areas, deeper inland. Second, related to gene flow, R. mangle has larger buoyant propagules 95 with longer period of longevity than A. germinans (Cerón-Souza et al., 2012). Finally, A. 96 germinans is pollinated by insects (e.g., bees) while R. mangle mainly by wind (Tomlinson, 97 2016). Thus, ecological differences exhibited by mangrove species can contribute to 98 genetic connectivity among populations. 99 100 The influence of land barriers and ocean currents on population subdivision have 101 been previously evaluated for R. mangle along Baja California Peninsula’s coasts 102 (Sandoval-Castro et al., 2012). In the coasts of Panamá contrasting patterns of genetic 103 diversity and structure between R. mangle and A. germinans were detected, indicating that 104 although life-history traits are important to predict genetic structure, ocean currents are 105 important to explain the observed patterns (Cerón-Souza et al., 2012). In this study, we 106 present a detailed population genetics analysis of the black mangrove (Avicennia germinans 107 (L.) L.) in both the Pacific and Atlantic (Gulf of Mexico and Caribbean Sea) coasts of 108 Mexico, to infer the role of some oceanic features as potential barriers to gene flow. Given 109 the urgency for mangroves conservation, we translate the previous genetic diversity 110 AQBOT_2019_112 82 findings in a way that can be easily understood by non-expert persons but who make 111 decisions with respect of mangrove conservation and/or restoration programs in Mexico. 112 113 2. Methods 114 115 2.1 Study system 116 Mangroves are a group of salt-tolerant trees that grow along the tropical and subtropical 117 intertidal zones (Tomlinson, 2016), which are subjected to environmental changes on a 118 daily (tides) and yearly (temperature and precipitation) basis. Avicennia germinans exhibits 119 a mixed mating system, with moderate levels of self-fertilization and evidence of biparental 120 inbreeding (Nettel-Hernanz et al., 2013; Mori et al., 2015). It produces cryptoviviparous 121 propagules that can float and survive in saltwater for up to 110 days (see references in 122 Nettel and Dodd, 2007), allowing long-distance dispersal (LDD). Propagules of A. 123 germinans are abscised from the parent tree between September and February, however, 124 more variation in abscission time was observed in the field. The species occurs in western 125 Africa, from Mauritania to Angola, and in the Americas, from Bermuda through 126 southeastern Brazil on the Atlantic coast, and from Baja California, Mexico through Peru, 127 on the Pacific coast. Mexican mangrove populations are patchily distributed, following 128 north-south clines reaching one of the northernmost latitudinal ranges for a mangrove 129 species in the Pacific coast. In Mexico, more than half of the mangrove forest cover is 130 found in the Yucatan Peninsula´s coasts; while only 25% of the total mangrove cover is 131 found on the Pacific coast, concentrated within a few localities (Valderrama-Landeros et 132 al., 2017) where in Chiapas are well-developed mangrove communities with trees up to 30 133 m tall (Núñez-Farfán et al., 1996). 134 135 2.2 Genotypic dataset 136 Microsatellites are short sequences of DNA abundantly distributed across species’ 137 genomes. They contain short tandem repeats of two to six nucleotides (e.g. TATATAn). 138 They are known to be hypervariable due to a rapid accumulation of length polymorphisms 139 by intra-allelic polymerase slippage during replication. This makes them very informative 140 and well suited for analyzing interpopulation variation and patterns of gene flow. 141 AQBOT_2019_112 83 142 We surveyed a dataset consisting of 1144 individuals from 29 populations from 143 both coasts of Mexico (Table 1; Fig. 1). The 10 nuclear microsatellite loci genotyped, 144 consisted of two, four and six tandem repeat motifs (Ochoa-Zavala et al., 2019b). 145 Microsatellites are codominant makers (heterozygotes can be distinguished from 146 homozygotes) that are generally considered to be adaptively neutral but sometimes could be 147 associated to genomic regions which are under natural selection. They are prone to 148 homoplasy as well as the occurrence of null alleles (Chistiakov et al., 2006). In fact, two 149 loci (Agerm_CT_003 and Agerm_GA_003) were excluded from analyses because of high 150 frequencies of null alleles (Ochoa-Zavala et al. 2019a). 151 152 2.3 Gene diversity and structure analyses 153 We estimated the geographic projection of the expected heterozygosity (HE) using the 154 function sHe from the R package ‘biotools’ (da Silva et al., 2017). This package allows to 155 predict spatial variation of the expected heterozygosity over a geographic grid using a table 156 of the geographic coordinates of genotype data. For this analysis, we projected genotypes 157 of1144 individuals from 29 populations and projected the results onto a spatial grid based 158 on the distribution of mangroves in Mexico at a resolution of 0.005 degrees (~500 m2, see 159 the GeneDiversity_Ag.R script in Appendix A). 160 161 In addition, we computed, for each coast independently, the effective number of 162 alleles (Ne), the information index (I), gene diversity (Hs), percentage of polymorphic loci 163 (P), and number of common alleles in £ 25 % and £ 50 %. For each locus, we calculated 164 the number of alleles (Na), Ne, I, the observed (HO) and unbiased expected heterozygosities 165 (uHE), and allele frequencies at individual loci across populations. All statistics were 166 computed using GENALEX version 6.503 (Peakall and Smouse, 2006, 2012), except for 167 population level Hs, which were obtained with FSTAT version 2.9.3.2 (Goudet, 2002). 168 169 All genetic structure analyses were computed for each coast independently. Genetic 170 differentiation between populations was evaluated by computing RST values (Slatkin, 1995) 171 following the Stepwise Mutation Model (SMM; Kimura & Ohta, 1978). Because allele 172 AQBOT_2019_112 84 distributions at each locus were patchy (Fig. B.1), we also considered an infinite allele 173 model (IAM; Kimura & Crow, 1964) and estimated FST (Wright 1965) between all pairs of 174 populations using ARLEQUIN (Excoffier and Lischer, 2010). Parameters included 10,000 175 permutations and a fixed significance level of 0.05. We visualized the corresponding 176 pairwise matrix following Lischer (2017). Genetic relationships among populations were 177 evaluated through a neighbor-joining tree (NJ-tree) on Nei’s genetic distance (DA, Nei et 178 al., 1983) using POPTREE2 (Takezaki et al., 2010). The reliability of each node was 179 estimated by the bootstrap with 100,000 replicates. 180 181 The apportionment of genetic variation within and among population groups was 182 tested through an analysis of molecular variance (AMOVA) using ARLEQUIN, with 10,000 183 permutations, both the SMM and IAM models were considered. This analysis was 184 performed based on population groups identified by Ochoa-Zavala et al., (2019a) for the 185 Pacific [Northern Gulf of California (PN04, PN02, PN06), North Pacific (PN03, PN05, 186 PN11, PN12, PN10), Central Pacific (PC15, PC11), Tropical Pacific (PC19, PS17, PS20) 187 and Tropical South Pacific (PS29, PS24)], and Atlantic [Northern Gulf of Mexico (GM37, 188 GM41), Gulf of Mexico (GM56, GM40, GM53, GM54, GM36, GM46, PY67, PY66) and 189 Yucatan Peninsula (PY70, PY81, PY65, PY77)], coasts of Mexico (Fig. B.2). We further 190 performed a principal coordinate analysis (PCoA) on a matrix of DA genetic distances 191 between population pairs using GENALEX. 192 193 2.4 Geographic barriers and gene flow rates 194 Potential genetic barriers associated with geography were visualized with the Monmonier’s 195 maximum-difference algorithm implemented in BARRIER version 2.2 (Manni et al., 2004) 196 for each coast. To obtain barrier statistical confidence, 100 replicates of DA genetic distance 197 matrices were calculated by resampling individuals within populations using the 198 Microsatellite Analyzer software (Dieringer and Schlötterer, 2003). We retained only those 199 barriers supported by bootstrap scores ³ 80 %. 200 201 We estimated rates of recent migration among populations using BAYESASS version 202 3.0.4 (Wilson and Rannala, 2003). This program implements a Bayesian approach using 203 AQBOT_2019_112 85 Markov chain Monte Carlo (MCMC) techniques to estimate migration rates over the last 204 two generations. We analyzed the Pacific and Atlantic datasets independently using runs of 205 3´108 iterations, after discarding 3´107 as burn-in and using a thinning interval of 15,000 206 iterations. The mixing parameters for estimating migration rates, inbreeding coefficient and 207 allele frequencies were m = 0.4, f = 1.0 and a = 0.8, respectively. We carried out five 208 different MCMC runs (differing in starting seeds) for each dataset and calculated Bayesian 209 deviance to measure the model fit for each one of the runs. We used the 210 calculateDeviance.R script from Meirmans (2014) and selected the run with the lowest 211 deviance to obtain estimates of migration rate for each dataset (Meirmans, 2014). 212 Convergence of parameters was assessed by examination of trace files with the same R-213 script. 214 215 3. Results 216 217 3.1 Gene diversity and genetic structure 218 The pattern of spatial distribution of HE shows a clear geographic trend along the Pacific 219 coast populations, lower in the north and higher in southern populations (Fig. 1). This trend 220 was also observed for the other genetic parameters computed: the tropical south Pacific 221 populations (PS29, PS24) had moderate values of effective number of alleles (Ne = 2.08, 222 2.11, respectively), gene diversity (Hs = 0.407, 0.374) and larger information index values 223 (I = 0.8, 0.7). In contrast, the northern Gulf of California populations (PN04, PN02, PN06) 224 had low values of effective number of alleles (Ne from 1.02 to 1.13) and information index 225 (ranged from 0.03 to 0.22) (Table 2). 226 227 For the Atlantic populations we did not observe a geographic trend in gene diversity 228 (Fig. 1). Most populations in this coast had moderate levels of gene diversity (e. g. Hs = 229 0.186 for GM41; Hs= 0.273 for GM53 and Hs = 0.379 for PY77), effective number of 230 alleles and information index (Table 3). 231 232 Loci showed low to moderate values of gene diversity, ranging from 0.006 to 0.351 233 for HO, and from 0.015 to 0.416 for uHE across the Pacific populations. For the Atlantic 234 AQBOT_2019_112 86 populations, values of these parameter were larger than in the Pacific: HO ranged from 235 0.002 to 0.509 and uHE from 0.009 to 0.505 (Table B.1, B.2; see also Fig. 2). 236 237 The allele frequency distribution showed strong differences between the Pacific and 238 Atlantic evolutionary units (Fig. 2). From the 54.32 % of the alleles that were shared 239 between coasts, some alleles showed divergent frequencies. Some alleles had low 240 frequency in the Atlantic and high frequency in the Pacific (e.g. locus Agerm-25, AgT9, 241 AgD6; Fig. 2), and some had high frequency in the Atlantic and low frequency in the 242 Pacific (e.g. locus Agerm-25, Agerm-18, AgD6; Fig. 2). Also, for many loci, some allele 243 frequencies changed drastically among populations, with one predominant allele in each 244 population, in some cases fixed or nearly fixed, especially in the northernmost populations 245 of both coasts (Fig. B.3). 246 247 Pairwise population FST and RST indicated strong genetic differentiation (FST = 0.046 - 248 0.69; RST = 0.039 - 0.926; P < 0.05) with substantial genetic divergence between the 249 northern Gulf of California and tropical south Pacific populations (Fig. 3a). Along the 250 Atlantic populations most FST and RST values were significantly different from zero (FST = 251 0.014 - 0.458; RST = 0.016 - 0.381; P < 0.05) and showed moderate values of gene 252 differentiation between the northern Gulf of Mexico and Yucatan Peninsula populations 253 (Fig. 3a). There was moderate divergence between PY70 and PY65 (FST = 0.138; RST = 254 0.229; P < 0.05) within the last region. The NJ-tree of genetic distances between 255 populations supports the strong divergence between north and south populations of black 256 mangrove in Mexico (Fig. 3b). Furthermore, the NJ-tree showed that populations from the 257 Yucatan Peninsula are divided into two subgroups, East and West (Fig. 3b). 258 259 On the Pacific coast, AMOVA revealed that 49.48 and 27.20 % (P < 0.001) of the 260 variance is contained in differences among populations groups, based on RST and FST, 261 respectively (Table B.3). Although the distribution of genetic variation for the two mutation 262 models differs slightly among groups, the IAM model detected moderate differentiation 263 among population groups. On the Atlantic coast, AMOVA results showed that most of the 264 genetic variation was within individuals (75 % for both models; P < 0.001; Table B.3) and 265 AQBOT_2019_112 87 that only 6.39 and 13.87 % (for both RST and FST, respectively; P < 0.01) of the variance 266 was contained in the differences among population groups (Table B.3). PCoA results 267 showed a cline of genetic variation indicating isolation by distance (IBD) along both coasts 268 (Fig. 3c). PCoA 1 axis explained a large fraction of genetic variation (73.69 and 67.68 % 269 for the Pacific and Atlantic, respectively) distinguishing north and south groups in both 270 coasts; PCoA 2 separated the intermediated groups explaining 17.06 % and 13.66 % of the 271 variation for the Pacific and Atlantic populations, respectively (Fig. 3c). 272 273 3.2 Geographic barriers and gene flow 274 Barrier analyses indicated two potential barriers to gene flow for both the Pacific and 275 Atlantic coasts of Mexico (Fig. 4). In the Pacific, the first barrier was found at the mouth of 276 the Gulf of California region, between populations PN05 and PN11, extending along the 277 Baja California Peninsula and isolating PN03 from PN02. The second barrier was located at 278 the Gulf of Tehuantepec, between populations PS20 and PS29 (Fig. 4). On the Atlantic 279 coast, a potential boundary between PY66 and PY70 separated the Yucatan Peninsula from 280 the Gulf of Mexico populations and a second barrier, between PY81 and PY65, divides the 281 Yucatan Peninsula into West and East region (Fig. 4). 282 283 Migration rates ranged from 0.0060 to 0.2375 on the Pacific coast and from 0.0061 284 to 0.2216 on the Atlantic coast. Migration results suggest that gene flow is more likely to 285 occur between geographically adjacent rather than between distant populations on each 286 coast (Table B.4, B.5). The highest migration rates on the Pacific coast were from PN04 to 287 PN02 and from PN04 to PN06. In the Gulf of California region, the migration pattern was 288 predominantly southwards; PC11 is apparently a source for populations further south along 289 the Pacific coast (Table B.4). On the Atlantic coast, gene flow seems to be more dynamic, 290 especially within the Gulf of Mexico population group (Table B.5). The highest migration 291 rates were also between neighbor stands (from GM36 to GM54 and from GM53 to GM40) 292 and decreased from PY65 to PY81 (Table B.5). On both coasts, no gene flow was detected 293 between northern and southern populations (e. g. among PN04 and PS29 or GM37 and 294 PY70 or vice versa). 295 296 AQBOT_2019_112 88 4. Discussion 297 298 4.1 Ocean currents prevent gene flow among populations of Avicennia germinans 299 Mexican black mangrove populations are separated by land barriers and interconnected by 300 ocean currents. Previous studies have suggested that marine currents play an important role 301 in genetically structuring mangroves species (Pil et al., 2011; Wee et al 2014; Yan et al., 302 2016) and marine fishes. They also participate in the dispersion of eggs and larvae that are 303 influenced by the same ocean currents that affect mangrove propagules (Ochoa-Zavala et 304 al., 2019a), preventing and/or diminishing gene flow and maintaining genetic divergence. 305 306 On the Pacific coast, we found evidence of two potential barriers to gene flow. The 307 first coincides with the Baja California Peninsula and was potentially associated with 308 surface water circulation at the mouth of the Gulf of California. Similar breaks in gene flow 309 has been documented for another mangrove species (R. mangle) in the same region 310 (Sandoval-Castro et al., 2012) and for marine taxa with larval dispersal (Saarman et al., 311 2010; Saavedra-Sotelo et al., 2013; García-de León et al., 2018). These geographic barriers 312 are consistent with values of genetic differentiation (FST and RST) and the general pattern of 313 gene flow detected in this study: higher migration rates among populations within the Gulf 314 of California (PN04, PN02, PN06) – in which was predominantly southward (probably 315 following Gulf of California current) – but diminished substantially between populations on 316 both sides of the barrier (e.g. From PN03 to PN02 and from PN05 to PN02). 317 318 The second genetic discontinuity in the Pacific coast was located at the Gulf of 319 Tehuantepec. In this geographic area, several oceanic gyres and fronts develop and 320 changing periodically their direction and strength (Müller-Karger and Fuentes-Yaco, 2000). 321 This may prevent gene flow between the tropical south Pacific and the rest of the Pacific 322 coast populations by transporting propagules offshore, into the open ocean, as has been 323 observed in invertebrate larvae (López-Chávez et al., 2016; Sandoval-Huerta et al., 2019). 324 Though other ecological factors as habitat discontinuities may also be important in 325 generating or maintaining genetic discontinuities among populations of marine taxa, recent 326 reports highlight that larval migration can also be prevented by ocean currents in the Gulf 327 AQBOT_2019_112 89 of Tehuantepec, contributing to the isolation of populations (Prieto-Ríos et al., 2014; 328 López-Chávez et al., 2016; Sandoval-Huerta et al., 2019). 329 330 On the Atlantic coast, we found evidence of two genetic barriers, one that isolates 331 the Yucatan’s Peninsula from the Gulf of Mexico and another that splits the Yucatan’s 332 Peninsula into Western and Eastern regions. Wind and ocean currents may explain this 333 pattern. During June and July, ocean currents flow westward from the Yucatan Peninsula to 334 the northern part of the Gulf of Mexico, however, from August to December currents 335 change in direction, returning close to the Yucatan Peninsula (Fernández-Eguiarte et al., 336 2018a). From October to January the surface wind runs predominantly southwest 337 (Fernández-Eguiarte et al., 2018b) restricting the propagules to the Yucatan Peninsula. In 338 addition, the Yucatan current flows from the southern Cozumel Island and across the 339 western side of the Yucatan Channel (Athié et al., 2011; Fernández-Eguiarte et al., 2018a). 340 Hence, the direction of wind and ocean current patterns may constrain propagule dispersal, 341 resulting in a genetic break that isolates the Yucatan Peninsula gene pool and producing a 342 second barrier that divides the Yucatan Peninsula into West and East region. Recently, a 343 similar break was reported for R. mangle in the same region, suggesting that the Yucatan’s 344 current prevents continuous migration of propagules (Cisneros-de la Cruz et al., 2018). 345 346 In the Pacific coast, while most migration rates are consistent with our genetic 347 analyses, an unexpectedly large migration rate of 0.1387 was estimated between two 348 populations (PN04 and PN03) separated by the Baja California Peninsula. Deviations from 349 the inference model assumed by BAYESASS (e.g. the assumption that genetic drift and 350 migration during the last few generations do not change subpopulation allele frequencies) 351 may have produced errors in the estimation of migration rates, as was stated by Meirmans 352 (2014 and references therein). Thus, results from BAYESASS should be interpreted with 353 caution. 354 355 Although A. germinans is capable of long-distance dispersal (Nettel and Dodd, 356 2007), our results suggest limited gene flow among populations caused by restrictions on 357 pollen and propagule dispersal. We found deep divergence between north and south 358 AQBOT_2019_112 90 populations supported by pairwise FST and RST, NJ-trees, patterns of gene flow, and IBD 359 inferred from PCoA analyses in both coasts. In addition to ocean currents and physical 360 barriers, more recent studies suggest that genetic drift and habitat discontinuity need to be 361 considered to explain spatial genetic variation in mangrove species (Hodel et al., 2018; 362 Binks et al., 2019). However, in the populations we studied, habitat discontinuity does not 363 appear to explain genetic subdivision, because the population groups do not correspond 364 geographically with habitat fragments (Fig. B.2). Instead, we hypothesize that ocean 365 currents and land barriers amplifying the effects of founder events after glaciations (Pil et 366 al., 2011), promoting differentiation by genetic drift, which result in a strong signal of 367 divergence between north and south populations, and likely evolving as independent 368 lineages. 369 370 4.2 Patterns of allele frequencies and conservation implications 371 The Pacific and Atlantic populations of A. germinans exhibit a range of gene diversity, 372 from low to moderate values (from 0.018 to 0.407) that are in agreement with previous 373 estimates (HO = 0.11; HE = 0.17, Pil et al., 2011; HO = 0.16; HE = 0.17, Sandoval-Castro et 374 al., 2012; HO = 0.27; HE = 0.37, Cisneros-de la Cruz et al., 2018). The distributions of allele 375 frequencies of A. germinans reported here are consistent with previous evidence supporting 376 the effect of the final closure of the CAI on the Pacific and Atlantic gene pools (Nettel and 377 Dodd, 2007; Takayama et al., 2013; Cerón-Souza et al., 2015; Ochoa-Zavala et al., 2019), 378 but they also reflect the important role of genetic drift in Mexican populations. Since their 379 appearance along the shores of the Tethys Sea during the Late Cretaceous (Ricklefs et al., 380 2006), mangroves have endure many climatic events, changes in population size according 381 to sea level and temperature shifts (Nettel and Dodd, 2007; Cerón-Souza et al, 2015; Xu et 382 al., 2017; Guo et al., 2018). Thus, the presence of fixed or nearly fixed alleles in many loci 383 across populations (Fig B.3), together with mean low gene diversity, are likely caused by 384 the cumulative effect of genetic drift due to the continual fluctuation in effective population 385 size and the vicariance effect on both gene pools. This supports the hypothesis that low 386 levels of genetic diversity is a common feature of mangroves (Yan et al., 2016), as shown 387 by several studies (Núñez-Farfán et al., 2002; Pil et al., 2011; Sandoval-Castro et al., 2012; 388 AQBOT_2019_112 91 Yan et al., 2016), even at the whole-genome level (Guo et al., 2016; Xu et al., 2017; Guo et 389 al., 2018). 390 391 Although there have been significant advances have occurred in phylogeographic 392 and phylogenetic studies of mangroves, little information has been translated into 393 conservation programs (Wee et al., 2019). We focus in the pattern of spatial distribution of 394 HE (Fig. 1) and complement diversity estimates as a tool to give more conservative 395 recommendation for restoration programs and conservation strategies of mangrove 396 ecosystems in Mexico. 397 398 Mexican black mangrove populations bearing high genetic diversity and 399 information index (e.g. PS24, PS29 and Yucatan Peninsula populations; Fig. 1) occur in 400 well-preserved mangrove communities that account for an important fraction of the total 401 forest cover. In some localities, of the Pacific coast (PS24, PS29), mangrove trees can reach 402 up to 30 m high, which is similar to those observed in well conserved and developed 403 mangrove ecosystems in Costa Rica (Núñez-Farfán et al., 1996; Valderrama-Landeros et 404 al., 2017). On the other hand, the northern mangrove communities (e. g. PN04, PN02, 405 PN03, PN05, PN06), are subject to strong selective pressures from environmental 406 instability at the species range limit. They are shorter in height (approximately 4 m high) 407 and have a more scatter geographic distributed; these and other characteristics of these 408 populations are hypothetically attributable, at least in part, to a unique set of genes that 409 have been shaped by this environmental instability. Those genes may confer advantages 410 when mangroves inevitably face the effects of global climate change, making the genetic 411 resources contained in the northern populations worth preserving. Finally, the degree of 412 divergence and migration patterns found in this study constitute a baseline to guide 413 restoration programs. Accordingly, we recommend the use of individual trees from a 414 neighboring population to avoid modifying the natural genetic structure in the recipient 415 population, thus preventing the disruption of potential local adaptation. 416 417 In conclusion, even though propagules of A. germinans are dispersed by ocean 418 currents, leading to the expectation of high genetic connectivity between populations, our 419 AQBOT_2019_112 92 results suggest that ocean currents and geographical barriers are more important for 420 maintaining and generating genetic discontinuities along the coasts of Mexico. However, 421 other factors such as genetic drift have also played an important role in determining the 422 distribution of genetic variation in black mangrove populations in Mexico. On average, the 423 Pacific and Atlantic populations exhibit low gene diversity, but there is a high degree of 424 heterogeneity in gene diversity within each coast. We strongly recommend increasing 425 conservation efforts in northern and southern mangrove communities. and the use of 426 individual plants from adjacent populations for restoration purposes. 427 428 Acknowledgements 429 The authors thanks to Tapia-López, R. and all members of the Laboratorio de Genética 430 Ecológica y Evolución for logistical and field support. M.O.-Z acknowledges the 431 scholarship granted for graduate studies by the National Council of Science and 432 Technology (CONACyT) and thanks the Graduate Program in Biological Sciences, 433 National Autonomous University of Mexico (UNAM). L. O.-O acknowledges PAPIIT-434 UNAM IN116018 and CONACyT postdoctoral fellowship program support for his 435 extended academic training and CONACyT-FORDECyT Project number 273646 for partial 436 funding. 437 438 Funding: This work was supported by the National Commission for the Knowledge and 439 Use of Biodiversity (CONABIO), grant KE008. 440 441 Declarations of conflicts of interest: none 442 443 Author contributions: Maried Ochoa-Zavala: Methodology, investigation, visualization, 444 writing - original draft preparation; Luis Osorio-Olvera: Methodology, writing - review and 445 editing; Daniel Piñero: funding acquisition, writing - review and editing; Juan Núñez-446 Farfán: funding acquisition, supervision, writing - review and editing. 447 448 AQBOT_2019_112 93 References 449 1. Athié, G., Candela, J., Sheinbaum, J., Badan, A. and Ochoa, J. 2011. Yucatan Current 450 variability through the Cozumel and Yucatan channels. Ciencias Marinas, 37(4A), 471-451 492. https://doi.org/10.7773/cm.v37i4A.1794 452 2. Binks, R.M., Byrne, M., McMahon K., Pitt G., Murray K. and Evans, R. D. 2019. 453 Habitat discontinuities form strong barriers to gene flow among mangrove populations, 454 despite the capacity for long-distance dispersal. Diversity and Distributions, 25, 298-455 309. https://doi.org/10.1111/ddi.12851 456 3. Castro, R., Mascarenhas, A. S., Durazo-Arvizu, R., Collins, C. A., 2000. Variación 457 estacional de la temperatura y salinidad en la entrada del Golfo de California, México. 458 Ciencias del Mar, 26, 561–563. http://dx.doi.org/10.7773/cm.v26i4.621 459 4. Cerón-Souza, I., Bermingham, E., Owen, M.W. and Jones, F.A. 2012. Comparative 460 genetic structure of two mangrove species in Caribbean and Pacific estuaries of 461 Panama. BMC Evolutionary Biology, 12, 205. http://www.biomedcentral.com/1471-462 2148/12/205 463 5. Cerón-Souza, I., Gonzalez, E. G., Schwarzbach, A. E., Salas-Leiva, D.E., Rivera-464 Ocasio, E., Toro-Perea, N., Bermingham, E. and McMillan, W.O. 2015. Contrasting 465 demographic history and gene flow patterns of two mangrove species on either side of 466 the Central American Isthmus. Ecology and Evolution, 5, 3486-3499. 467 https://doi.org/10.1002/ece3.1569 468 6. Chistiakov, D. A., Hellemans, B. and Volckaert, F. A. M. 2006. Microsatellites and 469 their genomic distribution, evolution, function and applications: A review with special 470 reference to fish genetics. Aquaculture, 255, 1-29. 471 doi:10.1016/j.aquaculture.2005.11.031 472 7. Cisneros-de la Cruz, D. J., Martínez-Castillo, J., Herrera-Silveira, J., Yáñez-Espinosa, 473 L., Ortiz-García, M., Us-Santamaria, R. and Andrade, J. L. 2018. Short-distance 474 barriers affect genetic variability of Rhizophora mangle L. in the Yucatan Peninsula. 475 Ecology and Evolution, 8, 11083-11099. https://doi.org/10.1002/ece3.4575 476 AQBOT_2019_112 94 8. da Silva, A. R., Malafaia, G. and Menezes, I. P. P. 2017. Biotools: an R function to 477 predict spatial gene diversity via an individual-based approach. Genetics and Molecular 478 Research 16, gmr16029655. doi: 10.4238/gmr16029655 479 9. Dieringer, D. and Schlötterer, C. 2003. Microsatellite analyser (MSA): a platform 480 independent analysis tool for large microsatellite datasets. Molecular Ecology Notes 3, 481 167-169. https://doi.org/10.1046/j.1471-8286.2003.00351.x 482 10. Excoffier, L. and Lischer, H. E. L. 2010. Arlequin suite ver 3.5: a new series of 483 programs to perform population genetics analyses under Linux and Windows. 484 Molecular Ecology Resources, 10, 564-567. https://doi.org/10.1111/j.1755-485 0998.2010.02847.x 486 11. Fernández-Eguiarte, A., Zavala-Hidalgo, J. and Romero-Centeno, R. 2018a. 487 Climatología mensual de velocidad geostrófica absoluta (1992-2012). Atlas Climático 488 Digital de México. Centro de Ciencias de la Atmósfera. Universidad Nacional 489 Autónoma de México. http://uniatmos.atmosfera.unam.mx/ 490 12. Fernández-Eguiarte, A., Zavala-Hidalgo, J. and Romero-Centeno, R. 2018b. Viento 491 climatológico superficial mensual en el mar (1999-2006). Atlas Climático Digital de 492 México. Centro de Ciencias de la Atmósfera. Universidad Nacional Autónoma de 493 México. http://uniatmos.atmosfera.unam.mx/ 494 13. Frankham, R. 2012. How closely does genetic diversity in finite populations conform to 495 predictions of neutral theory? Large deficits in regions of low recombination. Heredity, 496 108, 167-178. https://doi.org/10.1038/hdy.2011.66 497 14. García -De León, F., Galván-Tirado, C., Sánchez Velasco, L., Silva-Segundo, C., 498 Hernández-Guzmán R., Barriga-Sosa, I… Cruz-Hernández, P. 2018. Role of 499 oceanography in shaping the genetic structure on the north Pacific hake Merluccius 500 productus. Plos One, 13, e0194646. https://doi.org/10.1371/journal.pone.0194646 501 15. Goudet, J. 2002. FSTAT software (version 2.9.3.2): A program to estimate and test 502 gene diversities and fixation indices. Retrieved from 503 http://www.unil.ch/popgen/softwares/fstat.htm 504 AQBOT_2019_112 95 16. Guo, Z., Huang, Y., Chen, Y., Duke, N. C., Zhong, C. and Shi, S. 2016. Genetic 505 discontinuities in a dominant mangrove Rhizophora apiculata (Rhizophoraceae) in the 506 Indo-Malesian region. Journal of Biogeography, 43, 1856-1868. 507 https://doi.org/10.1111/jbi.12770 508 17. Guo, Z., Li, X., He, Z., Yang, Y., Wang, W., Zhong, C., … Duke, N.C. 2018. 509 Extremely low genetic diversity across mangrove taxa reflects past sea level changes 510 and hints at poor future responses. Global Change Biology, 24, 1741-1748. 511 https://doi.org/10.1111/gcb.13968 512 18. Hodel, R. G. J., Knowles L. L., McDaniel S. F., Payton, A. C., Dunaway, J. F., Soltis, 513 P, S., and Soltis, D, E. 2018. Terrestrial species adapted to sea dispersal: Differences in 514 propagule dispersal of two Caribbean mangroves. Molecular Ecology, 27, 4612-4626. 515 https://doi.org/10.1111/mec.14894 516 19. Kimura, M. and Crow, J. 1964. The number of alleles that can be maintained in a finite 517 population. Genetics, 49, 725-738. https://www.ncbi.nlm.nih.gov/pmc/PMC1210609/ 518 20. Kimura, M. and Ohta, T. 1978. Stepwise mutation model and distribution of allelic 519 frequencies in a finite population. Proceedings of the National Academy of Sciences of 520 the United States of America, 75, 2868-2872. https://doi.org/10.1073/pnas.75.6.2868 521 21. Lischer, H. 2017. FST pairwise table plot. In: Giannico, R. RpairwiseFST.R 522 https://github.com/rgiannico/RpairwiseFST 523 22. López-Chávez, F, J., Chassin-Noria, O., Ríos-Chávez, P., Rocha-Ramírez, V., Macip-524 Ríos R. and Oyama, K. 2016. Phylogeography of the purple snail Plicopurpura pansa 525 along Mexican Pacific coast. Ciencias Marinas, 42, 1-14. 526 https://doi.org/10.7773/cm.v42i1.2576 527 23. Manni, F., Guerard, E. and Heyer, E. 2004. Geographic patterns of (genetic, 528 morphologic, linguistic) variation: how barriers can be detected by using Monmoniers’s 529 algorithm. Human Biology, 76, 173-190. 10.1353/hub.2004.0034 530 24. Meirmans, P, G. 2014. Nonconvergence in Bayesian estimation of migration rates. 531 Molecular Ecology Resources, 14, 726-733. https://doi.org/10.1111/1755-0998.12216. 532 AQBOT_2019_112 96 25. Mori, G.M., Zucchi, M.I. and Souza, A.P. 2015. Multiple-geographic-scale genetic 533 structure of two mangrove tree species: the roles of mating system, hybridization, 534 limited dispersal and extrinsic factors. Plos One, 10, e0118710. 535 DOI:10.1371/journal.pone.0118710 536 26. Müller-Karger, F. E. and Fuentes-Yaco, C. 2000. Characteristics of wind generated 537 rings in the eastern tropical Pacific Ocean. Journal of Geophysical Research, 105, 1271-538 1284. https://doi.org/10.1029/1999JC900257 539 27. Nei, M., Tajima, F. and Tateno, Y. 1983. Accuracy of estimated phylogenetic trees 540 from molecular data. Journal of Molecular Evolution, 19, 153-170. 541 https://doi.org/10.1007/BF02300753 542 28. Nettel-Hernanz, A., Dodd, R.S., Ochoa-Zavala, M., Tovilla-Hernández, C. and Díaz-543 Gallegos, J.R. 2013. Mating system analyses of tropical populations of the black 544 mangrove, Avicennia germinans (L.) L. (Avicenniaceae). Botanical Sciences, 91, 115-545 117. 546 29. Nettel, A. and Dodd, R. S. 2007. Drifting propagules and receding swamps: Genetic 547 footprints of mangrove recolonization and dispersal along tropical coasts. Evolution, 548 61, 958-971. https://doi.org/10.1111/j.1558-5646.2007.00070.x 549 30. Núñez-Farfán, J., Domínguez, C. A., Dirzo, R., Eguiarte, L. E. and Quijano, M. 1996. 550 Estudio ecológico de las poblaciones de Rhizophora mangle en México. Final Report 551 No. B007, National Commission for Biodiversity (CONABIO), Mexico City, Mexico. 552 31. Núñez-Farfán, J., Domínguez, C. A., Eguiarte, L. E., Cornejo, A., Quijano, M., Vargas, 553 J. and Dirzo, R. 2002. Genetic divergence among Mexican populations of red mangrove 554 (Rhizophora mangle): Geographic and historic effects. Evolutionary Ecology Research, 555 4, 1049-1064. 556 32. Ochoa-Zavala, M., Jaramillo-Correa, J. P., Piñero, D., Nettel-Hernanz, A. and Núñez-557 Farfán, J. (2019a). Contrasting colonization patterns of black mangrove (Avicennia 558 germinans (L.) L.) gene pools along the Mexican coasts. Journal of Biogeography, 46, 559 884-898. https://doi.org/10.1111/jbi.13536 560 AQBOT_2019_112 97 33. [dataset] Ochoa-Zavala, M., Jaramillo-Correa, J. P., Piñero, D., Nettel-Hernanz, A. and 561 Núñez-Farfán, J. Data from: Contrasting colonization patterns of black mangrove 562 (Avicennia germinans (L.) L.) gene pools along the Mexican coasts. Dryad Digital 563 Repository. (2019b). doi:10.5061/dryad.4r0q7m5. 564 34. Peakall, R. and Smouse P. E. 2006. GENALEX 6: genetic analysis in Excel. Population 565 genetic software for teaching and research. Molecular Ecology Notes, 6, 288-295. 566 https://doi.org/10.1111/j.1471-8286.2005.01155.x 567 35. Peakall, R. and Smouse, P. E. 2012. GenAlEx 6.5: genetic analysis in Excel. Population 568 genetic software for teaching and research-an update. Bioinformatics, 28, 2537-2539. 569 https://doi.org/10.1111/j.1471-8286.2005.01155.x 570 36. Pil, M. W., Boeger, M. R. T., Muschner, V. C., Pie, M. R., Ostrensky, A., Boeger, W.A. 571 2011. Postglacial north-south expansion of populations of Rhizophora mangle 572 (Rhizophoraceae) along the Brazilian coast revealed by microsatellite analysis. 573 American Journal of Botany, 98, 1031-1039. doi:10.3732/ajb.1000392 574 37. Prieto-Rios, E., Solís-Marín, F. A., Borrero-Pérez, G. H and Díaz-Jaimes, P. 2014. 575 Phylogeography of Holothuria (Halodeima) inornata Semper, 1868 (Echinodermata: 576 Holothuroidea). Revista Peruana de Biología, 21, 155-162. doi: 577 http://dx.doi.org/10.15381/rpb.v21i2.9818 578 38. Ricklefs, R. E., Schwarzbach, A. E., Renner, S. S. 2006. Rate of lineage origin explains 579 the diversity anomaly in the world’s mangrove vegetation. The American Naturalist, 580 168, 805- 810. doi: 10.1086/508711 581 39. Saarman, N. P., Louie, K, D. and Hamilton, H. 2010. Genetic differentiation across 582 eastern Pacific oceanographic barriers in the threatened seahorse Hippocampus ingens. 583 Conservation Genetics, 11,1989-2000. https://doi.org/10.1007/s10592-010-0092-x 584 40. Saavedra-Sotelo, N. C., Calderón-Aguilera, L. E., Reyes-Bonilla, H., Paz-García, D. A., 585 López-Pérez, R. A., Cupul-Magaña, A… Rocha-Olivares, A. 2013. Testing the genetic 586 predictions of biogeographical model in a dominant endemic Eastern Pacific coral 587 (Porites panamensis) using a genetic seascape approach. Ecology and Evolution, 3, 588 4070-4091. doi: 10.1002/ece3.734 589 AQBOT_2019_112 98 41. Sandoval-Castro, E., Dodd, R.S., Riosmena-Rodríguez, R., Enríquez-Paredes, L. M., 590 Tovilla-Hernández, C., López-Vivas, J. M., Aguilar-May. B. and Muñiz-Salazar, R. 591 2014. Post-Glacial expansion and population genetic divergence of mangroves species 592 Avicennia germinans (L.) Stearn and Rhizophora mangle L. along the Mexican coast. 593 Plos One, 9, e93358. https://doi.org/10.1371/journal.pone.0093358 594 42. Sandoval-Castro, E., Muñiz-Salazar, R., Enríquez-Paredes, L. M., Riosmena-595 Rodríguez, R., Dodd, R. S., Tovilla-Hernández, C. and Arredondo-García, M. C. 2012. 596 Genetic population structure of red mangrove (Rhizophora mangle L.) along the 597 northwestern coast of Mexico. Aquatic Botany, 99, 20-26. 598 https://doi.org/10.1016/j.aquabot.2012.01.002 599 43. Sandoval-Huerta, E. R., Beltrán-López, R. G., Pedraza-Marrón, C. R., Paz-Velásquez, 600 M, A., Angulo, A., Robertson, D. R… Domínguez-Domínguez, O. 2019. The 601 evolutionary history of the goby Elacatinus puncticulatus in the tropical eastern Pacific: 602 Effects of habitat discontinuities and local environmental variability. Molecular 603 Phylogenetics and Evolution, 130, 269-285. 604 https://doi.org/10.1016/j.ympev.2018.10.020 605 44. Slatkin, M. 1995. A measure of population subdivision based on microsatellite allele 606 frequencies. Genetics Society of America, 139, 457-462. 607 https://www.ncbi.nlm.nih.gov/pubmed/7705646 608 45. Takayama, K., Kajita, T., Murata, J. and Tateishi, Y. 2006. Phylogeography and genetic 609 structure of Hibiscus tiliaceus–speciation of a pantropical plant with sea-drifted seeds. 610 Molecular Ecology, 15, 2871-2881. doi: 10.1111/j.1365-294X.2006.02963.x 611 46. Takayama, K., Tamura, M., Tateishi, Y., Webb, E.L. and Kajita, T. 2013. Strong 612 genetic structure over the American continents and transoceanic dispersal in the 613 mangrove genus Rhizophora (Rhizophoraceae) revealed by broad-scale nuclear and 614 chloroplast DNA analysis. American Journal of Botany 100, 1191-1201. 615 doi:10.3732/ajb.1200567 616 47. Takezaki, N., Nei, M. and Tamura K. 2010. Poptree2: Software for constructing 617 population trees from allele frequency data and computing other population statistics 618 AQBOT_2019_112 99 with Windows interface. Molecular Biology and Evolution, 27, 747-752. 619 https://doi.org/10.1093/molbev/msp312 620 48. Tomlinson, P.B. 2016. The botany of mangroves. Second edition. Cambridge, UK. 621 Cambridge University Press. 432 pp. 622 49. Valderrama-Landeros, L. H., Rodríguez-Zúñiga, M. T., Troche-Souza, C., Velázquez-623 Salazar, S., Villeda-Chávez, E., Alcántara-Maya, J. A…Ressl, R. 2017. Manglares de 624 México: actualización y exploración de los datos del sistema de monitoreo 1970/1980-625 2015. National Commission for Biodiversity (CONABIO), Mexico City, Mexico. 626 128pp 627 50. Wee, A. K., Takayama, K., Asakawa, T., Thompson, B., Onriza, Sungkaew, S… Webb, 628 E, L. 2014. Oceanic currents, not land masses, maintain the genetic structure of the 629 mangrove Rhizophora mucronata Lam. (Rhizophoraceae) in Southeast Asia. Journal of 630 Biogeography 41, 954–964. https://doi.org/10.1111/jbi.12263 631 51. Wee, A.K.S., Mori, G.M., Lira, C.F., Núñez-Farfán, J., Takayama, K., Faulks, L… 632 Kajita, T. 2019. The integration and application of genomic information in mangrove 633 conservation. Conservation Biology, 33, 206-209. DOI: 10.1111/cobi.13140 634 52. Wilson, G. A and Rannala, B. 2003. Bayesian inference of recent migration rates using 635 multilocus genotypes. Genetics, 163, 1177-1191. 636 https://www.ncbi.nlm.nih.gov/pubmed/12663554 637 53. Wright, S. 1965. The interpretation of population structure by F-statistics with special 638 regards to systems of mating. Evolution, 19, 395-420. https://doi.org/10.1111/j.1558-639 5646.1965.tb01731.x 640 54. Xu, S., He, Z., Zhang, Z., Guo, Z., Lyu, H., Li, J., … Shi, S. 2017. The origin, 641 diversification and adaptation of a major mangrove clade (Rhizophoreae) revealed by 642 whole-genome sequencing. Molecular Biology and Genetics, 4, 721-734. 643 https://doi.org/10.1093/nsr/nwx065 644 AQBOT_2019_112 100 55. Yan, Y-B., Duke, N. and Sun, M. 2016. Comparative analysis of the pattern of 645 population genetic diversity in three Indo-West Pacific Rhizophora mangrove species. 646 Frontiers in Plant Science, 7, 1434. https://doi.org/10.3389/fpls.2016.01434 647 648 AQBOT_2019_112 101 Figure captions 649 Fig. 1. Location of the 29 populations of Avicennia germinans and the geographic 650 projection of the expected heterozygosity of the black mangrove populations in Mexico 651 (Names of localities in Table 1). 652 Fig. 2. Allelic patterns over populations of Avicennia germinans across loci. Bar plots show 653 for each locus: Na, the number of different alleles with a frequency ³ 5 %; Ne, the effective 654 number of alleles (number of equally frequent alleles in an ideal population); I, information 655 index (equivalent to Shannon-Weaver Index of ecology). Pies chart show allele frequencies 656 for each locus across the Pacific and Atlantic populations. Different alleles are represented 657 by different colors for each locus. 658 Fig. 3. Population genetic differentiation of Avicennia germinans along the coasts of 659 Mexico. a) Pairwise FST (below diagonal) and RST values (above the diagonal) for the 660 Pacific and Atlantic populations, respectively; NS = Not significant (P > 0.05). b) 661 Neighbor-joining tree based on Nei's genetic distances for the Pacific and Atlantic 662 populations, respectively; bootstrap values > 60 percent are shown. c) The Pacific (upper) 663 and Atlantic (lower) principal coordinate analyses (PCoA). Symbols belong to a population 664 group identified within each coast; stars within the Atlantic gene pool (panel b) shown the 665 division into West (black stars) and East (gray stars) populations within the Yucatan 666 Peninsula. 667 Fig. 4. The most likely genetic barriers detected for Avicennia germinans populations in 668 Mexico. Red lines indicate geographic barriers. Voronoï tessellation is shown in gray and 669 the corresponding Delaunay triangulation of samples (red dots) in blue. Numbers indicate 670 bootstrap support. The green shading denotes mangrove distribution in Mexico. 671 672 102 Figures Fig 1. No color in print Fig 2. No color in print 103 Fig 3. No color in print 104 Fig 4. No color in print 105 Tables Table 1 Sampling sites of Avicennia germinans along Mexican coasts. Population identifier (ID), population names and geographic location. No. Population name ID Geographic location Latitude (N) Longitude (W) Pacific 1 El Sargento PN04 29.328517 -112.32918 2 Bahía Concepción PN02 26.762133 -111.88158 3 Topolobampo PN06 25.582191 -109.12353 4 Bahía Magdalena PN03 24.794183 -112.11565 5 La Paz PN05 24.181593 -110.29985 6 Bahía de Altata PN11 24.615667 -107.86983 7 El Caimanero PN12 22.882208 -106.06256 8 Marismas Nacionales PN10 22.374467 -105.64037 9 Punta Pérula PC15 19.592027 -105.12519 10 Barra de Navidad PC11 19.195532 -104.67225 11 Técpan de Galeana PC19 17.213935 -100.79211 12 Barra de Tecoanapa PS17 16.4978 -98.724133 13 Lagunas de Chacahua PS20 16.0114 -97.591833 14 Mar Muerto PS29 16.0104 -93.846033 15 La Encrucijada PS24 15.168367 -92.831017 Atlantic 16 Río Bravo GM37 25.946133 -97.15795 17 La Pesca GM41 23.776083 -97.74145 18 Tuxpan GM56 20.9807 -97.3415 19 La Mancha GM40 19.586167 -96.384783 20 Alvarado GM53 18.73285 -95.78225 21 Sontecomapan GM54 18.508417 -95.024717 106 22 Coatzacoalcos GM36 18.0971 -94.431383 23 Mecoacán GM46 18.419626 -93.115455 24 Pom-Atasta PY67 18.5745 -92.104 25 Los Petenes PY66 20.158917 -90.371583 26 Río Lagartos PY70 21.6102 -88.1282 27 Yumbalam PY81 21.443583 -87.202967 28 Cozumel PY65 20.555633 -86.913633 29 Sian Ka'an PY77 20.110967 -87.470917 107 Table 2. Gene diversity estimates for Pacific populations of Avicennia germinans in Mexico. Standard errors, where relevant, are given in parentheses. N % P Nea Ib Hs* Comm. Allelesc 25 % Comm. Allelesc 50 % PN04 40 12.5 1.020 0.033 0.018 0 0.125 (0.020) (0.033) (0.047) (0) (0.125) PN02 40 62.5 1.031 0.072 0.030 0 0.500 (0.012) (0.025) (0.029) (0) (0.189) PN06 40 75 1.133 0.221 0.107 0.250 0.750 (0.051) (0.079) (0.101) (0.164) (0.366) PN03 40 62.5 1.160 0.164 0.100 0.250 0.375 (0.113) (0.082) (0.157) (0.164) (0.183) PN05 40 37.5 1.181 0.189 0.108 0 0.250 (0.126) (0.104) (0.170) (0) (0.250) PN11 40 100 1.527 0.540 0.292 0.375 1.125 (0.166) (0.134) (0.208) (0.183) (0.398) PN12 40 87.5 1.215 0.318 0.156 0.625 1.250 (0.081) (0.104) (0.134) (0.324) (0.526) PN10 40 87.5 1.489 0.476 0.242 0.625 1.625 (0.246) (0.153) (0.222) (0.375) (0.596) PC15 40 50 1.287 0.287 0.166 0.125 0.875 (0.143) (0.130) (0.206) (0.125) (0.398) PC11 40 75 1.165 0.198 0.109 0.250 0.500 (0.106) (0.083) (0.148) (0.250) (0.378) PC19 40 50 1.442 0.416 0.222 0.625 1.250 (0.203) (0.178) (0.242) (0.375) (0.620) PS17 40 75 1.395 0.390 0.202 0.875 1.750 (0.216) (0.145) (0.216) (0.515) (0.726) 108 PS20 40 75 1.646 0.545 0.305 0.625 1.375 (0.270) (0.158) (0.236) (0.324) (0.460) PS29 40 100 2.083 0.806 0.407 1.250 2.250 (0.357) (0.214) (0.277) (0.526) (0.773) PS24 40 75 2.116 0.708 0.374 0.875 1.750 (0.440) (0.226) (0.305) (0.479) (0.620) Mean 68.33 1.393 0.357 0.189 (6.08) (0.058) (0.038) (0.226) N, number of individuals sampled; % P, percentage of polymorphic loci; Ne, number of effective alleles; I, information index; Hs, gene diversity; Comm. Alleles, number of common alleles found in £ 25 and £ 50 % of populations, respectively. *Standard deviations are shown for this parameter. a Number of equally frequent alleles in an ideal population. b A measure of genetic diversity equivalent to the Shannon-Weaver Index. c Number of alleles in a given population with a frequency greater than 5 % (i. e. present in more than 5 % of individuals sampled in that population) that are shared by that population with up to 25 % (Comm. Alleles 25 %) or 50 % (Comm. Alleles 50 %) of the rest of the sampled populations. 109 Table 3. Genetic diversity estimates for Atlantic populations of Avicennia germinans in Mexico. Standard errors are given in parentheses. N % P Nea Ib Hs* Comm. Allelesc 25 % Comm. Allelesc 50 % GM37 40 62.5 1.321 0.290 0.186 0 0.250 (0.145) (0.119) (0.211) (0) (0.164) GM41 40 75 1.303 0.314 0.185 0.125 0.375 (0.134) (0.107) (0.191) (0.125) (0.183) GM56 40 75 1.596 0.557 0.310 0.500 0.875 (0.191) (0.147) (0.222) (0.267) (0.295) GM40 34 87.5 1.527 0.476 0.271 0.375 0.750 (0.201) (0.144) (0.233) (0.263) (0.250) GM53 41 100 1.469 0.493 0.273 0.250 0.625 (0.155) (0.110) (0.188) (0.250) (0.375) GM54 34 75 1.376 0.374 0.210 0.250 0.375 (0.167) (0.138) (0.213) (0.164) (0.263) GM36 39 100 1.580 0.528 0.308 0.375 0.625 (0.182) (0.121) (0.213) (0.183) (0.324) GM46 35 87.5 1.744 0.621 0.373 0.125 0.625 (0.190) (0.145) (0.221) (0.125) (0.263) PY67 40 75 1.685 0.624 0.347 0.375 0.875 (0.188) (0.154) (0.222) (0.183) (0.295) PY66 41 87.5 2.029 0.700 0.397 0.500 0.750 (0.387) (0.181) (0.256) (0.189) (0.250) PY70 40 75 1.474 0.460 0.253 0.375 0.750 (0.182) (0.141) (0.225) (0.183) (0.250) PY81 40 87.5 1.502 0.495 0.276 0.500 0.875 (0.167) (0.141) (0.215) (0.267) (0.398) 110 PY65 40 75 1.760 0.631 0.351 0.625 1.000 (0.226) (0.171) (0.257) (0.183) (0.327) PY77 40 75 1.890 0.686 0.379 0.625 1.125 (0.306) (0.175) (0.250) (0.183) (0.350) Mean 81.25 1.590 0.518 0.294 (2.86) (0.057) (0.038) (0.234) N, number of individuals sampled; % P, percentage of polymorphic loci; Ne, number of effective alleles; I, information index; Hs, gene diversity; Comm. Alleles, number of common alleles found in £ 25 and £ 50 % of populations, respectively. *Standard deviations are shown for this parameter. a Number of equally frequent alleles in an ideal population. b A measure of genetic diversity equivalent to the Shannon-Weaver Index. c Number of alleles in a given population with a frequency greater than 5 % (i. e. present in more than 5 % of individuals sampled in that population) that are shared by that population with up to 25 % (Comm. Alleles 25 %) or 50 % (Comm. Alleles 50 %) of the rest of the sampled populations. 111 Inferring potential barriers to gene flow in tropical populations of Avicennia germinans Ochoa-Zavala, M., Osorio-Olvera, L., Piñero, D. and Núñez-Farfán, J. 112 Appendix A. R code used to predict the expected heterozygosity of the black mangrove in Mexico #----------------------------------------------------- # GeneDiversity_Ag.R; Ochoa-Zavala M., Osorio-Olvera L., Piñero D. and Núñez-Farfán J. Inferring barriers to gene flow of tropical populations of the black mangrove (Avicennia germinans (L.) L.). Supplementary material. # This script will obtain a genetic diversity heat map of the black mangrove onto a spatial grid based on the distribution of mangroves in Mexico. #----------------------------------------------------- #Set the working space setwd("~/Documents/LuisOsorio/Capas/Capas/conabio_mapa/") #‘biotools’ has the function sHe which computes the unbiased estimate of expected # heterozygosity (Nei, 1978) library(biotools) # Functions to work with raster and data frame library(raster) library(dplyr) # ‘maptools’ has wrld_simpl object which is # a SpatialPolygonsDataFrame of the world’s countries library(maptools) library(rgdal) data("wrld_simpl") #Coordinates with the genotyping data for the black mangrove data <- read.csv(file="AgDatabase.csv", header = TRUE, stringsAsFactors = F) data2 <- read.csv("Coord_UTM_corregida_LOsorio.csv", stringsAsFactors = F) names(data2)[2:3] <- c("Pop.ID","Ind.ID") clean_data <- dplyr::inner_join(data,data2,by=c("Pop.ID","Ind.ID")) # ---------------------------------------------------------------------------- # 1. Code to create the Spatial grid # ---------------------------------------------------------------------------- #Extract the polygons of mangroves distribution in Mexico (CONABIO, 2016) mapa_mangles <- readOGR("mx_man15gw/","mx_man15gw") g1 <- raster() extent(g1) <-extent(mapa_mangles) res(g1) <- 0.005 g1[]<- 1 # Cut and mask the raster using the extent of mangroves distribution in Mexico mex_ras <- crop(g1, mapa_mangles) mex_ras <- mask(mex_ras,mapa_mangles) #plot(mex_ras) #Table with spatial coordinates of the grid of Mexico cords_mex <- rasterToPoints(mex_ras) dim(cords_mex) #Run the analysis and compute the expected heterozygosity per individual #using sHe function (da Silva et al., 2017) 113 ex3 <- sHe(clean_data, coord.cols= 23:22, marker.cols= 5:20, marker.type = "codominant", grid = cords_mex[,1:2], radius = 110,latlong2km = T) #Transform results into a raster object cordsdiv <- ex3$diversity cordsdiv$uHe coordinates(cordsdiv) <- ~coord.x+coord.y gridded(cordsdiv) <- TRUE r_div <-raster(cordsdiv[,3]) pdf("diversidad_mangle_005_res.pdf",width = 8,height = 7) plot(r_div) plot(mapa_mangles,add=T) dev.off() #Write raster results writeRaster(r_div,"Mangles_diversidad_005_res.tif",overwrite=T) References CONABIO. 2016. Distribución de los manglares en México en 2015. Escala: 1:50000. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad. Sistema de Monitoreo de los Manglares de México (SMMM). Ciudad de México, México. http://www.conabio.gob.mx/informacion/metadata/gis/mx_man15gw.xml?_httpcache=y es&_xsl=/db/metadata/xsl/fgdc_html.xsl&_indent=no da Silva, A. R., Malafaia, G. and Menezes, I. P. P. 2017. Biotools: an R function to predict spatial gene diversity via an individual-based approach. Genetics and Molecular Research 16, gmr16029655. Nei, M. 1978. Estimation of average heterozygozity and genetic distance from a small number of individuals. Genetics, 89, 583-590. 114 Inferring potential barriers to gene flow in tropical populations of Avicennia germinans Ochoa-Zavala, M., Osorio-Olvera, L., Piñero, D. and Núñez-Farfán, J. 115 Appendix B. Gene diversity estimates, allele frequencies, populations groups and migration rates estimate for Avicennia germinans in Mexico Fig. B.1 Distribution of allele frequencies by locus across Pacific and Atlantic populations of Avicennia germinans in Mexico. 116 Table B.1. Gene diversity estimates by locus across populations of Avicennia germinans in the Pacific coast of Mexico. Standard errors are shown in parentheses. Agerm_CT_004 Agerm_GT_006 Agerm-25 Agerm-22 Agerm-18 AgT9 AgT4 AgD6 Grand Mean N 34.667 34.733 38.400 39.800 39.667 40 39.867 39.867 38.375 (1.545) (1.584) (0.970) (0.145) (0.270) (0.000) (0.091) (0.091) (0.357) HO 0.006 0.095 0.138 0.028 0.351 0.127 0.239 0.339 0.165 (0.003) (0.034) (0.044) (0.021) (0.067) (0.043) (0.061) (0.062) (0.020) uHE 0.015 0.124 0.164 0.040 0.416 0.155 0.253 0.343 0.189 (0.006) (0.041) (0.051) (0.030) (0.071) (0.044) (0.065) (0.054) (0.021) N, sample size; the observed (HO) and unbiased expected heterozygosity (uHE). 117 Table B.2. Gene diversity estimates by locus across populations of Avicennia germinans in the Atlantic coast of Mexico. Standard errors are shown in parentheses. Agerm_CT_004 Agerm_GT_006 Agerm-25 Agerm-22 Agerm-18 AgT9 AgT4 AgD6 Grand Mean N 37.286 38.429 38.357 36.857 38.000 38.857 38.857 38.857 38.188 (0.683) (0.677) (0.782) (1.084) (0.825) (0.670) (0.670) (0.670) (0.272) HO 0.450 0.373 0.013 0.287 0.328 0.002 0.509 0.347 0.289 (0.043) (0.039) (0.005) (0.058) (0.090) (0.002) (0.042) (0.056) (0.024) uHE 0.494 0.406 0.017 0.301 0.250 0.009 0.505 0.368 0.294 (0.037) (0.042) (0.006) (0.055) (0.055) (0.005) (0.029) (0.058) (0.022) N, sample size; the observed (HO) and unbiased expected heterozygosity (uHE). 118 Fig. B.2. Groups of populations recognized by Ochoa-Zavala et al. (2019) for both, the Pacific and Atlantic coasts of Mexico. Each colour represents one population group; the shading denotes the mangroves distribution in Mexico. Names of localities in Table 1. 119 Fig. B.3. Allele frequencies by locus across populations in the Pacific and Atlantic coasts. Names of localities in Table 1. Allele frequencies were plotted using standArich (Alberto, 2006). 120 Fig. B.3 Continued 121 Table B.3 Partition of genetic variability for Avicennia germinans based on the population’s groups defined for the Pacific and Atlantic coasts of Mexico. Source F-statistic Percentage variation F-statistic Percentage variation SMM IAM Pacific Among population groups FCT=0.495*** 49.482 FCT=0.272*** 27.204 Among populations within groups FSC=0.142*** 7.197 FSC=0.101*** 7.343 Among individuals within populations FIS=0.115*** 4.967 FIS=0.128*** 8.382 Within individuals FIT=0.616*** 38.353 FIT=0.429*** 57.071 Atlantic Among population groups FCT=0.064** 6.397 FCT=0.139*** 13.876 Among populations within groups FSC=0.101*** 9.484 FSC=0.105*** 9.083 Among individuals within populations FIS=0.096*** 8.125 FIS=0.018 1.384 Within individuals FIT=0.240*** 75.995 FIT=0.243*** 75.656 SMM, stepwise mutation model with RST; IAM, infinite allele model with FST; ** P < 0.01; *** P < 0.001. 122 Table B.4 Migration rates estimates between source and recipient populations of Avicennia germinans in the Pacific coast of Mexico; migration rates higher than 0.025 are shown. Names of localities in Table 1. Source Recipient Mean Standard deviation PN04 PN02 0.1557 0.0228 PN04 PN06 0.2375 0.0221 PN04 PN03 0.1387 0.0305 PN02 PN04 0.0571 0.0176 PN02 PN05 0.0448 0.0403 PN06 PN05 0.0424 0.0238 PN06 PN11 0.2088 0.0246 PN06 PN12 0.1222 0.0589 PN06 PN10 0.1691 0.0284 PN03 PN05 0.0631 0.0274 PN03 PC19 0.0378 0.0278 PN03 PS17 0.026 0.0239 PN03 PS20 0.0268 0.023 PN05 PN12 0.0302 0.0153 PN05 PN10 0.0283 0.0145 PC11 PN12 0.0505 0.0277 PC11 PN10 0.0363 0.0208 PC11 PC15 0.0291 0.0225 PC11 PC19 0.0369 0.0254 PC11 PS17 0.1783 0.0449 PC11 PS20 0.0547 0.0411 PS24 PS29 0.0254 0.0212 123 Table B.5 Migration rates estimates between source and recipient populations of Avicennia germinans in the Atlantic coast of Mexico; migration rates higher than 0.025 are shown. Names of localities in Table 1. Source Recipient Mean Standard deviation GM37 GM41 0.0358 0.0257 GM37 GM53 0.0298 0.0182 GM37 PY67 0.0457 0.0272 GM41 GM56 0.0422 0.0306 GM41 GM53 0.026 0.0218 GM41 GM36 0.0416 0.0294 GM41 GM46 0.0319 0.023 GM41 PY67 0.0274 0.0218 GM56 PY67 0.0277 0.0403 GM56 PY66 0.0321 0.0471 GM53 GM56 0.1195 0.051 GM53 GM40 0.2146 0.0273 GM53 GM36 0.0638 0.0424 GM53 GM46 0.0409 0.0299 GM36 GM56 0.0291 0.0269 GM36 GM53 0.1044 0.0432 GM36 GM54 0.2216 0.0236 GM36 GM46 0.0371 0.0255 GM36 PY67 0.0471 0.0338 PY67 GM56 0.0426 0.0297 PY67 GM46 0.044 0.029 PY67 PY66 0.1466 0.054 PY70 GM46 0.0421 0.0254 PY70 PY81 0.0705 0.0547 PY81 PY70 0.0359 0.0283 PY81 PY77 0.0286 0.0198 PY65 PY81 0.0331 0.0215 PY65 PY77 0.1013 0.0378 References Alberto, F. 2006. standArich v1.0: an R package to estimate populations allelic richness using standardized sample size. University of the Algarve, Portugal. Capítulo 3 Distancia al centroide del nicho como predictor de la diversidad genética de las poblaciones del mangle negro 125 Distance to the niche centroid as a predictor of genetic diversity in black mangrove 1 populations 2 3 Maried Ochoa-Zavala1+, Luis Alfredo Osorio-Olvera2,3+, Ivania Cerón-Souza4, Elsie 4 Rivera-Ocasio5, Vania Jiménez-Lobato6 and Juan Núñez-Farfán1. 5 6 1. Departamento de Ecología Evolutiva, Instituto de Ecología, Universidad Nacional 7 Autónoma de México. Circuito Exterior S/N anexo Jardín Botánico, Ciudad Universitaria, 8 C.P. 04510 Ciudad de México, México. 9 2. Biodiversity Institute, University of Kansas, Lawrence, KS 66045. 10 3. Centro del Cambio Global y la Sustentabilidad en el Sureste A.C, C.P. 86080 11 Villahermosa, Tabasco, México. 12 4. Centro de Investigación Tibaitatá, Corporación Colombiana de investigación 13 Agropecuaria – AGROSAVIA. Km 14 vía Bogotá-Mosquera, Cundinamarca, Colombia. 14 5. Department of Biology, University of Puerto Rico–Bayamon,Bayamon, PR 00959, USA. 15 6. Universidad Autónoma de Guerrero, Facultad de Desarrollo Sustentable. Carr. Federal 16 Acapulco Zihuatanejo Km. 106+900. Col. Las Tunas CP. 40900 Tecpan de Galeana, Gro. 17 + Both authors contributed equally. 18 19 Abstract 20 Recent theoretical and empirical works state that populations near to the ecological niche 21 centroid of species are expected to have higher fitness attributes than populations that are 22 closer to niche edges; this encompasses the so-called "niche-centroid hypothesis", which 23 has been mostly tested using abundance data and, more recently, with genetic diversity 24 data. Here, we coupled ecological niche modeling and population genetics data in order to 25 determine the set of environmental variables that better explain genetic variability of a 26 mangrove species, Avicennia germinans, test whether there is a relationship between 27 distance to the niche centroid and genetic diversity, and predict the spatial variation of gene 28 diversity of A. germinans along most of its natural geographic range. Variables that better 29 explained the geographic variation of gene diversity of A. germinans were related to soil 30 126 texture. As predicted by the niche-centroid hypothesis, we found a strong negative 31 correlation between observed (Ho) and expected heterozygosity (He) and the distance to the 32 niche centroid (rs = -0.626 P < 0.001 for Ho; rs = -0.605, P < 0.001 for He). Our models to 33 predict gene variation as a function to the distance to the niche-centroid showed 34 predictability of 40%. The results suggest the strong influence that soil texture (clay 35 content) have in the establishment, growth and distribution of mangroves, which in turn, 36 impact standing genetic variation. The capacity to predict the genetic variation across the 37 natural range of a species and predict future changes are powerful tools in conservation and 38 management efforts of wildlife species. 39 40 Keywords: Abundant center hypothesis, Avicennia germinans, clay, environmental niche-41 centroid, genetic diversity, niche modelling, soil texture. 42 127 Introduction 43 44 Genetic diversity is a central piece in conservation biology and evolution. It is the raw 45 material in which natural selection acts to produce adaptive evolutionary change, and it is a 46 strong predictor of fitness and extinction risk (Frankham, 1996; Reed and Frankham, 2003; 47 Spielman et al., 2004; Frankham, 2005; Frankham, 2012). The amount of genetic diversity 48 is governed by several factors at different levels, among species and within genomes 49 (Ellegren and Galtier, 2016). However, effective population size has undoubtedly a key role 50 in determining the genetic diversity of species (Frankham, 2012; Ellegren and Galtier, 51 2016). 52 53 The abundant-centre hypothesis predicts a peak of abundance in the geographic 54 centre of populations, while peripheral populations suffer a loss of genetic variation 55 because of reduced effective population size, drift, and inbreeding (Diniz-Filho et al., 56 2009). Intuitively, the abundant-centre hypothesis makes sense however many exceptions 57 exist (e. g. Pfeifer et al., 2009; Abeli et al., 2014; Pironon et al., 2015, Dallas et al., 2017; 58 Santini et al., 2018), reflecting the fact that suitability of a place for a species is a function 59 of environmental conditions, and it is not necessarily related to the distance to the 60 geographic centre of the species' distribution (Manthey et al., 2015). Thus, demographic 61 processes may relate more directly to the quality of local conditions, as expressed by the 62 fundamental ecological niche of species (Martínez-Meyer et al., 2013; Lira-Noriega and 63 Manthey, 2014). Hutchinson (1957) proposed that the ecological niche is a 64 multidimensional hyper-volume constructed with n-axes, each of which represents 65 fundamental variables for the population survival. Such hyper-volume has an internal 66 structure in which optimal conditions (e. g. the highest intrinsic population growth rate) are 67 in the centroid of the ecological niche (Maguire, 1973); outside this environmentally 68 defined niche, populations should show zero or negative population growth (Lira-Noriega 69 and Manthey, 2014). In this sense, the distances to the environmental niche centroid may 70 better reflect fitness attributes (like population abundance and genetic diversity) than the 71 distance to the geographic range center; this is the so-called "niche-centroid" hypothesis 72 (Martínez-Meyer et al., 2013). 73 128 Recent papers have shifted to such a niche-based view, finding that distance to the niche 74 centroid (instead of the geographic centre) better explain trends in population abundance 75 (Yañez-Arenas et al., 2012; Martínez-Meyer et al., 2013; Osorio-Olvera et al., 2016) and 76 the standing genetic variation (Lira-Noriega and Manthey 2014; Micheletti and Storfer, 77 2015). From this niche-based point of view, the largest genetic variation is observed in 78 locations with high environmental suitability, which in turn occur in regions close to the 79 niche-centroid. Thus, genetic variability declines as populations are located 80 environmentally farther from this niche-centroid (Lira-Noriega and Manthey 2014). Hence, 81 a significant and negative correlation between heterozygosity and niche-centroid distance 82 would reflect the effects of variable population size in geographical space, which, under 83 distinct environmental conditions, leads to a well-known pattern in which larger 84 populations are able to maintain more genetic diversity (Diniz-Filho et al., 2009; Martínez-85 Meyer et al., 2013; Diniz-Filho et al., 2015). 86 87 Ecological niche modelling coupled with population genetics has the capacity to 88 show how habitat suitability can affect the genetic diversity of species (Micheletti and 89 Storfer, 2015). Predicting genetic variability on the geographical space and define the 90 environmental variables that better explain these patterns has important implications in 91 species ecology and therefore, in the success of conservation programs. The majority of 92 studies applying niche modelling techniques concentrate on temperate terrestrial species, 93 and in a lesser extent, on tropical and marine species (Eckert et al., 2008). Despite 94 mangroves play a critical role in maintaining species diversity, store more carbon than most 95 other forest types, and provide protection from erosion caused by storm and tsunamis 96 (Constanza et al., 1997; Tomlinson 2016; Donato et al., 2011), a few studies using niche 97 modelling are available for mangroves species (e.g. Crase et al., 2013; Record et al., 2013). 98 99 Since niche modelling has the potential of using the environmental suitability 100 estimates to infer demographic data, which in turn, can be related to the genetic variability 101 of a species across geographical space (Hedrick, 2011; Micheletti and Storfer, 2015; Diniz-102 Filho et al., 2015), we used ecological niche modelling approach to evaluate whether 103 habitat quality (measured as the distance to the niche centroid) might be a good predictor of 104 129 gene diversity in an important mangrove species, Avicennia germinans, determine the set of 105 environmental variables that better explain genetic variability in this taxon and predict the 106 spatial variation of heterozygosity along most of its natural geographic distribution. 107 108 Materials and methods 109 110 Study system 111 Mangroves are flowering trees that live along the coastlines of tropical and subtropical 112 regions of the world (Tomlinson, 2016). Global distribution of mangroves is mainly limited 113 by physiological tolerance to low temperature (Duke et al., 1998). However, the range of 114 conditions over which mangroves occur naturally encompasses other environmental 115 extremes (Clough, 1992). Mangroves are adapted to tidal conditions and a special 116 combination of factors that characterize coastal and estuarine shorelines, like seawater, 117 periodic inundation and exposure to waves, wind, and offshore currents (Duke, 2017). 118 Mangroves also grow on a variety of soil types, including heavy consolidated clays, 119 unconsolidated silts, calcareous and mineral sands, coral rubble, and organic peats. They 120 also tolerate varying levels of salinity, up to 90 ppt (Clough, 1992). Soils inhabited by 121 mangroves are anaerobic, and may be acid or alkaline with varying values of nitrogen, 122 phosphorous, potassium, calcium, magnesium, sodium and chlorine, some of which are 123 frequent in tidal water (Hossian and Nuruddin, 2016). Mangrove forests are best developed 124 on tropical shorelines with low-energy and abundant supply of fine-grained sediments, and 125 are most luxuriant in areas of high rainfall or abundant freshwater supply through river 126 discharge, in areas where conditions are persistently cloudy, the ratio of precipitation to 127 evaporation is low, solar radiation fluxes are not extreme, and seasonal variability in 128 climate is small. In general, high wave energy and sandy substrate are not favorable 129 conditions for mangrove establishment (Clough, 1992; Woodroffe, 1992). 130 131 The black mangrove, Avicennia germinans, is a common mangrove species that 132 grows in the intertidal zone of the tropical and subtropical coastlines of America and West 133 Africa; it is one of the most tolerant species to high salinity, aridity and low temperatures 134 (Clough, 1992). Avicennia germinans is commonly found in sympatry with Rhizophora 135 130 species in most of its distribution, growing in fibrous soil (high percentage of the organic 136 matter in the form of undecomposed remains of Rhizophora rootles); however, Avicennia 137 seems to prefer firmer and more sandy soils. The fibrous soils are siltier and have higher 138 content of clay than non-fibrous sandy soils (Jordan, 1964). It is possible that soft soils are 139 not suitable for the growth of Avicennia; its distribution is favored in sandy substrate 140 instead of clayish substrates, which benefit Rhizophora (Jordan, 1964; Diop et al., 1997). 141 142 Genetic data 143 Molecular genetic data was gathered for 1419 individuals from 40 populations covering 144 most of black mangrove’s natural distribution (Table 1). Genotypes were scored for two 145 datasets of eight microsatellite loci obtained from Ochoa-Zavala et al. (2019b) and Cerón-146 Souza, V. and Rivera-Ocasio, E. (personal communication; Appendix 1). The first dataset 147 included 1144 individuals from 29 natural populations located in Mexico. The number of 148 alleles per locus ranged from 2 to 4.5 with a sample size of 40 individuals per site. The 149 second dataset included 275 individuals distributed in 11 populations located mainly in 150 Central America (Table 1). The number of alleles per locus ranged from 9.8 to 3.3, and 151 sample sizes within local populations ranged from 13 to 40 individuals (Table 1). Both 152 databases share three microsatellite loci (AgT4, GT_006, and CT_004). Mean observed and 153 expected heterozygosity (Ho, He, respectively) were estimated for each population and for 154 each data set using Arlequin version 3.5 (Excoffier and Lischer, 2010). 155 156 Ecological niche modelling 157 To characterize environmental variation for the ecological niche modeling, we used a total 158 of 76 environmental variables, 19 of which were gathered from WorldClim (Hijmans et al., 159 2005), and 57 additional soil variables were obtained from the SoilGrids database 160 (https://www.isric.org/explore/soilgrids; Hengl et al., 2014). The bioclimatic variables were 161 derived from monthly temperature and precipitation reports of weather data between 1960 162 and 1990. On the other hand, soil variables included: depth to the bedrock, organic carbon 163 stock, bulk density, clay, silt, and sand content, soil pH and cation exchange capacity 164 (Hengl et al., 2014). Models were calibrated within the M region, which is the area 165 hypothesized to be accessible for a species (Barve et al., 2011). Because the black 166 131 mangrove has high dispersal capacity (Nettel and Dodd, 2007), we defined M, as the area 167 occupied by mangroves forests (Giri et al., 2011) along the coastlines of America and West 168 Africa applying a buffer of 0.5 degrees. We obtained a polygon which comprised areas that 169 A. germinans could potentially reach. Environmental layers were masked to the extent of M 170 by using the 'crop and mask' functions in the 'raster' R package (Hijmans et al., 2019). We 171 obtained 3710 occurrence data for A. germinans from the global biodiversity information 172 facility (GBIF); we eliminated incorrectly georeferenced records and thinned the data using 173 a spatial filter of 0.008333333 degrees (~1 km); to do so, we used the ‘ntbox’ R package 174 (Osorio-Olvera et al., 2019). Data curation resulted in 594 occurrence records. We were 175 unable to include other further variables to improve the ecological niche modeling 176 algorithm, including sea surface salinity and sea surface temperature, because they were at 177 a different spatial resolution than we used for modeling. 178 179 For ecological niche modelling, we used the ‘cov_center’ function from ntbox. We 180 first searched all possible combinations of environmental variables (in sets of three) that 181 described the niche, rejecting variables that showed strong correlations with each other 182 (multi-collinearity). For each combination of variables, we built a minimum volume 183 ellipsoid (Van Aelst and Rousseeuw, 2009; Qiao et al., 2016) that represented the 184 ecological niche. This function computes the shape and centroid of the ellipsoid model 185 using values of niche variables at each occurrence’s points. Then, to searched for 186 combinations of environmental variables that better fit Ho and He (P < 0.001), for each 187 model, we extracted environmental data from each geographic location for which Ho and 188 He values were available. To evaluate the relationship between the distance to the niche 189 centroid (DNC) and population genetic diversity, we estimated DNC (Yañez-Arenas et al., 190 2012; Martínez-Meyer et al., 2013) with the Mahalanobis distance measure and performed 191 an exponential regression with Spearman’s rank (rs) between DNC and gene diversity 192 estimates. We projected the fundamental niche of A. germinans on the geographic space 193 using niche models (for Ho and He) that include the combination of variables with the 194 highest rs at P < 0.001 (best fitting models). 195 196 132 We used the equation of correlation coefficient between DNC and Ho and He, 197 independently, to predict the spatial variation of genetic diversity on the geographic space. 198 To evaluate the quality of these predictions, we performed linear regression between the 199 observed and predicted genetic estimates using as a predictable variable (x) from the 200 simulated genetic estimates. 201 202 Results 203 204 Resulting models indicate that the ecological niche of A. germinans mostly explained by 205 soil’s texture properties, including: the percentage of clay particles (< 0.0002 mm) at soil 206 surface, 1 m and 2 m standard depth (Table S2.1 in Appendix 2). As expected, regression 207 coefficients indicated a strong negative relationship between genetic diversity (rs = -0.626 208 P < 0.001 for Ho; rs = -0.605, P < 0.001 for He) and the distance to the centroid of its 209 ecological niche (Fig. 1). Spatial predictions of genetic diversity were estimated with the 210 regression coefficients of the DNC models, based on the equations y = 0.73/ (0.69 + x 0.48) 211 and y = 0.76/ (0.62 + x 0.44) for Ho and He, respectively (Fig. 2). Our quality test indicated 212 that both simulated models have considerable explanatory power of the observed data (Fig. 213 2). 214 215 We generated a total of 140, 600 models of the fundamental niche of A. germinans, of 216 which only 70 models (with a mean rs of -0.5602 and -0.5595 for Ho and He, respectively) 217 resulted with P < 0.001 (Table S2.1). These models were characterized by a total of 18 and 218 15 environmental variables (for Ho and He, respectively) related mainly to temperature and 219 clay content at different soil deeps (Fig. 3). Among these models, seven specific 220 environmental variables highlighted by the frequency of which they appear over the 70 221 models with P < 0.001 (black stars in Fig. 3). These variables were annual mean 222 temperature, mean temperature of the three wettest months, mean temperature of the three 223 driest months, mean temperature of the three coldest months, mean percentage of the clay 224 particles at soil surface (0 m) and at standard deep of 0.05 m and 2 m (Fig. 3). Note that 225 clay particles at soil surface (0 m) and at standard deep of 2 m were variables included in 226 the better fit niche models. 227 133 228 In spite that variables that defined black mangrove fundamental niche (clay content) did not 229 show a clear pattern, we observed that populations located close to the niche-centroid 230 exhibited lower median value of clay content than those in the periphery (Fig. S2.1 in 231 Appendix 2). Additionality, four other variables (annual mean temperature, mean 232 temperature of the three wettest, driest and coldest months) showed a trend in which 233 centrality populations appear to experience in average, warmer temperatures than 234 populations farther from the central-niche, with exception of mean temperature of the three 235 wettest months (Fig. S2.1). 236 237 Discussion 238 239 Here, we evaluated whether there is a correlation between niche-centroid distance and 240 genetic variation (Lira-Noriega and Manthey 2014) and defined the set of environmental 241 variables that better predict the spatial genetic variation of a tropical tree along the 242 coastlines of America. 243 244 ¿how habitat quality may affect demography of populations? 245 The abundant-centre hypothesis postulates that the populations occurring at the edge of 246 their geographic range tend to be smaller, less dense and have less genetic diversity than 247 central populations. Recent studies have found that distances to the niche centroid, instead 248 of distances to geographic centre, better explain trends in population abundance, gene 249 diversity and genetic structure (Yañez-Arenas et al., 2012; Martínez-Meyer et al., 2013; 250 Lira-Noriega and Manthey, 2014; Micheletti and Storfer, 2015; Osorio-Olvera et al., 2016). 251 Under this niche-based point of view, distance to the ecological niche centroid represents 252 an important determinant of the spatial patterns of genetic variation, because environmental 253 variables impact on populations dynamics. Thus, a close link exists between the set of 254 habitable conditions and geographic genetic variation (Lira-Noriega and Manthey, 2014). 255 The strong correlation between genetic variation and soil texture variables, may reflect the 256 influence of this factor for mangroves’ populations dynamics. There are two mechanism by 257 which the percentage of the clay particles may act on mangrove growth and establishment. 258 134 Ukpong (1997) found that clay being a favourable substrate for rooting of mangroves 259 propagules (seedlings density and soil texture r = 0.62; p < 0.01) and also that, soil 260 properties are an important components of the mangroves environment, because soils 261 texture (silt, clay and sand compositions) are important determinants of tree high and 262 density (stems per hectares) in Avicennia and other mangrove species. Earlier, McMillan 263 (1975) found that seedlings of A. germinans and Laguncularia racemosa grown in 264 hypersaline conditions in 100 % sand composition failed to survive. However, seedlings 265 grown in soil composed of 90 % sand and 10 % clay had 100 % survival in hypersaline 266 conditions but showed some leaf discoloration. Seedlings grown in soil composed of 75 % 267 sand and 25 % clay, there was 100 % survival with no observable effect on leaves. In spite 268 of Avicennia showed a preference for sandy soils (Jordan, 1964; Diop et al., 1997), it is 269 possible that there is an optimum sand-clay composition for developing larger and vigorous 270 populations of A. germinans capable to bear higher genetic diversity, outside of this range 271 of soil components, constrain population developing, resulting in more shorter in height 272 and patchy populations with lower genetic variation, as can be observed in the north-273 western Mexico (Ochoa-Zavala et al., 2019a). Our results coupled with earlier findings 274 provide insights about the key role of soil texture in mangroves’ populations dynamics, 275 which in turn contribute to explain, at least in part, the standing genetic variation of 276 populations of black mangrove. 277 278 Temperature and precipitation are known to influence global patterns of distribution 279 in a variety of taxa (Woodward and Williams, 1987), and mangroves are not the exception 280 (Tomlinson, 1986; Duke et al., 1998). In addition to soil texture, we observed that 281 populations farther from the niche centroid experience lower temperatures during driest, 282 coldest and annual mean temperature. These variables might be relevant in the fundamental 283 niche of A. germinans for two reasons. First, a reduction of growth rate in mangroves is 284 related to the decrease of air and soil temperatures (Clough, 1992); and second, extreme 285 cold events such as frost, which are numerically incorporated in the mean temperature of 286 the coldest month, have a negative impact on mangroves (e. g. dieback) (Woodrooffe and 287 Grindrod, 1991; Krauss et al., 2008). Thus, we argued that higher temperatures during such 288 135 periods might be important for determining the habitat suitability as they may exert less 289 stress, allowing them to grow and develop better. 290 291 The relationship between genetic diversity and niche-centroid distance 292 We found strong non-linear relationships between heterozygosity and niche-centroid 293 distance (Fig. 1). Non-linear relationships between abundance and distance to the niche 294 centroid implies that optimal niche conditions are relatively narrow; thus, few sites hold 295 suitable conditions for maintaining large populations (Martínez-Meyer et al., 2013). In A. 296 germinans, few populations with suitable conditions, mostly located along the Atlantic 297 coast, bore high genetic diversity. Linear regression analysis showed that our predicted 298 models of observed and expected heterozygosity indicated considerable explanatory power 299 of observed values (R2 = 0.40; P < 0.001; Fig. 2) compared to other related studies 300 (Martínez-Meyer et al., 2013; Lira-Noriega and Manthey, 2014), thus correlation analysis 301 between DNC and genetic variation are useful for predict the spatial variation of 302 heterozygosity along most of its natural geographic distribution. These results are relevant 303 to plant conservation and evolutionary biology because the evolutionary potential of a 304 species is in large part a function of its genetic variation (Reed and Frankham, 2003), 305 modelling such genetics parameters provide a useful way to support and envisage 306 management plans. In addition, facing one of the major threats to biodiversity, using 307 current niche models coupled with general circulation models to project probable future 308 changes in gene diversity, recognizing the effects of uncertainties and assumptions in the 309 niche modelling (Wiens et al., 2009), provide a valuable tool for conservation and 310 management efforts. Although for mangroves species there have been a number of 311 predictions about the future in the face of climate changes (Record et al., 2013; Alongi, 312 2015), given the multi-ecosystem services they provide, the capacity of anticipate genetic 313 diversity changes are needed for more accurate information about which places are likely to 314 be most at risk for making decisions by conservationist and resource managers. 315 316 317 318 319 136 Conclusions 320 321 Black mangrove genetic diversity meets with the environmental based hypothesis of 322 abundant centre. We did an effort to contribute to the understanding of how habitat quality 323 may affect demography of populations of an important mangrove species. The strong 324 correlation between heterozygosity estimates and DNC illustrate how the environmental 325 variables that defined the fundamental niche (clay content) impact on the establishment, 326 growth and develop of individuals, meaning that soil compositions have an important role 327 in populations dynamics of A. germinans, explaining part of the genetic variation we 328 observed. The habitat suitability in A. germinans are spatially structured with few 329 populations mainly located along the Atlantic coast, able to bear high genetic diversity. The 330 capacity to predict the genetic variation across the natural range of a species and modeled 331 the future changes in the face of climate change, are powerful tools in conservation and 332 management efforts of wildlife species. 333 334 Acknowledgements 335 336 The authors thanks to all members of the Laboratorio de Genética Ecológica y Evolución 337 and Tapia-López, R. for logistical and field support. We thank to Rodriguez-Medina, K for 338 her useful advice to improve the ecological niche modeling section. M.O.-Z acknowledges 339 the scholarship granted for graduate studies by the National Council of Science and 340 Technology (CONACyT) and thanks the Graduate Program in Biological Sciences, 341 National Autonomous University of Mexico (UNAM). L. O.-O acknowledges PAPIIT 342 UNAM IN116018 and CONACyT postdoctoral fellowship program support for his 343 extended academic training and CONACyT-FORDECyT Project number 273646 for partial 344 funding. 345 137 References 346 Abeli, T., Gentili, R., Mondoni, A., Orsenigo, S. and Rossi, G. 2014. Effects of marginality 347 on plant population performance. Journal of Biogeography, 41, 239-249. 348 Alongi, D.M. 2015. The impact of climate change on mangrove forest. Current Climate 349 Change Reports, 1, 30-39. 350 Barve, N., Barve, V., Jiménez-Valverde, A., Lira-Noriega, A., Maher, S. P., Peterson, T... 351 Villalobos, F. 2011. The crucial role of the accessible area in ecological niche modeling 352 and species distribution modeling. Ecological Modeling, 22, 1810-1819. 353 Clough, B. F. 1992. Primary productivity and growth of mangroves forest. In: Robertson, 354 A. I. and Alongi, D. M. (Eds.), Tropical mangrove ecosystem, pp 225-249, vol. 41, 355 American Geophysical Union, Washington, DC. USA. 356 Costanza, R., d’Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B... and van den 357 Belt, M. 1997. The value of the world’s ecosystem services and natural capital. Nature, 358 387, 253–260. 359 Crase, B., Liedloft, A., Vest, P. A., Burgman, M. A. and Wintle, B. A. 2013. Hydroperiod 360 is the main driver of the spatial pattern of the dominance in mangrove communities. 361 Global Ecology and Biogeography. 22, 806-817. 362 Dallas, T., Decker, R. R. and Hastings, A. 2017. Species are not most abundant in the 363 centre of their geographic range or climatic niche. Ecology Letters, 20, 1526-1533. 364 Diniz-Filho, J. A. F., Nabout, J. C., Bini, L. M., Soares, T. N., Pires de Campos Telles, M., 365 de Marco, P. Jr., Collevatti, R. G. 2009. Niche modelling and landscape genetics of 366 Caryocar Brasiliense (“Pequi” tree: Caryocaraceae) in Brazilian Cerrado: an integrative 367 approach for evaluating central-peripheral population patterns. Tree Genetics & 368 Genomes, 5, 617-627. 369 Diniz-Filho, J. A., Rodrigues, H., Pires de Campos Telles, M., de Oliveira, G., Terribile, L. 370 C., Soares, T. N. and Nabout, J. C. 2015. Correlation between genetic diversity and 371 environmental suitability: taking uncertainty from ecological niche models in to account. 372 Molecular Ecology Resources, 15, 1059-1066. 373 Diop, E. S., Soumare, A. Diallo, N. and Guisse, A. 1997. Recent changes of mangroves of 374 the Saloum River Estuary, Senegal. Mangroves and Salt Marshes, 1, 163-172. 375 138 Donato, D.C., Kauffman, J.B., Murdiyarso, D., Kurnianto, S., Stidham, M. and Kanninen, 376 M. 2011. Mangroves among the most carbon-rich forests in the tropics. Nature 377 Geoscience, 4, 293–297. 378 Duke, N. C. 2017. Mangrove Floristics and Biogeography Revisited: Further Deductions 379 from Biodiversity Hot Spots, Ancestral Discontinuities, and Common Evolutionary 380 Processes. In: Rivera-Monroy, V. H., Lee, S.Y., Kristensen, E. and Twilley R.R. (Eds.), 381 Mangrove Ecosystems: A Global Biogeographic Perspective, pp 17-53, Springer 382 International Publishing AG. 383 Duke, N.C., Ball, M.C. and Ellison, J.C. 1998. Factors influencing biodiversity and 384 distributional gradients in mangroves. Global Ecology and Biogeography Letters, 7, 27–385 47. 386 Eckert, C. G., Samis, E. and Lougheed, S. C. 2008. Genetic variation across species’ 387 geographical ranges: the central-marginal hypothesis and beyond. Molecular Ecology, 388 17, 1170-1188. 389 Ellegren, H. and Galtier, N. 2016. Determinants of genetic diversity. Nature Reviews 390 Genetics, 17, 422-433. 391 Excoffier, L. and Lischer, H. E. L. 2010. Arlequin suite ver 3.5: a new series of programs to 392 perform population genetics analyses under Linux and Windows. Molecular Ecology 393 Resources, 10, 564-567. 394 Frankham, R. 1996. Relationship of genetic variation to population size in wildlife. 395 Conservation Biology, 10, 1500-1508. 396 Frankham, R. 2005. Genetics and extinction. Conservation Biology, 126, 131-140. 397 Frankham, R. 2012. How closely does genetic diversity in finite populations conform to 398 predictions of neutral theory? Large deficits in regions of low recombination. Heredity, 399 108, 167-178. 400 Giri, C., Ochieng, E., Tieszen, L. L., Zhu, Z., Singh, A., Loveland, T., Masek, J. and Duke, 401 N. 2011. Status and distribution of mangrove forest of the world using earth observation 402 satellite data. Global Ecology and Biogeography, 20, 154-159. 403 Hedrick, P. W. 2011. Genetics of populations. 4th edition. Jones & Bartlett Publishers, 404 Sudbury, Massachusetts. 405 139 Hengl, T., de Jesus, J. M., MacMillan, R. A., Batjes, N. H., Heuvelink, G. B. M., Ribeiro, 406 E… Gonzalez, M. R. 2014. SoilGrids1km — Global Soil Information Based on 407 Automated Mapping. PLoS ONE, 9, e105992. 408 Hijmans, R. J., Cameron, S. E., Parra, J. L., Jones, P. G. and Jarvis, A. 2005. Very high 409 resolution interpolated climates surfaces for global land areas. International Journal of 410 Climatology, 25, 1965-1978. 411 Hijmans, R. J., van Etten, J., Summer, M., Cheng, J., Bevan, A., Bivand, R… Wueest, R. 412 2019. Raster R package, Geographic data analysis and modelling. https://cran.r-413 project.org/web/packages/raster/raster.pdf 414 Hossain, M. D. and Nuruddin, A. A. 2016. Soil and Mangrove: A Review. Journal of 415 Environmental Science and Technology, 9, 198-207. 416 Hutchinson, G. E. 1957. Concluding remarks. Cold Spring Harbor Symposia on 417 Quantitative Biology, 22, 415-427. 418 Jordan, H. D. 1964. The relation of the vegetation and soil to development of mangroves 419 swamps for rice growing in Sierra Leone. Journal of Applied Ecology, 1, 209-212. 420 Krauss, K. W., Lovelock, C. E., McKee, K. L., Lopez-Hoffman, L., Ewe, S. M. L., Sousa, 421 W. P. (2008) Environmental drivers in mangrove establishment and early development: 422 a review. Aquatic Botany, 89, 105–127. 423 Lira-Noriega, A. and Manthey, J. D. 2014. Relationship of genetic diversity and niche 424 centrality: A survey and analysis. Evolution, 68, 1082-1093. 425 Maguire, B. l973. Niche response structure and the analytical potentials of its relationship 426 to the habitat. Global Ecology and Biogeography, 17, 145-151. 427 Manthey, J. D. Campbell, L. P., Saupe, E. E., Soberón, J., Hensz, C. M. Myers, C. E., 428 Owens, H. L., Ingenloff, K., Peterson, A. T., Barve, N., Lira-Noriega, A. and Barve, V. 429 2015. A test of niche centrality as determinant of population trends and conservation 430 status in threatened and endangered North American birds. Endangered Species 431 Research, 26, 201-208. 432 Martínez-Meyer, E., Díaz-Porras, D. F., Peterson, A. T. and Yañez-Arenas, C. 2013. 433 Ecological niche structure and rangewide abundance patterns of species. Biology 434 Letters, 9, 20120637. 435 140 McMillan, C. 1975. Interaction of soil texture with salinity tolerences of black mangrove 436 (Avicennia) and white mangrove (Laguncularia) from North America. In: Walsh, G.E., 437 Snedaker, S.C., and Teas, H.J., (Eds.), Proceedings of the International Symposium on 438 the Biology and Management of Mangroves, pp. 561-566, University of Florida, 439 Gainesvill. 440 Micheletti, S. J. and Storfer, A. 2015. A test of the central-marginal hypothesis using 441 population genetics and ecological niche modelling in an endemic salamander 442 (Ambystoma barbourin). Molecular Ecology, 24, 967-979. 443 Nettel, A. and Dodd, R. S. 2007. Drifting propagules and receding swamps: Genetic 444 footprints of mangrove recolonization and dispersal along tropical coasts. Evolution, 61, 445 958-971. 446 Ochoa-Zavala, M., Jaramillo-Correa, J. P., Piñero, D., Nettel-Hernanz, A. and Núñez-447 Farfán, J. 2019a. Contrasting colonization patterns of black mangrove (Avicennia 448 germinans (L.) L.) gene pools along the Mexican coasts. Journal of Biogeography, 46, 449 884–898. 450 Ochoa-Zavala, M., Jaramillo-Correa, J. P., Piñero, D., Nettel-Hernanz, A. and Núñez-451 Farfán, J. Data from: Contrasting colonization patterns of black mangrove (Avicennia 452 germinans (L.) L.) gene pools along the Mexican coasts. Dryad Digital Repository. 453 (2019b). doi:10.5061/dryad.4r0q7m5. 454 Osorio-Olvera, L., Barve, V., Barve, N., Soberón, J. and Falconi, M. 2019. Nichetoolbox: 455 from getting biodiversity data to evaluating species distribution models in a friendly 456 GUI environment R package version 0.4.5.2. https://luismurao.github.io/. 457 Osorio-Olvera, L., M. Falconi, and J. Soberón. 2016. Sobre la relación entre idoneidad del 458 hábitat y la abundancia poblacional bajo diferentes escenarios de dispersión. Revista 459 Mexicana de Biodiversidad 87:1080-1088. 460 Pfeifer, M., Schatz, B., Picó, F. X., Passalacqua, N.G., Fay, M. F., Carey, P. D. and Jeltsch, 461 F. 2009. Phylogeography and genetic structure of the orchid Himantoglossum hircinum 462 (L.) Spreng. across its European central–marginal gradient. 36, 2353-2365. 463 Pironon, S., Villellas, J., Morris, W. F., Doak, D. F. and García, M. B. 2015. Do 464 geographic, climatic or historical ranges differentiate the performance of central versus 465 peripheral populations? Global Ecology and Biogeography, 24, 611-620. 466 141 Qiao, H., Escobar, L. E., Saupe, E. E., Ji, L. and Soberón, J. 2016. A cautionary note on the 467 use of hypervolume kernel density estimators in ecological niche modelling. Global 468 Ecology and Biogeography, 26, 1066-1070. 469 Record, S., Charney, N. D., Zakaria, R. M. and Ellison, A. M. 2013. Projecting global 470 mangrove species and community distributions under climate change. Ecosphere. 4, 34. 471 Reed, D. H. and Frankham, R. 2003. Correlation between fitness and genetic diversity. 472 Conservation Biology, 17, 230-237. 473 Santini, L., Pironon, S., Maiorano, L. and Thuiller, W. 2018. Addressing common pitfalls 474 does not provide more support to geographical and ecological abundance-centre 475 hypotheses. Ecography, 42, 696-705. 476 Spielman, D., Brook, B. W. and Frankham, R. 2004. Most species are not driven to 477 extinction before genetic factors impact them. PNAS, 101, 15261-15264. 478 Tomlinson, P.B. 1986. The botany of mangroves. Cambridge, UK. Cambridge University 479 Press. 419 pp. 480 Tomlinson, P.B. 2016. The botany of mangroves. 2nd Edition. Cambridge, UK. Cambridge 481 University Press. 419 pp. 482 Ukpong, I, E. 1997. Vegetation and its relation to soil nutrient and salinity in the Calabar 483 mangrove swamp, Nigeria. Mangroves and Salt Marshes, 1, 211-218. 484 Van Aelst, S. and Rousseeuw, P. 2009. Minimum volume ellipsoid. Advanced Review, 1, 485 71-82. 486 Wiens, J. A., Stralberg, D., Jongsomijit, D., Howell, C. A. and Snyder, M. A. 2009. Niches, 487 model, and climate change: Assessing the assumptions and uncertainties. PNAS, 106, 488 19729-19736. 489 Woodroffe, C. 1992. Mangroves sediments and geomorphology. In: Robertson, A. I. and 490 Alongi, D. M. (Eds.), Tropical mangrove ecosystem, pp 7-41, vol. 41, American 491 Geophysical Union, Washington, DC. USA. 492 Woodroffe, C. D. and Grindrod, J. (1991) Mangrove biogeography—the role of quaternary 493 environmental and sea-level change. Journal of Biogeography, 18, 479–492. 494 Woodward, F. I. and Williams, B. G. 1987. Climate and plant distribution at global and 495 local scales. Vegetatio, 69, 189-197. 496 142 Yañez-Arenas, C., Martínez-Meyer, E., Mandujano, S. and Rojas-Soto, O. 2012. Modelling 497 geographic patterns of population density of the white-tailed deer in central Mexico by 498 implementing ecological niche theory. Oikos, 121, 2081-2089. 499 .500 1 4 3 F ig u res 5 0 1 5 0 2 5 0 3 F ig u re 1 . R elatio n sh ip b etw een g en etic d iv ersity o f A vicen n ia g erm in a n s an d d istan ce to th e n ich e cen tro id . L eft p an el sh o w th e 5 0 4 relatio n sh ip b etw een o b serv ed an d ex p ected h etero zy g o sity an d n ich e-cen tro id d istan ce. T h e m ap is co lo red as d istan ce to th e n ich e 5 0 5 cen tro id , w ith p in k to n es rep resen tin g fu rth er d istan ce fro m n ich e cen tro id . P o in ts in d icate g eo g rap h ic p o sitio n o f p o p u latio n s w ith 5 0 6 g en etic estim ates. P o p u latio n s n am es, lo catio n an d estim ates o f g en etic d iv ersity are av ailab le in T ab le 1 . 5 0 7 1 4 4 5 0 8 F ig u re 2 . P red ictio n s o f th e sp atial v ariatio n o f g en etic d iv ersity in A vicen n ia g erm in a n s (left p an el). R elatio n sh ip s b etw een th e 5 0 9 o b serv ed an d sim u lated d ata o f o b serv ed an d ex p ected h etero zy g o sity (rig h t p an el). M ap s sh o w s th e p red ictio n o f o b serv ed an d 5 1 0 ex p ected h etero zy g o sity fo r each p ix el. P o p u latio n s d etails av ailab le in T ab le 1 . 5 1 1 145 512 513 Figure 3. Environmental variables included in the best fitting models for observed and 514 expected heterozygosity. Black stars indicate the most shared variables between models 515 516 1 4 6 T a b le 1 . P o p u latio n s n am es, g eo g rap h ic co o rd in ates an d estim ates o f o b serv ed (H o ) an d ex p ected h etero zy g o sity (H e) co m p u ted fro m 5 1 7 m icro satellite d ata. 5 1 8 N o . N a m es o f p o p u la tio n s S a m p le size L o n g itu d e L a titu d e H o sta n d a rd d ev ia tio n H e sta n d a rd d ev ia tio n R eferen ce 1 E l S arg en to 4 0 -1 1 2 .3 2 9 1 8 2 9 .3 2 8 5 1 7 0 .0 5 0 0 .1 4 0 O ch o a-Z av ala et al., 2 0 1 9 a 2 B ah ía C o n cep ció n 4 0 -1 1 1 .8 8 1 5 8 2 6 .7 6 2 1 3 3 0 .0 3 2 0 .0 1 1 0 .0 4 7 0 .0 2 4 O ch o a-Z av ala et al., 2 0 1 9 a 3 T o p o lo b am p o 4 0 -1 0 9 .1 2 3 5 3 2 5 .5 8 2 1 9 1 0 .1 4 1 0 .1 1 4 0 .1 4 2 0 .1 O ch o a-Z av ala et al., 2 0 1 9 a 4 B ah ía M ag d alen a 4 0 -1 1 2 .1 1 5 6 5 2 4 .7 9 4 1 8 3 0 .1 1 7 0 .1 9 4 0 .1 5 9 0 .1 9 2 O ch o a-Z av ala et al., 2 0 1 9 a 5 L a P az 4 0 -1 1 0 .3 3 2 8 2 4 .1 7 0 9 0 .2 4 1 0 .1 8 0 .2 8 7 0 .1 9 5 O ch o a-Z av ala et al., 2 0 1 9 a 6 B ah ía d e A ltata 4 0 -1 0 7 .8 6 9 8 3 2 4 .6 1 5 6 6 7 0 .2 0 2 0 .1 6 1 0 .2 9 1 0 .2 2 1 O ch o a-Z av ala et al., 2 0 1 9 a 7 E l C aim an ero 4 0 -1 0 6 .0 6 2 5 6 2 2 .8 8 2 2 0 8 0 .1 6 3 0 .1 4 3 0 .1 7 7 0 .1 3 8 O ch o a-Z av ala et al., 2 0 1 9 a 8 M arism as N acio n ales 4 0 -1 0 5 .6 4 0 3 7 2 2 .3 7 4 4 6 7 0 .2 1 4 0 .1 7 6 0 .2 7 6 0 .2 3 3 O ch o a-Z av ala et al., 2 0 1 9 a 9 P u n ta P éru la 4 0 -1 0 5 .1 2 5 1 9 1 9 .5 9 2 0 2 7 0 .3 0 7 0 .2 2 1 0 .3 3 1 0 .2 O ch o a-Z av ala et al., 2 0 1 9 a 1 0 B arra d e N av id ad 4 0 -1 0 4 .6 7 2 2 5 1 9 .1 9 5 5 3 2 0 .1 3 2 0 .2 3 4 0 .1 4 5 0 .1 7 1 O ch o a-Z av ala et al., 2 0 1 9 a 1 1 T écp an d e G alean a 4 0 -1 0 0 .7 9 2 1 1 1 7 .2 1 3 9 3 5 0 .4 3 7 0 .1 8 0 .4 4 3 0 .1 6 O ch o a-Z av ala et al., 2 0 1 9 a 1 2 B arra d e T eco an ap a 4 0 -9 8 .7 2 4 1 3 3 1 6 .4 9 7 8 0 .2 7 0 .2 5 5 0 .2 6 9 0 .2 3 O ch o a-Z av ala et al., 2 0 1 9 a 1 3 L ag u n as d e C h acah u a 4 0 -9 7 .5 9 1 8 3 3 1 6 .0 1 1 4 0 .3 5 8 0 .1 4 8 0 .4 0 5 0 .1 9 7 O ch o a-Z av ala et al., 2 0 1 9 a 1 4 M ar M u erto 4 0 -9 3 .8 4 6 0 3 3 1 6 .0 1 0 4 0 .3 4 9 0 .2 9 3 0 .4 0 5 0 .2 9 5 O ch o a-Z av ala et al., 2 0 1 9 a 1 5 L a E n cru cijad a 4 0 -9 2 .8 3 1 0 1 7 1 5 .1 6 8 3 6 7 0 .4 7 6 0 .2 8 6 0 .4 9 8 0 .2 7 2 O ch o a-Z av ala et al., 2 0 1 9 a 1 6 M o n tijo G u lf 4 0 -8 1 .0 5 6 7 4 9 7 .8 2 2 9 5 6 0 .4 9 5 0 .2 1 3 0 .6 4 9 0 .1 4 8 C eró n -S o u za et al., 2 0 1 2 1 7 Ju an D íaz 1 4 -7 9 .4 4 0 1 0 7 9 .0 2 0 0 5 1 0 .6 2 3 0 .2 8 6 0 .7 0 .1 9 4 C eró n -S o u za et al., 2 0 1 5 1 8 S an M ig u el G u lf 4 0 -7 8 .2 5 7 1 8 .3 9 3 6 1 0 .3 7 2 0 .2 9 5 0 .4 5 1 0 .3 1 1 C eró n -S o u za et al., 2 0 1 2 1 4 7 1 9 R ío B rav o 4 0 -9 7 .1 5 7 9 5 2 5 .9 4 6 1 3 3 0 .2 3 0 .1 7 8 0 .2 9 6 0 .2 1 7 O ch o a-Z av ala et al., 2 0 1 9 a 2 0 L a P esca 4 0 -9 7 .7 4 1 4 5 2 3 .7 7 6 0 8 3 0 .2 1 2 0 .1 8 2 0 .2 4 5 0 .1 9 9 O ch o a-Z av ala et al., 2 0 1 9 a 2 1 T u x p an 4 0 -9 7 .3 4 1 5 2 0 .9 8 0 7 0 .4 3 6 0 .1 5 7 0 .4 1 3 0 .1 6 4 O ch o a-Z av ala et al., 2 0 1 9 a 2 2 L a M an ch a 3 4 -9 6 .3 8 4 7 8 3 1 9 .5 8 6 1 6 7 0 .2 7 4 0 .2 0 1 0 .3 0 9 0 .2 4 O ch o a-Z av ala et al., 2 0 1 9 a 2 3 A lv arad o 4 1 -9 5 .7 8 2 2 5 1 8 .7 3 2 8 5 0 .2 3 1 0 .2 0 9 0 .2 7 1 0 .2 O ch o a-Z av ala et al., 2 0 1 9 a 2 4 S o n teco m ap an 3 4 -9 5 .0 2 4 7 1 7 1 8 .5 0 8 4 1 7 0 .2 9 5 0 .2 5 2 0 .2 8 0 .2 2 1 O ch o a-Z av ala et al., 2 0 1 9 a 2 5 C o atzaco alco s 3 9 -9 4 .4 3 1 3 8 3 1 8 .0 9 7 1 0 .2 8 7 0 .2 5 8 0 .3 0 7 0 .2 2 8 O ch o a-Z av ala et al., 2 0 1 9 a 2 6 M eco acán 3 5 -9 3 .1 1 5 4 5 5 1 8 .4 1 9 6 2 6 0 .4 1 2 0 .1 9 0 .4 2 5 0 .1 9 6 O ch o a-Z av ala et al., 2 0 1 9 a 2 7 P o m -A tasta 4 0 -9 2 .1 0 4 1 8 .5 7 4 5 0 .5 1 4 0 .1 6 2 0 .4 6 3 0 .1 2 1 O ch o a-Z av ala et al., 2 0 1 9 a 2 8 L o s P eten es 4 1 -9 0 .3 7 1 5 8 3 2 0 .1 5 8 9 1 7 0 .4 3 5 0 .2 5 6 0 .4 5 2 0 .2 4 O ch o a-Z av ala et al., 2 0 1 9 a 2 9 R ío L ag arto s 4 0 -8 8 .1 2 8 2 2 1 .6 1 0 2 0 .3 3 3 0 .2 7 3 0 .3 3 6 0 .2 1 4 O ch o a-Z av ala et al., 2 0 1 9 a 3 0 Y u m b alam 4 0 -8 7 .2 0 2 9 6 7 2 1 .4 4 3 5 8 3 0 .3 5 8 0 .2 7 5 0 .3 1 6 0 .2 1 7 O ch o a-Z av ala et al., 2 0 1 9 a 3 1 C o zu m el 4 0 -8 6 .9 1 3 6 3 3 2 0 .5 5 5 6 3 3 0 .4 9 1 0 .3 0 .4 6 9 0 .2 0 2 O ch o a-Z av ala et al., 2 0 1 9 a 3 2 S ian K a'an 4 0 -8 7 .4 7 0 9 1 7 2 0 .1 1 0 9 6 7 0 .4 9 3 0 .1 6 9 0 .5 0 .1 5 3 O ch o a-Z av ala et al., 2 0 1 9 a 3 3 F lo rid a 2 1 -8 2 .7 5 2 8 .2 5 0 .4 6 0 .2 1 2 0 .5 4 8 0 .2 2 5 C eró n -S o u za et al., 2 0 1 5 3 4 P u erto R ico 1 6 -6 5 .7 8 1 8 .4 1 0 .3 7 8 0 .2 0 .4 5 8 0 .1 7 C eró n -S o u za et al., 2 0 1 5 3 5 G u ad elo u p e 1 8 -6 1 .4 7 1 6 .2 0 .3 4 6 0 .3 0 1 0 .3 6 2 0 .1 9 4 C eró n -S o u za et al., 2 0 1 5 3 6 T rin id ad 1 9 -6 1 .0 2 2 2 1 1 0 .3 9 3 2 0 .5 9 3 0 .1 7 1 0 .7 1 0 .2 0 1 C eró n -S o u za et al., 2 0 1 5 3 7 F ren ch G u ian a 1 4 -5 2 .6 3 5 .1 5 0 .7 0 7 0 .1 0 6 0 .7 5 8 0 .1 0 5 C eró n -S o u za et al., 2 0 1 5 3 8 H o n d u ras 1 3 -8 3 .7 1 0 4 7 1 5 .3 8 8 2 5 0 .5 5 2 0 .2 1 2 0 .5 3 8 0 .2 1 3 C eró n -S o u za et al., 2 0 1 5 1 4 8 3 9 B o cas d el T o ro 4 0 -8 2 .1 9 5 9 9 9 .1 5 5 4 8 0 .5 0 1 0 .1 2 1 0 .6 2 8 0 .1 4 3 C eró n -S o u za et al., 2 0 1 2 4 0 G aleta 4 0 -7 9 .8 4 1 8 1 5 9 .3 9 8 9 3 0 .7 3 1 0 .1 5 4 0 .7 6 3 0 .0 7 2 C eró n -S o u za et al., 2 0 1 2 5 1 9 149 APPENDIX 2 Table S1. Results of the model fitting for the observed and expected heterozygosity with P < 0.001. Bold letters show the combination of variables we used to project an approximation of the fundamental niche of Avicennia germinans on the geographic space. No Set of environmental variables p-value rho Observed heterozygosity 1 BDRICM, bio_08, CLYPPT_M_sl1 0.00076109 -0.542475 2 BDRICM, bio_08, CLYPPT_M_sl2 0.00067915 -0.5466769 3 bio_01, bio_03, CLYPPT_M_sl1 0.00022695 -0.5844947 4 bio_01, bio_03, CLYPPT_M_sl2 0.00027499 -0.5781918 5 bio_01, bio_04, CLYPPT_M_sl1 0.00098358 -0.5328104 6 bio_01, bio_04, CLYPPT_M_sl2 0.00096571 -0.5335108 7 bio_01, bio_08, bio_09 0.00021377 -0.5793166 8 bio_01, bio_08, CLYPPT_M_sl1 0.0006896 -0.5461167 9 bio_01, bio_08, CLYPPT_M_sl2 0.00053623 -0.555221 10 bio_01, bio_08, CLYPPT_M_sl3 0.00089378 -0.5364521 11 bio_01, bio_09, CLYPPT_M_sl1 0.00016851 -0.5940192 12 bio_01, bio_09, CLYPPT_M_sl2 0.00017537 -0.5927586 13 bio_01, bio_09, CLYPPT_M_sl3 0.0001746 -0.5928987 14 bio_01, bio_09, CLYPPT_M_sl4 0.00093085 -0.5349114 15 bio_01, bio_09, CLYPPT_M_sl5 0.00064863 -0.5483577 16 bio_01, bio_09, CLYPPT_M_sl7 0.00068698 -0.5462567 17 bio_01, bio_11, CLYPPT_M_sl2 0.00074686 -0.5431753 18 bio_01, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00072657 -0.5441939 19 bio_03, bio_08, CLYPPT_M_sl1 0.00069754 -0.5456965 20 bio_03, bio_08, CLYPPT_M_sl2 0.00064863 -0.5483577 21 bio_03, bio_08, CLYPPT_M_sl3 0.00094814 -0.5342111 22 bio_03, bio_09, CLYPPT_M_sl1 0.00083594 -0.5389733 23 bio_03, bio_09, CLYPPT_M_sl2 0.0005935 -0.5515792 24 bio_03, bio_11, CLYPPT_M_sl1 0.00065866 -0.5477975 25 bio_03, bio_11, CLYPPT_M_sl2 0.00066884 -0.5472372 26 bio_04, bio_09, CLYPPT_M_sl1 0.00048777 -0.5585825 27 bio_04, bio_09, CLYPPT_M_sl2 0.00048391 -0.5588627 28 bio_04, CLYPPT_M_sl6, CLYPPT_M_sl7 0.0008971 -0.5363121 29 bio_06, bio_08, CLYPPT_M_sl1 0.00096218 -0.5336508 30 bio_06, bio_08, CLYPPT_M_sl2 0.00089378 -0.5364521 31 bio_06, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00058435 -0.5521395 32 bio_08, bio_09, CLYPPT_M_sl2 0.00097998 -0.5329505 33 bio_08, bio_11, CLYPPT_M_sl2 0.00067398 -0.5469571 150 34 bio_08, CLYPPT_M_sl1, CLYPPT_M_sl6 0.00066628 -0.5473773 35 bio_08, PHIHOX_M_sl5, PHIHOX_M_sl6 0.00064863 -0.5483577 36 bio_08, PHIHOX_M_sl5, PHIHOX_M_sl7 0.0008485 -0.5384131 37 bio_09, bio_11, CLYPPT_M_sl1 0.00010788 -0.6077456 38 bio_09, bio_11, CLYPPT_M_sl2 0.00011512 -0.6057847 39 bio_09, bio_11, CLYPPT_M_sl3 0.00020364 -0.5879964 40 bio_09, bio_11, CLYPPT_M_sl4 0.00039589 -0.565866 41 bio_09, bio_11, CLYPPT_M_sl5 0.00026136 -0.5798725 42 bio_09, bio_11, CLYPPT_M_sl6 0.00027967 -0.5776315 43 bio_09, bio_11, CLYPPT_M_sl7 0.00015973 -0.5957 44 bio_11, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00028925 -0.576511 45 CLYPPT_M_sl1, CLYPPT_M_sl6, CLYPPT_M_sl7 2.00E-05 -0.6263097 46 CLYPPT_M_sl2, CLYPPT_M_sl6, CLYPPT_M_sl7 6.39E-05 -0.5953333 47 CLYPPT_M_sl3, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00011971 -0.5771839 Expected heterozygosity 48 bio_01, bio_03, CLYPPT_M_sl1 0.00046134 -0.5605435 49 bio_01, bio_03, CLYPPT_M_sl2 0.00049753 -0.5578822 50 bio_01, bio_08, bio_09 0.00080085 -0.5337538 51 bio_01, bio_09, CLYPPT_M_sl1 0.00022793 -0.5843546 52 bio_01, bio_09, CLYPPT_M_sl2 0.00029538 -0.5758106 53 bio_01, bio_09, CLYPPT_M_sl3 0.00023189 -0.5837944 54 bio_01, bio_09, CLYPPT_M_sl4 0.00096218 -0.5336508 55 bio_01, bio_09, CLYPPT_M_sl5 0.00048392 -0.5588627 56 bio_01, bio_09, CLYPPT_M_sl7 0.00063873 -0.548918 57 bio_01, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00048051 -0.5591119 58 bio_03, bio_09, CLYPPT_M_sl2 0.00099808 -0.5322502 59 bio_08, PHIHOX_M_sl5, PHIHOX_M_sl6 0.00067398 -0.5469571 60 bio_08, PHIHOX_M_sl5, PHIHOX_M_sl7 0.00061219 -0.5504587 61 bio_09, bio_11, CLYPPT_M_sl1 0.00055765 -0.5538203 62 bio_09, bio_11, CLYPPT_M_sl2 0.00028683 -0.5767911 63 bio_09, bio_11, CLYPPT_M_sl3 0.00054899 -0.5543806 64 bio_09, bio_11, CLYPPT_M_sl5 0.00046876 -0.5599832 65 bio_09, bio_11, CLYPPT_M_sl6 0.00099443 -0.5323902 66 bio_09, bio_11, CLYPPT_M_sl7 0.00052167 -0.5562014 67 bio_11, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00032109 -0.5730093 68 CLYPPT_M_sl1, CLYPPT_M_sl6, CLYPPT_M_sl7 4.44E-05 -0.6054057 69 CLYPPT_M_sl2, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00013528 -0.5735182 70 CLYPPT_M_sl3, CLYPPT_M_sl6, CLYPPT_M_sl7 0.00022519 -0.5577488 151 BDRICM = Deep to the bed rock; CLYPPT_ M_sl1 = Mean percentage of the clay particles at deep of 0 m; CLYPPT_ M_sl2 = Mean percentage of the clay particles at deep of -0.05 m; CLYPPT_ M_sl3 = Mean percentage of the clay particles at deep of -0.15 m; CLYPPT_ M_sl4 = Mean percentage of the clay particles at deep of -0.30 m; CLYPPT_ M_sl5 = Mean percentage of the clay particles at deep of -0.60 m; CLYPPT_ M_sl6 = Mean percentage of the clay particles at deep of -1.0 m; CLYPPT_ M_sl7 = Mean percentage of the clay particles at deep of -2.0 m; PHIHOX_M_sl5 = pH index at standard deep of -0.60 m; PHIHOX_M_sl6 = pH index at standard deep of -1.0 m; PHIHOX_M_sl7 = pH index at standard deep of -2.0 m; bio_01 = Annual mean temperature; bio_03 = Isotermality; bio_04 = Temperature seasonality; bio_06 = Mean temperature of the coldest month; bio_08 = Mean temperature of the wettest quarter; bio_09 = Mean temperature of the driest quarter; bio_11 = Mean temperature of the coldest quarter. 1 5 2 F ig u re S 2 .1 . R an g e v alu es reco rd ed in p o p u latio n s clo se to th e cen tro id -n ich e (cen tre) an d p o p u latio n s farth er fro m th e cen tral-n ich e (p erip h eral). R ed o u tlin e refers to en v iro n m en tal v ariab les th at u sed to p ro ject th e fu n d am en tal n ich e o f A v ic e n n ia g e rm in a n s; b lack o u tlin e sh o w s th e v alu es o f o th er relev an t v ariab les fo r A v ic e n n ia g e rm in a n s. 153 Discusión General La distribución de los alelos y los genotipos en el espacio y tiempo es afectada por numerosos factores. En este trabajo evaluamos la variación genética de Avicennia germinans con microsatélites nucleares dentro y entre poblaciones a lo largo de las costas del Pacífico y del Atlántico en México, para dilucidar los factores históricos y contemporáneos que han modelado la distribución de su variación genética. Contrario a lo que podríamos pensar, y tomando en cuenta que los linajes de ambas costas se distribuyen más o menos en las mismas latitudes, hemos encontrado importantes diferencias en la estructura filogeográfica de A. germinans en México; lo cual sugiere diferencias en sus dinámicas demográficas. Con base en lo que se sabe de la evolución reciente de A. germinans y nuestros resultados, hemos planteado varias líneas de investigación para ampliar estos resultados que detallamos a continuación. Predicción de la respuesta de ambos linajes frente a los cambios climáticos La estructura genética de Avicennia germinans está organizada en diferentes escalas geográficas con relativamente poco flujo genético entre poblaciones. Los linajes Pacífico y Atlántico se separaron hace 0.75 Ma y a pesar de que la diferenciación con microsatélites es muy alta (FCT = 0.878 basa en RST, Tabla 1), nuestros resultados mostraron cierta ancestría compartida. Recientemente, se ha observado un patrón similar en haplotipos de cloroplasto (Mori et al., 2015a), lo que sugiere que la separación entre linajes no es absoluta. Ahora bien, aunque ambas costas muestran signos de subestructura, también se observan diferencias en la profundidad de la divergencia dentro de estos grupos, siendo mucho más conspicua en la costa Pacífica que en la costa Atlántica. Estas diferencias pueden deberse a que ambos linajes responden de manera distinta a los cambios climáticos, donde la magnitud de los mismos también varía geográficamente, y a la composición genética de las poblaciones locales (Sork et al., 2010). Para entender mejor las respuestas a los cambios climáticos se requiere de un mayor conocimiento de los patrones geográficos adaptativos. Una de las formas de identificar locus que estén posiblemente bajo selección es buscar aquellos que están fuertemente asociados a un gradiente ambiental y detectar cambios en sus frecuencias alélicas (Manel et 154 al., 2010; Jones et al., 2013). Esta información ayuda a entender mejor cómo las poblaciones se adaptan localmente y su potencial para responder a los cambios climáticos futuros. Actualmente, están disponibles varios métodos que nos permiten evaluar si los datos ambientales tienen una estructura de agrupamiento, es decir, si los datos que corresponden a un mismo grupo climático son diferentes con respecto a otro grupo (Mirkin, 2016; Osorio- Olvera et al., en revisión). Recientemente, pocos estudios han asociado información de variación genética en la modelación de nicho ecológico, dichos estudios han demostrado tener el poder de predecir las respuestas de grupos genéticamente distintos frente a varios escenarios de cambio climático (e. g. Ikeda et al., 2017; Marcer et al., 2016). Por lo tanto, si asociamos la estructura genética adaptativa (en lugar de la estructura neutral) en el análisis de nicho ecológico y modelamos los cambios en las condiciones ambientales futuras de dichos grupos adaptativos, podremos predecir con mayor exactitud la respuesta de ambos linajes. Debido a la vulnerabilidad de los manglares frente al cambio climático, varias predicciones se han realizado referente al impacto que pueden tener los cambios en los patrones de lluvia, salinidad y temperatura en el futuro próximo (al año 2080; Record et al., 2013; Alongi, 2015). En general, se espera una contracción de la distribución de los manglares y una severa reducción en la riqueza de especies en el Caribe y Centroamérica (Record et al., 2013; Alongi, 2015). Debido al incremento en la salinidad y al aumento en las temperaturas máximas, se predice una disminución de la cobertura de los manglares en las costas de Baja California, Sonora, Sinaloa, Tamaulipas y gran parte de Veracruz (Alongi, 2015). Mientras que en el Caribe, se predice una disminución del bosque de mangle por un aumento en el nivel del mar. Sin embargo, en América del Norte se espera una expansión de sus límites latitudinales hacia el norte en ambas costas. Los patrones de introgresión e hibridación en Avicennia Un resultado muy interesante es que a pesar de que la costa Pacífica es en promedio menos diversa que la del Atlántico, poblaciones como la Encrucijada en el estado de Chiapas (PS24) presentan altos niveles de diversidad genética, comparables con su contraparte, Sian Ka’an, o con cualquiera del Atlántico mexicano, incluso con las poblaciones genéticamente más diversas en Panamá (Cerón-Souza et al., 2012). 155 Tabla 1. Resultados del análisis de varianza molecular (AMOVA) de Avicennia germinans entre costas (poblaciones del Pacífico vs. Atlántico de México) basadas en RST (modelo de mutación por pasos) y FST (modelo de alelos infinitos); P < 0.001. Fuente de variación Estadísti co F Porcentaje de variación Estadístic o F Porcentaje de variación RST FST Entre costas (Pacífico vs. Atlántico) FCT = 0.88 75.23 FCT = 0.63 62.74 Entre poblaciones dentro de costas FSC = 0.45 11.06 FSC = 0.24 9.07 Entre individuos dentro de poblaciones FIS = 0.11 1.5 FIS = 0.06 1.81 Dentro de individuos FIT = 0.88 12.22 FIT = 0.74 26.38 Estudios anteriores en América del Sur han sugerido una posible hibridación entre A. germinans y A. bicolor en la zona de simpatría (Sureste de México (Chiapas) hasta Panamá) como posible explicación a la alta diversidad genética de esta zona. Nettel et al. (2008) encontraron evidencia de hibridación con secuencias de espaciadores internos transcritos (ITS), pero no a nivel nuclear en marcadores microsatélites y tampoco en ADN de cloroplasto, por lo que la explicación más plausible fue una diferenciación intraespecífica en antiguos sitios refugio (Cerón-Souza et al., 2005, 2012; Nettel et al., 2008), apoyando el patrón que encontramos en el Pacífico de México. Aunque el desarrollo de los marcadores genéticos ha facilitado los estudios de hibridación e introgresión debido a su relativo poder para identificar dichos patrones, varias consideraciones se deben tener en cuenta. Estudios clásicos de hibridación combinan diferentes marcadores genéticos con distintas tasas de mutación y modos de herencia. Uno de los marcadores más populares para inferir relaciones filogenéticas entre taxones son los espaciadores internos transcritos de los genes ribosomales. El arreglo del ADN ribosomal del núcleo de un genoma eucariótico consiste en cientos de copias repetidas en tándem. Una de las características que ha sobresalido en estos genes es que están bajo evolución concertada, por lo que las copias individuales en el genoma mantienen secuencias muy similares (Álvarez y Wendel, 2003). Sin embargo, diferentes estudios han reportado individuos con copias polimórficas de ITS, por lo que se ha sugerido que la evolución concertada es incompleta en varios grupos de plantas (Xu et al., 2017). Esto es relevante porque la ausencia de una completa homogeneización de las copias repetidas 156 conducirá a una (malinterpretada) incongruencia filogenética (Álvarez y Wendel, 2003; Xu et al., 2017), que es exactamente el mismo patrón usado para inferir hibridación (Linder y Rieseberg, 2004). Por otro lado, los marcadores de orgánulos celulares como el cloroplasto, con una tasa de mutación menor a los genes nucleares, de herencia predominantemente uniparental (materna en la mayoría de los angiospermas) y su naturaleza no recombinante, proveen de información adicional a los estudios de hibridación (e. g. dirección de hibridación). Sin embargo, cuando la hibridación es rara, muy reciente o de baja magnitud, la identificación de individuos que lleven ese fragmento incorporado se restringe (Rosenzweig et al., 2016), por lo que se debe incluir suficientes individuos en el muestreo para detectar capturas raras de ADN de cloroplasto en el rango de distribución de la especie (Twyford y Ennos, 2012). En el caso de los microsatélites, empleados para identificar eventos más recientes de hibridación (e. g. Mori et al., 2015b), estos poseen una alta tasa de mutación por lo que están sujetos a homoplasia, lo que difumina la señal de hibridación. Si bien nuestros resultados con microsatélites probablemente no han sido afectados por eventos de hibridación – por lo que no afecta nuestras conclusiones – el grado y el patrón de flujo genético interespecífico entre especies del género Avicennia en el nuevo mundo aún no se han estudiado. Para esto, la tecnología de secuenciación de próxima generación promete mejorar enormemente nuestra comprensión acerca de las consecuencias evolutivas de los eventos de hibridación e introgresión, y nuestra habilidad para detectarla. Una buena estrategia que nos permite obtener resultados más concluyentes es estudiar la variación de polimorfismos de un solo nucleótido (SNP) distribuidos a lo largo del genoma y combinar la genómica de poblaciones con métodos de comparación de modelos evolutivos. Para determinar la extensión geográfica del flujo genético interespecífico se deben incluir individuos colectados en la zona de simpatría y extenderse más allá del límite norte y sur de la zona de contacto. Además, la genómica de poblaciones nos permite evaluar la estructuración que hay entre las especies, el grado de diferenciación y consecuentemente, estimar la tasa de flujo genético entre las mismas. Uno de los argumentos que se contraponen frente a los escenarios de hibridación e introgresión, son los polimorfismos compartidos y la separación incompleta de linajes (Twyford y Ennos, 2012; Linder y Rieseberg, 2004). Los análisis ABC (Approximated Bayesian Computation; Cornuet et al., 2014) nos permiten evaluar el ajuste de nuestros datos a distintos escenarios, por ejemplo, un modelo de 157 aislamiento con admixia frente a un escenario donde los linajes se hayan separado al mismo tiempo (e. g. Baena-Díaz et al 2018). ¿Las tasas de entrecruzamiento pueden ser influenciadas por la distribución espacial de Avicennia germinans? Diferentes estudios han sugerido que el flujo genético de A. germinans mediado por el propágulo es similar al del polen (Cerón-Souza et al., 2012; Mori et al., 2015a). El patrón de flujo genético obtenido a partir de las tasas de migración inferidas con BayesAss sugieren que las poblaciones del Atlántico están mejor conectadas entre sí que las poblaciones del Pacífico, explicando en parte los niveles de diversidad genética de esta costa. Uno de los argumentos que apoya el sistema de entrecruzamiento de A. germinans es la protandria (los órganos masculinos maduran antes que los femeninos). Sin embargo, la ausencia de autopolinización (autogamia) no es equivalente a la autoincompatibilidad, porque la polinización puede ocurrir entre diferentes flores de la misma planta (geitonogamia), un patrón que se ha encontrado en A. marina, A. officinalis y A. schaueriana (Clarke y Myerscough, 1991; Cerón-Souza et al., 2012; Nadia et al., 2013). Además, existen varios factores que afectan el movimiento del polinizador y, por lo tanto, la tasa de intercambio de polen dentro y entre flores. Entre estos factores están la densidad de las plantas con flores y el número de flores abiertas (Makino et al., 2007); así, los parches más densos pueden ser más atractivos para los polinizadores y facilitar el intercambio de polen entre flores de diferentes plantas, sin embargo, el incremento en el atractivo floral puede conducir también a la geitonogamia o al apareamiento con parientes cercanos. Además, se ha observado que las abejas visitan más frecuentemente aquellos sitios donde la densidad y la oferta floral es mayor (Nattero et al., 2011). La relación planta-polinizador tiene un impacto directo en la variación genética de A. germinans. Comparado con el Atlántico, en la costa Pacífico la distribución espacial de A. germinans es menos densa y discontinua, dispuesta en parches que varían en tamaño (Valderrama-Landeros, 2017). Por lo tanto, al haber diferencias en la distribución espacial de A. germinans, es posible que la tasa de intercambio de polen entre poblaciones difiera entre costas, por lo que las poblaciones del Atlántico pueden tener cierta tendencia a la fertilización cruzada más que a la endogamia biparental o autofecundación, 158 incrementando la diversidad genética y disminuyendo la diferenciación y la estructura espacial, lo cual explica la variación y estructura de esta costa. Una de las formas de probar esta hipótesis es estimando las tasas de entrecruzamiento de diferentes árboles madre con su respectiva progenie, en al menos una población muestreada en el norte, centro y sur de cada costa. La comparación directa entre los genotipos maternos y los genotipos de la progenie – usualmente a partir de microsatélites como en Nettel-Hernanz et al. (2013) y Mori et al. (2015b) – es usada para estimar la media de la tasa de entrecruzamiento y el coeficiente de endogamia de los individuos adultos (árboles madre) en una población (Ritland, 2002). Debido a la poca variación genética en el Pacífico Norte, es necesario buen muestreo del genoma El uso de marcadores como RAD-Seq pueden proporcionar mayor resolución de los parámetros del sistema de apareamiento por individuo, de tal manera que se puede determinar si la descendencia es producto de entrecruza o autocruzamiento (Colicchio et al., 2019). Adicionalmente, realizar cruzas manuales en unas 60 flores de 10 individuos diferentes nos ayudará a probar la hipótesis de autoincompatibilidad en A. germinans, de los cuales podemos usar para autopolinización, fertilización cruzada y grupos control, 30 flores cada uno (e. g. Nadia et al., 2013). La dispersión a larga distancia y el hábitat Poblaciones en diferentes partes del rango de distribución de la especie experimentan diferentes condiciones ambientales; algunas pueden permanecer bajo condiciones óptimas mientras que otras pueden experimentar condiciones ambientales más extremas (Sork et al., 2010). La relación entre las distancias al centroide del nicho de A. germinans y la diversidad genética mostró que hay pocas poblaciones con las condiciones adecuadas y que albergan alta diversidad genética, ubicadas la mayoría de ellas, dentro de la costa Atlántica. De hecho, más del 65 % de las poblaciones del Pacífico mexicano se ubican en la periferia del nicho fundamental de la especie (Fig. 3). El 80 % de estas poblaciones ambientalmente periféricas se ubican en el norte del país (e. g. Sonora, Sinaloa y Baja California), coincidiendo con las poblaciones que se habrían establecido durante el calentamiento del Holoceno (Capítulo 1). Este resultado resalta la importancia de la calidad del hábitat como un eje importante en el entendimiento de la distribución de la variabilidad genética en A. germinans. Bajo un 159 escenario de expansión poblacional, el efecto fundador disminuye la diversidad genética (Austerlitz et al., 1997); sin embargo, la baja calidad del hábitat podría potenciar su efecto, incrementando las diferencias entre las áreas recolonizadas y los sitios refugio, por lo que se ha enfatizado en la relevancia de considerar la historia demográfica de las especies en la dinámica de las poblaciones periféricas y en las desviaciones a los patrones esperados (Abeli et al., 2014; Michelletti y Storfer, 2015). Figura 3. Ubicación de las 29 poblaciones de Avicennia germinans con respecto al centroide del nicho ecológico de la especie. Círculos grises, poblaciones en la periferia del ellipsoide y círculos blancos, poblaciones más cercanas al centroide. Varias líneas de evidencia sugieren que los manglares tienen la capacidad de dispersarse a varios miles de kilómetros (Takayama et al., 2013; Lo et al., 2014), por lo que no sorprende que haya flujo genético entre las poblaciones del Oeste de África y Sudamérica (Nettel y Dodd, 2007; Cerón-Souza et al., 2015; Mori et al., 2015a;). Sin duda, la frecuencia con la que ocurren los eventos de dispersión a larga distancia (LDD) debe estar limitada por la frecuencia con la cual alcanzan las corrientes oceánicas (Nettel y Dodd et al., 2007; Cerón- Souza et al., 2012; Mori et al., 2015a). Sin embargo, aunque los individuos sean capaces de 160 llegar a sitios geográficamente distantes, su establecimiento y posterior reproducción está dictado por la selección natural (Orsini et al., 2013; Sexton et al., 2014). Entre el Oeste de África y Sudamérica existen similitudes florísticas (Graham, 2006) que surgen bajo condiciones climáticas similares. De hecho, de acuerdo con nuestros análisis (información no publicada), estas poblaciones están aproximadamente a la misma distancia del centroide del nicho de la especie, por lo que las condiciones climáticas no difieren significativamente entre ambas costas. Por lo que, aunque se separen por más de 4000 km, las condiciones ambientales han permitido a los migrantes transferir sus genes y, por lo tanto, la identificación de eventos de LDD. No obstante, para probar esta hipótesis necesitaríamos estudiar, por un lado, la distribución de la variación adaptativa de A. germinans y la correspondencia ambiental; bajo esta hipótesis esperamos encontrar firmas adaptativas similares (e. g. mismo patrón de asociación entre SNPs y variables ambientales) entre poblaciones de Brasil y el Oeste de África, pero, distinta entre Brasil y Texas, lo cual es ambientalmente distinta que Brasil, de acuerdo con las distancias al centroide del nicho; y por otro, realizar experimentos de transplantes recíprocos para estudiar la sobrevivencia de plántulas y su establecimiento (Blanquart et al., 2013) entre las poblaciones mencionadas. Perspectivas En los últimos años el conocimiento de la genética de manglares se ha incrementado de manera significativa y se espera que esta tendencia se extienda a varios taxones para los cuales aún no hay estudios. Un ejemplo son Laguncularia racemosa y Conocarpus erectus, especies con amplia distribución y presentes en las costas de México, pero prácticamente ausentes de los estudios de genética de poblaciones. Estudios detallados de ambas especies con poblaciones muestreadas a lo largo del área de su distribución podrían confirmar los patrones de diversidad, estructura genética y dispersión que han sido observados en R. mangle y A. germinans. Encontrar el mismo patrón de variación genética – baja dentro y alta entre poblaciones – implicaría que los mismos procesos evolutivos se comparten entre especies que coexisten en un nicho ecológico similar. Empleando marcadores de herencia biparental (microsatélites e ITS) y uniparental (ADN de cloroplasto) podemos determinar si el cierre del Istmo Centroamericano pudo promover la divergencia entre poblaciones de 161 ambas costas; si la diversidad genética de ambas especies pudo haber sido impactada de una manera similar a A. germinans después del LGM; o si las secuencias de ITS y ADN cloroplasto respaldan la hipótesis de LDD (en lugar de vicarianza) como el mecanismo más probable para el establecimiento de los mangles en las costas del continente americano (Nettel y Dodd, 2007). Es importante desarrollar más estudios para conocer los procesos involucrados en la definición del sistema de apareamiento en las poblaciones, y específicamente en la tasa de autofecundación de las cuatro especies de manglares presentes en México. Las tasas de entrecruzamiento y auto-fertilización pueden diferir de manera significativa tanto entre como dentro de las costas, teniendo un impacto significativo en la diferenciación y diversificación de las poblaciones. Usando marcadores microsatélites, Nettel-Hernanz et al. (2013) abordaron las diferencias en el sistema de apareamiento de A. germinans entre poblaciones de Chiapas y Baja California, sin embargo, el bajo nivel de diversidad genética de las poblaciones en esta última impidió la estimación de las tasas de autofecundación (información no publicada). Desafortunadamente, la baja diversidad genética parece ser una característica de los manglares (Yan et al., 2016), por lo que es importante emplear tecnología de secuenciación de próxima generación para garantizar un buen muestreo de todo el genoma y capturar suficiente variación que permita analizar las tasas de entrecruzamiento. Por otro lado, poco se sabe de los efectos de las perturbaciones humanas sobre la estructura y diversidad genética de las poblaciones naturales de manglares (e. g. Hasan et al., 2018; Martínez-García et al., en preparación). Teóricamente, las poblaciones fragmentadas son más susceptibles a la extinción local debido a la pérdida de variación genética ocasionada por los efectos negativos de la endogamia, la deriva genética y la reducción del flujo genético. Sabemos que A. germinans tiene baja diversidad genética y relativamente poco flujo genético entre poblaciones; la pérdida de diversidad genética debido a la fragmentación por actividades antropogénicas supondría un grave riesgo para la persistencia y viabilidad de las sus poblaciones, porque puede limitar la capacidad de respuesta a los cambios en las presiones de selección e incrementar el riesgo de extinción local (Lienert, 2004; Frankham, 2005). Un estudio típico de fragmentación se lleva acabo comparando los estimados de 162 diversidad genética, endogamia y estructura genética (e. g. número de alelos, heterocigosidad observada y esperada, diferenciación entre pares de poblaciones, etc.) entre individuos adultos (pre-fragmentación) y juveniles (post-fragmentación). Si se observa un cambio significativo de una generación a otra (de adultos a juveniles), se infiere un efecto de la fragmentación. De esta manera, los estudios de fragmentación pueden ayudar a priorizar aquellas poblaciones que resulten más vulnerables, es decir, aquellas en donde la reducción en la diversidad genética en individuos juveniles sea más fuerte. Una de las principales preocupaciones es la integración de la información disponible en los proyectos de manejo y políticas de conservación (Laikre 2010; Shafer et al., 2015), por lo que distintos autores han enfatizado en la urgencia de cerrar la brecha que existe entre los tomadores de decisiones y los investigadores para asegurar la conservación y viabilidad de los manglares a largo plazo (Mori y Kajita, 2016; Wee et al., 2019). La Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO) implementó un sistema de monitoreo de los manglares de México (https://www.biodiversidad.gob.mx/ecosistemas/manglares2013/smmm.html), en este proyecto se detallan el estado de degradación, funcionamiento y tendencias de cambio en la cobertura de manglar en el país. Además de esto, se identifican sitios prioritarios de conservación; cada sitio tiene una ficha de caracterización en la cual podemos encontrar información de las características físicas, socioeconómicas, amenazas, procesos de transformación, así como una lista de especies de plantas y animales, entre otras. Sin embargo, estas fichas carecen de información genética. Incorporar esta información podría facilitar la accesibilidad y manejo de las poblaciones. Conclusiones Generales En este trabajo abordamos distintos factores que contribuyen a la distribución de la variación genética de Avicennia germinans en México. Resaltamos la importancia de los cambios climáticos pasados y cómo éstos pueden impactar de distinta manera a poblaciones que se distribuyen en latitudes similares, pero bajo condiciones ambientales distintas; cómo influyen las características del paisaje en las tasas de flujo genético y el importante rol que ha jugado 163 la deriva genética, así como, la importancia de la calidad del hábitat en la configuración de la diversidad genética. Las poblaciones de ambas costas muestran diferencias notables en los patrones filogeográficos como en la variación genética y endogamia, siendo la costa Atlántico la más diversa y menos endogámica (Ho = 0.287, He = 0.292, FIS = 0.018). Los linajes del Pacífico y Atlántico de México se separaron mucho después del cierre del CAI, hace unos 0.75 Ma. Encontramos una fuerte estructura espacial de la variación genética de A. germinans que difiere entre costas. Aunque encontramos patrones compartidos (e. g. aislamiento por distancia), también encontramos evidencia de que ambos linajes han estado bajo distintas dinámicas poblacionales. El flujo genético entre poblaciones de A. germinans es limitado y varios factores de la configuración del paisaje como barreras físicas (e. g. Península de Baja California) y el patrón de corrientes oceánicas y la prevalencia de la deriva genética dentro de poblaciones influyen en la conectividad y la variación genética entre poblaciones. Las variables que definen el nicho fundamental de A. germinans probablemente estén relacionadas con el establecimiento, sobrevivencia y crecimiento de las plántulas, los cuales determinan en parte, los patrones de diversidad observados. Es posible que exista una composición óptima de arena y arcilla que permita el desarrollo de poblaciones más vigorosas capaces de tener mayor diversidad genética. La costa Atlántico alberga mayor número de poblaciones bajo condiciones ambientalmente más adecuadas que la costa Pacífico. La calidad del hábitat constituye un factor importante en el entendimiento de la distribución de la variabilidad genética en A. germinans. Nuestras predicciones de heterocigosidad mostraron un poder predictivo de un 40 % por lo que son útiles en planes de conservación y manejo. 164 Referencias Abeli, T., Gentili, R., Mondoni, A., Orsenio, S. & Rossi, G. (2014) Effects of marginality o plant population performance. Journal of Biogeography, 41, 239- 249. Alongi, D.M. (2015) The impact of climate change on mangrove forest. Current Climate Change Reports, 1, 30-39. Álvarez, I. & Wendel, J. F. (2003) Ribosomal ITS sequences and plant phylogenetic inference. Molecular Phylogenetics and Evolution, 29, 417-434. Austerlitz, F., Jung-Muller, B., Godelle, B. & Pierr-Henri, G. (1997) Evolution of coalescence times, genetic diversity and structure during colonization. Theoretical Population Biology, 51, 148-164. Baena-Díaz, F., Ramírez-Barahona, S. & Ornelas, J. F. (2018) Hybridization and differential introgression associated with environmental shifts in a mistletoe species complex. Scientific Reports, 8, 5591. Blanquart, F., Kaltz, O., Nuismer, S. L. & Gandon, S. (2013) A practical guide to measuring local adaptation. Ecology Letters, 16, 1195-1205. Cerón-Souza, I., Bermingham, E., Owen, M.W. & Jones, F.A. (2012). Comparative genetic structure of two mangrove species in Caribbean and Pacific estuaries of Panama. BMC Evolutionary Biology, 12, 205. Cerón-Souza, I., Toro-Perea, N., Cárdenas-Henao, H. (2005) Population genetic structure of Neotropical mangrove species on the Colombian Pacific Coast; Avicennia germinans (Avicenniaceae). Biotropica, 37, 258–265. Clarke, P. J. & Myerscough, P. J. (1991) Floral biology and reproductive phenology of Avicennia marina in South-eastern Australia. Australian Journal of Botany, 39, 283-293. Colicchio, J., Monnahan, P. J., Wessinger, C. A., Brown, K., Kern, J. R. & Kelly, J. K., Individualized mating system estimation using genomic data. Molecular Ecology Resources, 2019, 00, 1–15. Cornuet, J-M., Pudlo, P., Veyssier, J., Dehne-García, A., Gautier, M., Leblos, R... Estoup, A. (2014). DIYABC v2.0: a software to make approximate Bayesian computation inferences about population history using single nucleotide polymorphism, DNA sequence and microsatellite data. Bioinfomatics, 30, 1187-1189. Frankham, R. (2005) Genetics and extinction. Biological Conservation, 126, 131-140. 165 Graham, A. 2006. Modern processes and historical factors in the origin of the african element in Latin America. Annals of the Missouri Botanical Garden Press, 93, 355-339. Hasan, S., Triest, L., Afrose, S., De Rick, D. J. R. (2018) Migrant pool model of dispersal explains strong connectivity of Avicennia officinalis within Sundarban mangrove areas: Effect of fragmentation and replantation. Estuarine, Coastal and Shelf Science, 214, 38- 47. Ikeda, D., Max, T., Allan, G. J., Lau, M. K., Shuster, S. M. & Whitman, T. G. (2017). Genetically informed ecological niche models improve climate change predictions. Global Change Biology, 23, 164-176. Jones, M. R., Forester, B. R., Teufel, A. I., Adams,R. V., Anstett, D. N., Goodrich, B. A... Manel, S. (2013) Integration landscape genomics and spatially explicit approaches to detect loci under selection in clinal populations. Evolution, 67, 3455-3468. Laikre, L. (2010) Genetic diversity is overlooked in international conservation policy implementation. Conservation Genetics, 11, 349–354. Lienert, J. (2004) Habitat fragmentation effects on fitness of plant populations – a review. Journal of Nature Conservation, 12, 53-72. Linder, C. R. & Rieseberg, L. H. (2004) Reconstructing patterns of reticulate evolution in plants. American Journal of Botany, 91, 1700-1708. Lo, E. Y. Y., Duke, N. C. & Sun, M. (2014) Phylogeographic pattern of Rhizophora (Rhizophoraceae) reveals the importance of both vicariance and long-distance oceanic dispersal to modern mangrove distribution. BMC Evolutionary Biology, 14, 83. Makino, T. T., Ohashi, K. & Sakai, S. (2007) How do floral display size and the density of surrounding flowers influence the likelihood of bumble bee revisitation to a plant? Functional Ecology, 21, 87-95. Manel, S., Joost, S., Epperson, B. K., Holderegger, R., Storfer, A., Rosenberg, M. S... Fortin, M-J. (2010). Perspectives on the use of landscape genetics to detect genetic adaptive variation in the field. Molecular Ecology, 19, 3760-3772. Marcer, A., Méndez-Vigo, B., Alonso-Blanco, C. & Picó, X. (2016) Tackling intraespecific genetic structure in distribution models better reflects species geographical range. Ecology and Evolution, 6, 2084-2097. 166 Micheletti, S. J. & Storfer, A. (2015). A test of the central-marginal hypothesis using population genetics and ecological niche modelling in an endemic salamander (Ambystoma barbourin). Molecular Ecology, 24, 967-979. Mirkin, B. (2016) Clustering: a data recovery approach. 2a edición. CRC Press. 374 pp. Mori, G.M. & Kajita, T. (2016) Mangrove conservation genetics. Journal of integrated field science, 13, 13-19. Mori, G.M., Zucchi, M.I. & Souza, A.P. (2015b). Multiple-geographic-scale genetic structure of two mangrove tree species: the roles of mating system, hybridization, limited dispersal and extrinsic factors. Plos One, 10, e0118710. Mori, G.M., Zucchi, M.I., Sampaio, I. and Souza, A. P. (2015a). Species distribution and introgressive hybridization of two Avicennia species from the Western Hemisphere unveiled by phylogeographic patterns. BMC Evolutionary Biology, 15, 16. Nadia, T. D. L., Menezes, N. L., Machado, I. C. (2013) Floral traits and reproduction of Avicennia schaueriana Moldenke (Acanthaceae): a generalist pollination system in the Lamiales. Plant Species Biology, 28, 70-80. Nattero, J., Malerba, R., Medel, R. & Cocucci, A. (2011) Factors affecting pollinator movement and plant fitness in a specialized pollination system. Plant Systematics and Evolution, 296, 77-85. Nettel-Hernanz, A., Dodd, R.S., Ochoa-Zavala, M., Tovilla-Hernández, C. & Días-Gallegos, J.R. (2013) Mating system analyses of tropical populations of the black mangrove, Avicennia germinans (L.) L. (Avicenniaceae). Botanical Sciences, 91, 115-117. Nettel, A. & Dodd, R.S. (2007) Drifting propagules and receding swamps: Genetic footprints of mangrove recolonization and dispersal along tropical coasts. Evolution, 61, 958-971. Nettel, A., Dodd, R.S., Afzal-Rafii, Z. & Tovilla-Hernández, C. (2008). Genetic diversity enhanced by ancient introgression and secondary contact in East Pacific black mangroves. Molecular Ecology 17, 2680-2690. Orsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J. & de Meester, L. (2013) Drivers of population genetic differentiation in the wild: isolation by dispersal limitation, isolation by adaptation and isolation by colonization. Molecular Ecology, 22, 5983-5999. Osorio-Olvera, L., Lira-Noriega, A., Soberón, J., Peterson, A. T., Falconi, M., Contreras- Díaz, R., Martínez-Meyer, E. En Revisión. NicheToolBox: An R package with a GUI for 167 accessing biodiversity data and modeling and evaluating multidimensional ecological niches. Record, S., Cherney, N. D., Zakaria, R. M. & Ellison, A. M. (2013). Projecting global mangrove species and community distributions under climate change. Ecosphere, 4, 34. Ritland, K. (2002) Extensions of models for the estimation of mating systems using n independent loci. Heredity, 88, 221-228. Rosenzweig, B. K., Pease, J. B., Besansky, N. J. & Hahn, M.W. (2016) Powerful methods for detecting introgressed regions from population genomic data. Molecular Ecology, 25, 2387-2397. Sexton, J.P., Hangartner, S.B., & Hoffmann, A.A. (2014). Genetic isolation by environment or distance: Which pattern of gene flow is most common? Evolution, 68, 1–15. Shafer, A.B.A., Wolf, J.B.W., Alves, P.C., Bergström, L., Bruford, M.W., Brännström, I... Zielinski, P. (2015) Genomics and the challenging translation into conservation practice. Trends in Ecology and Evolution, 30, 78–87. Sork, V.L., Davis, F.W., Westfall, R., Flint, A., Ikegami, M., Wang, H. & Grivet, D. (2010). Gene movement and genetic association with regional climate gradients in California valley oak (Quercus lobata Née) in the face of climate change. Molecular Ecology, 19, 3806-3823. Takayama, K., Tomura, M., Tateishi, Y., Webb, E. L. & Kajita, T. (2013). Strong genetic structure over the American continents and transoceanic dispersal in the mangrove genus Rhizophora (Rhizophoraceae) revealed by broad-scale nuclear and chloroplast DNA analysis. American Journal of Botany, 100, 1191-1201. Twyford, A. D. & Ennos, R. A. (2012) Next-generation hybridization and introgression. Heredity, 108, 179-189. Valderrama-Landeros, L. H., Rodríguez-Zúñiga, M. T., Troche-Souza, C., Velázquez- Salazar, S., Villeda-Chávez, E., Alcántara-Maya, J. A…Ressl, R. (2017). Manglares de México: actualización y exploración de los datos del sistema de monitoreo 1970/1980- 2015. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad (CONABIO), Ciudad de México, México. 128pp 168 Wee, A.K.S., Mori, G.M., Lira, C.F., Núñez-Farfán, J., Takayama, K., Faulks, L… Kajita, T. (2019) The integration and application of genomic information in mangrove conservation. Conservation Biology, 33, 206-209. Xu, B., Zeng, X-M., Gao, X-F., Jin D-P. & Zhang, L-B. (2017) ITS non-concerted evolution and rampant hybridization in the legume genus Lespedeza (Fabaceae). Scientific Reports, 7, 40057. Yan, Y-B., Duke, N. C & Sun, M. (2016) Comparative analysis of the pattern of population genetic diversity in three Indo-West Pacific Rhizophora mangrove species. Frontiers in Plant Science, 7, 1434.