Universidad de Concepción

Dirección de Postgrado

Centro de Ciencias Ambientales EULA-Chile

Programa de Doctorado en Ciencias Ambientales

Distribución, estructura comunitaria y poblacional de Galáxidos en Patagonia: aspectos determinantes históricos y actuales amenazas para su conservación.

JORGE FABIAN GONZALEZ GONZALEZ

CONCEPCIÓN-CHILE

2012

Profesor Guía: Evelyn Habit Conejeros

Unidad de Sistemas Acuáticos

Centro de Ciencias Ambientales EULA-Chile

Universidad de Concepción INDICE Pág.

AGRADECIMIENTOS……………………...……………………….……………………..2

RESUMEN……………………….…………...…………………………….………………3

ABSTRACT…………………………………...………………………………….………...6

INTRODUCCIÓN………………………………………………………………….……….8

Hipótesis de trabajo…………………...……………………………………….……….…..14

Objetivos……………………………………………………………………………….…..14

CAPÍTULO I: Changes in the distribution of native fishes in response to introduced species and other anthropogenic effects………………………………………………..…..15

CAPÍTULO II: Native and introduced fish species richness in lacustrine Chilean Patagonia: inferences on species invasion from salmonid free lakes………………....……41

CAPÍTULO III: Freshwater fish in southernmost Chilean Patagonia: salmonids and galaxiids coexisting within protected areas…………………………………………...……67

CAPÍTULO IV: Trophic interference by Salmo trutta on Aplochiton zebra and Aplochiton taeniatus in southern Patagonian lakes………………………….….………………..… ..93

DISCUSIÓN………………………………………………………………….…….….…114

CONCLUSIONES……………………………………………………………………..…120

REFERENCIAS…………………………………………………………………………..121

1

AGRADECIMIENTOS

Quiero agradecer sinceramente a todos/as las personas e instituciones que ayudaron de una u otra forma durante mi paso por este programa de doctorado y en la realización de esta tesis doctoral.

A mi profesora tutora Dra. Evelyn Habit, por su constante apoyo y dedicación. A mis compañeros de laboratorio por su amistad y ayuda. A mis compañeros del programa de doctorado. También, a todos los funcionarios del centro Eula, que de una u otra forma me han apoyado en este estudio.

A mi familia, muy especialmente, a mi hijo Emilio y a Tamara. A mis padres y hermanos. A mis amigos de la vida. A todos ellos muchas gracias por su incondicional amor y cariño.

Además, tanto este programa de doctorado como el presente trabajo de tesis fueron realizados con el apoyo en fianciamiento de: CONICYT, a traves de la beca Doctorado Nacional y beca Apoyo a realización de tesis; la direccion de Postgrado de la Universidad de Concepción y del Centro Eula-Chile. A todas estas instituciones les agradezco sinceramente su apoyo.

2

RESUMEN.

Una de las principales amenazas a la integridad de la fauna nativa, a nivel mundial, es la pérdida de biodiversidad resultante de la introducción de especies exóticas. Sin embargo, no es fácil desvincular los efectos de estas introducciones sobre la distribución y patrones geográficos de riqueza de especies, de otras múltiples variables físicas, biológicas y antrópicas que afectan a esta fauna. Por otro lado, estos efectos varían espacialmente y dependen de la escala de análisis. El impacto de la introducción de salmónidos sobre la fauna de peces nativos en zonas protegidas de la Patagonia se espera que sea menor, sin embargo, se ha visto que las áreas silvestres protegidas no son muy eficaces en la conservación de esta fauna. Los Galáxidos son uno de los grupos de peces más afectados por los salmónidos introducidos en todo el mundo, sin embargo, no existen suficientes estudios que examinen los mecanismos responsables de estos efectos negativos.

En esta investigación se documentan en un amplio rango latitudinal (28º-54ºS) los cambios recientes en la distribución de cinco especies de Galáxidos, mediante la comparación de las distribuciones históricas y actuales basada en la más extensa base de datos de peces de agua dulce de Chile, y poniendo a prueba las relaciones de la distribución y abundancia de los peces nativos, con la incidencia de las especies introducidas. Por otro lado, estudiamos los gradientes de riqueza de especies de peces de agua dulce nativos e introducidos en lagos de la Patagonia chilena (39°-54°S). Aquí nos enfocamos en el rol del entorno físico (temperatura, tamaño del ecosistema, conectividad actual e histórica, actividad humana como vías de acceso y uso de suelo), en explicar los patrones de riqueza de nativos, en tanto, para los patrones de riqueza y dominancia de salmónidos introducidos, se ponen a prueba las hipótesis de resistencia biótica y actividad humana. Todo esto, con el objetivo de identificar los factores que mejor explican la persistencia de lagos sin salmónidos en la Patagonia. A menor escala, se analiza la composición, distribución y estructura de las comunidades de peces nativos en relación con la invasión de salmónidos y la protección de áreas silvestres protegidas en la cuenca del río Serrano, Patagonia (50° S). Además, se comparan relación la longitud/peso, dieta y composición isotópica ( 15 N y 13 C) de Galáxidos, en sistemas invadidos y libres de salmónidos.

Encontramos que los rangos de distribución de la mayoría de las especies de Galáxidos se ha reducido significativamente en los últimos años ( Galaxias maculatus : 26%, Brachygalaxias bullocki : 17%, Aplochiton taeniatus : 7%), excepto Galaxias platei especie que mostró un aumento considerable de su distribución, producto de un aumento en el esfuerzo de muestreo desplegado en la Patagonia sur. Esta especie es dominante en los sistemas de alta elevación, mientras que en sistemas de elevaciones intermedias, especialmente ríos, están dominados actualmente por salmónidos. Se sugiere que las interacciones negativas entre nativos y salmónidos, son responsables de algunas de las

3 reducciones de rango en la distribución de Galáxidos en Chile. Sin embargo, la gravedad de los impactos varía con la latitud y la altitud, y está probablemente relacionado con la temperatura. Los peces nativos parecen haber encontrado refugio de los salmónidos en sistemas más cálidos del norte y en la zona costera, así como en altura, en ambientes relativamente fríos. Los peces nativos también parecen ser menos vulnerables a los salmónidos en lagos que en ríos.

En ecosistemas lacustres de la Patagonia, encontramos que la riqueza de especies introducidas se correlacionó positivamente con la riqueza de especies nativas, y que la riqueza de nativos e introducidos se correlaciono negativamente con la latitud, y positivamente con la temperatura y el tamaño de los ecosistemas (área del lago). Además, las variaciones en la riqueza de nativos se relacionaron con las conexiones de drenaje históricos, mientras que la riqueza y la dominancia de salmónidos fueron significativamente afectados por la conectividad del hábitat actual. Encontramos un total de 15 lagos libres de salmónidos, todos ellos situados en zonas remotas al sur de los 45°S, y todos ubicados aguas arriba de barreras físicas de origen natural. No se encontró evidencias a favor de la existencia de resistencia biótica ejercida por especies nativas a la expansión de salmónidos, y a pesar de que las introducciones originales fueron mediadas por el hombre, los patrones actuales de riqueza no están relacionados con actividades antropogénicas, al menos aquellas medidas a través del acceso de caminos y uso de suelo. Por el contrario, los factores ambientales, especialmente la conectividad del hábitat y la temperatura, parecen limitar la expansión de salmónidos en aguas continentales.

A una escala menor, intracuenca, en la cuenca hidrográfica del río Serrano (51 ⁰ S), encontramos que, a pesar de que la invasión de los salmónidos en las zonas prístinas estudiadas no ha generado cambios regionales en la riqueza de especies nativas, sí esta impactando sobre algunas de ellas. Las interacciones negativas con salmónidos, como depredación y/o exclusión competitiva, parecen estar desplazando a las especies nativas a hábitats menos productivos y provocando cambios en su conducta alimentaria, lo que consecuentemente, se ve reflejado en variaciones en la composición de la dieta, un deterioro de la condición física y una disminución de su posición trófica . Este resultado se obtuvo consistentemente para especies del género Aplochiton y Galaxias platei , al comparar poblaciones en sistemas con y sin salmónidos. Dichas interferencias podrían dar lugar a disminuciones de los tamaños poblacionales y explicarían la disminución de rango de distribución de algunas especies y explicarían la marcada distribución disjunta entre nativos y truchas.

Asociado a los resultados anteriores, un aporte interesante de esta tesis es que se encontró que las áreas silvestres protegidas estudiadas (Parque Nacional Torres del Paine y Bernardo O’Higgins) ofrecen poca protección de la invasión de salmónidos a las especies

4 de Galáxidos y, además, sus limites no se ajustan a los patrones actuales de distribución y diversidad de esta fauna nativa.

Por otro lado, los resultados de esta tesis han permitido identificar cuencas hidrográficas y sistemas acuáticos particulares que son críticos para la conservación de la biodiversidad de peces nativos de agua dulce de Chile. Los principales sistemas que se proponen son la cuenca del río Maullín, Valdivia, los ríos del norte de la isla de Chiloé y varios lagos en la región de Aysén, la mayoría de las cuales incluye la totalidad de las especies de Galáxidos. Además, se identificaron 15 lagos libres de salmónidos de las cuencas del rio Aysén (1), del rio Cuervo (2), del Baker (5) y del rio Serrano (7), los cuales consideramos como prioritarios para la conservación de estas poblaciones únicas de peces nativos.

Finalmente, esperamos que toda esta información sea un aporte significativo a entender el efecto de la introducción de salmónidos, al manejo áreas silvestres protegidas y a la conservación de la ictiofauna nativa en la Patagonia.

5

ABSTRACT.

One of the main threats to the integrity of native wildlife, worldwide, is the loss of biodiversity resulting from the introduction of exotic species. However, it is easy to disentangle the effects of these introductions on the distribution and geographic patterns of species richness of other multiple variables that affect this fauna. Furthermore, these effects vary spatially and depend on the scale of analysis. The impact of the introduction of salmonids on native fish fauna in protected areas of Patagonia is expected to be lower, however, protected wilderness areas seem to be not very effective for the conservation of this fauna. The Galaxiids are one of the fish groups most affected by introduced salmonids worldwide however, there are not enough studies examining the underlying mechanisms responsible for those negative effects.

In this research we documented in a wide latitudinal range (28º - 54º S) recent changes in the distribution of five species of Galaxiids, by comparing historical and current distributions based on the most extensive database of freshwater fishes of Chile, and testing the relationship of the distribution and abundance of native fish, with the incidence of introduced species. In addition, we studied the species richness gradients of freshwater fish native and introduced in lakes of Chilean Patagonia (39° - 54° S). Here we focus on the role of the physical environment (temperature, size of the ecosystem, current and historical connectivity, human activity such as access roads and land use), to explain patterns of native wealth, while, for patterns of wealth and dominance of introduced salmonids, are tested biotic resistance hypotheses and human activity. Our main goal was to identify which factors best explain the persistence of salmonid free lakes in Patagonia. On a smaller scale, we analyze the composition, distribution and structure of native fish communities in relation to the invasion of salmonids and the protection of conservation areas inside the Serrano River basin (50° S). Then, we compare the length / weight, diet and isotopic composition ( 15 N and 13 C) of Galaxiids in invaded systems and free of salmonids.

We found that distribution ranges of most Galaxiids species have been reduced in recent years, except for the distribution of Galaxias platei which increased considerably due to the high sampling effort. This species is dominant in high-lift systems, while intermediate elevations systems, especially rivers, are now dominated by salmonids. It is suggested that negative interactions between native species and salmonids, are responsible for some of the reductions in the distribution range of Galaxiids in Chile. However, the severity of impacts varies with latitude and altitude, and is probably related to temperature. Native fish seem to have found shelter for salmonids in warmer systems in the northern range and in coastal zone and height, in relatively cold. Native fish also seem to be less vulnerable to salmonids in lakes than in rivers. In lake ecosystems of Patagonia, we found that introduced species richness was positively correlated with native species richness, and richness of native and

6 introduced decreased with latitude, increasing the temperature and size of ecosystems. In addition, variations in the richness of native related to the historical drainage connections, while the richness and the dominance of salmonids were significantly affected by the current habitat connectivity. We found a total of 15 free salmonid lakes, all located in remote areas south of 45° S, and all of them above natural physical barriers. There was no evidence of biotic resistance of native species for the expansion of salmonids, and although the original introductions were human-mediated, current patterns seem not to be related with human activities as measured by road access or land use. By contrast, environmental factors, particularly habitat connectivity and temperature, appear to limit the expansion of salmon in inland waters. On a smaller scale, within a watershed, we found that despite the invasion of salmonids in pristine areas studied did not generate regional changes in native species richness, it does change life history of galaxiids. Negative interactions with salmonids, such as predation and / or competitive exclusion result in the displacing of native species to less productive habitats, causing changes in their feeding behavior which consequently is reflected in variations in the composition of the diet, a decrease in welfare and trophic position. Such trophic interferences can lead to decreases in population sizes and explain the species distribution range decline, and may explain the strong disjunct distribution between native and trout.

Associated with the above results, an interesting contribution of this thesis is that it was found that the protected areas studied (Torres del Paine and Bernardo O'Higgins National Parks) offer little protection from the invasion of salmonid to galaxiids species and also their boundaries do not conform to current patterns of distribution and diversity of this native fauna.

Besides, the results of this thesis have allowed identifying watersheds and particular aquatic systems that are critical for biodiversity conservation of freshwater fish native to Chile. The main systems proposed are: Maullín River, Valdivia River, rivers in the north of Chiloe Island and several lakes in the Aysén district, most of which include all the Galaxiid species. In addition, we identified 15 salmonids free lakes of Aysén River (1), Cuervo River (2), Baker River (5) and the Serrano River basins (7), which we consider as priorities for conservation of these unique native fish populations.

Finally, we hope that this information is a significant contribution to understanding the effect of the introduction of salmonids, the protected areas management and conservation of native fauna in Patagonia.

7

INTRODUCCION

Globalmente, la invasión de especies exóticas constituye una de las principales amenazas para la biodiversidad acuática (Dudgeon et al. 2006; Light & Marchetti 2007; Mack et al. 2000; Saunders et al. 2002), produciendo homogeneización de comunidades de peces y cambios en patrones biogeográficos de las especies nativas (Leprieur et al. 2008; 2009).

En Chile, la fauna de peces nativos continentales tiene una distribución restringida y baja abundancia en comparación con especies introducidas como las truchas (Soto et al. 2006). Igualmente, el estado de conservación de estas especies según la ultima clasificación del Ministerio del Medio Ambiente de Chile es de alta vulnerabilidad, con un ~70% de las especies en algún grado de amenaza (Vulnerable, En Peligro o En Peligro Critico; MMA 2012). Además, si bien es cierto que las áreas de conservación de la vida silvestre chilenas cubren un gran porcentaje del territorio nacional (~20%), éstas no están diseñadas en concordancia con la diversidad y distribución de la fauna, o no poseen la información biológica y ecológica para realizar una conservación eficiente (Tognelli et al. 2008, Martínez-Harms & Gajardo 2008).

En el caso de la Patagonia chilena, el valor de conservación de su ictiofauna dulceacuícola radica en su alta singularidad, derivada de su historia geológica particular (Ruzzante et al. 2008). Particularmente, la elevación de la cordillera de los Andes y los repetidos ciclos glaciales han determinado una ictiofauna de baja riqueza específica, pero alta diversidad morfológica y genética (Cussac et al. 2004, Ruzzante et al. 2008). Por otro lado, intervenciones de origen antrópico, como la introducción de salmónidos, son una amenaza para esta fauna, ya que estarían alterando sus patrones naturales de distribución y diversidad (Baigún & Ferriz 2003, Cussac et al. 2004, Pascual et al. 2007). El entendimiento de cómo los procesos históricos y actuales influyen en el estado de la fauna de peces nativos es de vital importancia para la conservación de los mismos.

Ictiofauna patagónica.

En el continente sudamericano la fauna de peces dulceacuícolas está distribuida en dos grandes regiones biogeográficas, caracterizadas por poseer gran diferencia en diversidad: la región Brasílica representada por 38 familias y 550 géneros, y la región Austral con solo 10 familias y 16 géneros (Arratia 1997). Para esta región, Dyer (2000) propone la existencia de tres Provincias ictiogeográficas (de norte a sur): Atacama, Chilena y Patagónica. Entre ellas, destaca la Provincia Chilena por su alta diversidad de especies mientras que las de Atacama y Patagónica, se caracterizan por poseer una alta singularidad.

La Provincia Patagónica se encuentra separada en dos vertientes por la cordillera de los Andes y se extiende desde Chiloé continental por la vertiente Oeste (Chile, 40ºS) y desde el

8 río Colorado por la vertiente Este (Argentina, 36ºS) hasta Tierra del Fuego en el extremo sur del continente (Baigún & Ferriz 2003, Dyer 2000, Pascual et al. 2007). Actualmente la ictiofauna nativa de aguas continentales de la Patagonia se compone de un total de 20 especies (Pascual et al. 2002), nueve de las cuales se encuentra en la vertiente Oeste (Habit et al . 2006). Esta diversidad resulta baja si se compara con otras provincias como la Chilena. Sin embargo, la diversidad morfológica y genética de esta fauna es alta (Milano et al. 2002, Zattara & Premoli 2005, Ruzzante et al. 2006, 2008).

Entre la ictiofauna patagónica destaca el grupo de los Galáxidos (Familia Galaxidae) representada por cuatro especies: Galaxias maculatus, Galaxias platei, Aplochiton zebra y Aplochiton taeniatus . Este grupo se caracteriza por poseer diversas historias de vida entre especies, relacionadas a la diadromía, lo que explicaría su éxito en diversidad y abundancia . De este grupo la especie de más amplia distribución es G. maculatus , descrita como una especie catádroma, capaz de formar poblaciones diádromas y encerradas o “landlocked” (McDowall 2006). Además, se ha descrito gran variabilidad intra e interpoblacional, relacionadas con reproducción (épocas reproductivas desfasadas), uso de hábitat (lagos, ríos y estuarios), hábitos alimentarios, capacidad de desplazamiento y caracteres morfológicos asociados principalmente a la conducta trófica (Barriga et al. 2006, Battini et al. 2000, Boy et al. 2007, Chapman et al. 2006, McDowall & Charteris 2006). Esta alta plasticidad en historia de vida podría explicar su amplia distribución, siendo más exitosas y menos vulnerables ante cambios ambientales (Barriga et al. 2006). Galaxias platei en cambio, es una especie que formaría exclusivamente poblaciones encerradas o “landlocked”, asociadas a ambientes lacustres (Cussac et al. 2004). Esta especie se ha descrito como típicamente bentófaga, piscívora, y con adaptaciones morfológicas asociada a ambientes profundos y al riesgo de depredación (Macchi et al. 1999, Milano et al. 2002, Zama 1986). Además, es una especie tolerante a condiciones ambientales extremas como bajas temperaturas y bajas concentraciones de oxígeno, lo que explicaría su mayor presencia en cuencas de la Patagonia (Cussac et al. 2004, Dyer 2000, Ruzzante et al. 2008). Menos diverso en la Patagonia es Percichthyidae, con solo una especie, Percichthys trucha , abundante en ambas vertientes de la Patagonia (Ruzzante et al. 2006, 2008). Esta especie habita diversos lagos, donde los ambientes someros, de mayor temperatura y con abundante vegetación son de vital importancia para su reproducción, realizando grandes desplazamientos asociados a ella (Buria et al. 2007). Por otro lado, esta especie posee una conducta trófica piscívora depredando sobre otros nativos como G. maculatus pero en mucho menor grado que especies Salmonídeas. Además, posee una amplia dieta de origen bentónico lo que la convierte en un potencial competidor de especies nativas como G. platei (Macchi et al. 1999, 2007).

En resumen, la ictiofauna nativa de la Patagonia chilena posee características que la hacen altamente singular en comparación con otras regiones biogeográficas de Sudamérica, como

9 son la baja riqueza de especies, dominancia de Galáxidos, distribución restringida en cuencas de menor tamaño, reducido rango de tamaño corporal y alta diversidad morfológica y genética, (Arratia 1981, Vila et al. 1999, Dyer 2000, Habit et al. 2006a, Milano et al. 2002, Zattara & Premoli 2005, Ruzzante et al. . 2006, 2008). Estas características podrían atribuirle una gran vulnerabilidad a diversas alteraciones generadas por actividades antrópicas (Soto et al. 2006, Habit et al. 2006b).

Patrones de distribución de Galáxidos

Los patrones de distribución de esta fauna a macroescala o escala biogeográfica han sido determinados naturalmente por eventos ambientales a escala de tiempo geológico, sin embargo, no todas las especies responden de igual manera, resultando en patrones filogegráficos que pueden ser especie específicos (Cussac et al. 2004, Ruzzante et al. 2008). Por otro lado, patrones de mesoescala como distribución espacial dentro de una cuenca, uso de hábitat, y sus variaciones temporales asociadas a la reproducción y/o búsqueda de alimento, responden a variables ambientales actuales. Por ello, es esperable que estos patrones de distribución, así como la historia de vida de las especies nativas puedan estar siendo alteradas por efectos de la intervención antrópica, como por ejemplo por la masiva introducción de especies invasivas como los salmónidos (Basulto 2003, Cussac et al. 2004, Soto et al. 2006).

Las evidencias de los procesos de colonización de ictiofauna de las áreas posglaciadas, muestran que esta recolonización fue especie o grupo específica. Entre los Galáxidos, G. maculatus , que posee la capacidad de formar poblaciones diádromas y encerradas o “landlocked”, tendría distintas rutas alternativas de colonización. Una de ellas habría sido la recolonización por la vertiente Oeste, vía marina (poblaciones diádromas; Cussac et al. 2004), y otra desde zonas refugiales como grandes paleolagos desde la vertiente Este. En esta última zona, de acuerdo a Zattara & Premoli (2005), la mayoría de las poblaciones lacustres son “landlocked”. Para el caso de G. platei se sabe por su distribución casi exclusiva en zonas que estuvieron glaciadas, que habría permanecido en estas áreas, en poblaciones encerradas, gracias a sus adaptaciones para soportar condiciones extrema (Cussac et al. 2004). Sin embargo, esto provocó que la especie pasara por una marcada disminución en su tamaño poblacional y “cuello de botella”, en una época coincidente con el Ultimo Maximo Glacial (UMG, Ruzzante et al. 2008). En cambio, la especie P. trucha , de acuerdo a su distribución más amplia, que incluye áreas que no estuvieron glaciadas y más ligada a la zona norte de la Patagonia, habría persistido en esas zonas menos frías, recolonizando a medida que retrocedieron los hielos. Por ello, no habría resultado poblacionalmente tan afectada como G. platei (Ruzzante et al. 2008). Información más reciente, basada en la diversidad genética de G. platei , muestra que en la Patagonia existirían dos grupos de poblaciones con orígenes distintos: uno exclusivo de la zona norte

10 de la vertiente Oeste, más antiguo y relacionado con procesos geológicos pretéritos como la elevación de la cordillera de los Andes; y otro en la zona sur, compartido entre ambas vertientes y relacionado con las glaciaciones, en que la reversión de cuencas habría sido de gran importancia (Zemlak et al. 2008). Todos estos estudios han incluido poblaciones de la Patagonia chilena norte (hasta los 47°Sur o cuenca del Baker). Sin embargo, considerando los patrones descritos para las distintas especies en la vertiente Este y norte de la vertiente Oeste, es esperable para el extremo sur de la Patagonia chilena que las vías más probables de colonización postglacial hayan sido marinas para Galáxidos diádromos como Galaxias maculatus o a partir de refugios periglaciales para el caso de Galaxias platei . Así mismo, esperamos que la diversidad genética de esta especie posea mayor similitud con poblaciones de la Patagonia argentina que con poblaciones ubicadas en cuencas chilenas hacia el norte. Por otra parte, P. trucha parece estar en la Patagonia chilena sólo en cuencas que han revertido su drenaje (Ruzzante et al. 2006), por lo cual podría ser esperable que estuviera presente en ciertas cuencas de la Patagonia sur. Por ende, para el extremo sur de la Patagonia, no se espera la colonización desde poblaciones chilenas del norte, sino vía reversión de cuencas (Este-Oeste).

Finalmente, se espera que todos estos patrones de recolonización y distribución podrían estar siendo modificados por acciones antropogénicas como la invasión de especies salmonídeas (Cussac et al. 2004).

Invasión de salmónidos y efectos sobre Galáxidos.

Las especies salmonídeas han sido descritas como especies invasoras a nivel mundial (Kitano 2004, Korsu et al. 2008, 2010, Irz et al. 2004, McDowall 2003, 2006, McIntosh 2010, Sanderson et al. 2009) y son capaces de provocar cambios a distintos niveles de organización biológica en los ecosistemas que colonizan (Simon & Townsend 2003, Townsend 2003). A nivel individual pueden provocar cambios de hábitos conductuales o segregación en hábitat subóptimos (McIntosh 2000, McIntosh et al. . 1992); a nivel poblacional cambios como extirpación, reducción o fragmentación poblacional (Townsend & Crowl 1991); a nivel comunitario pueden suprimir la presión de pastoreo por invertebrados (Flecker & Townsend 1994) y a nivel ecosistémico aumentando significativamente la productividad primaria (Simon & Townsend 2003, Townsend 2003). Estas especies son consideradas depredadores tope dentro de las tramas tróficas acuáticas, así como voraces competidores por alimento. Por ello, los efectos sobre la fauna nativa pueden ser distintos dependiendo de la posición trófica y del uso de hábitat de cada especie.

Actualmente, la fauna íctica introducida en ambientes acuáticos continentales de Chile, está representada aproximadamente por 22 especies, de las cuales 20 pertenecen a familias no- nativas (Dyer 2000). De las introducciones de especies, la más importante ha sido la de

11 salmónidos, en un proceso que ha llevado aproximadamente más de un siglo, con los primeros éxitos a principios del 1900 (Basulto 2003). Estas introducciones se realizaron con ayuda de los gobiernos de la época, fundamentalmente con fines de incentivar la pesca deportiva, pero sin estudios previos que diagnosticaran sus potenciales efectos (Campos 1970). Hoy en día, las especies introducidas más abundantes, principalmente en el centro y sur de Chile, son la trucha arcoíris ( Oncorhynchus mykiss ) y la trucha café ( Salmo trutta ), las que conforman poblaciones asilvestradas en la mayoría de las cuencas de esta zona (Soto et al. 2001, 2006). Ambas especies son depredadoras y competidoras de las especies nativas (Soto et al. 2006, Ortíz Sandoval 2007). A éstas, se suman otros salmónidos como el salmón Coho ( Oncorhynchus kisutch ) y salmón del Atlántico ( Salmo salar ) escapados de la salmonicultura (Soto et al. 2001). Estas podrían estar conformando poblaciones autosustentables y colonizando ríos, como lo ha realizado el salmón Chinook (Oncorhynchus tshawytscha ) en la mayoría de las cuencas de la Patagonia (Becker et al. 2007, Correa & Gross 2007, Ibarra et al. 2011, Soto et al. 2007). Estos salmones pasan la mayor parte de su vida en el mar hasta su época de reproducción, cuando remontan ríos y esteros para desovar y luego morir (semelparía). Ello genera el ingreso de gran cantidad de nutrientes marinos a las cabeceras de las cuencas, lo que podría provocar alteraciones comunitarias y ecosistémicas, modificando las tramas tróficas (Naiman et al. 2002, Soto et al. 2007).

En cuanto a los impactos producidos por salmónidos sobre Galáxidos, en la Patagonia existen evidencias de interferencias. Por ejemplo, se ha encontrado una correlación negativa entre las abundancia de ambos grupos en varias cuencas de Chile (Soto et al. 2006). Esta correlación negativa puede estar explicada por depredación sobre especies nativas (Macchi et al. . 1999, 2007, Ortiz-Sandoval 2007, Vigliano et al. . 2009) e interferencia en la conducta dietaria (González 2005, Lattuca et al. . 2008, Macchi et al. . 1999, 2007, Ortiz-Sandoval 2007, Soto et al. . 2001, 2006). Por ejemplo, se ha encontrado sobreposición de dietas entre G. platei y O. mykiss , con una mayor intensidad de piscivoría de O. mykiss (Ortiz-Sandoval 2007). Igualmente, las especies nativas G. platei y B. bullocki cambian significativamente su dieta en presencia de altas densidades de truchas en ríos y lagos (Ortiz 2007, González 2005). Además, se han citado cambios en historia de vida de G. platei como disminución en la tasa de crecimiento y la edad máxima alcanzada en poblaciones de lagos con dominancia de salmónidos versus lagos donde la dominante fue la especie nativa (Ortiz 2007). Esto afecta no solo la diversidad (riqueza y abundancia) sino, la posición trófica, la diversidad morfológica (Milano et al. 2002, 2006) y genética de las especies nativas.

12

Análisis con enfoque multiescalar.

En este estudio utilizamos distintas escalas de análisis de los posibles efectos de los procesos de invasión de salmónidos sobre la fauna nativa en Chile.

A gran escala, analizamos la distribución de Galáxidos para probar sus cambios a en el centro y sur de Chile. Para esto, evaluamos la magnitud de los cambios en la distribución (expansiones o contracciones) de las especies más abundantes mediante la comparación de su distribución histórica y actual. Además analizamos las relaciones de la distribución y abundancia de Galáxidos, con la incidencia de los salmónidos. Nuestro objetivo es proporcionar la información necesaria para entender cómo las introducciones de peces han afectado a la distribución de especies y para comparar las respuestas entre las especies. Además, analizamos la ictiofauna en los lagos en un amplio rango latitudinal para determinar los gradientes latitudinales y de altitud de la riqueza de especies, y así, determinar si los patrones difieren entre especies nativas e introducidas. Finalmente nos preguntamos si el dominio de salmónidos y la conectividad del hábitat pueden explicar los patrones de riqueza de especies y la composición de la comunidad, y si sus funciones varían con la latitud. Esperamos que la información resultante de este estudio contribuya a retrasar los procesos de homogeneización de esta fauna única, y ayudar en el diseño de planes de conservación y acciones para las especies de peces nativos en el centro y sur de Chile.

A una menor escala, se estudió la cuenca del río Serrano en la Patagonia chilena (50º30’ a 51º30’), para determinar los efectos de la invasión de salmónidos sobre la distribución local e historia de vida de Galáxidos, además de la relación con la protección de áreas silvestres protegidas. En la cuenca existen evidencias de la presencia tanto de Galáxidos como de salmónidos (Correa y Gross 2007, DGA 2004, McDowall 1971a, 1971b, 2006, Soto et al. 2006). La reciente expansión de la acuicultura hacia el Sur (Buschmann & Fortt 2005) y el creciente fomento a la pesca deportiva, sumado a los nuevos registros de invasión de salmón Chinook (Correa & Gross 2007), hace esperable un incremento en la carga de especies salmonídeas en la cuenca, aumentando también las amenazas para la conservación de las especies nativas. Finalmente, esta cuenca se encuentra en gran parte protegida (51% de su superficie total) por dos grandes reservas naturales los Parques Nacionales Bernardo O’Higgins y Torres del Paine (DGA, 2004). Este último, fue reconocido internacionalmente en 1978 por la UNESCO como Reserva de la Biosfera, siendo su principal función la conservación de paisajes, ecosistemas, especies y diversidad genética. Esta condición de área protegida podría facilitar la conservación de la fauna íctica nativa, sin embargo, poco se sabe del estado de esta fauna, ni de los criterios necesarios para implementar planes eficientes de protección; sobre todo en un área enfocada al resguardo de la fauna terrestre más que la acuática. Este escenario, hace de esta cuenca un lugar muy interesante para la investigación de la fauna íctica dulceacuícola nativa .

13

HIPÓTESIS DE TRABAJO:

Basado en que la distribución y diversidad de la ictiofauna nativa en la Patagonia fueron determinadas por procesos naturales, los cuales estarían siendo modificados por procesos de invasión de salmónidos, esperamos encontrar una fuerte relación entre niveles de dominancia de salmónidos y distintos patrones de distribución, comunitarios y poblacionales de la fauna íctica nativa. Específicamente, esperamos encontrar a gran escala, cambios en la distribución y una directa relación entre la diversidad de especies nativas y la dominancia de salmónidos. A una escala menor, esperamos encontrar menor diversidad, abundancia, amplitud de nicho trófico y cambios en la posición trófica de Galáxidos, producto de la dominancia de salmónidos. Por otro lado, esperamos que las áreas silvestres protegidas no muestren efectos significativos sobre la dominancia de salmónidos y abundancia de especies nativas, respecto a las no protegidas.

OBJETIVOS.

Objetivo general:

La presente investigación tiene por objetivo determinar el efecto de la invasión de salmónidos sobre la ictiofauna nativa en la Patagonia chilena, en un enfoque multiescalar, analizando la diversidad, distribución y parámetros poblacionales de Galáxidos en relación con la dominancia de las especies invasoras. Además, poner a prueba la efectividad de las áreas silvestres protegidas en la conservación de la fauna nativa, identificar las principales amenazas y proponer medidas de manejo.

Objetivos específicos:

1.- Determinar la influencia de la invasión de salmónidos en la distribución de Galáxidos en la Patagonia chilena.

2.- Determinar patrones de diversidad ictiofaunistica a macroescala y su relación con la invasión de salmónidos en sistemas lacustres de la Patagonia chilena.

3.- Evaluar la coexistencia de salmónidos y Galáxidos en la cuenca del río Serrano, y la eficiencia de las áreas silvestres protegidas en la conservación de la fauna nativa.

4.- Evaluar las potenciales interferencias troficas entre salmónidos y Galaxidos en la cuenca del rio Serrano.

14

CAPITULO I: CHANGES IN THE DISTRIBUTION OF NATIVE FISHES IN RESPONSE TO INTRODUCED SPECIES AND OTHER ANTHROPOGENIC EFFECTS.

ABSTRACT

Globally one of the major threats to the integrity of native faunas is the loss of biodiversity that can result from the introduction of exotics. Here we document recent changes in the distribution of five common fish species that are linked to introductions in Chile. Location: Chile from 28ºS to 54 ºS. We assess the extent of changes in distribution of galaxiid species by comparing their historical and current distributions based on the results of the most extensive survey of freshwater fishes in Chile to date, a range that encompasses the full latitudinal and elevational range of the Galaxiidae in Chile. We test for relationships of the distributions and abundances of native fishes with the incidence of introduced species. Latitudinal range of Galaxias maculatus has declined 26%, and most of this reduction occurred in the northern part of its range. Aplochiton taeniatus and Brachygalaxias bullocki , have experienced reductions (8 to 17 % loss) in total drainage area occupied, and they have disappeared from, or are now extremely difficult to find, in latitudes 36º to 41ºS, coincidently with areas of urban growth and intense economic activities. The distribution range of Galaxias platei has, instead, increased considerably. In northern basins, G. maculatus has apparently been replaced by an introduced poecilid Gambusia sp. High elevation systems remain dominated by native G. platei , whereas systems at intermediate elevations, especially rivers, are now dominated by introduced salmonids. Within drainages, native galaxiids remain abundant where exotic salmonid abundance is low. We suggest that negative interactions between introduced and native fish are responsible for some of the range reductions among Galaxiidae in Chile. The severity of the impacts varies with latitude and altitude and is probably related to temperature. The effects of Gambusia sp. are restricted to warmer systems. Native fish also appear to have found temperature refugia from salmonids; impacts are low in the warmer northern and coastal, as well as in high altitude, relatively cold systems. Native fish also appear less vulnerable to salmonids in lakes than in rivers. This study identifies watersheds critical for the conservation of biodiversity within the Galaxiidae.

Evelyn Habit, Priscila Piedra, Daniel Ruzzante, Sandra Walde, Mark Belk, Victor Cussac, Jorge Gonzalez & Nicole Colin. 2010. Ecology and Biogeography, 19(5), 697–710.

15

INTRODUCTION

Globally, one of the major threats to freshwater fish diversity is the homogenization that results from the introduction of non-native species into isolated, evolutionarily naive faunas (Rahel, 2002). Most successfully introduced species are top predators and/or strong competitors with broad environmental tolerances. In contrast, freshwater fishes from isolated faunas are often poor competitors and evolutionarily naive prey (Mills et al., 2004; Nannini & Belk, 2006). In the western U.S., for example, the introduction of salmonids, carp, and mosquitofish has caused declines in abundance or local extinctions of many native freshwater fishes via predation and competition (Rahel, 2000). Thus, homogenization results from the negative effects of introduced species on populations and distributions of native fishes and the eventual dominance of a relatively small number of introduced species over large areas (Rahel, 2002).

A large number of species have been introduced to freshwater ecosystems worldwide in the first half of the 20th century, long before adequate census information on the native fauna was available. Many studies have now documented changes in abundance and range or distribution of native fauna due to introduced species, and there is good evidence of such effects in the temperate waters of North American, European and New Zealand/Australian systems. Much less is known about the effects of introduced species on the native fish of temperate South America, where almost all lakes and rivers now have one or more exotic species of fish.

The native fish fauna of temperate South America is relatively depauperate (approximately 40 species), and the highest biodiversity within the region is found west of the Andes, in Chile. The Galaxiidae (Osmeriformes), the most species-rich family of freshwater fish in Chile, comprises over 50 species distributed throughout the southern hemisphere (South America, South Africa, Australia and New Zealand; McDowall, 1971, 2006). Eight species belonging to three genera ( Galaxias , Brachygalaxias and Aplochiton ) are found in Chile, five of which are endemic. Two of the endemics, Brachyglaxias bullocki and Aplochiton taeniatus , are widely distributed, but the other three ( G. globiceps, A. marinus, B. gothei ) have very restricted distributions (McDowall, 1971; Campos, 1973; Berra et al., 1995; Dyer 2000). Three species ( G. platei, G. maculatus, A. zebra ) also occur east of the Andes (Cussac et al., 2004; see e.g. Ruzzante et al., 2008; Zemlak et al., 2008). Chilean freshwaters contain a historically isolated fauna, with the Atacama Desert forming a barrier in the north, the Andes to the east, and the Pacific Ocean on the west and south. This system thus offers the opportunity to examine the response of species that evolved in species-poor environments to the introduction of broad generalist exotics such as salmonids. It also allows for a comparison of the response of the Chilean fauna to that of a

16 similar fauna in New Zealand, and to that of the much less isolated systems of North America.

Freshwater ecosystems today face challenges worldwide, and in Chile, the impact of humans currently extends to almost all aquatic systems. Many lakes and rivers are under continuing threat from pollutants and intensive forestry (Campos et al., 1998; Goodwin et al., 2006; Habit et al., 2006b). Several major river systems have been dammed, and increasing demands for energy to sustain economic growth are fueling plans for new dams, with some planned for the more pristine rivers in the south. Various exotic salmonids have been introduced into Chilean waters over the past century. The first recorded introduction of salmonids occurred in 1883, when Oncorhynchus mykiss (rainbow trout) were stocked in a small coastal basin at 37ºS. After 1905, there were additional and frequent introductions of O. mykiss , from the Rapel River to the Valdivia River (34 º – 39 º S). The first recorded introductions of Salmo trutta (brown trout) were in 1910, when both brown and rainbow trout were introduced from the Aconcagua to the Bueno Rivers (32º - 40º S). Since that time both species have been introduced into nearly all basins, including those in the southern region (Magallanes, 53º) beginning in 1927, and in northern basins such as the Loa River basin (21º S) beginning in 1949. Stocking has continued to the present in most systems (Basulto, 2003).

The history of fish introductions in Chile is not restricted to salmonids. Gambusia sp., a cyprinodontid, was introduced to the region from unknown sources for mosquito control in the early part of the twentieth century, and current populations in Chile are known to have been present by 1937 (Welcomme, 1988). Along with O. mykiss and S. trutta , Gambusia affinis is included among the 100 "World's worst" invaders (Lowe et al., 2000).

Impacts of introduced salmonids are just starting to be evaluated, but several lines of evidence suggest they are likely to be substantial, particularly in Patagonia, where other human activities are not yet significantly developed. Firstly, salmonid biomass in some systems exceeds that of native fish (Soto et al., 2006; Pascual et al., 2002, 2007), and secondly, trout have been documented to prey on galaxiids (Macchi et al., 1999, 2007; Arismendi et al., 2009). Local impacts of salmonids may vary among habitat types and fish communities. For example, the impact of introduced trout on native fish may be greater in streams than in lakes (Soto et al., 2006) and greater in landlocked than in anadromous populations (Jowett et al., 1998).

In this study we use newly acquired information on the distribution of the Chilean Galaxiidae to test for large scale changes in this important component of the endemic fish fauna of central and southern Chile. We first assess the extent of changes in distribution (expansions or contractions) of the most abundant species by comparing their current and

17 historical distributions. We then test for relationships of the distributions and abundances of native fishes with the incidence of introduced salmonids. We conduct this detailed analysis for two regions: the Aysén Region (Lat. 43°45' to 49°15' S) and the Valdivia river basin (~Lat. 39°50' S), an important basin of the South-Central region of Chile. Our goal is to provide information necessary to understand how fish introductions have affected species distributions and to compare responses among species. Such knowledge should aid in the delay of the process of homogenization of this unique fauna, and help in the design of informed conservation plans and actions for native freshwater fishes both worldwide and in particular in central and southern Chile.

METHODS

Historical and current distribution of the Galaxiidae

We constructed historical distributions by assembling geo-referenced data for all species of Galaxiidae in Chile, using data from (1) the recent compilation by Cussac et al. (2004), (2) additional records cited by Arratia (1981) and Vila et al. (1999b), (3) unpublished records (Hugo Campos et al., unpublished) for Torres del Paine National Park (51ºS). The time frame of the historical data was 1967 to 2002.

Current distributions were determined through extensive sampling (2003 to 2009) from the Huasco River in the north (28ºS) to Tierra del Fuego in the south (54ºS). Sampling was stratified by microhabitats (e.g. zones of varying substrates, water depths and current velocities within the river channel, generally on a scale of tens of meters in extent; Maddock, 1999). We used multiple gear types to increase the capture efficiency in all microhabitats, counting individuals of all species (native and non-native) captured at each sample site. We used a backpack electroshocker and seine net (2 mm mesh) to sample rivers, streams and lake littoral zones, and gillnets (15, 20, 30, 50, 60, 70 and 120 mm mesh) placed on the bottom at different depths, to sample lakes. Differences in sampling efficiency of gears may arise in different habitats and could confound estimates of abundance; we thus only use presence and relative abundance data for these analyses. Our dataset thus consisted of a total of 287 sampling localities: (1) 211 locations within 24 main Andean (Chilean) drainage systems, (2) 31 locations within 22 coastal basins, (3) 10 rivers and lakes from Chiloe Island, (4) 14 locations within 8 drainages in Tierra del Fuego, and 21 locations in islands of Magallanian Fjords. In addition to our data, we used reports from Arismendi et al., (2006) for Tierra del Fuego (53º to 54ºS). Despite the broad geographic coverage, we failed to collect individuals of the three rare species: Galaxias globiceps , Aplochiton marinus and Brachygalaxias gothei . Collection of these species will require additional, targeted sampling.

18

Data were incorporated to the GIS software Arc View 3.2. (ESRI), on digitized maps of river drainages at 1:50,000 scale. For each species, we plotted all historic and all current locations. We distinguish locations with historical records that have not been visited in our recent sampling from those that were visited and the focal species was not found. The distributional range of a species was measured in two ways. First, we measured the area of each drainage system where the species was recorded (historically and currently). Total area of occupancy was the sum of the occupied drainage areas. The difference between historic and current drainage area occupied was used to infer range expansion or contraction. Second, we looked for north-south shifts in distribution, measuring the distance from the northern-most occupied drainage to the southern-most occupied drainage.

Distribution of native fishes relative to that of introduced salmonids

We used three different data sets and analyses to examine the distribution of the native Galaxiidae in relation to that of the introduced salmonids. We first examined the latitudinal variation in dominance of native fish versus salmonids, by plotting proportions of galaxiids and trout over latitudinal gradients for coastal systems (34º to 42ºS) and Andean basins (36º to 47ºS). We pooled data within 2 degrees latitude. We also examined latitudinal variation in biomass of galaxiids versus salmonids in Andean basins from 39º to 47ºS. For these analyses we included four galaxiid species ( G. maculatus , G. platei , A. taeniatus and A. zebra ), since they are widespread along Chile like salmonids ( S. trutta and O. mykiss ).

Secondly, we used data from the Aysén region (43°45' to 49°15'S, 109,025 km 2) to test for variation in salmonid dominance in lakes vs. rivers and as a function of elevation. To standardize estimates of abundance and biomass we only use relative abundance to avoid differences in sampling gear efficiency and site effort (Clarke & Gorley, 2005). We compiled data for 23 lakes with 32 sampling locations and 19 river systems with 69 sampling locations from the Palena, Cisnes, Aysén, Baker and Pascua drainages. Data were from our survey, and from recent publications (CH2MHILL 2000, Niklitsheck & Aedo 2002; CEA 2005). We classified lakes into four elevation categories (m.a.s.l.): (1) 0 – 199, (2) 200 – 299, (3) 300 – 399, and (4) > 400. Elevations were obtained directly in the field by an Etrex GPS or using the software Google Earth 4-beta (http://earth.google.es/).

Because of differences in effort among sampling sites, we analyzed relative rather than absolute abundances. We used ANOSIM (Clarke et al., 2005), a procedure that uses permutation and randomization methods based on similarity matrices, to test for differences in species composition between lakes and rivers and among elevation categories. Proportions were fourth root-transformed, and similarity matrices were calculated using the Bray-Curtis measure. The Similarity Percentage procedure (SIMPER) (standardized, 4th

19 root transformed data) was used to identify the species most important in generating differences among elevations (PRIMER v.6, Clarke & Gorley, 2005).

Thirdly, we used samples from 35 locations along the San Pedro River (40 km length) from the Valdivia basin (39°52’ to - 39°50’S, 10,245 km 2) to examine the relationship between abundance of the native Galaxiidae and that of the two most common salmonids in Chile (S. trutta and O. mykiss ) at a smaller spatial scale. This river contains all the galaxiid species present in Chile except the rare G. globiceps, A. marinus and B. gothei. All locations were sampled weekly between November 2005 and February 2008. Samples were collected by electrofishing, and abundance data are expressed as capture per unit effort (CPUE in N * 100 / hr * m 2).

RESULTS

Species distributions and habitat

Galaxias maculatus : A comparison of historical and current ranges indicates a reduction of 26% in the latitudinal range of G. maculatus , from a historical north-south distribution of 3,118 km, to a current range of only 2,319 km (Table 1). The loss occurred mainly in the north; G. maculatus was found in the Huasco and Elqui Rivers 30 years ago, but we did not find it in any of 6 locations in the region (Fig. 1). Our sampling, however, did register several new locations for the species in Chile, mostly between the Maullín (41ºS) and Serrano (51ºS) River basins. The total drainage area currently occupied by G. maculatus is higher than that registered by historic records (Table 1), due to the more intensive sampling. We found potentially landlocked populations in the middle of the latitudinal range, mainly in lakes of the Valdivia River basin (Fig. 2).

Habitat: In rivers, G. maculatus were found in shallow, slow current zones along river edges, on substrates ranging from mud to boulders, with or without aquatic vegetation. Larger individuals (> 12 cm SL) tended to inhabit slightly deeper (> 50 cm) zones. During periods of high discharge, G. maculatus used river floodplains extensively. In lakes, G. maculatus inhabited littoral zones, sometimes forming large shoals, and were absent, or nearly so, from lake tributaries.

Galaxias platei : New records from Chiloé Island, and from the Puelo, Palena, Cisnes, Cuervo, Baker, Serrano and Tierra del Fuego River basins, increased the total drainage area known to be occupied by G. platei by 243% (Table 1, Fig. 1). We also increased its latitudinal range (Table 1) in 459 km due to new records in Chilean Tierra del Fuego. Arismendi et al., (2006) reported this species as absent in Tierra del Fuego, after sampling the littoral areas of Lakes Lynch and Rasmussen by electrofishing. We suspected, however, that G. platei might be present at greater depths in lakes of this region, as it was found at 30

20 to 40 m depth in Pacific draining lakes on the Argentinean side of the island (Yehuín, Margarita and Escondido) (Milano et al., 2006; Ruzzante et al., 2008). We found G. platei in the benthic zone of lakes Blanco and Deseado at 50 m depth.

Habitat: Galaxias platei inhabits rivers and lakes. In rivers, G. platei usually occupies shallow (< 50 cm depth), low current habitats (< 0.3 m s-1). Larger individuals tend to use deeper habitats, though the largest individual captured in the San Pedro River (16.8 cm SL) was in a shallow pool of 50 cm depth. Galaxias platei occurs on varied substrata, but appears to prefer habitats with refuges such as trunks and big boulders. During high flow, it moves to the floodplains. In the south (Aysén) G. platei is relatively uncommon in rivers but abundant in lakes, where it is commonly found in littoral zones with abundant large woody debris and coarse particulate organic matter from the native riparian forest. Only large adults (> 20 cm) were found on muddy bottoms at depths greater than 8 m. The largest individuals (35 cm TL) were caught in the Cochrane Lake (Aysén River basin) at 80 m depth.

Brachygalaxias bullocki : A comparison of historical and current ranges indicates a reduction of 17% in the latitudinal range of B. bullocki (Table 1), due to its current absence in the Biobío, Imperial, Toltén and Bueno River basins (Fig. 1). Locations in the Biobío River with historic records of B. bullocki are currently dominated by an introduced species, Gambusia holbrooki (Habit et al., 2006b). Brachygalaxias bullocki remains abundant in the Maullín River basin and in the rivers of Chiloe Island.

Habitat: B. bullocki typically inhabits streams and rivers with high levels of humic acids, lowland rivers with native riparian forest and large floodplains. It is found in riverine habitats with little or no current, and is completely absent from rapids or riffles. We found only one specimen in all the lakes we sampled, in Riñihue Lake (Valdivia River basin).

Aplochiton taeniatus : We collected A. taeniatus from the Tirúa River (38º 22`S), extending the northern limit for this species from the historical limit in the Toltén River (39º 17´S; McDowall 1971). However, the drainage area occupied by A. taeniatus (Table 1) has declined in approximately 8%. Most of the loss is due to its absence (or extreme rarity) in the Bueno River basin (Fig. 1). A. taeniatus was not found in Chilean Tierra del Fuego. In south-central Chile, A. taeniatus is currently most abundant in the Valdivia and Maullín River basins and the northern rivers of Chiloé Island, but it is more frequent in the southernmost distribution range, including the Serrano River Basin and islands of the Magallanic Fjords.

Aplochiton zebra : We increased the area occupied by A. zebra in approximately 33% of what was reported historically (Fig.1, Table 1), due mainly to new records in the Serrano River Basin and island of the Magallanic Fjords. It seems to be absent from the Biobío

21

River Basin, where it was reported in the Galletué Lake by Campos et al., (1993) and where we did not sample. The abundance of A. zebra was also low in other central river basins (e.g., Valdivia River), but numbers remained high in the lakes of the Aysén Region and Serrano River basin. Aplochiton zebra tended to be more abundant than A. taeniatus in the south, while the latter was more common in the north.

Habitat: Both species of Aplochiton use shallow rapids or riffles in rivers as juveniles, and move to lakes or deep river pools as adults. Both species were rare in lake tributaries, but common in lake littoral zones close to tributaries. Neither species uses floodplain habitats. Specimens of A. taeniatus found in rivers were always close to lakes or estuaries, and those captured very close to the estuaries of the Maullín and Tirúa Rivers were probably diadromous.

In sum, four of the five species of Galaxiidae that we surveyed in Chile show evidence of range contraction or reduction in abundance in some regions over the past few decades. G. maculatus is no longer found in the northern part of its historical range, A. zebra and A. taeniatus appear to have reduced abundances and B. bullocki , the only galaxiid species restricted to the central area of Chile (north of Patagonia), showed a severe contraction in occupied range. The significant expansions that we document are almost certainly, due to the high intensity of the recent sampling effort. The drainage area currently occupied by G. maculatus and G. platei is larger than that reported historically, and we report several new locations for A. zebra and A. taeniatus .

Latitudinal patterns

Species richness within the Galaxiidae varies with latitude, and the highest species richness occurs in the central part of the group’s range, between 39º and 42ºS (Fig. 3).

Coastal range (34º to 42ºS): Of the three native species present in the low elevation coastal basins, G. maculatus was numerically dominant at all latitudes (Fig. 3), and more common than the introduced salmonids. Both species of Aplochiton were present, but only in the central coastal drainages (34º to 36ºS). Galaxias platei was absent from the coastal rivers. Rainbow trout ( O. mykiss ) was more abundant than brown trout ( S. trutta ).

Andean basins (36o to 47oS): Strong latitudinal patterns were found for the Andean basins (Fig. 3). The four most common species of Galaxiidae co-occur between 39º and 44ºS. Galaxias maculatus was the only galaxiid found in northern drainages, and G. platei dominated the native fish assemblage in southern basins (Fig. 3). The two species of Aplochiton were most common in the centre of the range, with only A. zebra found south of 44ºS.

22

Introduced salmonids were more abundant than galaxiids in both the most northern and southern basins, making up to 85% of the entire fish community south of 44º S. (Fig. 3). Within the salmonids, O. mykiss was relatively more abundant in the north, while S. trutta dominated southern basins. Salmonids also dominate in terms of biomass. In Andean basins, salmonids (rainbow and brown trout) make up 75% of the total fish biomass from 39º to 41ºS, 89% from 42º to 44ºS and 92% from 45º to 47ºS. In the same latitudinal range, G. maculatus , A. taeniatus and A. zebra reduce their cumulative biomass from north to south; meanwhile G. platei increases its own biomass from 2% to 7% in the same gradient.

Relationships with introduced salmonids

Aysén region: A total of seven native and six introduced species were collected in the Aysén region (Table 2). The most widespread native species was G. platei , present in 41% of the sampled locations, and the most common introduced species, S. trutta , was found in 82% of the locations.

Individual lakes had between 1 and 7 species, and higher elevation lakes tended to have fewer species (Fig. 4). Species composition also varied significantly with elevation

(ANOSIM, R global = 0.22 p = 0.001), with native species, especially G. maculatus and G. platei , dominating at the lowest elevations, and G. platei dominating in high elevation lakes (Table 3). Salmonids made up more than half of the total fish community at intermediate elevations (Table 3). Pairwise comparisons showed significant differences in species composition between the first two elevation categories (ANOSIM, R=0.26; p = 0.002), and between categories 2 and 4 (ANOSIM, R=0.4; p=0.007).

Salmonids dominated rivers, while lakes had similar numbers of native and introduced species (Fig. 5). Rivers and lakes differed in species composition (ANOSIM, R global =0.24 p=0.001); with higher abundances of G. platei in lakes, and more S. trutta and O. mykiss in rivers.

San Pedro River: Four species of Galaxiidae ( G. platei , G. maculatus , A. taeniatus and B. bullocki ) were collected in sufficient numbers from the San Pedro River to examine their relationships with salmonids (Fig. 6). Locations with high abundances of salmonids ( S. trutta or O. mykiss ) had very few galaxiids, and high numbers of Galaxiidae were only found in locations with few salmonids (Fig. 6).

DISCUSSION

In this study we present evidence for important changes in the distributions of the five most abundant and widespread species within the Galaxiidae present in Chile. Distributions were mapped using historical records obtained from the literature (1960's and onward) and

23 current records obtained largely from our own sampling efforts (2003 to 2009). The latitudinal range of Galaxias maculatus has shrunk by approximately 26%, and this reduction has taken place in the northern part of its range. Two other species, Aplochiton taeniatus , and Brachygalaxias bullocki , have experienced reductions (8 to 17 % loss) in total drainage area occupied, and they have disappeared from, or are now extremely difficult to find, in the Biobío, Imperial, Toltén, and Bueno river basins (latitudes 36o to 41oS). The distribution range for G. platei has instead, increased significantly, primarily as a result of our extensive sampling, mainly in the Patagonian Region, where historical records were scarce.

Our results suggest that introduced species are generating local effects on galaxiids, but the relative role played by the presence and abundance of exotic species on the distribution range constriction of the native species is difficult to estimate. Locally, we found that the current patterns of native and exotic fish distributions support the hypothesis that introduced species are having a negative effect on native fish abundances, though effects vary spatially. Low elevation, coastal systems, common between latitudes 34o and 42oS, remain dominated by the native G. maculatus , though O. mykiss constitutes up to 20% of the current fish assemblage. However in northern Andean basins, G. maculatus has apparently been replaced by introduced mosquitofish, Gambusia spp. High elevation (Andean) systems also remain dominated by native galaxiids ( G. platei ), whereas systems at intermediate elevations, especially rivers, are now dominated by salmonids. Within drainages, galaxiids remain abundant where salmonid abundances are low. Below we examine the implications of these findings for the conservation and management of aquatic biodiversity in general and of southern Chile in particular.

Efforts to predict and document the consequences of human activity for native fish in Chilean watersheds have historically been hampered by lack of data and inadequate geographical coverage (but see Habit et al., 2006b; Soto et al., 2006; Arismendi et al., 2009). This study represents the most extensive survey of freshwater fishes in Chile to date, with samples from a total of 287 lakes and streams, from 56 drainages, a range that encompasses the full latitudinal and elevational range of the Galaxiidae west of the Andes. While some lines of evidence strongly suggest that the past introduction of exotic species has had local impacts on native fish communities in Chilean waters, introductions in most areas coincide with many other human activities (Campos, 1998; Vila et al., 1999a; Habit et al., 2006a, b; Soto et al., 2006, 2007), making it difficult to clearly delineate the relative importance of each factor (Leprieur et al,. 2008). Negative effects of multiple human activities within a region (e.g., pulp and mill industry, irrigation, replacement of native forest, hydropower) have been shown at individual (Orrego et al., 2005) and community levels (Goodwin et al., 2006; Habit et al., 2007). These impacts are of special concern as

24 the zone of highest native fish species richness (34º to 38º S) in Chile coincides with areas of urban growth and intense economic activities related to resource extraction.

Two regions of decreased species richness

Our study identified two regions of particular concern. One significant distributional change was the shift in the northern limit of the range of G. maculatus . Historic records for this species in northern areas come from Arratia (1981), who cited a personal communication of Cabrera in 1975 for populations inhabiting the estuaries of the Huasco and Elqui rivers (28º and 29º S). Although we sampled both estuaries and also areas upstream in both rivers, we did not find G. maculatus . The cause of this disappearance is not known, but we suggest three non-exclusive hypotheses. First, G. maculatus is a temperate species, and the northern limit for coastal, diadromous populations may fluctuate, depending on sea water temperature. The Chilean coast north of Lat. 43º S is dominated by the cold north-flowing Humboldt Current (Acha et al., 2004), the strength of which is affected by ENSO (El Niño Southern Oscillation) fluctuations. Strong positive phases can result in the displacement of tropical water southward (McPhaden, 2001), and the 4 ENSO events between 1980 and 2000, which included two very intense events (1982-83 and 1997- 98; Bello et al., 2004), could have raised estuary temperatures in northern rivers. A second contributing factor could be changes in flow, sediment and temperature regimes following the construction of dams for irrigation purposes in both rivers. The dams have significantly altered the aquatic environment, causing changes in river temperature and decreased macro- invertebrate abundances (Habit, personal observations). The third factor may be the presence of exotic species. Almost all the northern rivers we sampled were dominated by the exotic Gambusia spp., which occupies the same habitat types as G. maculatus . In New Zealand, Gambusia affinis was shown to cause high mortality of G. maculatus in experimental trials (Rowe et al., 2007). These hypotheses are clearly non-exclusive and the factors may have acted synergistically to produce dramatic reductions in the abundance of G. maculatus in the northern part of its range.

The second region of concern lies between 36º and 41º S, where several major watersheds (Biobío to Bueno) appear to have lost at least two species of Galaxiidae over the past few decades ( Aplochiton taeniatus and Brachygalaxias bullocki ). This region, which includes the cities of Concepción and Temuco, has experienced multiple pressures from rapidly expanding economic activity, much of which occurs in or near rivers: (i) conversion of native forest to plantations with exotic species, (ii) increasing pulp and mill production and (iii) hydropower generation (Parra et al., 2004). Introduced species probably also contribute to species reductions and losses. High numbers of salmonids now occupy headwaters. The disappearance of Brachygalaxias bullocki from several drainages near Concepción (36º-38º S) is most easily explained by the presence of Gambusia sp. Within the Biobío River basin

25

Gambusia appears to have completely replaced B. bullocki in its favored habitats (Habit et al., 2006b); similar patterns were seen in New Zealand and Australia after the introduction of Gambusia (Rowe et al., 2007).

Introductions and their effect on the native fish fauna

Non-native fish species have been introduced into Chilean freshwaters since the early twentieth century, doubling the total number of freshwater fish species in the country. Most were introduced for recreational purposes or as an accidental side-effect of aquaculture activity (Welcomme, 1988), and are now established as self-sustaining populations. Gambusia was introduced for mosquito control in Central Chile (Campos, 1973; Welcomme, 1988). Our analysis suggests significant effects of two groups of exotic fish on galaxiids in Chile, the poecilid Gambusia spp. and salmonids. Gambusia is restricted to warmer waters, but its effects on G. maculatus and B. bullocki appear dramatic. The watersheds formerly occupied by these species are subject to the effects of other human activities, however, and whether extirpation from affected drainages would have occurred in the absence of these other impacts is uncertain.

Salmonids are a much more widely distributed exotic group, and the most abundant species in Chile are rainbow trout ( Oncorhynchus mykiss ) and brown trout ( Salmo trutta ), both of which sustain recreational fisheries. Salmonids have been introduced into cool temperate waters world-wide, and negative effects on native fish, especially on galaxiids, have been reported or inferred in virtually all locations (McDowall, 2006), but especially in New Zealand (Townsend & Crowl, 1991; Flecker & Townsend, 1994; McIntosh, 2000; Simon & Townsend, 2003) and Tasmania (Hardie et al., 2006). One of the most important effects on river Galaxias in New Zealand was population fragmentation, with local extirpation in larger rivers and viable populations remaining only in isolated headwaters inaccessible to salmonids (Townsend, 1996). Various mechanisms contribute to the dominance of salmonids in southern hemisphere fish communities, including competition, predation, and indirect effects (Simon & Townsend 2003). There is good evidence for predation by introduced trout on native galaxiids from both sides of the Andes (Ruiz & Berra, 1994; Macchi et al., 1999, 2007). In our surveys we found G. maculatus and G. platei in the stomachs of O. mykiss and S. trutta as small as 12 cm and in the stomachs of juvenile O. tshawytscha no larger than 9 cm long .

We suggest that negative interactions between introduced and native fish are likely responsible in part of the range reductions of Galaxiidae in Chile, but these effects are difficult to separate from those of other human activities. For example, we found A. zebra in lower abundance in lakes with salmonids in the Serrano River drainage in the Torres del Paine region (Gonzalez et al., unpublished observations). This region has experienced very

26 little direct human impact; it is a remote region in the extreme south of the country, with limited road access, sparse human settlement and little economic activity. The only major environmental change over the past 100 years in rivers and lakes of this region is the introduction and establishment of salmonids, especially S. trutta. Galaxiids, where they are still present in other southern watersheds, are already far less abundant than salmonids. Future development in the region such as expansion of salmonid aquaculture and plans to build five hydroelectric dams in the Baker and Pascua Rivers (see www.hidroaysen.cl), is likely to push abundances of remaining populations even lower.

Distributional patterns within watersheds suggest that there is considerable spatial variation in vulnerability to salmonid impacts. The disproportionate dominance of salmonids in streams and rivers compared to lakes is similar to that reported in another Chilean study (Soto et al., 2006), and coincides with evidence from New Zealand and Australia where the strongest negative effects of salmonids on Galaxiidae also occur in streams (McDowall, 2003). Trout, especially S. trutta , preferentially lay their eggs in the riffles of small to medium-sized streams (Raleigh et al., 1986), and exclusion of galaxiids from stream habitats may be due to interactions early in the life cycle. Galaxiids that spawn in the littoral zones of lakes probably experience lower predation rates, and the shallow littoral areas and deep benthic zones of lakes likely provide refugia from trout predation at other life stages. Rivers with large floodplains, like the San Pedro River, probably also provide refugia for galaxiids. The variation in the degree of salmonid dominance among lakes at different elevations indicates that galaxiids may have a temperature refuge; summer temperatures reached in many low elevation lakes may be sub-optimal for salmonids, and low visibility and extreme cold may reduce salmonid impacts in high elevation glacial-fed lakes. A temperature refuge probably also explains lower abundances of salmonids in central and northern watersheds (Soto et al., 2006), and coastal basins (Fig. 3). Galaxias maculatus dominates in coastal rivers despite the presence of trout, and G. platei is abundant in both rivers and lakes in central and northern watersheds. In the cooler south, however, G. platei has become predominantly a lake species, as is currently also seen east of the Andes in southern Argentina (Milano et al., 2006; Ruzzante et al., 2008; Zemlak et al., 2008). Thus salmonids greatly reduce the abundance of galaxiids in certain habitats – especially cool running water. Galaxiid species persist by occupying less favored habitats; G. maculatus remains abundant in coastal systems, G. platei in warmer central rivers or in southern lakes. It is important to note, however, that if these habitats are to maintain healthy galaxiid populations, they must be protected from the effects of other types of human activity. The cause of the range reductions in A. taeniatus is not clear, especially as little is known about their ecology. Microhabitat use overlaps greatly with that of S. trutta and O. mykiss ; all inhabit shallow rapids or riffles in rivers as juveniles and move to lakes or deep river pools as adults. None use floodplain habitats. The near absence of Aplochiton from the

27 small tributaries of rivers and lakes suggests interference from trout, especially as it remains common in habitats not used by trout, i.e., the littoral zones of lakes.

The evidence above collectively suggests that galaxiids have temperature refugia from salmonids in both, the northern portion of their range, where summer water temperatures are relatively warm and in the southern portion of their range at high altitude, where winter water temperatures may be too low for salmonids. Galaxiids also appear to have found refugia from salmonids in lakes, as has been found by others in northern and Argentinean Patagonia (Soto et al., 2006; Pascual et al., 2007; Arismendi et al., 2009). In general, however, the various freshwater fish species introductions in Chile (Salmonids, Poeciliids) seem to have caused declines in the abundance and distribution of native galaxiids, partially restricting their presence to more marginal habitats and potentially leading to increases in population fragmentation but not to total extinctions. Worldwide evidence of the primary role for invasive aliens in total extinctions remains limited (Gurevitch & Padilla, 2004), but synergistic effects of introduced species and other human actions could increase the likelihood of extinction.

Four areas of special conservation concern

Our study also identified four geographic areas of special conservation interest, where the effects of salmonids and human activity have been low enough that abundant galaxiid populations remain. These include the (1) coastal areas of the Maullín River basin and the Maullín River itself, which still hosts abundant populations of G. maculatus , B. bullocki , A. zebra and A. taeniatus , and probably G. globiceps , (2) the Valdivia River basin, which contains all the Chilean Galaxiidae, except the rare B. gothei and G. globiceps, (3) the northern rivers of Chiloé Island, and (4) several lakes in the Aysén region, including Lakes Yulton and Meullín, which are still salmonid free (Soto et al., 2006), Lake Thompson, where G. platei remains dominant to salmonids (Habit et al., 2006a), and Riesco Lake, which has a high species richness for its latitude (45ºS). None of these areas are currently protected and only one (coastal area of Maullín) has been proposed as an important area for vertebrate conservation (Tognelli et al., 2008).

REFERENCES

Acha, E., Hermes, M., Mianzan, W., Guerrero, R. A. , Favero, M. & Bava, J. (2004) Marine fronts at the continental shelves of austral South America Physical and ecological processes. Journal of Marine Systems, 44, 83– 105.

Arismendi, I., Gonzalez, J., Penaluna, B. & Sanzana, J. (2006) Preliminary assessment of the potential for sustainable trout fisheries in the Cóndor, Grande, and Sánchez basins of

28

Chilean Tierra del Fuego. Chapter: water quality, fish abundance and potential for fisheries. Laboratory Report. Universidad Austral de Chile, Puerto Montt, Chile.

Arismendi, I., Soto, D., Penalula, B., Jara, C., Leal, C. & León-Muñoz, J. (2009) Aquaculture, non-native salmonid invasions and associated declines of native fishes in Northern Patagonian lakes. Freshwater Biology, 54, 1135-1147.

Arratia, G. (1981) Géneros de peces de aguas continentales de Chile. Museo Nacional Historia Natural Chile, Publicación Ocasional, 4, 3-108.

Basulto, S. (2003) El largo viaje de los salmones: una crónica olvidada, Propagación y cultivo de especies acuáticas en Chile. Maval Limitada, Santiago, Chile.

Bello, M., Castillo, M., Maturana, J., Valenzuela, C. & Barbieri, A. (2004) Featuring ENSO 1997-2000 in central Chile. Gayana, 68, 48-53

Berra, T., Feltes, R.M., & Ruiz, V.H. (1995) Brachygalaxias gothei from South-Central Chile, a synonym of B. bullocki (Osteichthyes: Galaxiidae). Ichthyological Explorations of Freshwaters, 6(3), 227-234.

Campos, H. (1973) Lista de peces de aguas continentales de Chile. Noticiero Mensual Museo Nacional Historia Natural, 17, 3-14.

Campos, H., Dazarola, G. , Dyer, B, Fuentes, L., Gavilán, J., Huaquín, L., Martínez, G. Meléndez, R., Pequeño, G. , Ponce, F., Ruiz, V., Sielfeld, W., Soto, D., Vega, R. & Vila I. (1998) Categorías de conservación de peces nativos de aguas continentales de Chile. Boletín del Museo Nacional de Historia Natural (Chile), 47,101-122.

CEA (2005). Evaluación potencial de pesca y capacidad de carga de poblaciones de peces de interés para la pesca deportiva, en propiedades fiscales ribereñas en lagos y ríos de la XI Región de Aysén del General Carlos Ibáñez del Campo. Informe Técnico del Centro de Ecología Aplicada, Chile.

CH2MHILL, 2000. Informe de Línea de Base Proyecto Alumysa. In: Estudio de impacto ambiental proyecto Alumysa. Alumysa project. Coyhaique, Chile.

Clarke, K.R., Warwick, R.M., Somerfield, P.J. & Gorley, R.N. (2005) Change in marine communities: an approach to statistical analysis and interpretation, 3rd edition PRIMER-E Ltda, Plymouth, UK.

Clarke, K. & Gorley, R. (2005) PRIMER v6: User Manual/Tutotial. PRIMER-E Ltd., Plymouth, UK.

29

Cussac, V., Ortubay, S., Iglesias, G. , Milano, D., Lattuca, M., Barriga, J., Battini, M. & Gross, M. (2004) The distribution of South American galaxiid Fishes: the role of biological traits and post-glacial history. Journal of Biogeography, 31,103–121.

Dyer, B. (2000) Systematic review and biogeography of the freshwater fishes of Chile. Estudios Oceanológicos (Chile), 19, 77-98.

Flecker, A. S. &. Townsend, C.R. (1994) Community-wide consequences of trout introduction in New Zealand streams. Ecological Applications, 4, 798-807.

Goodwin P., Jorde, K., Meier, C. & Parra. O. (2006). Minimizing environment impacts of hydropower development: transfering lessons of past projects to a proposed strategy for Chile. Journal of Hydroinformatics, 8(4), 253-270.

Gurevitch J. & Padilla, D.K. (2004). Are invasive species a major cause of extinctions? TRENDS in Ecology and Evolution, 19(9), 470-474.

Habit, E., Dyer, B. & Vila, I. (2006a) Estado de conocimiento de los peces dulceacuícolas de Chile: estado de su conocimiento. Gayana, 70(1), 100-112.

Habit, E. Belk, M., Tuckfield, C. & Parra, O. (2006b) Response of the fish community to human-induced changes in of the Biobío River in Chile. Freshwater Biology, 51, 1-11.

Habit, E., Belk, M. & Parra, O. (2007). Response of the riverine fish community to the construction and operation of a diversion hydropower plant in central Chile. Aquatic Conservation-Marine and Freshwater Ecosystems, 17, 37-49.

Hardie, S., Barmuta, L. & White, R. (2006) Comparison of Day and Night Fyke Netting, Electrofishing and Snorkelling for Monitoring a Population of the Threatened Golden Galaxias ( Galaxias auratus) . Hidrobiologia, 1, 145-158.

Jenyns, L. (1842). Beagle, under the Command of Captain Fitzroy, during the years 1832 to 1836. The Zoology of the Voyage of H.M.S. Part IV Fish (ed. by CH. Darwin, E. Smith, and C. Cronhill), London.

Jowett, I. G. , Hayes, J. W., Deans, N. & Eldon, G. A. (1998) Comparison of fish communities and abundance in unmodified streams of Kahurangi National Park with other areas of New Zealand. New Zealand Journal of Marine and Freshwater Research, 32, 307- 322.

Leprieur, F., Beauchard, O., Blanchet, S., Oberdorff, T. & Brosse, S. (2008) Fish Invasions in the World’s River Systems: When Natural Processes Are Blurred by Human Activities. PLoS Biology, 6, 1-28.

30

Lowe S., Browne M., Boudjelas S. & De Poorter M. (2000) 100 of the World’s Worst Invasive Alien Species A selection from the Global Invasive Species Database. Aliens, 12, 1 - 12

Macchi, P., Cussac, V.E., Alonso, M.F. & Denegri, M. A. (1999) Predation relationships between introduced salmonids and the native fish fauna in lakes and reservoirs in northern Patagonia. Ecology of Freshwater Fish, 8, 227-236.

Macchi, P., Pascual, M. & Vigliano, P. (2007) Differential piscivory of the native Percichthys trucha and exotic salmonids upon the native forage fish Galaxias maculatus in Patagonian Andean lakes. Limnologica, 37, 76–87.

Maddock, I. (1999) The importance of physical habitat assessment for evaluating river health. Freshwater Biology, 41, 373–391.

McDowall, R.M. (1971) The galaxiid fishes of South America. Zoological journal of the Linnean Society, London, 50, 33-73.

McDowall, R.M. (2003). Impacts of Introduced Salmonids on Native Galaxiids in New Zealand Upland Streams: A New Look at an Old Problem. Transactions of the American Fisheries Society, 132, 229–238.

McDowall, R.M. (2006) Crying wolf, crying foul, or crying shame: alien salmonids and a biodiversity crisis in the southern cool-temperate galaxioid fishes? Reviews in Fish Biology and Fisheries, 16, 233-422.

McIntosh, A. R. (2000) Habitat- and size-related variations in exotic trout impacts on native galaxiid fishes in New Zealand streams. Canadian Journal of Fisheries and Aquatic Sciences, 57, 2140–2151.

McPhaden, M. (2001) El Niño & La Niña: causes and global consequences. Encyclopedia of global environmental change 1 (ed by M. Cracken and J. Perry), pp. 353-370. John Wiley & Sons, New York.

Milano, D., Ruzzante, D., Cussac, V., Macchi, P., Ferriz, R., Aigo, J., Barriga, J.P., Lattuca, M.E. & Walde, S. (2006) Latitudinal and ecological correlates of morphological variation of Galaxias platei (Pisces, Galaxiidae) in Patagonia . Biological Journal of the Linnean Society, 87, 69–82.

Mills, M.D., Rader, R.D., & Belk, M.C. (2004) Complex interactions between native and invasive fish: the simultaneous effects of multiple negative interactions. Oecologia, 141, 713-721.

31

Nannini, M. A. , & Belk, M.C. (2006) Antipredator responses of two native stream fishes to an introduced predator: does similarity in morphology predict similarity in behavioral response? Ecology of Freshwater Fish, 15, 453-463.

Niklitschek, E. & Aedo, E. (2002) Estudio Del Ciclo Reproductivo de las Principales Especies Objetivo de la Pesca Deportiva en la XI Región. Universidad Austral de Chile. Centro Universitario de la Trapananda, Chile.

Orrego R., Jimenez B., Bordajandi L.R., Gavilan J.F., Inzunza B., Abad E., Gonzalez M.J., Rivera J. & Barra R. (2005) EROD induction and PCD/F levels in fish liver from the Biobío river in Chile. Chemosphere, 60, 829–835.

Pascual, M., Macchi, P., Urbanski, J., Marcos, F., Riva Rossi, C., Novara, M., & Dell’ Arciprete, P. (2002). Evaluating potential effects of exotic freshwaters fish from incomplete species presence-absence dat a. Biological Invasions, 4, 101–113.

Pascual, M.A. & Ciancio, J.E. (2007) Introduced anadromous salmonids in Patagonia: risks, uses, and a conservation paradox. Ecological and Genetic Implications of Aquaculture Activities (ed by T.M. Bert), pp.1-18 . Springer, New York, USA.

Pascual, M.A., Cussac, V., Dyer, B., Soto, D., Vigliano, P.H. Macchi, P.J. & Ortubay S. (2007). Freshwater fishes of Patagonia in the 21st Century after a hundred years of human settlement, species introductions, and environmental change. Aquatic Ecosystem Health & Management, (10)2: 212–227.

Rahel, F.J. (2000) Homogenization of fish faunas across the United States. Science, 288, 854-856.

Rahel, F.J. (2002) Homogenization of freshwater faunas. Annual Review of Ecology and Systematics, 33, 291-315.

Raleigh, R.F., Zuckerman, L.D. & Nelson, P.C. (1986) Habitat suitability index models and instream flow suitability curves: Brown trout, revised. Fish Wildl. Serv. Biol. Rep. 82(10.124), U.S.

Rowe D. K., J. P. Smith & Baker C. (2007) Agonistic interactions between Gambusia affinis and Galaxias maculatus : implications for whitebait fisheries in New Zealand rivers. Journal of Applied Ichthyology, 23, 668–674.

Ruiz, V.H. & Berra, T. (1994) Fishes of the high Biobío River of south-central Chile with notes on diet and speculations on the origin of the ichthyofauna. Ichthyological Exploration Freshwaters, 5, 5-18.

32

Ruzzante, D.E., Walde, S.J., Gosse J.C., Cussac V.E., Habit E, Zemlak T.S. & Adams E.D. (2008) Climate control on ancestral population dynamics: insight from patagonian fish phylogeography. Molecular Ecology. 17: 2234 - 2244.

Simon, K.S. & Townsend, C.R. (2003). Impacts of freshwater invaders at different levels of ecological organisation, with emphasis on salmonids and ecosystem consequences. Freshwater Biology, 48, 982-994.

Smith-Ramirez, C. (2004) The Chilean coastal range: a vanishing center of biodiversity and endemism in South American temperate rainforests. Biodiversity and Conservation, 13, 373–393.

Soto, D., Arismendi, I., González, J., Sanzana, J., Jara, F., Jara, C., Guzmán, E. & Lara, A. (2006) Sur de Chile, país de truchas y salmones: patrones de invasión y amenazas para las especies nativas. Revista Chilena de Historia Natural, 79(1), 97-117.

Soto, D., Arismendi, I., Di Prinzio, C., Jara, F. (2007) Establishment of Chinook salmon (Oncorhynchus tshawytscha ) in Pacific basins of southern South America and its potential ecosystem implications. Revista Chilena de Historia Natural, 80, 81-98.

Tognelli, M., Ramírez de Arellano, P. & Marquet, P. (2008). How well do the existing and proposed reserve networks represent vertebrate species in Chile? Diversity and Distributions, 14 , 148–158.

Townsend, C.R., &. Crowl, T. A. (1991) Fragmented population structure in a native New Zealand fish: An effect of introduced brown trout?. Oikos, 61, 348-354.

Vila, I., Fuentes, L. & Contreras, M. (1999a). Peces límnicos de Chile. Boletín del Museo Nacional de Historia Natural (Chile), 48, 61-75.

Vila I., Fuentes, L. & Saavedra, M. (1999b) Ictiofauna en los sistemas límicos de la Isla Grande, Tierra del Fuego, Chile. Revista Chilena de Historia Natural, 72, 273-284.

Villagrán, C. & Hinojosa, L.F. (1997) Historia de los bosques del sur de Sudamérica, II: Análisis fitogeográfico de la flora. Revista Chilena de Historia Natural, 70, 241-267.

Welcomme, R.L. (1988) International introductions of inland aquatic species. FAO, Fisheries Technical Paper, 294, 1 – 318.

Zemlak T.S., Habit, E.M., Walde, S.J., Battini, M. A. , Adams, E. & Ruzzante, D.E. (2008). Across the southern Andes on fin: Glacial refugia, drainage reversals & a secondary contact zone revealed by the phylogeographic signal of Galaxias platei in Patagonia. Mol. Ecol., 17: 5049-5061.

33

Table 1. Historic and current area (in 1000 Km 2) and North-South linear extent (in 1000 Km) of the distribution of galaxiids. Increases in area and North-South linear extent reflect higher sampling effort in the current study, however, decreases occurred despite the increased sampling effort and reflect anthropogenic impact.

Distribution range in historic Distribution range in current records records North-South North-South Area Extent Area Extent Species (1000 Km 2) (1000 Km) (1000 Km 2) (1000 Km) 215.5 3.1 2.3 Galaxias maculatus 304.3 82.6 1.4 200.5 1.9 Galaxias platei 122.7 1.4 112.6 1.5 Aplochiton taeniatus 120.5 1.9 160.5 1.5 Aplochiton zebra Brachygalaxias 94.5 0.7 79.0 0.8 bullocki

34

Table 2. Species contributions to differences (dissimilarity) between lakes and rivers in the Aysén Region, based on SIMPER analyses for relative abundance. Average dissimilarity between lakes and rivers considering all species is 72%.

Species Average relative Average relative abundance Mean dissimilarity and abundance in Lakes in Rivers percentage of contribution (%)

Lake v/s Rivers

G. platei 39.7 2.6 19.5 (27.0)

G. maculatus 9.5 2.9 4.8 (6.8)

A. zebra 3.4 3.7 5.0 (6.9)

H. macraei 3.0 1.6 2.0 (2.8)

P. trucha 5.6 0.0 3.1 (4.3)

G. australis 0.5 0.1 0.98 (1.4)

O. hatcheri 0.02 0.0 0.4 (0.6)

S. trutta 28.1 54.9 17.6 (24.4)

S. salar 0.01 0.0 0.4 (0.5)

O. mykiss 9.5 28.9 14.4 (20.1)

O. kisutch 0.0 0.3 0.2 (0.3)

O. nerka 0.3 0.0 0.4 (0.5)

O. tshawytscha 0.0 1.1 1.0 (1.4)

35

Table 3. Species contributions (cut off low contributions equal 90%) to differences (dissimilarity) among elevation categories (1: 0-199, 2: 200-299, 3: 300-399, 4: >400) based on SIMPER analyses for relative abundance.

Average relative abundance Mean dissimilarity and percentage of in each elevation category (m.a.s.l) contribution (%)

2 v/s 4 1 v/s 2 Avg. Avg. Dissimilarity = Dissimilarity = Species 1 2 3 4 69.4 71.7

S.trutta 10.8 59.7 38.6 8.7 16.6 (23.8) 26.6 (37.1)

G.platei 29.9 15.4 31.5 84.6 13.5 (19.5) 24.1 (33.6)

G.maculatus 30.6 0.0 0.0 0.0 11.2 (16.1) -

O.mykiss 14.9 4.2 13.0 6.24 9.2 (13.2) 8.0 (11.2)

P.trucha 10.3 2.5 0.0 0.0 7.4 (10.7) 4.1 (5.7)

A.zebra 0.9 0.0 17.1 0.0 5.0 (7.1) 4.1 (5.7)

H.macraei 0.0 9.8 0.0 0.0 - 3.9 (5.4)

36

b a c d e

Figure 1. Distribution of a) Galaxias maculatus , b) G. platei , c) Brachygalaxias bullocki, d) Aplochiton zebra , and e) A. taeniatus in Chile. Green and Red indicate presence. Green: No historical record for the species in location and species currently present. Red: Species historically and currently present. Purple: indicate absence, the species was historically present but is currently absent.

37

6.5e+6 R.Topocalma

6.0e+6 R.Cholguan L.Calafquen L.Riñihue L. Panguipulli 5.5e+6

5.0e+6

Latitude (UTM) 4.5e+6 L. Sarmiento

4.0e+6 Rivers R.Condor Lakes

3.5e+6 0.0 0.5 1.0 1.5 2.0 2.5

Proportional Distance to sea Figure 2. Distribution of G. maculatus in relation to latitude (from North to South) and distance to sea (proportional to the Country width). Black circles: rivers, white circles: lakes. Names of some systems as references.

Figure 3. Proportion of four Galaxiidae species in the two dominant salmonids in the latitudinal gradient, both for Coastal and Andean basins.

38

4.0

3.5

3.0

2.5

Species Richness 2.0

1.5

1.0 1 2 3 4 Elevation Category

Figure 4. Lacustrine species richness in the elevation gradient the Aysén Region (northern Chilean Patagonia), including native and non-native species.

100

80

60

40

Relative Abundance(%) 20

0 Salmonids Natives Salmonids Natives LAKES RIVERS

Figure 5. Relative abundance of native and non-native species (salmonids) in rivers and lakes of the Aysén Region (northern Chilean Patagonia).

39

1200 1100 O. mykiss a) S. trutta 1000 900 800 700 600 maculatus G. 500 400

CPUE CPUE 300 200 100 0 0 10 20 30 40 50 60 70 80 90 100 110

400

O. mykiss 350 S. trutta

b) 300

250

200 platei G. 150 CPUE CPUE 100

50 0 0 10 20 30 40 50 60 70 80 90 100 110

50 O. mykiss 45 S. trutta c) 40

35

30

25 A. taeniatus A. 20

15 CPUE 10 5

0 0 10 20 30 40 50 60 70 80 90 100 110

400 d) O. mykiss 350 S. trutta

300

250

200 B. bullocki B. 150 CPUE CPUE 100

50

0 0 10 20 30 40 50 60 70 80 90 100 110 CPUE Salmonids Figure 6. Abundance relationship among galaxiids and salmonids in the San Pedro River. Abundance is expressed as Capture per Unit Effort (in number of individuals * 100 / sampling time * area in m 2). a) Galaxias maculatus , b) Galaxias platei , c) Aplochiton taeniatus , d) Brachygalaxias bullocki .

40

CAPITULO II: NATIVE AND INTRODUCED FISH SPECIES RICHNESS IN CHILEAN PATAGONIAN LAKES: INFERENCES ON INVASION MECHANISMS USING SALMONID-FREE LAKES.

ABSTRACT

Geographic patterns of species richness have been linked to many physical and biological drivers. In this study we document and explain gradients of species richness for native and introduced freshwater fish in Chilean lakes. We focus on the role of the physical environment to explain native richness patterns. For patterns of introduced salmonid richness and dominance, we also examine the biotic resistance and human activity hypotheses. We were particularly interested in identifying the factors that best explain the persistence of salmonid-free lakes in Patagonia, Chile (39° to 54° S). We conducted an extensive survey of 63 lakes, over a broad latitudinal range. We tested for the importance of temperature, ecosystem size, current and historic aquatic connectivity as well as measures of human activity (road access and land use) in determining patterns of native and introduced richness. Introduced species richness was positively correlated with native richness. Native and introduced richness declined with latitude, increased with temperature and ecosystem size. Variation in native richness was related to historic drainage connections, while introduced richness and salmonid dominance were significantly affected by current habitat connectivity. We found a total of 15 salmonid free lakes, all located in remote areas south of 45°S, and all upstream of major naturally-occurring physical barriers. Temperature, as a correlate of latitude, and lake size were key determinants of native and introduced species richness in Chilean lakes, and were responsible for the positive correlation between native and introduced richness. We found no evidence for biotic resistance by native species to salmonid expansion, and although the original introductions were human-mediated, current patterns of introduced richness were not related to human activity, as measured by road access or land use. Rather, environmental factors, especially habitat connectivity and temperature, appear to limit salmonid expansion within Chilean freshwaters.

Evelyn Habit, Jorge González, Daniel Ruzzante & Sandra Walde. 2012. Diversity and Distributions. 1–13. DOI: 10.1111/j.1472-4642.2012.00906.x

41

INTRODUCTION

Effective conservation of biodiversity requires knowledge of the distribution of the flora and fauna within a region, information that is still incomplete for many regions of the world, including many of those harboring high levels of diversity (Amarasinghe & Welcomme, 2002). Chile, a narrow strip of land between the Pacific and the southern Andes that stretches from 17° S in the Atacama Desert to 54° S in Tierra del Fuego, contains an unusually high number of endemic species, products of current and past climate and geology. The biota of the region is becoming better known, but considerable gaps still remain, particularly with respect to the aquatic faun a. Two areas with high levels of freshwater fish endemism have been recognized, the South Chilean area (39° to 41° S) and Patagonia (42° to 54° S) (Dyer, 2000). Many short, steep rivers are found along the latitudinal gradient; all flow east to west, draining the Andes Mountains and emptying into the Pacific Ocean. Lakes are present in watersheds between 39° and 54° S. The landscape has been shaped by tectonics, volcanism and the Pleistocene glaciations, with consequences for the distribution of species. Many of the lake fish populations are relatively young, as the southern lakes formed and were colonized after the most recent glacial retreat (Ruzzante et al., 2006, 2008, 2011; Zemlak et al., 2008, 2010, 2011).

Latitudinal gradients of species richness are common, though not ubiquitous (Gaston, 2000; Colwell & Lees, 2000), and proposed mechanisms include stochastic factors, historic perturbation, environmental stability, habitat heterogeneity, productivity and species interactions (Gaston, 2000). Latitudinal gradients are also common for freshwater biodiversity. Climatic, biological, geographical and historical factors have been suggested as causes of variation in species richness for freshwater fish along both latitudinal and altitude gradients (Fu et al., 2004). Oberdorff et al. (1995) showed that variation in richness of native riverine fish was strongly related to area and energy, and concluded that historical events had little explanatory power at the global scale. Lake area, mean depth and altitude were linked to species richness of fish in North American lakes (Griffiths, 1997), while Irz et al. (2004) found that the richness of native fish in French lakes was mostly controlled by regional richness. Some groups do not show clear latitudinal gradients. For example, diadromous fish often show high diversity at temperate latitudes, probably linked to the productivity of temperate oceans (Gross et al., 1988). Thus no single mechanism adequately accounts for all gradients of biodiversity, and multiple factors likely contribute to many of the patterns (Willig et al., 2003).

Much of the native flora and fauna of Chile is currently under threat, in part due to the introduction and establishment of non-native species. Southern South America is one of six global "hotspots" of freshwater fish invasions (Leprieur et al., 2008) and shows significant homogenization of its fish fauna (Villéger et al., 2011). Salmonids were first introduced to establish sports fisheries more than 100 years ago, and some species have

42 established more recently as a by-product of aquaculture activities. Introduced salmonids are influencing the distribution, diversity and life history of native Patagonian fish (Macchi et al., 1999; Pascual et al., 2002, 2007; McDowall 2006; Soto et al., 2006, 2007; Habit et al., 2010), and now dominate most Chilean riverine systems (Habit et al., 2010). In contrast to north temperate lakes, which have shown little resistance to invasion (Irz et al., 2004), southern lakes seem to be acting as refuges for native species, especially the Galaxiidae, in Argentina (Pascual & Ciancio, 2007) as well as Chile (Soto et al., 2006; Habit et al., 2010). Lakes may thus be the key to conserving the biodiversity of the native fish fauna of Chile.

The invasion process can be viewed as a series of barriers or filters that must be overcome by potential invaders (Colautti and MacIsaac, 2004; Blackburn et al., 2011), and multiple factors, including species traits, habitat characteristics and those affecting arrival, contribute to observed richness patterns of invasive species (Vander Zanden & Olden, 2008; Pyšek et al., 2010). Most factors fall into three general categories: the physical environment (environmental suitability), dispersal/propagule pressure (often related to human activity), and the existing community (biotic resistance), each of which may limit invasion and thus influence patterns of species richness (Levine, 2000; Taylor & Irwin, 2004; Fridley et al., 2007). Environmental suitability, or the 'biotic acceptance' hypothesis (Fridley et al., 2007), predicts that native and introduced richness should co- vary, i.e., good environments are good for both native and exotic species. Measures of human activity are good predictors of non-native species richness in some (Lepreiur et al., 2008, Pyšek et al., 2010), though not all cases (e.g. Blanchet et al., 2009; Roura- Pascual et al. 2011). Species richness patterns and the relative importance of different causal mechanisms vary with spatial scale, and the importance of local abiotic and biotic factors may be more evident at the local or community scale (Levine & D'Antonio 1999).

In this study we use an extensive sampling survey of fish communities in lakes over a large latitudinal range (39° to 54° S) to document gradients of richness for native and introduced species of freshwater fish. We test for the effects of natural environmental predictors (temperature, lake area, habitat connectivity, historic drainage connection) in explaining patterns of native richness. We then evaluate three general hypotheses (Fridley et al., 2007), that may have contributed to current patterns of species richness and dominance by introduced salmonids in Chilean lakes: (1) Biotic acceptance (environmental suitability), (2) Human activity (road access and land use), and (3) Biotic resistance (native species richness).

43

METHODS

Measurement of natural environmental variables

A total of 63 lakes were sampled between 2006 and 2011, from the Valdivia river basin in the north (Lat. 39° S) to Chilean Tierra del Fuego in the south (54° S) (Fig. 1). We obtained the geographical position and altitude of each lake with GPS in the field, or by using the software Google Earth 4-beta (http://earth.google.es/), incorporated into the GIS software Arc View 3.2. (ESRI), on 1:50,000 scale digitized maps of river drainages. We estimated lake area (km2) and distance to the sea (km) digitally using 1:50,000 scale maps. For each lake we estimated maximum, minimum and average annual temperature, as well as the average annual precipitation using the largest available climatic dataset from the Chilean Agricultural Research Institute (INIA, Novoa & Villaseca, 1989). This data base contains daily statistics collected over a period of 40 years at polygon sizes of approximately 50 km2. To estimate statistics for each lake we used Arc Map in ArcGis 9.0 (ESRI).

For each lake, we calculated an index of connectivity (CI) using: (i) the presence of natural barriers (NB) between the lake and the sea, i.e., waterfalls or cascades, river reaches with slopes >40°, and (ii) hydrological connectivity (HC) between the focal lake and other lakes in the drainage. Highly isolated lakes (NB = 0) had a large barrier or several successive barriers that prevent upstream migration of fish from the sea, moderately isolated lakes (NB = 1) had barriers that reduce but do not prevent upstream migration, and highly connected lakes had no barriers (NB = 2). The hydrological connectivity index ranged from 0 to 4, where HC = 0 indicates a lake with no surficial streams from the lake, and HC = 1 through 4 reflects the stream order of the lake's outlet stream. Stream order was assigned using the Strahler (1952) system, based on 1:50,000 scale maps, and lake order was assigned based on the inflow stream order.

The connectivity index, CIi, was then calculated as:

  = 1 NB i + HC i CI i   2  2 4  in which NB and HC are standardized to have equal weight, and the index CI ranges from 0 for an isolated lake to 1 for a highly connected lake. The index was intended to represent the degree of difficulty experienced by a fish attempting to reach a lake.

We also classified lakes based on whether their headwaters lie west (Pacific drainages) or east of the Andean divide (trans-Andean drainages). Trans-Andean drainages reversed direction from the east (Atlantic drainage) to the west (Pacific drainage) following the Last Glacial Maximum , thus providing a historical east to west colonization route for some species (Zemlak et al. 2008, 2010).

44

Measurement of human-related variables

In Chile, the level of human activity generally declines from north to south, with the more populated areas and more intensive natural resource use concentrated in the northern part of the study region. Data on human population size and population density are available by administrative region (4 regions within the study area), but not at the scale of lake or watershed (INE or National Statistic Institute of Chile, www.ine.cl).

Human population size and density are often correlated with impacts on the water quality of aquatic ecosystems, effects which are often the result of human-related changes in land use within watersheds, and especially near the water bodies themselves. We used measures of human-related land use in the immediate vicinity of the sampled lakes as an indicator of human influence. We delimited a 5 km wide band around each lake (using ArcMap in ArcGis 9.0), and then used the Chilean National Forest Commission database (land use estimates from 1996 - 2005) to determine the proportions of five categories of land use surrounding each lake: native forest, native steppe-type vegetation, plantation forest, agriculture and urban development. We chose to measure near-shore land use rather than land use at the level of the entire watershed, since littoral areas usually have the strongest influence on aquatic ecosystems and are often the most heavily utilized areas (Tockner & Standford, 2002).

Human population size is not likely to be closely correlated with historic or current salmonid propagule pressure in Chilean waterways. While specific data pertaining to historic salmonid seeding rates or current human visitation rates at the scale of lakes or watersheds are not available, it is known that rainbow (O. mykiss) and brown ( S. trutta ) trout were repeatedly seeded in most low elevation rivers between the latitudes of 37° and 53°S, starting in the early 1900's (Basulto, 2003; Habit et al., 2010). There is also local anecdotal evidence of informal human-assisted transport of trout among water bodies. Roads greatly facilitate human access to lakes, and we therefore categorized each lake in terms of road access (no roads, gravel or dirt roads only, paved road access).

Species richness and community composition

Species lists for each lake were assembled based on the sampling described in this study, together with supplemental information from earlier studies. Lakes were sampled during the austral summer (December to March) between 2006 and 2011, and more than half were sampled twice. Each lake was intensively sampled using several types of fishing gear, and included the littoral, limnetic, shallow benthic and deep benthic zones of each lake. Two sets of gillnets of eight mesh sizes (10, 15, 20, 30, 50, 60, 70 and 120 mm bar mesh size) in the water column and at different depths on the bottom. Fish were also collected from shallow littoral areas using a seine net (5 mm mesh size) and a backpack electroshocker. Previous species lists exist for some of the sampled lakes (Campos et al., 1985; Niklitshek & Aedo, 2002; Soto et al., 2006; Arismendi et al.,

45

2009). In three cases (Lakes Rupanco, Llanquihue and General Carrera) these species lists include two additional salmonid species that we did not find during our fieldwork. This is the most extensive and intensive sampling of native fish conducted to date in Chile, and we are quite certain that our species lists for the sampled lakes are very close to complete.

Salmonids vary in abundance and in number of species present in any given lake. We used gill net data to place lakes into one of four salmonid dominance classes: (i) salmonid-free lakes, (ii) native-dominated lakes (> 60% of individuals captured were native), (iii) co-dominated lakes (40-60% of individuals caught were native), and (iv) salmonid-dominated lakes (> 60% of fish were salmonids). There were no lakes without native species. We used gillnet data only to define salmonid dominance; these catches should best reflect relative abundances within the lakes as effort and location were standardized.

Analyses

As many of the predictor variables are correlated with latitude (and thus also with one another), we used the following approach to tease out the best predictor variables. We first tested for latitudinal gradients in species richness, and then determined if the significant relationships with latitude could be accounted for by temperature gradients. We then statistically removed the dependence on temperature by regressing richness or dominance on mean annual temperature, and used the residuals from the regression as the dependent variable in further tests. We tested for residual relationships of native species richness with lake size (square root of lake surface area), lake connectivity (CI) and historic drainage connection (Pacific vs. trans-Andean). We tested for residual relationships of introduced species richness with lake size, lake connectivity, road access and human-related land use variables (percentages of land used for plantations, agriculture and urban development). We also tested for a residual relationship between introduced and native species, after relationships with temperature and lake size were removed.

Finally, we looked at how community composition varied with latitude and altitude, using presence-absence matrices for native and introduced species. We conducted two way analyses of similarity (ANOSIM), with altitude nested within latitudinal band. Lakes were grouped into latitudinal bands: (8 categories: 39°S, 41°S, 42°S, 44°S, 45°S, 47°S, 50°S and 54°S) and into four altitude categories (masl): (1) 1 – 100, (2) 101 – 200, (3) 201 – 300, (4) > 300, following Habit et al. (2010). There are no lakes between 48° and 49° Lat. S, due to the presence of the Patagonian ice fields. Differences among factors and levels were assessed using permutation and randomization methods based on the Bray-Curtis similarity matrix, which takes the same form as the Sorensen index for matrices of presence-absence (Clarke et al., 2005). To determine which species were most important in generating the resulting significant

46 patterns we used the similarity percentage procedure (SIMPER; Primer v.6.1.5, Clarke & Gorley, 2005).

We identified the environmental variables that best explained the variation in fish community composition across latitude using the BioEnv routine in Primer v6.1.5 (Clarke & Gorley, 2005) based on the Bray-Curtis similarity matrix for presence – absence of all species and normalized geographic variables. This procedure uses a multiple regression approach to determine which environmental variables best explain the multivariate relationship of the fish community structure based on pair-wise rank similarity matrices (Clarke et al., 2005).

RESULTS

The 63 sampled lakes occur in 15 independent river basins located between latitudes 39°S and 54°S (Fig. 1). The lakes vary in altitude, from sea level to over 800 masl, in area from 0.18 to ca. 1800 km 2, and occur at distances of 0.27 to 319 km from the sea. Human population size and density were much higher in the northern administrative districts than in the south.

We found a total of 11 native and 9 introduced species throughout the study region (Table 1). Maximum species richness (native plus introduced) for an individual lake was 9 (Lake Riñihue in the north). The number of native species per lake ranged from 1 to 7 and that of introduced species from 0 to 5. The highest numbers of introduced species (5) occurred in lakes with salmonid aquaculture activity (Lakes Rupanco and Llanquihue) (Table 1).

Gradients of species richness and composition

Average richness (number of species per lake) declined with latitude for native (slope = -0.12, R = 0.359, P = 0.004) and introduced species (slope = -0.18, R = 0.597, P < 0.001) (Fig. 2a). Many predictor variables co-varied with latitude, including average temperature, precipitation, lake connectivity, road access and several land use categories. Southern lakes were colder, had lower natural connectivity, and had more steppe-type vegetation and less forest (native and plantation) in their littoral zones.

Average native richness was strongly correlated with average temperature (R = 0.438, P < 0.001), (Fig. 2b), and after removing the relationship with temperature, native richness was no longer correlated with latitude (P = 0.768). Two factors, lake size and historic drainage connection, explained significant portions of the remaining variation in species richness (Table 2a). Larger lakes had more native species, as did lakes located in trans-Andean drainages. In particular, some lakes in southern trans-Andean drainages such as the Baker had populations of Percichthys trucha (Valenciennes, 1833), Odontesthes hatcheri Eigenmann 1909, and Hatcheria macraei (Girard, 1855), species not found in purely Pacific drainages at the same latitude. Native richness was not related to current patterns of lake connectivity.

47

Introduced species richness (number of salmonid species per lake) was significantly and positively correlated with native species richness (R = 0.400, P = 0.001) (Fig. 2c). Introduced richness was also strongly correlated with lake temperature (R = 0.564, P < 0.001), and after removing the temperature relationship, salmonid richness showed a much weaker correlation with native richness (R = 0.206, P = 0.105). Introduced species richness was higher in larger lakes and in those with greater current lake connectivity (Table 2a). Introduced richness was unrelated to native species richness once the relationships with temperature and lake area were accounted for (regression of residuals: R = 0.069, P = 0.589). Road access, a measure of the potential for human- mediated propagule pressure, and human-related land use showed no relationship to salmonid richness (Table 2b).

The composition of the native fish communities (identity and frequency of occurrence) varied significantly among latitude (ANOSIM, R = 0.41, P < 0.001) but not among altitudinal bands (R = 0.33, P > 0.05). The differences were due mainly to the very high incidence of Galaxias platei Steindachner, 1898 in southern lakes, to the decreased incidence of Galaxias maculatus (Jenyns, 1842) and P. trucha at higher latitudes and to the absence of Basilichthys australis (Eigenmann, 1927) south of 40° Lat.S (Fig. 2c). The composition of introduced species also varied significantly with latitude (ANOSIM, R = 0.42, P < 0.001), driven mostly by the near absence of species other than Salmo trutta in southern lakes. Salmonid species composition did not vary with altitude (ANOSIM R = -0.01, P > 0.05).

Salmonid dominance

Fifteen (24%) of the 63 lakes sampled were salmonid-free (Table 1). All salmonid-free and native-dominated lakes were south of 45°S, and occurred in only four river basins: the Aysén (1 of 11sampled lakes), Cuervo (2 of 2), Baker (5 of 11 lakes) and Serrano (7 of 12 lakes) systems (Fig. 3). Twelve of the 16 salmonid-free lakes were single species lakes, and the sole species was G. platei in 11 of these lakes.

Salmonid dominance was negatively related to latitude (R = 0.502, P < 0.001; Fig. 4) and positively correlated with temperature (R = 0.380, P = 0.002). A weaker relationship with latitude (R = 0.237, P = 0.062) remained after removing the relationship between salmonid dominance and temperature. Lake connectivity was the only significant predictor of the residual variation in salmonid dominance, though there was a tendency for greater dominance in larger lakes (Table 3). Variation in salmonid dominance was not related to road access or to any land use variable.

Average annual temperature and salmonid dominance were the best predictors of the variation in fish community composition across lakes (multivariate BioEnv, rSpearman=0.46, p≤0.01). In lakes with high salmonid dominance, large adults of the

48 native species normally found the limnetic zone are usually absent ( Aplochiton zebra and Aplochiton taeniatus Jenyns, 1842), and adults of G. platei are limited to deeper benthic habitats. Juveniles or small adults of the native species are still found in salmonid-dominated lakes if the littoral zones have good refuges, like large woody debris or macrophytes.

DISCUSSION

The freshwater environments of southern Chile are currently subjected to numerous stressors (Nilsson et al., 2005; Goodwin et al., 2006; Vince, 2010), but the invasion of northern hemisphere salmonids presents one of its most serious challenges. Recent studies (Arismendi et al., 2009; Young et al., 2009, 2010; Penaluna et al., 2009; Vargas et al., 2010) suggest similar impacts of introduced salmonids on temperate South American freshwater ecosystems to those well-documented for New Zealand streams and rivers (Simon & Townsend, 2003; McDowall, 2006; Townsend, 2003). While salmonids now dominate most southern Chilean rivers, lakes may act as refugia for some native species (Soto et al., 2006; Habit et al., 2010). We found strong latitudinal gradients in species richness of native and introduced fish in Chilean lakes, with the numerical dominance of salmonids also declining from north to south. The positive correlation between introduced and native richness could be accounted for by a common response to temperature and lake size, and there was no evidence for biotic resistance by native fish communities. Most important from a conservation perspective, we found that lake connectivity and the presence of natural barriers was a key predictor of salmonid invasions. Below we discuss these patterns in more detail.

Patterns of native species richness

The species richness of native fish in Chilean lakes was negatively related to latitude. The northenmost basin (Valdivia) had a total of 8 native species, and the maximum number of natives (7) in a single lake was found in this basin (Lake Riñihue). In contrast, the southernmost basin on the mainland (Serrano) had a total of 4 native species (mean of 1.6 per lake) and only one native ( G. platei ) was found in sampled lakes on the island of Tierra del Fuego. The latitudinal gradient appears to be largely driven by climate, especially temperature. Warmer temperatures and the associated higher productivity of lower latitude lakes likely help maintain higher species diversity in the north. Some of the warm water species (e.g. Cheirodon australe or Basilichthys australis ) found only in the north likely cannot tolerate the harsher southern conditions. Warmer lakes often have more physical complexity which may also contribute to higher richness (Mason et al., 2008). The higher productivity and shallow mixing depth of some of the northern lakes in our study (Riñihue, Llanquihue) has been associated with relative high abundances of Daphnia, zooplankters that are extremely rare in more southern Chilean lakes (de los Rios & Soto 2006), and may be key to supporting some fish species. The positive relationship between native species richness and lake size was

49 not unexpected; many studies have obtained similar results (Tonn & Magnuson, 1982; Pierce et al., 1994; Griffiths 1997; Magnuson et al., 1998; Zhao et al., 2006), and the relationships are likely due to the greater structural and thermal complexity of larger lakes.

Quaternary glacial cycles almost certainly also influenced species richness gradients, as has been suggested for north temperate lakes (e.g., Mandrak, 1995). Much of the biodiversity in the north has likely been preserved in the continuously habitable, northern watersheds (Turner et al., 2005). Never-glaciated regions also tend to preserve beta diversity, contributing to their higher diversity (Leprieur et al., 2011). In contrast, most southern Chilean watersheds were covered by ice, and limited time (< 15K yrs) and opportunity has limited the colonization of post-glacial lakes. Evidence for the importance of dispersal limitation comes from the higher native richness in lakes located in trans-Andean drainages; where historic drainage reversals allowed for colonization across the Andean divide. Galaxias maculatus appears to have colonized several high altitude lakes in the upper Valdivia basin from Argentina (Zemlak et al., 2010), and the presence of A. zebra in high altitude lakes in the Futaleufu basin, and of P. trucha and O. hatcheri in the Baker basin lakes (Table 1), are most easily explained by dispersal from headwater lakes east of the Andean divide that contain these species (Ruzzante et al. 2003). Marine dispersal also influenced post-glacial colonization of southern low elevation lakes (Zemlak et al., 2010); three of the four native species found in the south are at least partially diadromous ( A. zebra , A. taeniatus and G. maculatus ).

Habitat connectivity was not related to patterns of native richness, suggesting that current dispersal routes within watersheds contribute little to the overall geographic pattern of richness. Specific barriers do appear to have limited the upstream dispersal of specific taxa, however. For example, it seems likely that a very large waterfall, “El Saltón”, has prevented upstream dispersal by Aplochiton zebra in the Baker basin.

Galaxias platei has the broadest distribution range and the greatest variation in habitat use of any native species in Chilean Patagonia (Habit et al., 2010). It was found throughout the entire latitudinal range of this study. Galaxias platei inhabits large transparent oligotrophic lakes and highly turbid glacial lakes, small mesotrophic lagoons and completely isolated (endorheic) lakes (e.g. Lake Thompson). It appears physiologically and morphologically adapted to deep benthic habitats (cold-tolerance, slow metabolism, tubiform shape, reduced eyes) (Cussac et al., 2004). However, especially in salmonid-free lakes, it shows a remarkable capacity to use all the available habitats, from littoral shallow areas with or without vegetation, to deep (> 80 m) benthic zones where individuals appear to thrive buried in muddy sediments. Galaxias platei reaches high abundances, large sizes and becomes strongly cannibalistic at a young age in some salmonid-free lakes (unpublished data). Its physical tolerances and ecological plasticity were likely critical to its survival through glacial cycles in local refugia near

50 or within the limits of the LGM (Zemlak et al., 2008, 2011) and are key to the broad distribution of the species today.

Introduced species

The broad geographic patterns of richness for introduced species were similar to those for native species; richness decreased with latitude and increased with temperature and lake size. Introduced species richness was significantly and positively correlated with native richness, a pattern reported for many taxa at large spatial scales (Levine and D'Antonio, 1999), and found also for fish communities (Marchetti et al., 2004; Blanchet et al., 2009). Positive correlations are usually interpreted as a common response to environmental conditions, a prediction of the 'biotic acceptance' hypothesis (Fridley et al., 2007). The common response for native and introduced fish in Chile appears to be to temperature and lake size and their correlates (most likely productivity and physical complexity). After accounting for these responses, introduced species richness showed no relationship with native species richness, indicating that native Patagonian fish communities have shown little or no biotic resistance to invading salmonids.

We also found no link between salmonid richness or dominance and road access, in contrast to some studies of north temperate lakes (e.g. Kaufman et al., 2009). It may be that intentional seeding of salmonids (formal and informal) was so intensive and extensive that current distributions bear little resemblance to past propagule pressure. However, it is also possible that more accurate measures of propagule pressure (e.g., lake visitation rates or local interest in promoting sport fisheries) would show a stronger correlation with salmonid invasion patterns in Chilean lakes. For example, anecdotal accounts of seeding by helicopter suggest that road access may not have consistently limited human-assisted invasion. Seeding statistics are not available, but observations such as the very high introduced species richness in lakes with aquaculture activities, and the disproportionate number of salmonid-free lakes in national parks, suggests that human activity may play a more important role than our analysis revealed.

Habitats that have already experienced anthropogenic stress are often more easily invaded (Strayer, 2010). Changing land use, especially the replacement of native forest with pine and eucalyptus plantations in the riparian zones of south-central Chile streams, has been linked to changes in water retention, nitrogen dynamics, soil erosion, and to effects on native fish and introduced trout densities (Oyarzún et al., 2007; Little et al. 2009; Lara et al., 2009). However, anthropogenic impacts have not been found for all Chilean watersheds (Little et al., 2008), and the importance of human impacts on non-native fish richness have been shown to vary across biogeographic realms (Blanchet et al., 2009). We detected no effect of land use on species richness of native or introduced fish across Chilean lakes, nor were salmonids more likely to dominate numerically in lakes with more human-related land use, suggesting that anthropogenic effects related to land use are likely small compared to other factors.

51

Salmonids now dominate almost all rivers within Chile, and are likely responsible for density reductions and local displacement of native species in rivers (Arismendi et al., 2009; Habit et al., 2010). Lakes may serve as refuges for native species, as native species still share dominance with salmonids in many lakes, even in those invaded many decades ago. Most salmonid-free lakes contain only G. platei , a species not considered threatened, and these lakes are thus not critical to the conservation of native fish biodiversity. From a conservation point of view, however, some of these populations may represent evolutionarily distinct populations, and in addition, offer one of the few remaining opportunities to observe the feeding ecology, behavior and population dynamics of native taxa in the absence of salmonids.

Perhaps the most significant result of our study, from a conservation perspective, is the importance of habitat connectivity in explaining variation among lakes in salmonid species richness and dominance. Natural connectivity is higher in the north and all the salmonid free lakes that we located occurred south of 45° S. Furthermore, these lakes occur only in remote areas, mostly with limited human access. However, even in remote areas, only the low natural connectivity of these lakes has prevented colonization by salmonids from other lakes and rivers within the basin. Brown trout ( S. trutta ) is capable of very rapid expansion within drainage systems that lack physical barriers (Launey et al., 2010; Westley & Fleming, 2011), and it is clear that natural waterfalls are what prevent the invasion of the remaining salmonid-free lakes in the remote regions of southern Chile.

The areas of southern Chile that still hold salmonid free lakes are not only unpopulated, but most are contained within large conservation areas (Tognelli et al., 2008; Pauchard & Villarroel 2002). More than half of the salmonid free lakes occur in the Serrano River basin, half of which is under protection by the Torres del Paine and Bernardo O’Higgins National Parks, the high attractiveness and high visitation rates of parks such as the Torres del Paine Park (~150,000 visitors per year, personal communication, Jovito Gonzalez, National Forest Corporation, CONAF Magallanes), is likely to increase the risk of invasion. Of greater concern is the fact that the new Chilean “Sport Fisheries Law” (Ley 20.259 de Pesca Recreativa, 2008, Ministerio de Economía, Fomento y Reconstrucción) does not restrict salmonid stocking inside National Parks. In addition, there is huge pressure to expand hydropower development in the remote regions of southern Chile (Vince, 2010), and the new roads and increased human presence associated with the construction of dams at these latitudes are likely to challenge currently uninvaded systems. Recognition of the conservation importance of these still pristine ecosystems, coupled with the protection of National Park status, undoubtedly offers their best chance of preservation. While conservation areas perhaps do not protect all of Chilean biodiversity as effectively as they could (Tognelli et al., 2008), their presence in unpopulated areas that have natural barriers have helped to limit salmonid invasions and are likely to continue to contribute in the future.

52

Our results show that patterns of salmonid invasion in Chilean lakes are best explained by environmental factors, including temperature, ecosystem (lake) size and habitat connectivity. Species richness and dominance by introduced salmonids was high under conditions that also lead to high native species richness, i.e., low latitude, warmer temperatures, larger lake size. The principal difference between native and introduced species distributions is the role of dispersal: current habitat connectivity is linked to salmonid but not native species distributions, while historic dispersal pathways have played a greater role in determining native distributions. Finally, while we found no link between salmonid invasions and human activity as measured by road access or land use, we want to emphasize that the remaining salmonid free lakes in Chile are all found in remote regions and mostly within conservation areas. The continued preservation of these unique ecosystems will depend on maintaining low levels of human activity in the region and on protection of the natural barriers within the watersheds.

REFERENCES

Amarasinghe, U.S. & Welcomme, R.L. (2002) An analysis of fish species richness in natural lakes. Environmental Biology of Fishes, 65(3), 327-339.

Arismendi, I., Soto, D., Penaluna, B., Jara, C., Leal, C. & León-Muñoz, J. (2009) Aquaculture, non-native salmonid invasions, and associated declines of native fishes in lakes of the northern Chilean Patagonia. Freshwater Biology, 54, 1135-1147.

Basulto, S. (2003) El Largo Viaje De Los Salmones Una Crónica Olvidad a. Editorial Autoedición, Santiago, Chile.

Blackburn, T.M., Pyšek, P., Bacher, S., Carlton, J.T., Duncan, R.P., Jarošik, V., Wilson, J.R.U, & Richarson, D.M. (2011) A proposed unified framework for biological invasions. Trends in Ecology and Evolution, 26, 333-339.

Blanchet, S., Leprieur, F., Beauchard, O., Staes, J., Oberdorff, T., & Brosse, S. (2009) Broad-scale determinants of non-native fish species richness are context-dependent. Proceedings of the Royal Society London B, 276, 2385-2394.

Campos, H. (1985) Distribution of the fishes in the Andean rivers in the South of Chile. Archives Hydrobiologie, 104(2), 169-191.

Clarke, K. R. & Gorley, R. N. (2005) PRIMER v.6: User Manual / Tutorial. PRIMER-E Ltda., Playmouth, UK.

Clarke, K. R., Warwick, R. M., Somerfield, P. J., Gorley, R. N. (2005) Change in marine communities: an approach to statistical analysis and interpretation. 3rd edition. PRIMER-E Ltda, Plymouth, UK.

Colautti, R. I., & MacIsaac, H. J. (2004) A neutral terminology to define 'invasive' species. Diversity and Distributions, 10, 135-141.

53

Colwell, R. K., & Lees, D. C. (2000) The mid-domain effect: geometric constraints on the geography of species richness. Trends in Ecology and Evolution, 15, 70-76.

Correa, C. & Gross, M. (2007) Chinook salmon invade southern South America. Biological Invasions, 10, 615.

Cussac, V., Ortubay, S., Iglesias, G. , Milano, D., Lattuca, M. E., Barriga, J. P., Battini, M. & Gross M. (2004) The distribution of South American galaxiid fishes: the role of biological traits and post-glacial history. Journal of Biogeography, 31, 103-121.

De los Rios, P. & Soto, S. (2006) Effects of the availability of energetic and protective resources on the abundance of daphniids (Cladocera, Daphniidae) in Chilean Patagonian lakes (390 - 510S). Crustaceana 79, 23-32.

Dyer, B. (2000) Systematic review and biogeography of the freshwater fishes of Chile. Estudios Oceanologicos, 19, 77–98.

Fridley, J.D., Stachowicz, J.J., Naeem, S., Sax, D.F., Seabloom, E.W., Smith, M D., Stohlgren, T.J., Tilman D.& Von Holle B. (2007) The invasion paradox: reconciling pattern and process in species invasions. Ecology, 88, 3–17.

Fu, C., Wu. J., Wang, X., Lei, G. & Chen, J. (2004) Patterns of diversity, altitudinal range and body size among freshwater fishes in the Yangtze River basin, China. Global Ecology and Biogeography, 13, 543–552.

Gaston, K. J. (2000) Global patterns in biodiversity. Nature, 405, 220-227

Goodwin, P., Jorde, K., Meier, C. & Parra, O. (2006) Minimizing environmental impacts of hydropower development: transferring lessons from past projects to a proposed strategy for Chile. Journal of Hydroinformatics, 8(4), 253-270

Griffiths, D. (1997) Local and regional species richness in North American lacustrine fish. Journal of Animal Ecology, 66, 49–56.

Gross, M.R., Coleman, R.M. McDowall, R.M. (1988) Aquatic productivity and the evolution of diadromous fish migration. Science, 239, 1291–1293.

Habit, E., Piedra, P., Ruzzante, D. E., Walde, S. J., Belk, M. C., Cussac, V. E., Gonzalez, J. & Colin, N. (2010) Changes in the distribution of native fishes in response to introduced species and other anthropogenic effects. Global Ecology and Biogeography, 19, 697–710.

Habit, E., Dyer, B. & Vila, I. (2006) Estado de conocimiento de los peces de agua dulce de Chile. Gayana, 70, 100-113.

54

Irz, P., Argillier, C. & Oberdorff, T. (2004) Native and introduced fish species richness in French lakes: local and regional influences. Global Ecology and Biogeography, 13, 335–44.

Kaufman, S.D., Snucins, E., Gunn, J.M., & Selinger, W. (2009) Impacts of road access on lake trout ( Salvelinus namaycush ) populations: regional scale effects of overexploitation and the introduction of smallmouth bass ( Micropterus dolomeiue ). Canadian Journal of Fisheries and Aquatic Science, 66, 212-223.

Lara, A., Little, C., Urrutia, R., McPhee, J., Alvarez-Garretón, C., Oyaruzún, C., & Soto, D. (2009) Assessment of ecosystem services as an opportunity for the conservation and managemnet of native forests in Chile. Forest Ecology and Management, 258, 415-424.

Launey, S., Brunet, G. , Guyomard, R. & Davaine, P. (2010) Role of introduction history and landscape in the range expansion of brown trout ( Salmo trutta L.) in the Kerguelen Islands. Journal of Heredity, 1010, 270–283.

Leprieur, F., Beauchard, O., Blanchet, S., Oberdorff, T. & Brosse, S. (2008) Fish invasions in the world's river systems: when natural processes are blurred by human activities. PLoS Biology, 6(2), 404-410.

Leprieur, F., Tedesco, P.A., Hugueny, B., Beauchard, O., Dürr, H.H., Brosse, S., & Oberdorff, T. (2011) Partitioning global patterns of freshwater fish beta diversity reveals contrasting signatures of past climate changes. Ecology Letters, 14, 325-334.

Levine, J.M. (2000) Species diversity and biological invasions: relating local process to community pattern. Science, 288, 852–854.

Levine, J.M., & D'Antonio, C.M. 1999. Elton revisited: a review of evidence linking diversity and invasibility. Oikos, 87, 15-26.

Little, C., Soto, D., Lara, A. , & Cuevas, J. G. (2008) Nitrogen exports at multiple-scales in a southern Chilean watershed (Patagonian Lakes district). Biogeochemistry, 87, 297- 309.

Macchi, P.J., Cussac V., Alonso M. & Denegri M. (1999) Predation relationships between salmonids and the native fish fauna in lakes and reservoirs in northern Patagoni a. Ecology of Freshwater Fish, 8, 227-236.

Magnuson, J.J., Tonn, W.M., Bannerjee, A. , Toivonnen, J., Sanchez, O., & Rask, M. (1988) Isolation vs. extinction in the assembly of fishes in small northern lakes. Ecology, 79, 2941-2956.

55

Mandrak, N.E. (1995) Biogeographic patterns of fish species richness in Ontario lakes in relation to historical and environmental factors. Canadian Journal of Fisheries and Aquatic Science, 52, 1462-1474.

Marchetti, M. P., Light, T., Moyle, P. B., & Viers, J. H. (2004) Fish invasions in California watersheds: testing hypotheses using landscape patterns. Ecological Applications, 14, 1507-1525.

Mason, N.W.H., Irz, P., Lanoiselée, C., Mouillot, D., & Argillier, C. (2008) Evidence that niche specialization explains species-energy relationships in lake fish communties. Journal of Animal Ecology, 77, 285-296.

McDowall, R. M. (2006) Crying wolf, crying foul or crying shame: alien salmonids and a biodiversity crisis in the southern cool-temperate galaxioid fishes. Reviews in Fish Biology and Fisheries, 16, 233-422

Niklitshek, E. & Aedo, E. (2002) Estudio del ciclo reproductivo de las principales especies objetivo de la pesca deportiva en la XI Región. Informe Técnico Código FIP 2000-25. Universidad Austral de Chile, Puerto Montt.

Nilsson, C., Reidy C.A., Dynesius, M., & Revenga, C. (2005) Fragmentation and flow regulation of the World’s large river systems. Science, 308, 405–408.

Novoa, R., & S. Villaseca (1989) Mapa Agro-climático de Chile. Ediciones Instituto de Investigaciones Agropecuarias, INIA, Ministerio de Agricultura, Santiago, Chile.

Oberdorff, T., Guégan, J. F. & Hugueny, B. (1995) Global scale patterns of fish species richness in rivers. Ecography, 18, 345–352.

Oyarzun, C., Aracena, C. Rutherford, P., Godoy, R., & Deschrijver, A. (2007) Effects of land use conversion from native forests to exotic plantations on nitrogen and phosphorus retention in catchments of southern Chile. Water Air Soil Pollution, 179, 341-350.

Pascual, M., Macchi, P., Urbanski, J., Marcos, F., Riva Rossi, C., Novara, M. & Dell’Arciprete, P. (2002) Evaluating potential effects of exotic freshwater fish from incomplete species presence–absence dat a. Biological Invasions, 4, 101–113.

Pascual, M.A. & Ciancio J.E. (2007) Introduced anadromous salmonids in Patagonia: risks, uses, and a conservation paradox. Ecological and Genetic Implications of Aquaculture Activities (ed. by T.M. Bert), pp. 1-18. Springer, New York City, US A.

Pauchard, A. & Villarroel, P. (2002) Protected areas in Chile: history, current status and challenges. Natural Areas Journal, 22, 318-330

56

Penaluna, B.E., Arismendi, I. & Soto, D. (2009) Evidence of interactive segregation between introduced trout and native fishes in northern Patagonia Rivers, Chile. Transactions of the American Fisheries Society, 138, 839-845.

Pierce, C.L., Rasmussen, J.B., & Leggett, W.C. (1994) Littoral fish communities in southern Quebec lakes: relationships with limnological and prey resource variables. Canadian Journal of Fisheries and Aquatic Science, 51, 1128-1138.

Pyšek, P., Jarosík, V., Hulme, P. E., Kühn, I., Wild, J., Arianoutsou, M., Bacher, S., Chiron F., Didžiulis V., Essl F., Genovesi P., Gherardi F., Hejda M., Kark S., Lambdon P.W., Desprez-Loustau A. -M., Nentwig W., Pergl J., Poboljšaj K., Rabitsch W., Roques A., Roy D.B., Solarz W., Vila M. & Winter M. (2010) Disentangling the role of environmental and human pressures on biological invasions across Europe. Proceedings of the National Academy of Science of the United States of America, 107(27), 12157- 62.

Roura-Pascual, N., Hui, C., & Ikeda, T. et al. (2011) Relative roles of climatic suitability and anthropogenic influence in determining the pattern of spread in a global invader. Proceedings of the National Academy of Science of the United States of America, 108, 220-225.

Ruzzante D.E., Walde, S.J., Cussac, V.E., Macchi, P.J., Alonso, M., & Battini, M. 2003. Trophic polymorphism in Percichthys trucha (Pisces, Percichthyidae) in the Andes: phenotypic evidence for variation among lakes. Biological Journal of the Linnean Society, 78, 497-515.

Ruzzante, D. E., Walde, S. J., Gosse, J. C., Cussac, V. C., Habit, E., Zemlak, T. S. & Adams, E. D. M. (2008) Climate control on ancestral population dynamics: insight from Patagonian fish phylogeography. Molecular Ecology, 17, 2234 - 2244.

Ruzzante, D.E., Walde, S.J., Cussac, V.E., Dalebout, M.L., Seibert, J., Ortubay, S. & Habit, E. (2006) Phylogeography of the Percichthyidae in Patagonia: roles of orogeny, glaciation, and volcanism. Molecular Ecology, 15(10), 2949 - 2968.

Ruzzante, D.E., Walde, S.J., Macchi, P.J., Alonso, M. & Barriga, M. (2011) Phylogeography and phenotypic diversification in the Patagonian fish Percichthys trucha : the roles of Quaternary glacial cycles and natural selection. Biological Journal of the Linnean Society, 103(2): 514-529.

Simon, K.S. & Townsend, C.R. (2003) Impacts of freshwater invaders at different levels of ecological organisation, with emphasis on salmonids and ecosystem consequences. Freshwater Biology, 48, 982–994.

57

Soto, D., Arismendi I., González, J., Sanzana J., Jara, F., Jara, C., Guzmán E. & Lara, A. (2006) Sur de Chile, país de truchas y salmones: patrones de invasión y amenazas para las especies nativas. Revista Chilena de Historia Natural, 79(1), 97-117.

Soto, D., Arismendi, I., Prinzio, C. & Jara, F. (2007) Establishment of Chinook salmon (Oncorhynchus tshawytscha ) in Pacific basins of southern South America and its potential ecosystem implications. Revista Chilena de Historia Natural. 80, 81-98.

Strahler, A. N. (1952) Dynamic basis of geomorphology. Geological Society of America Bulletin, 63, 923-938.

Strayer, D.L. (2010) Alien species in fresh waters: ecological effects, interactions with other stressors, and prospects for the future. Freshwater Biology, 55 (Suppl. 1), 152- 174.

Taylor, B.W. & Irwin, R.E. (2004) Linking economic activities to the distribution of exotic plants. Proceedings of the National Academy of Science of the United States of America, 101: 17725–17730.

Tockner, K. & Stanford, J. A. (2002) Riverine flood plains: present state and future trends. Environmental Conservation, 29: 308-330.

Tognelli, M. F., Ramirez de Arellano, P. I. & Marquet, P. A. (2008) How well do the existing and proposed reserve networks represent vertebrate species in Chile? Diversity and Distributions, 14, 148–158.

Tonn, W.M. & Magnuson, J.J. (1982) Patterns in the species composition and richness of fish assemblages in northern Wisconsin lakes. Ecology, 63,1149-1166.

Townsend, C. R. (2003) Individual, population, community and ecosystem consequences of a fish invader in New Zealand streams. Conservation Biology, 17, 38– 47.

Turner K. J., Fogwill C.J., McCulloch R.D. & Sugden D.E. (2005) Deglaciation of the eastern flank of the North Patagonian Icefield and associated continental-scale lake diversions. Geografiska Annaler Series A-Physical Geography, 87A, 363–374.

Vander Zanden, M. J., & Olden, J. D. (2008). A management framework for preventing the secondary spread of aquatic invasive species. Canadian Journal of Fisheries and Aquatic Science, 65(7), 1512-1522.

Vargas, P.V., Arismendi, I., Lara, G. , Millar, J. & Peredo, S. (2010) Evidence of microhabitat overlap between juvenile of introduced salmon Oncorhynchus tshawytscha and the native fish Trichomycterus areolatus in the Allipén River, Chile. Revista de Biología Marina y Oceanografía, 45(2), 285-292.

58

Villéger, S., Blanchet, S., Beauchard, O., Oberdorff, T., Brosse, S. (2011) Homogenization patterns of the world's freshwater fish faunas. Proceedings of the National Academy of Science of the United States of America, 108, 18003-18008.

Vince, G. (2010) Dams for Patagonia. Science, 329, 382-385.

Westley, P. A. & Fleming, I. A. (2011) Landscape factors that shape a slow and persistent aquatic invasion: brown trout in Newfoundland 1883-2010. Diversity and Distributions, 17, 566-579

Young, K.A., Dunham, J. B., Stephenson, J.F., Terreau, A., Thailly, A.F., Gajardo, G. & Garcia de Leaniz, C. (2010) A trial of two trouts: comparing the impacts of rainbow and brown trout on a native galaxiid. Animal Conservation, 13, 399-410.

Young, K.A., Stephenson, J.F., Terreau, A., Thailly, A.F., Gajardo, G. & Garcia de Leaniz, C. (2009) The diversity of juvenile salmonids does not affect their competitive impact on a native galaxiid. Biological Invasions, 11, 1955–1961.

Zemlak, T.S., Habit, E.M.,Walde, S.J., Battini, M.A., Adams, E. & Ruzzante, D.E. (2008) Across the southern Andes on fin: glacial refugia, drainage reversals and a secondary contact zone revealed by the phylogeographic signal of Galaxias platei in Patagonia. Molecular Ecology, 17, 5049–5061.

Zemlak, T. S., Habit, E. M., Walde, S. J., Carrea, C., & Ruzzante, D. E. (2010) Surviving historical Patagonian landscapes and climate: molecular insights from Galaxias maculatus . BMC Evolutionary Biology, 10, 67.

Zemlak, T.S., Walde, S.J., Habit, E.M. & Ruzzante, D.E. (2011). Climate-induced changes to the ancestral population size of two Patagonian galaxiids : the influence of glacial cycling. Molecular Ecology, 5280-5294.

Zhao, S., Fang, J., Peng, C., Tang, Z., & Piao, S. (2006) Patterns of fish species richness in China's lakes. Global Ecology and Biogeography, 15, 386-394.

59

Table 1 . Species composition of the 63 sampled lakes. Native species; Ga: Geotria australis , Ba: Basilichthys australis , Omau: Odontesthes mauleanum , Oh: Odontesthes hatcheri , Pt: Percichthys trucha , Cha: Cheirodon australe , Gm: Galaxias maculatus , Gp: Galaxias platei , Az: Aplochiton zebra , At: Aplochiton taeniatus , Hm: Hatcheria macraei . Introduced species; St: Salmo trutta , Ss: Salmo salar , Om: Oncorhynchus mykiss , Ok: Oncorhynchus kisutch , Ot: Oncorhynchus tshawytscha , Omas: Oncorhynchus masou , Sf: Salvelinus fontinalis .

Latitude Basin Lake system Native species Introduced species 39° Valdivia Calafquén Ba,Gm,Omau,Gp St,Om Panguipulli Ba,Gm,Omau,Gp St,Om Neltume At,Ba,Gm,Gp,Pt,Gp St,Om Pirehueico Gm,Pt St,Om Riñihue Az,Ba,Gm,Omau,Pt,Cha,Gp St,Om Pellaifa Pt,Gp St,Om 40º Bueno Rupanco Ba,Gm,Omau St,Om,Ok,Ot,Ss 41º Maullin Llanquihue Gm,Gp,Omau St,Om,Ok,Ot,Ss Puelo Tagua-tagua Pt St,Om,Ok,Ot Azul Gp,Pt St,Om Candelaria Gm,Pt St,Om Victoria At,Gm,Gp,Pt St,Om,Ok,Ot Las Rocas Gp St,Om,Sf Inferior Gp,Pt St,Om,Sf 42º Blanco Blanco Ga,Gm Om,Ot Chiloe Cucao Gm St Huillinco Gm,Omau St,Om Natri Gm St,Om Tarahuin Gm St,Om Tepuhueico Gm St,Om 43º Futaleufu Lonconao Az Sf Yelcho Gm St,Om,Ot,Ok 44º Cisnes Las Torres Gp St Palena Risopatron Gm,Gp St Aysen Escondida Gp St

60

Latitude Basin Lake system Native species Introduced species 45º Aysen Caro Az St Elizalde Az,Gp St La Paloma Az St Riesco Az,Ga,Gm,Gp St Thompson Gp Om Pollux Gp St Los Palos Ga,Gm,Gp St,Om Alta Gm,Gp Atravezado Gp St Verde Gp St Condor Az,Gm,Gp St Cuervo Yulton Gp Meullín Gp 46º Baker Gral Carrera Gp,Hm,Oh,Pt St,Omas,Ss Jeinimeni Gp St Silvia Pt 47º Baker Cochrane Gp,Hm,Pt St,Om Largo Gp,Hm Esmeralda Gp,Oh,Pt Om Juncal Oh,Pt St Maldonado Gp,Pt Tortel Gp Pullin Gp Ñadis Pt St,Om 50º Serrano Paine Gp 51º Serrano Las Mellizas E Gp Las Mellizas O Gp Nordenskjold Gp Toro At,Gm,Gp St,Om,Ot Sarmiento Gp St Pehoe Gp Porteño Az,At,Gm,Gp St Vision Mundo Gp Azul Az,At,Gp Dickson Gp Hollemberg Balmaceda Gm St 54º Grande Blanco Sur Gp St Sanchez Deseado Gp St

61

Table 2 . Results of multiple regression models testing for (a) the effects of lake size, current lake connectivity and historic drainage connection (trans-Andean vs. Pacific) on native richness, and (b) the effects of lake size, lake connectivity, and measures of human activity on introduced species richness. Dependent variables (native and introduced richness) were the residuals of regressions of richness on average temperature.

(a) Dependent variable: Native species richness Model R 2 = 0.191, P = 0.006 Independent variable Coefficient t P-value Lake size 0.044 2.115 0.039 Lake connectivity 0.627 1.201 0.235 Historic drainage - 2.436 0.018

(b) Dependent variable: Introduced species richness Model R 2 = 0.327, P = 0.001 Independent variable Coefficient t P-value Lake size 0.062 3.473 0.001 Lake connectivity 1.060 2.499 0.015 Road access -0.308 -0.958 0.342 Agriculture -0.002 -0.188 0.851 Plantations 0.108 0.883 0.381 Urban -0.224 -0.357 0.723

62

Table 3. Results of multiple regression model testing for the effects of lake size, lake connectivity, and measures of human activity on salmonid dominance after temperature effects were removed (dependent variable: residuals of regression of salmonid domance on average temperature).

Dependent variable : Salmonid dominance

Model R 2 = 0.334, P = 0.001 Independent variable Coefficient t P-value Lake size 0.036 1.876 0.066 Lake connectivity 1.744 3.950 <0.001 Road access 0.151 -0.270 0.668 Agriculture -0.007 -0.188 0.522 Plantations -0.032 -0.883 0.812 Urban -0.970 -0.357 0.162

63

Figure 1. Lakes sampled for this study, ranging from latitude 39° to 54° S.

64

Figure 2. Average species richness (native and introduced) declined with latitude (A), increased with average temperate (B). Average introduced richness was positively correlated with average native richness (C).

65

100

80

60

40 Percentage (%) 20

0 39°-41° 42°-44° 45°-47° 50°-54° Latitude South Figure 3. Percentage of sampled lakes in each latitudinal band that were salmonid- dominated (black), co-dominated (grey), native-dominated (diagonal lines) and salmonid- free (white). Dominance categories as defined in the text.

3,5

3

2,5

2

nI d x e 1,5

1

0,5

0

39°-41° 42°-44° 45°-47° 50°-54° Latitude South Figure 4. Average lake connectivity (empty circles) and salmonid dominance (solid circles) for each latitudinal band. Error bars are standard errors.

66

CAPITULO III: FRESHWATER FISH IN SOUTHERNMOST CHILEAN PATAGONIA: SALMONIDS AND GALAXIIDS COEXISTING WITHIN PROTECTED AREAS.

ABSTRACT

The diversity and distribution of the freshwater fish fauna of Patagonia is being affected by the introduction of salmonids, an activity that has historically been, and is currently promoted by sport fishing and aquaculture and protected wild areas are not effective at conserving this native fish fauna. The impact of the above activities on the native fish fauna in pristine and protected zones of southern Patagonia is expected to be minor. Herein, we analyze the composition, distribution, and community structure of native fishes in relation to the invasion of salmonids and the protection of wild areas in the Serrano River basin, Patagonia (50° S). Although the invasion of salmonids in the studied pristine zones has not generated regional changes in the native species richness, this invasion is causing significant changes at the population level. For example, the displacement of native species to less productive habitats is leading to at least a decline in their well-being and trophic position. Moreover, the protected wild areas studied were found to offer little protection to Galaxias species and did not conform to their current distribution or community patterns.

Jorge González, Daniel Ruzzante, Sandra Walde & Evelyn Habit. Enviado. Biological Invasion.

67

INTRODUCTION

Invasion by exotic species constitutes one of the main threats to freshwater biodiversity worldwide (Dudgeon et al. 2006; Light & Marchetti 2007; Mack et al. 2000; Saunders et al. 2002), producing local extinctions, homogenizing fish communities and distorting or eliminating native biogeographic patterns (Leprieur et al. 2008; 2009). The main driver behind invasions in most regions of the world has been human economic activity (Leprieur et al. 2008). However, freshwater fish, and salmonids in particular, have also successfully invaded unpopulated and relatively pristine regions. The goal of this study is to examine the patterns and mechanisms underlying invasions that have not been assisted by humans, and to assess their impacts on the pattern of native biodiversity.

Patagonia is often considered one of the last pristine regions on the planet. It is largely unpopulated and a high percentage of the land is under protection for the conservation of biodiversity (Brooks et al. 2006; Mittermeier et al. 2003). Despite this, Patagonian freshwater ecosystems have not been exempt from salmonid invasion. The diversity of native freshwater fishes in Patagonia is naturally low (Dyer 2000) and highly vulnerable (Campos et al. 1998; Habit et al. 2006; 2010). The glacial cycles of the Quaternary are in part responsible for this low species richness (Cussac et al 2004) and have been linked to patterns of genetic and morphological diversity (Ruzzante & Rabassa 2011; Ruzzante et al. 2006; 2008; 2011; Zemlak et al. 2008; 2010; 2011). The Galaxiidae is one of the most widespread families of fish in Patagonia, and the ranges of some galaxiid species extend further south than any other South American fish. Of the seven species in Chilean Patagonia, three are diadromous ( Aplochiton zebra Jenyns 1842, Aplochiton taeniatus Jenyns 1842, Galaxias maculatus Jenyns 1842) and four are restricted to freshwater (Galaxias platei Steindachner 1898, G. globiceps , Brachygalaxias bullocki and B. gothei ; McDowall 1971a; 1971b; 2006).

Most of the evidence for salmonid effects on native species, including the Galaxiidae, comes from studies conducted in northern Patagonia (Arismendi et al. 2009; 2011; Habit et al. 2010; Pascual et al. 2007; Soto et al. 2006), where the introduction of salmonids has been associated with major changes in abundance and distribution of native species. Impacts have tended to be stronger in fluvial than in lacustrine habitats (Habit et al. 2010). Correlations between salmonid densities and the many other human impacts on the landscape sometimes make clear separation of effects difficult (Habit et al. 2010), however much less is known about salmonid impacts on the freshwater ecosystems of the far south, a region characterized by low human population densities and by a high percentage of protected lands. Under conditions of low human impact, the distribution and abundance patterns of native species would be expected to reflect the life history traits of the species and the geomorphological characteristics and history of the basins they inhabit. However, given the magnitude of salmonid effects in northern Patagonia (e.g. Pascual et al. 2002; 2007; Baigún & Ferriz 2003; Cussac et al. 2004; Soto et al. 2006; 2007; Correa & Gross

68

2007), and on the Galaxiidae of New Zealand (Flecker & Townsend 1994; McIntosh et al. 1994, 2010; McIntosh 2000; Glova 2003; McDowall 2003; Simon & Townsend 2003; Townsend 2003), one might expect to observe effects on southern native species even in the absence of other human impacts.

The establishment of protected areas often falls short of its goal to conserve most or all of the local native biodiversity, due to poor representation of species or ecosystem types within reserve boundaries (Araujo 1999; Wells et al. 2003; Rodrigues et al. 2004). Chilean Patagonia is no exception. A large percentage of Chilean Patagonia is protected in wildlife conservation areas, including the National System of State Protected Areas and UNESCO Biosphere Reserves (Weber 1986; Jax & Rozzi 2004). Unfortunately, patterns of distribution and diversity of the native fauna and their endemism were not used to advise the design of these reserves, usually because the required biological and ecological information was absent or deficient. This is especially true for the aquatic fauna inhabiting reserves designed for the preservation of terrestrial biodiversity (Simonetti & Armesto 1991; Pauchard & Villarroel 2002; Martínez-Harms & Gajardo 2008; Rodriguez-Cabal et al. 2008; Tognelli et al. 2008), and this has led some to suggest that existing parks and reserves in Patagonia may be doing a better job of protecting salmonids than native fishes (Soto et al. 2006; Pascual et al. 2007).

At present, the most abundant salmonid species in Chile are rainbow and brown trout (Oncorhynchus mykiss and Salmo trutta , respectively). Both species occur in most river basins (Soto et al. 2001; 2006) and both prey on, and/or compete with native species (Macchi et al 1999; 2007; Soto et al. 2001; 2006; Iriarte et al. 2005; Penaluna et al. 2009; Vargas et al. 2010). Trout are indeed considered to be among the most dangerous invasive species on the planet (Lowe et al. 2000; Global Invasive Species Database 2005) and their presence in Patagonia largely stems from intentional stocking conducted beginning in the first half of the twentieth century. The Chinook salmon ( Oncorhynchus tshawytscha ) is a more recent invader that escaped from salmon farms (Soto et al. 2001), but has also established self-sustaining populations that are spreading throughout watersheds (Becker et al. 2007; Correa & Gross 2007; Soto et al. 2007).

In salmonid-invaded freshwater systems of Patagonia are known to have low abundances of native species in relation with salmonids (Habit et al. 2010; Soto et al. 2006). However, as no study thus far has compared salmonid-invaded and salmonid-free freshwater systems, it is not known whether the low abundance of native species in salmonid –invaded systems is caused by the salmonid presence or is due to the harsh environment or low productivity of southern systems. In the present study we address this question for the Serrano river system of southernmost Chile. Approximately half of the Serrano river basin is under protection by the National Parks Service, and the remainder is very thinly populated with little human-related activity. We first examine the extent to which salmonids have invaded protected versus unprotected reaches of the basin. We then look for evidence of salmonid

69 effects by comparing the composition of the fish communities in invaded and non-invaded fluvial and lacustrine habitats. We thus document the effectiveness of current protected areas in conserving the native fish fauna of southern Chile, and we test for salmonid impacts in a region with very low levels of anthropogenic impact. We also provide detailed information on the spatial distributions of the native Galaxiidae.

METHODS

Study area

The study was conducted in the Serrano River drainage (50º S) in southern Chilean Patagonia (Fig. 1). The Serrano River basin covers 667,300 ha. Its waters are of glacial, pluvial and nival origin (DGA 2004) and are thus cold and highly turbid. The large lakes (>3 km2) in this basin are polymictic and oligotrophic, limited by nutrients and light, whereas the smaller lakes are highly productive but mostly lacking in fishes (Soto et al. 1994; Soto & De los Rios 2006; De los Rios & Soto 2007; 2009).

Sampling design

Approximately half of the Serrano River basin falls within Torres del Paine National Park or the Bernardo O’Higgins National Park; the remaining land has no protection status and is used mainly for livestock grazing (DGA 2004; Fig. 1). We sampled a total of 34 sites, 13 fluvial sites from 9 rivers and 21 lacustrine sites in 11 lakes. Nine sites were on unprotected land and 25 sites within national parks jurisdiction (Fig. 1, Table 1). We sampled at one site in small lakes (<3 km2), at two sites in medium-sized lakes (3 to 86 km2), and at four sites in the largest lake Toro (202 km2) (Table 1). Most rivers were sampled in one location, but three sites were sampled for the two largest rivers.

Fish sampling was conducted during the austral summer of 2009 (January to March). Rivers, streams, and shallow littoral areas of lakes were sampled by electrofishing and/or with seine nets. Deeper sites in lakes were sampled with multi-mesh gill nets (15, 20, 30, 50, 60, 70, 120 mm openings), placed at the surface and on the bottom to a maximum depth of 60 m. Fish were identified to species, measured (total length, to 0.1 cm), and weighed (0.01 g) in the field. We also preserved muscle tissue in ethanol from Galaxias platei , the most common native species, for stable isotope (δ 15 N and δ 13 C) analysis.

Stable nitrogen isotope (δ 15 N) concentration was used as an indicator of trophic position and carbon isotope (δ 13 C) concentration was used to indicate carbon sources (Post 2002). Stable isotope analysis was conducted by the UC Davis Stable Isotopes Facility and fossil calcite (Pee Dee Belemnite of Vienna or V-PDB) used as the standard for carbon and atmospheric N used as the nitrogen standard.

70

For each lake and each river sampling site, we determined connectivity (C), an index ranging from 0 to 5. Endorheic lakes were classed as C=0 (no connectivity). Values between 1 and 5 were assigned based on the number of waterfalls > 5m between the site and the se a. Sites with C=5 had 0 waterfalls, sites with C=4 had 1 waterfalls, etc., and the least connected sites (C=1) were separated from the sea by 4 waterfalls. Waterfalls, water distance to the sea (upstream from estuary) and altitude were recorded with GPS in the field, or by using the software Google Earth 4-beta (http://earth.google.es/) and data were subsequently incorporated into ArcView 3.2. (ESRI), on 1:50,000 scale digitized maps of river drainages (Fig. 2). For each sampling site within lakes and rivers we recorded various water physico-chemical characteristics (temperature, dissolved solids, conductivity, pH, turbidity) (Table 1). We also characterized habitat structure in rivers measuring average depth, width, dominant substrate, presence of aquatic vegetation and shore area characteristics.

Statistical Analysis

Species richness and composition in protected and unprotected areas

We tested for differences between protected and unprotected sites in species richness and community composition for river sites and for lake sites. Differences in species richness (average number of species per site) were tested with nonparametric analysis of variance (Kruskal-Wallis ANOVA). We used analysis of similarity (ANOSIM with Bray-Curtis similarity matrices) to test for differences in community composition. As effort was not standardized among sites, we analyzed only relative abundances, which were fourth root transformed to down weight dominant species. Species characterizing each area were then identified using the similarity percentage procedure (SIMPER; Clarke et al. 2005). Finally, we used the BIOENV and Global Best analyses to determine the environmental variables (geographic and water quality) that best explained community patterns (Clarke & Gorley 2005). Community analyses were conducted with the program PRIMER V6.1.5 (Clarke & Gorley 2005).

Relationship between native fishes and salmonids

Only one species, Galaxias platei , was consistently abundant in salmonid-invaded and salmonid-free lakes. We tested for differences in body condition and in stable isotopic composition of individuals from lakes sites with and without salmonids. Condition (K) was estimated as K=W/L3 (Ricker 1968), and ANOVA was used to test for differences in body condition (log [K+1]) for lakes with and without salmonids. Tissue samples from 280 G. platei individuals from 3 invaded lakes (n Porteño =33, n Sarmiento =20 and n Toro =12 individuals) and 5 salmonid free lakes (n Azul =48, n Dickson =59, n Mellizas =29, n Nordenskjold =40 and n Pehoe =39 individuals) were analyzed for stable isotope δ 13 C and δ 15 N content. We used ANOVA to

71 compare the size (total length) and the isotopic composition (δ 13 C and δ 15 N) of G. platei from salmonid-invaded and non-invaded systems.

For rivers where native species and salmonids were allopatric, we analyzed habitat characteristics at the local scale. Turbidity, temperature, pH, conductivity, total dissolved solids and habitat characteristics (width, depth, dominant substrate size, aquatic vegetation and dominant component riparian zone) were all compared through a multivariate analysis of variance (MANOVA) between invaded and salmonid free rivers on transformed log (x+1) variables. Given the same local physical conditions for invaded and non-invaded systems, the presence of native fishes in rivers should primarily be determined by the presence of invasive species.

RESULTS

We found a total of four native species of the family Galaxiidae within the Serrano river drainage. Galaxias platei was the most abundant species at distant and higher altitude sites, including those with very low connectivity, but it was absent from sites near the sea (Table 1). In salmonid-free lakes, it was a habitat generalist, found from the littoral zone to deep benthic areas. Galaxias platei was only present in rivers without salmonids, all distant from the sea and above waterfalls (Fig. 2). Galaxias maculatus was the most abundant native species at lower altitudes, and was rarely present at higher altitudes or distant from the sea (Table 1). It tended to be most abundant in littoral zones (or other shallow, slow flow environments with submerged vegetation), where it often coexisted with S. trutta and juveniles of O. tshawytscha. The two Aplochiton species ( A. zebra and A. taeniatus ) were found in 4 of the 11 sampled lakes; all were clear and ultra-oligotrophic sites (Table 1). Aplochiton also tended to be found in littoral habitats, often near small inlet tributaries (like Porteño river), though larger individuals were sometimes captured in the pelagic environment.

Three introduced species (all salmonids) were also present: Rainbow trout, ( Oncorhynchus mykiss ), Chinook salmon ( Oncorhynchus tshawytscha ) and Brown trout ( Salmo trutta ).

Brown trout ( Salmo trutta ) was the most widely distributed and abundant, followed in abundance by Chinook salmon ( O. tshawytscha ). Salmonids were absent from most sites above waterfalls (low connectivity, Fig. 2). Two important exceptions were Lake Porteño, which is upstream of an important barrier and the endorheic Lake Sarmiento (Table 1).

Protected versus unprotected areas

Lakes outside the protected areas had significantly higher average native species richness than did lakes within the national parks (Fig. 3a) (Kruskal-Wallis ANOVA: H(1, N=21)= 6.07, p<0.05). The average number of salmonid species did not differ between lakes within

72 versus outside protected areas (Kruskal-Wallis ANOVA: H(1, N=21)= 0.33, p>0.05) (Fig. 3a).

In contrast, rivers within park boundaries had higher native species richness than river sites outside the protected area (Fig. 3b) (Kruskal-Wallis ANOVA: H (1, N=13)=8.75, p<0.05). Again, salmonid species richness did not differ between protected and unprotected river sites (Kruskal-Wallis ANOVA: H (1, N=13)=0.00, p>0.05).

Species composition (relative abundances) of the fish community in lakes differed significantly between lakes in protected and unprotected areas (ANOSIM

Rprotected/unprotected =0.20, p<0.05). G. maculatus and A. zebra tended to be relatively abundant in the unprotected area while G. platei usually dominated in the protected area (Table 2). The environmental variables that best explained these differences were turbidity and connectivity (BEST, r Spearman =0.3, p<0.05) (Table 3).

Community composition also differed significantly between protected and unprotected areas river sites (ANOSIM R protected-unprotected =0.23, p<0.05). River sites within park boundaries usually had populations of G. maculatus and G. platei , as well as brown trout (S. trutta ), while unprotected rivers were inhabited almost exclusively by this specie (Table 2). River sites above waterfalls tended to have only G. platei . Differences in connectivity between protected and unprotected river sites best explained the differences in community structure (BEST, r Spearman =0.79, p<0.05; Table 3).

Galaxias platei and salmonids

Galaxias platei were present in both salmonid-invaded and salmonid-free lakes. Average size (total length) of captured Galaxias platei did not differ significantly between lakes with and without salmonids (ANOVA: F(1)=1.53, p>0.05, Table 4). However, body condition was significantly lower in lakes with salmonids (ANOVA: F(1)=24.86, p<0.05, Table 4).

There were also differences in the isotopic composition of G. platei collected from salmonid-free vs. invaded systems (ANOVA: F(1)=17.03p<0.05). Individuals from 15 salmonid-free lakes had significantly higher average δ N (δN salmonid-free =10.4±0.2; 13 δN salmonid =8.9±0.2; Figure 4), but did not differ in average δ C (Table 4).

Galaxias platei were found in rivers only when these were salmonid-free, and all salmonid- free rivers occurred inside national park boundaries and above waterfalls (Fig. 2, Table 1). Rivers in unprotected areas were inhabited almost exclusively by S. trutta. The absence of G. platei in rivers occupied by salmonids is unlikely to be due to differences in the physico- chemical characteristics of the water (turbidity, temperature, pH, conductivity and total dissolved solids) or in the physical characteristics of the river habitats (width, depth dominant substrate size, aquatic vegetation and dominant component riparian zone) since

73 no significant differences in these variables were found (MANOVA F(1)=1.4, p≥0.05). However, river sites with and without salmonids did differ in connectivity; rivers without salmonids were located exclusively above natural barriers (Index C<3).

DISCUSSION

In this study we found that the richness of native fish communities in Chilean Patagonia varied between sites on land with protected status (i.e., within national parks) and sites located outside park boundaries. The direction of the difference varied between lakes and rivers, however; native species richness was higher at protected river sites, but lower at protected lake sites. Below we argue that these differences are most easily explained by the distribution of introduced salmonids and their effects on native fish.

Distribution of the native fish fauna

A significant proportion of the total number of South American Galaxiidae (4 of 7 species) occurs in the Serrano river basin in southern Chile. Three of the species ( G. maculatus , A. zebra and A. taeniatus ) are facultative diadromous and can be found in rivers associated with neighboring fjords (González et al. unpublished data). In addition, they are often seen in estuaries and along the coast of much of Chilean and Argentine Patagonia (Sielfield et al. 2006; Sielfeld and Vargas 1999; Landaeta and Castro 2006; Pequeño and Olivera 2007). Postglacial colonization of southern rivers by these species was likely via coastal dispersal, as has been shown for G. maculatus (Zemlak et al. 2010), and their presence in coastal waters today suggests that the potential for dispersal among rivers remains. Their current distributions also suggest a marine colonization route; all are usually found at low elevations and near the sea. One notable exception, however, is the presence of both Aplochiton species in Lake Azul, a high altitude lake (223 masl) with low connectivity to the sea. Neither is present at sites immediately downstream. There are at least two non- exclusive explanations for this disjunct distribution. First, it is possible that the lake was colonized from downstream at some time in the past, but downstream populations were later eliminated, perhaps by glacial lake outburst floods, which occur as glacial walls melt, releasing massive amounts of previously dammed water. Such "lake emptying" events appear to have once been common in the Serrano River basin and occurred as recently as 1982 (Peña & Escobar 1983a; 1983b; 1987). Alternatively, it is possible that Aplochiton colonized Lake Azul through past connections to other river basins, connections that no longer exist (Juan Luis García, Universidad de Chile, personal communication). Other disjunct high altitude populations of Aplochiton are known to exist: A. zebra was reported from the headwater lakes of the Biobío River (1160 masl and 380 km from the sea) (Campos et al. 1993), high altitude populations of A. zebra and A. taeniatus occur in the Valdivia basin, and A. zebra was found at over 400 masl in the Futaleufú basin (Habit et al. 2010). Aplochiton sp have also been reported from the headwater lakes of trans-Andean, Pacific draining systems east of the Andes (Ruzzante et al. 2003). A molecular approach

74 will probably be necessary to determine when and how these high altitude lakes were colonized by Aplochiton . However, their distribution is very different from that of the "typical" high altitude species, Galaxias platei ; Aplochiton is not found in turbid, glacial lakes, probably due to their visual feeding behavior (Jönsson et al. 2011).

The distribution of G. platei corresponds closely to the area covered by ice during the Last Glacial Maximum (Cussac et al. 2004). Post-glacial lakes and rivers were likely colonized from multiple glacial refugia (including proglacial lakes) west and east of the Andes (Ruzzante et al. 2008; Zemlak et al. 2008). Fragmentation of such proglacial lakes would explain the current presence of G. platei in endorheic lakes and ponds of glacial origin, including the Serrano basin lakes Sarmiento, Mellizas, and Visión de Mundo. Galaxias platei occurs in a wide range of environmental conditions, from the harsh environment of cold, turbid lakes fed by glacial streams to transparent, deep and ultra-oligotrophic lakes. In this study we also collected G. platei , usually considered a lacustrine species, from high altitude, steep, fast flowing streams (Victorina and Asencio).

Salmonid invasion processes and effects on native fauna

There is abundant evidence pointing to negative effects of introduced salmonids on the native Galaxiidae, including reductions in abundance and local extinctions of various species in New Zealand (Flecker & Townsend 1994; McIntosh et al. 1994; 2010; McIntosh 2000; Glova 2003; McDowall 2003; Simon & Townsend 2003; Townsend 2003). Field and experimental work conducted thus far in Chilean systems also suggest significant impacts, particularly on the Aplochiton species (Young et al. 2009; 2010; Habit et al. 2010). One result from the current study, our failure to find A. zebra in the Tres Pasos River, where it was apparently once at least moderately abundant (McDowall 1971a; 1971b), suggests that local extinctions may be occuring in Chile as well. The Tres Pasos sites are currently inhabited only by S. trutta.

However, the most definitive impacts of salmonids in this study were negative effects on G. platei , including likely range reductions/local extinctions, and apparent competitive effects at sites where the G. platei coexists with S. trutta. Current distributions of G. platei and S. trutta are markedly allopatric; G. platei was present in every lake and river that S. trutta has not yet been able to colonize. There appear to be no important environmental differences (habitat structure, water quality) between these lakes/rivers that can account for this, other than habitat connectivity. Salmonids have invaded all sites within the Serrano (and in Chile as a whole) with moderate to high connectivity. Galaxias platei persists in only a few salmonid-invaded lakes and in no salmonid-invaded rivers. Lakes and rivers without salmonids are now restricted to high altitudes and lie upstream of significant waterfalls. This distribution is very similar to that reported for the Taieri River in New Zealand, where native galaxiids persist only in salmonid-free streams above waterfalls (Townsend 1996) and for A. zebra in the Malvinas (McDowall 2001).

75

Evidence of negative effects probably linked to competition came from the comparison of body condition and stable nitrogen isotope concentrations for G. platei in salmonid-invaded versus salmonid-free lakes. Body condition was significantly lower in lakes shared with S. trutta (i.e., individuals were thinner for a given length), suggesting lower feeding rates. In addition, they had less enriched stable isotope ratios for δN 15 , suggesting that they were, on average, feeding lower on the food chain. Both results are consistent with a habitat shift by G. platei in the presence of salmonids. In previous work on G. platei (in salmonid-invaded lakes), G. platei is usually considered a deep benthic species with adaptations for cold and dark/turbid environments (Cussac et al. 2004). However, the present study suggests that while G. platei is able to utilize such environments, it is only restricted to the deep benthic habitat in lakes with salmonids. In salmonid-free lakes G. platei was found to be abundant in the littoral shore environments, and present at all depths. Littoral habitats are usually highly productive as they are warmer, often have structure in the form of vegetation, and also tend to be subsidized by terrestrial ecosystems (Cole et al. 2006; Vander-Zanden & Gratton 2011). Elimination from these habitats is thus likely to affect body condition due to a reduction in feeding opportunities. The littoral zone is likely to be especially important for inhabitants of ultra-oligotrophic lakes such those common in Patagonia (Modenutti et al. 2010). Displacement or restriction to less productive, deeper habitats likely underlies the observed changes in feeding as reflected in stable isotope ratios as well as the lower body condition. Absent of an equivalent refuge from salmonids in rivers probably explains their greater impact in rivers compared to lakes (Habit et al. 2010 and this paper). Whether G. platei will be able to persist over the long term in salmonid-invaded lakes, while restricted to this marginal habitat, remains to be seen. Its absence in several of the salmonid-occupied lakes sampled within the Serrano basin does not provide much optimism.

The distribution of salmonid species is expected to grow, as aquaculture activities expand and sport fishing is promoted as a means of local economic growth. The number of salmonid species in Chilean freshwater systems, and their abundance, is thus expected to continue to increase (Arismendi et al. 2009; Pascual et al. 2009). Within the Serrano basin, half of which lies within the boundaries of national parks, the dynamics of the invasion process has been largely determined by natural patterns of connectivity. The highest number of exotic species occur in low altitude lakes and rivers, in part because this is where they were originally introduced, and in part due to invasion by anadromous species such as the Chinook salmon that escaped from the “ocean-ranching” experiments of the 1980s (Becker et al. 2007; Correa & Gross 2007; Soto et al. 2007). Natural barriers (waterfalls) have thus far prevented salmonids from reaching many of the remote, high altitude lakes and rivers.

76

Thus, although salmonids have not yet eliminated any native species from the relatively pristine Serrano River basin, there is strong evidence for population level effects on G. platei , the most wide-spread and abundant species endemic to Patagonia. Its absence from salmonid-inhabited river sites, and its apparent displacement to suboptimal habitat within lakes (with consequences for feeding and growth), indicate a potentially very significant impact. Consequences for survival and reproductive success – and thus impacts on persistence– of the native Galaxiidae need to be examined.

Effectiveness of protected areas and advice for native fish conservation

Although a large part (20%) of Chilean territory is already under some type of protection, and a number of new sites have recently been proposed as conservation priorities, most of the areas with the greatest species richness have not been included to date, and the currently designated land is considered by many to be deficient for the protection of endemic and endangered species (Tognelli et al. 2008). Our results show that the protected wild areas within the Serrano basin in southern Chile do not coincide with the species richness patterns for native fishes. Some of the systems with the highest number of native species (e.g. Lake Porteño) are not protected. In fact, our results indicate that the systems most invaded by salmonids (highest salmonid species richness) are those found within protected areas. Only the protected in high altitude systems shelter exclusively native species ( G. platei ) due, as noted earlier, not to the protection afforded the site but to the coincidental existence of natural barriers within the parks. Thus, protection alone has not prevented salmonid invasion, but in concert with natural barriers, it may be able to prevent further invasion in the future, i.e. natural barriers are needed, but also the protection so people do not seed trout in the high altitude lakes.

Taking into account the fact that half of the native fish species present in these conservation areas are classified as endangered (Campos et al. 1998), their conservation should be a priority. The areas (lakes and rivers) we consider to have the greatest conservation value (i.e., areas free of, or with limited presence of salmonids), are those lakes located upstream of waterfalls and their respective small tributary streams. Therefore, we recommend prohibiting the seeding or repopulation of trout or salmon throughout the basin and the control or even eradication of resident salmonid populations from certain zones where the conservation design so requires (Saunders et al. 2002). Specifically, the area above waterfalls in the Serrano River basin, where lakes and rivers are dominated by G. platei should ideally be designated areas of high conservation priority where the potential future seeding with exotic salmonids is prevented.

REFERENCES

77

Araújo M B (1999) Distribution patterns of biodiversity and the design of a representative reserve network in Portugal. Divers Distrib 5:151-163.

Arismendi I, Penaluna B, Jara C, Leal C & León-Muñoz J (2009) Aquaculture, non-native salmonid invasions and associated declines of native fishes in Northern Patagonian lakes. Freshwater Biol 54:1135-1147.

Arismendi I, González J, Soto D & Penaluna B (2011) Interactions among introduced species, Piscivory and diet overlap between two successful non-native fishes in southern Chilean streams. Austral Ecol DOI: 10.1111/j.1442-9993.2011.02282.x.

Baigún C & Ferriz R (2003) Distribution patterns of native freshwater fishes in Patagonia (Argentina). Org Divers Evol 3:151-159.

Becker L A, Pascual M A. & Basso N G (2007) Colonization of the Southern Patagonia Ocean by Exotic Chinook salmon. Conserv Biol 21(5):1347–1352.

Brooks T M, Mittermeier R A, da Fonseca G A B, Gerlach J, Hoffmann M, Lamoreux JF, Mittermeier CG, Pilgrim JD & Rodrigues ASL (2006) Global Biodiversity Conservation Priorities. Science 313:58-61.

Campos H, Ruiz VH, Gavilán JF & Alay F. (1993) Peces del Río Bío-Bío. Serie Publicaciones de Divulgación EULA, Universidad de Concepción, Chile, 100 pp.

Campos H, Dazzarola G, Dyer B, Fuentes L, Gavilán J F, Huaquín L, Martínez G, Meléndez R, Pequeño G, Ponce F, Ruiz V H, Sielfeld W, Soto D, Vega R & Vila I (1998) Categorías de conservación de peces nativos de aguas continentales de Chile. Bol Mus Nac Hist Nat 47:101-122.

Cole JJ, Carpenter SR, Pace ML, Van de Bogert MC, Kitchell JL & Hodgson JR (2006) Differential support of lake food webs by three types of terrestrial organic carbon. Ecol Lett 9:558–568.

Correa C & Gross M (2007) Chinook salmon invade southern South Americ A. Biol Invas 10(5):615.

Clarke KR & Gorley R N (2005) PRIMER v.6, User Manual / Tutorial. PRIMER-E Ltd A. , Playmouth, UK.

Clarke KR, Warwick RM, Somerfield PJ & Gorley R N (2005) Change in marine communities, an approach to statistical analysis and interpretation. 3rd edition. PRIMER-E Ltda, Plymouth, UK.

78

Cussac V, Ortubay S, Iglesias G, Milano D, Lattuca ME, Barriga JP, Battini M & Gross M (2004) The distribution of South American galaxiid fishes, the role of biological traits and post-glacial history. J Biogeography 31:103-121.

De los Ríos P & Soto D (2007) Crustacean (Copepoda and Cladocera) zooplankton richness in Chilean Patagonian lakes. Crustaceana 80:285-296.

De los Ríos P & Soto D (2009) Estudios limnológicos en lagos y lagunas del Parque Nacional Torres del Paine (51° S, Chile). Anal Inst Patagonia 37:63-71.

DGA (2004) Diagnostico y clasificación de los cursos y cuerpos de agua según objetivos de calidad. Cuenca del río Serrano. Chile. Documento de la Direccion General de Aguas.

Dudgeon D, Arthington AH, Gessner MO, Kawabata Z, Knowler DJ, Lévêque C, Naiman RJ, Prieur-Richard A, Soto D, Stiassny MLJ & Sullivan CA (2006) Freshwater biodiversity, importance, threats, status and conservation challenges. Biol Rev 81:163-182.

Dyer B (2000) Systematic review and biogeography of the freshwater fishes of Chile. Estud Oceanol 19:77-98.

Flecker AS & Townsend CR (1994) Community-wide consequences of trout introduction in New Zealand streams. Ecol Appl 4:798-807.

Glova GJ (2003) A test for interaction between brow trout ( Salmo trutta ) and inanga (Galaxias maculatus ) in an artificial stream. Ecol Freshwater Fish 12:247–253.

Garcia de Jalón D, Mayo M, Hervella F, Barceló E. & Fernández T (1993) Principios y técnicas de gestión de la pesca de aguas continentales. Mundi Prens A. Madrid.

Habit E, Dyer B & Vila I (2006) Estado de conocimiento de los peces dulceacuícolas de Chile. Gayana (Concepc) 70(1):100-113.

Habit E, Piedra P, Ruzzante DE, Walde SJ, Belk MC, Cussac VE, Gonzalez J & Colin N (2010) Changes in the distribution of native fishes in response to introduced species and other anthropogenic effects. Global Ecol Biogeogr 19(5):697-710.

Iriarte JA, Lobos GA & Jaksic FM (2005) Invasive vertebrate species in Chile and their control and monitoring by governmental agencies. Rev Chil Hist Nat 78:143-154.

Jax K & Rozzi R (2004) Ecological theory and values in the determination of conservation goals, examples from temperate regions of Germany, United States of America and Chile. Rev Chil Hist Nat 77:349-366.

Jönsson M, Ranåker l, Nicolle A, Ljungberg P, Fagerberg T, Hylander S, Jephson T, Lebret K, Einem J, Hansson LA, Nilsson PA, Balseiro E & Modenutti B (2011) Glacial clay

79 affects foraging performance in a Patagonian fish and cladoceran. Hydrobiologia 663:101- 108.

Landaeta MF & Castro LR (2006) Variabilidad estacional en los patrones espaciales de las asociaciones ictioplanctónicas de la zona de fiordos de Chile austral. Cienc Tecnol Mar 29(2):25-44.

Leprieur F, Beauchard O, Blanchet S, Oberdorff T & Brosse S (2008) Fish Invasions in the World's River Systems, When Natural Processes Are Blurred by Human Activities. PLoS Biol 6(2):404-410.

Leprieur F, Olden JD, Lek S & Brosse S (2009) Contrasting patterns and mechanisms of spatial turnover for native and exotic freshwater fishes in Europe. J Biogeography 36:1899- 1912.

Light T & Marchetti MP (2007) Distinguishing between invasions and habitat changes as drivers of diversity loss among California’s freshwater fishes. Conserv Biol 21(2):434-446.

Lowe S, Browne M, Boudjelas S & De Poorter M (2000) 100 of the World’s Worst Invasive Alien Species A selection from the Global Invasive Species Database. In, ISSG, Auckland. http,//www.iss G. org/pdf/publications. Cited 5 March 2011.

Macchi PJ, Cussac V, Alonso MF & Denegri MA (1999) Predation relationships between introduced salmonids and native fish fauna in lakes and reservoirs of northern Patagoni A. Ecol Freshwater Fish 8:227-236.

Macchi PJ, Pascual MA & Vigliano PH (2007) Differential piscivory of the native Percichthys trucha and exotic salmonids upon the native forage fish Galaxias maculatus in Patagonian Andean lakes. Limnologica 37:76-87.

Mack, R., Simberloff, D., Lonsdale, W., Evans, H., Clout, M. & Bazzaz, F. A. (2000) Biotic invasions. Causes, epidemiology, global consequences and control. Ecol Appl 10:689-710.

Martínez-Harms, M. J. & Gajardo, R. (2008) Ecosystem value in the Western Patagonia protected areas. J Nat Conserv 16:72-87.

McDowall, R. M. (1971a) The galaxiid fishes of South Americ A. Zool J Linn Soc-Lond 50:33-73.

McDowall, R. M. (1971b) Fishes of the family Aplochiton idae. J Roy Soc New Zeal 1:31- 52.

McDowall, R. M. (2003) Impacts of introduced salmonids on native galaxiids in New Zealand upland streams – a new look at an old problem. Trans Am Fish Soc 139:229-238.

80

McDowall, R. M. (2006) Crying wolf, crying foul or crying shame, alien salmonids and a biodiversity crisis in the southern cool-temperate galaxioid fishes. Rev Fish Biol Fisheries 16:233-422

McDowall, R. M., Allibone, R. M. & Chadderton, W. L. (2001) Issues for the conservation and management of Falkland Islands freshwater fishes. Aquatic Conserv Mar Freshw Ecosyst 11:473-486.

McIntosh, A. R. (2000) Habitat and size-related variations in exotic trout impacts on native galaxiid fishes in New Zealand streams. Can J Fish Aquat Sci 57:2140-2151.

McIntosh, A. R., Crowl, T. A. & Townsend, C. R. (1994) Size related impacts of introduced brown trout on the distribution of native common river Galaxias . N Z J Mar Freshwater Res 28:135-144.

McIntosh, A. R., McHugh, P., Dunn, N. R., Goodman, J., Howard, S., Jellyman, P. G. , O’Brien, L.K., Nystrom, P. & Woodford, D.J. (2010) The impact of trout on galaxiid fishes in New Zealand. N Z J Ecol 34:195-206

Mittermeier, R. A. , Mittermeier, C. G. , Brooks, T. M., Pilgrim, J. D., Konstant, W. R., da Fonseca, G. A. B. & Kormos, C. (2003) Wilderness and biodiversity conservation. PNAS 100(18):10309-10313.

Modenutti B, Albariño R, Bastidas Navarro M, Díaz Villanueva V, Souza M S, Trochine, C., Laspoumaderes C, Cuassolo F, Mariluán G, Buria L & Balseiro E (2010) Structure and dynamic of food webs in Andean North Patagonian freshwater systems, organic matter, light and nutrient relationships. Ecología Austral 20:95-114.

Pauchard A & Villarroel P (2002) Protected areas in Chile, history, current status and challenges. Nat Area J 22:318-330.

Pascual, M., Macchi, P., Urbanski, J., Marcos, F., Riva Rossi, C., Novara, M. & Dell’Arciprete, P. (2002) Evaluating potential effects of exotic freshwater fish from incomplete species presence–absence dat A. Biol Invas 4:101-113.

Pascual, M. A. , Cussac, V., Dyer, B., Soto, D., Vigliano, P., Ortubay, S. & Macchi, P. (2007) Freshwater fishes of Patagonia in the 21st Century after a hundred years of human settlement, species introductions, and environmental change. Aquat Ecosyst Health Manage 10(2):212 – 227.

Pascual, M. A. , Lancelotti, J. L., Ernst, B., Ciancio, J. E., Aedo, E. & Garcia-Asorey, M. (2009) The importance of scale, connectivity, and incentives in conservation programs, the case of exotic Salmonids of Patagonia. Front Ecol Environ DOI:10.1890/070127.

81

Penaluna B, Arismendi I & Soto D (2009) Evidence of Interactive Segregation between Introduced Trout and Native Fishes in Northern Patagonian Rivers, Chile. Trans Am Fish Soc 138: 839-845.

Peña H & Escobar F (1983a) Análisis de una crecida por vaciamiento de una represa glacial. Publicación Interna E.H. N. 83/4. DGA (Dirección General de Aguas), Departamento de Hidrología, Santiago. 1-20.

Peña H & Escobar F (1983b) Análisis de las crecidas del río Paine, XII Región. Publicación Interna E.H. N. 83/7. DGA (Dirección General de Aguas), Departamento de Hidrología, Santiago. 1-78.

Peña H & Escobar F (1987) Aspects of glacial hydrology in Patagonia. Bull Glac Res 4:141-150.

Pequeño G & Olivera F (2007) Peces litorales de los canales de Aysén, capturados durante los dos cruceros del Proyecto CIMAR 9 Fiordos, en el año 2003. Cienc Tecnol Mar 30(1):105-113.

Post DM (2002) Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83: 703–718.

Ricker, W. E. (ed.). 1968. Methods for Assessment of Fish Production in Fresh Waters. IBP Handbook N° 3 Int. Biol. Prog. 7 Marylebone Road London NW1 Blackwell Scientific Publications. 313 p

Rivera, A. & Casassa, G. (2004) Ice Elevation, Areal, and Frontal Changes of Glaciers from National Park Torres del Paine, Southern Patagonia Icefield. Arct Antarct Alp Res 36(4):379-389

Rodrígues ASL, Andelman SJ, Bakarr M I, Boitani L, Brooks TM, Cowling RM, Fishpool LDC, da Fonseca GAB, Gaston KJ, Hoffmann M, Long JS, Marquet PA, Pilgrim JD, Pressey RL, Schipper J, Sechrest W, Stuart S N, Underhill LG, Waller RW, Watts MEJ & Yan X (2004) Effectiveness of the global protected area network in representing species biodiversity. Nature 428:640–3.

Rodriguez-Cabal MA, Nuñez MA & Martinez AS (2008) Quantity versus quality, Endemism and protected areas in the temperate forest of South America. Austral Ecology 33:730-736.

Ruzzante DE, Walde SJ, Cussac VE, Macchi PJ, Alonso & Battini M. (2003) Trophic polymorphism in Pecichthys trucha (Pisces, Percichthyidae) in the Andes, phenotypic evidence for variation among lakes. Biol J Linn Soc 78:497-515.

82

Ruzzante DE, Walde SJ, Cussac VE., Dalebout ML, Seibert J, Ortubay S & Habit E (2006) Phylogeography of the Percichthyidae in Patagonia, roles of orogeny, glaciation, and volcanism. Mol Ecol 15(10):2949 -2968.

Ruzzante DE, Walde S J, Gosse JC, Cussac VC, Habit E, Zemlak TS & Adams EDM (2008) Climate control on ancestral population dynamics, insight from patagonian fish phylogeography. Mol Ecol 17:2234-2244.

Ruzzante DE, Walde SJ, Macchi PJ, Alonso M, Barriga JP (2011) Phylogeography and phenotypic diversification in the Patagonian fish Percichthys trucha: the roles of Quaternary glacial cycles and natural selection. Biol J Linn Soc 103:514-529.

Ruzzante D & Rabassa J (2011) Palaeogeography and palaeoclimatology of Patagonia: effects on biodiversity. Biol J Linn Soc 103:221–228.

Saunders D L, Meeuwig J J & Vincent ACJ (2002) Freshwater protected areas, strategies for conservation. Conserv Biol 16(1):30-41.

Sharp Z D (2005) Principles of Stable Isotope Geochemistry (Prentice Hall). 344pp.

Sielfeld W & Vargas M (1999) Review of marine fish zoogeography of Chilean Patagonia (42º - 57ºS). Scientia Marina, 63(1):451-463.

Sielfeld, W., Guzman, G. & Amado, N. (2006) Distribución de peces del litoral rocoso de los canales patagónicos occidentales (48°37'S - 53°34'S). Anal Inst Patagonia 34:21-32.

Simon, K. & Townsend, C. R. (2003) Impacts of freshwater invaders at different levels of ecological organization, with emphasis on salmonids and ecosystem consequences. Freshwater Biol 48:982–994.

Simonetti, J. & Armesto, J. J. (1991) Conservation of Temperate Ecosystems in Chile, Coarse vs. Fine Filter Approaches. Rev Chil Hist Nat 64:615-626.

Soto, D., Campos, H., Steffen, W., Parra, O. & Zúñiga, L. (1994) The Torres del Paine lake district (Chilean Patagonia), A case of potentially N-limited lakes and ponds. Arch Hydrobiol 99:181–197.

Soto, D., Jara, F. & Moreno, C. (2001) Escaped Salmon in the Inner Seas, Southern Chile, Facing Ecological and Social Conflicts. Ecol Appl 11(6):1750-1762.

Soto, D., Arismendi, I., González, J., Sanzana, J., Jara, F., Jara, C., Guzman, E. & Lara, A. (2006) Southern Chile, trout and salmon country: conditions for invasion success and challenges for biodiversity conservation. Rev Chil Hist Nat 79(1):97-117.

83

Soto, D. & De los Rios, P. (2006) Influence of trophic status and conductivity on zooplankton composition in lakes and ponds of Torres del Paine National Park (Chile). Biología, Bratislava, 61(5), 541-546.

Soto, D., Arismendi, I., Di Prinzio, C. & Jara, F. (2007) Establishment of Chinook salmon (Oncorhynchus tshawytscha) in Pacific basins of southern South America and its potential ecosystem implications. Rev Chil Hist Nat 80(1):81-98.

Townsend, C. R. (1996) Invasion biology and ecological impacts of brown trout Salmo trutta in New Zealand. Biol Conserv 78: l 3-22.

Townsend, C. R. (2003) Individual, population, community and ecosystem consequences of a fish invader in New Zealand streams. Conserv Biol 17:38–47.

Tognelli, M. F., Ramirez de Arellano, P. I. & Marquet, P. A. (2008) How well do the existing and proposed reserve networks represent vertebrate species in Chile? Divers Distrib 14, 148-158.

Vargas, P. V., Arismendi, I., Lara, G. , Millar, J. & Peredo, S. (2010) Evidence of microhabitat overlap between juvenile of introduced salmon Oncorhynchus tshawytscha and the native fish Trichomycterus areolatus in the Allipén River, Chile. Rev Biol Mar Oceanogr 45(2):285-292,

Vander-Zanden, M. J. & Gratton, C. (2011) Blowin’ in the wind, reciprocal airborne carbon fluxes between lakes and land. Can J Fish Aquat Sci 68:170-182.

Weber, C. (1986) Conservación y uso racional de la naturaleza en áreas protegidas. Ambiente y Desarrollo, 2(1):165-181.

Wells, R.W., Bunnell, F.L., Haag, D., & Sutherland, G. (2003) Evaluating ecological representation within differing planning objectives for the central coast of British Columbia. Can J For Res 33(11):2141-2150.

Young, K.A., Stephenson, J.F., Terreau, A., Thailly, A.F., Gajardo, G. & Garcia de Leaniz, C. (2009) The diversity of juvenile salmonids does not affect their competitive impact on a native galaxiid. Biol Invas 11:1955-1961.

Young, K.A., Dunham, J. B., Stephenson, J.F., Terreau, A., Thailly, A.F., Gajardo, G. & Garcia de Leaniz, C. (2010) A trial of two trouts, comparing the impacts of rainbow and brown trout on a native galaxiid. Anim Conserv 13:399-410.

Zemlak, T. S., Habit, E. M., Walde, S. J., Battini, M. A., Adams, E. & Ruzzante, D. E. (2008) Across the southern Andes on fin, glacial refugia, drainage reversals and a secondary contact zone revealed by the phylogeographic signal of Galaxias platei in Patagonia. Mol Ecol 17:5049-5061.

84

Zemlak, T. S., Habit, E. M., Walde, S. J., Carrea, C., & Ruzzante, D. E. (2010) Surviving historical Patagonian landscapes and climate, molecular insights from Galaxias maculatus . BMC Evol Biol 10: 67.

Zemlak, T., Walde, S., Habit, E. & Ruzzante, D. (2011) Climate induced changes to the ancestral population size of two Patagonian galaxiids: the influence of glacial cycling. Mol Ecol . doi: 10.1111/j.1365-294X.2011.05352.x

85

Figure 1 . Sampling systems (L: Lake, R: River) in the study area of the Serrano River basin. Protected correspond to the area preserved by the National Parks, and Unprotected are areas with no protection for conservation. .

86

a 1 1

1 1 1

1

0 1 1 0 1

b

1

0 1

1

0 0

0 1

1

Figure 2. Distribution of native and salmonid species by altitude, connectivity and distance from the sea within protected (1) and unprotected (0) areas. a) Lakes, and b) Rivers inhabited by salmonid and native species ( ), only salmonids ( ), only by native species G. platei, A. zebra and A. taeniatus ( ) and only G. platei ( ).

87

a 3.0 N=9 N=2 n=18 a n=3 2.5

2.0

1.5 a

1.0

richness Species

0.5

0.0 Protected Unprotected 3.0 b

2.5

N=5 2.0 N=4 n=7 n=6

1.5

a 1.0 richness Species 0.5 a

0.0

Protected Unprotected

Figure 3. Mean species richness ± standard error for native (open bars) and salmonid (solid bars) species of protected and unprotected areas by a) lakes, and b) rivers. Same letters over bars correspond to significant differences (p<0.05) between protected and unprotected areas. The numbers above bars (N) indicate the number of lakes and rivers within each category, and (n) indicate the number of sampling sites, which differed among lakes or rivers according to size, i.e., larger lakes were sampled in more locations than small lakes (See Table 1 for details). Means and standard errors were calculated per sampling site.

88

11.0

10.5

10.0

15 9.5

δN

9.0

8.5 Salmonids invaded lakes Salmonids free lakes 8.0 -24.0 -23.5 -23.0 -22.5 -22.0 -21.5 -21.0 13 δC

Figure 4 . Mean (±error standard) stable isotope signature (δC 13 , δN 15 ) of G. platei in 3 invaded (n samples =65) and 5 free salmonid lakes (n samples =215).

89

TABLE 1 . Average (± standard error) of environmental variables included in the community analyses, pooled by system type (lakes and rivers) and zones. Az: A. zebra, At: A. taeniatus, Gp: G. platei, Gm: G. maculatus, St: S. trutta, Ot: O. tshawytscha, Om: O. mykiss

Number Connectivity Distance of category from Altitude Lake area Turbidity Temperature Conductivity TDS Native sites (0-5) sea (km) (masl) (km 2) (NTU) (°C) pH (µS/cm) (ppm) species Salmonids LAKES PROTECTED Azul 2 1 92 223 3 1.9±0.4 11.7±0.04 8.6±0.06 158±0.6 79±0.6 Gp, At, Az. - Dickson 2 1 112 204 28 41.8±15.2 10.3±3.4 7.1±0.2 20±9.0 10±4.0 Gp - Mellizas 1 0 57 91 0.1 4.8 13.4 9.5 1228 614 Gp - Nordenskjold 2 2 65 64 28 20.9±4.3 10.1±0.2 8.6±0.2 41.7±0.3 20.7±0.3 Gp - Pehoe 2 3 53 34 22 14.9±0.2 11.3±0.02 8.0±0.04 42±0.0 21±0.0 Gp - Sarmiento 2 0 77 74 86 2.0±0.55 9.6±0.07 10.0±0.33 1211.5±67.5 605.5±33.5 Gp St Vision 1 0 60 83 0.1 4.6 14.3 9.1 1170 585 Gp - Grey 2 4 56 30 36 67.2±0.07 5.3±0.02 8.0±0.1 29.3±0.3 14.3±0.3 - St, Ot Toro 4 4 42 23 202 9.3±1.5 9.1±0.2 8.0±0.04 60.8±4.9 31.7±2.8 Gp, Gm, Az St, Ot, Om Gm, Gp, At, UNPROTECTED Porteño 2 3 75 31 24 0.8±0.03 10.5±0.0 8.5±0.2 45.7±0.3 22.7 Az. St Maravilla 1 4 66 25 2 0.2 10.1 8.5 40 20 Az St RIVERS PROTECTED Paine 1 2 73 114 - 5.8 10.4 8.4 66 33 Gp - Victorina 1 2 77 70 - 51.7 10.7 8.8 41 20 Gp - Asencio 1 1 100 164 - 51.7 10.7 8.8 41 20 Gp - Serrano 3 4 36.4±1.0 22.4±0.4 - 1.1±0.0 7.374±0.0 8.056±0.0 72±0.0 36±0.0 Gm St, Ot, Om Serrano Estuary 1 5 <1 1 - 1.1 7.4 8.0 72 36 Gm St, Ot, Om UNPROTECTED Chinas 3 4 124.3±22.4 128.7±61.4 - 25.8±9.6 13.0±0.8 8.9±0.2 117.3±8.3 59±4.0 - St Porteño 1 3 75 31 - 0.8 10.5 8.1 46 23 Az St Tres Pasos 1 4 95 156 - 5 21.5 9.0 250 124 - St Maravilla 1 4 66 25 - 0.8 10 8.5 40 20 - St

90

TABLE 2 . Species contributions to differences (dissimilarity) between high and low zones in unprotected and protected areas, in lakes and rivers, based on similarity percentage procedure (SIMPER) for relative abundance. n.s.: non significant.

AVERAGE ABUNDANCE CONTRIBUTION TO (% CONTRIBUTION) DISSIMILARITY (%) PROTECTED/ PROTECTED UNPROTECTED UNPROTECTED LAKES G. platei 56.3 (85.6) - 23.5 S. trutta 17.0 (10.1) 19.0 (15.0) 16.6 G. maculatus - 42.3 (54.4) 26.4 A. zebra - 21.7 (25.9) 18.7 RIVERS S. trutta 22.3 (46.0) 90.0 (100.0) 30.9 G. maculatus 35.6 (25.6) - 23.5 G. platei 27.3 (14.6) - 20.4 O. tshawytcha 11.8 (13.9) - 15.7

91

TABLE 3. Spearman correlation coefficient (Rho) for BIOENV analysis, showing the variables that best explain the community patterns in both lakes and rivers. In bold the best significant set of variables.

NºVar. Variables Rho Best Set (p)

LAKES 1 Turbidity 0.20 2 Turbidity, connectivity, 0.30 0.05 3 Turbidity, connectivity, temperature 0.24 RIVERS 1 Connectivity 0.79 0.01 2 Connectivity, pH 0.79 3 Connectivity, pH, total dissolved solids 0.75

TABLE 4. Comparison of individuals of G. platei analyzed from invaded and salmonid free lakes. K represents the condition factor.

Total Length (mm) K ± SE δC13 ± SE δN15 ± SE Lakes n Mín-Max Mean ± SE Invaded by 138.9 ± 0.61 ± -22.68 ± Salmonid 65 55-395 9.33 0.02 0.37 8.86 ± 0.21 126.7 ± 0.73 ± -21.97 ± 10.44 ± Salmonid Free 215 45-375 4.60 0.02 0.22 0.20

92

CAPITULO IV: TROPHIC INTERFERENCE BY SALMO TRUTTA ON APLOCHITON ZEBRA AND A. TAENIATUS IN SOUTHERN PATAGONIAN LAKES

ABSTRACT

The length weight ratio, diet and isotopic composition of Aplochiton zebra and A. taeniatus inhabiting a salmonid invaded ( Salmo trutta ) and a salmonid free lake in southern Patagonia were compared. Results indicate that S. trutta exercises important trophic interference over A. zebra and A. taeniatus , causing changes in their dietary composition and diet-related behavior, reducing the consumption of winged Diptera, feeding behavior that involves jumping out of the water to catch prey. This effect is significantly higher in A. zebra than in A. taeniatus which has a highly specialized diet. The dietary changes of A. zebra and A. taeniatus in sympatry with S. trutta lead to an impoverishment of their isotopic nitrogen signals (δ15N), suggesting a reduction of their trophic position. In the case of A. zebra , this translates into a significant decrease in its body condition factor. Such interference could lead to a population decline of this species and would explain the current distribution range decline and allopatry with S. trutta in fluvial systems.

Anaysa Elgueta, Jorge González, Daniel Ruzzante, Sandra Walde & Evelyn Habit. Enviada. Journal of Fish Biology.

93

INTRODUCTION

Biological invasions constitute one of the most important threats to the conservation of biodiversity in general and of freshwater biodiversity in particular, as introduced species interfere with the use of resources by native species (Pilger et al., 2010). In the Southern Hemisphere, interactions of introduced salmonids with native galaxiids have generated negative effects at distinct levels of ecological organisation from individuals to ecosystemic changes (Simon & Townsend, 2003). As a result, galaxiids are one of the most severely threatened fish groups worldwide (McDowall, 2006). At present, three of the five most common galaxiid species in Chile have experienced important reductions in their distributional ranges due, at least in part, to the effect of salmonid invasions (Habit et al., 2010).

The introduction of salmonid fish in Chile began in the early 1900s, with the stocking of brown and rainbow trout ( Salmo trutta Linnaeus, 1758 and Oncorhynchus mykiss (Walbaum 1792), respectively; Soto et al., 2001; Basulto, 2003). Although it is generally accepted that these two species have negatively affected and continue to affect the distribution and abundance of native fish species (Soto et al., 2006, 2007; Correa & Hendry, 2012; Habit et al., 2010, 2012), few studies have examined the corresponding mechanisms in Chile (Penaluna et al., 2009; Young et al., 2010; Ibarra et al., 2011; Correa et al., 2012). The most cited mechanisms for explaining these negative interactions include competition for food and space, and predation (McIntosh et al., 1994; Macchi et al., 1999; McIntosh, 2000). To avoid predation by S. trutta , the native galaxiids tend to occupy suboptimal habitats, thereby reducing their grazing efficiency (McIntosh et al., 1992; Edge et al., 1993; Lattuca et al., 2007). Salmonids have a competitive advantage in the use of food resources and may be ichthyophagous from small sizes. Juvenile Chinook salmon, Oncorhynchus tshawytscha (Walbaum, 1792), as small as 8.5 cm in total length (LT) are known to prey on Galaxias maculatus (Jenyns, 1842) (Ibarra et al., 2011), and rainbow trout (O. mykiss) have been estimated to consume 2.5 kg and 2.0 kg of G. maculatus larvae and adults per year, respectively (Vigliano et al., 2009). Salmonids, both juvenile and adults, also consume the same benthic prey as galaxiids but at much greater rates (McIntosh, 2000; Simon & Townsend, 2003; Eby et al., 2006). Thus, trophic interference is one of the most plausible mechanisms explaining the negative effects of invading salmonids on native galaxiids (García de Leaniz et al., 2010). One of the expected results from these interferences is the reduction of the trophic position (as measured by 15N), as was reported by Vander Zanden et al. (1999) for the native species Salvelinus namaycush (Walbaum in Artedi, 1792) in presence of the introduced Micropterus dolomieu Lacepéde, 1802 and Ambloplites rupestris (Rafinesque, 1817).

Among the galaxiids present in Chile, the genus Aplochiton is represented by two species Aplochiton zebra (Jenyns, 1842) and Aplochiton taeniatus (Jenyns, 1842), (McDowall, 2006). Both species appear to be heavily affected by salmonids in southern Chile (Soto et al., 2006; Vanhaecke et al., 2012) where their local (Young et al., 2010) and regional distributions (Habit et al., 2010) have been severely reduced. In this study the potential trophic intereference of S. trutta on the two Aplochiton species, A. zebra and A. taeniatus , was analized by comparing their diet in

94 two lakes in Southern Patagonia. In one lake the two Aplochiton species coexist with S. trutta but in the other, S. trutta is absent. The presence of these two Aplochiton species in a single lacustrine system is extremely rare and has only been reported in this far-southern watershed in the Patagonia (Habit et al., 2012). Moreover, salmonid-free lakes tend to be single-species; most such lakes are inhabited exclusively by Galaxias platei (Steindachner, 1898) (Habit et al., 2012). Thus, the occurrence of two lakes in a single watershed, both lakes inhabited by A. zebra and A. taeniatus , with only one of them invaded by a salmonid, S. trutta , constitutes a unique, non- repeatable, natural experiment.

It is proposed that interference by S. trutta with the feeding habits of Aplochiton sp. should result in changes in the dietary compositions and isotopic nitrogen and carbon signals of A. zebra and A. taeniatus between the two lacustrine systems. In particular, given that the diet of S. trutta juveniles broadly overlaps with that of Aplochiton sp ., sympatry among these three species should translate into alterations in the diet of Aplochiton, which then feed on suboptimal prey. This would be reflected in lower condition factors and a reduced isotopic signature of δ15N, leading to a lower trophic position of Aplochiton sp. in a salmonid invaded lake than in a salmonid free lake.

MATERIALS AND METHODS

Study area and fish sample

This study considers two lakes – Azul (50º 52` 36” S, 72º 44 `34” W) and Porteño (51º 21` 47” S, 72º 47` 40” W) – in the Serrano River watershed. Lake Azul is located within the Torres del Paine National Park and Biosphere Reserve at 223 m.a.s.l. and at 92 km waterway distance from the sea. This lake has low connectivity with the fluvial system of the Paine River to which it is connected via a small stream with one waterfall and is free of salmonids. Lake Porteño is outside the boundaries of the National Park at 31 m.a.s.l. and at 75 km waterway distance from the sea. It is connected to the Serrano River through two lakes. This lake was stocked with Salmo trutta, probably prior to 1950.

The fishes were caught using multi-mesh gill nets (10, 15, 20, 30, 50, 60, 70, 120 mm bar mesh size) placed at the surface and anchored at the bottom in the littoral zone of each lake. Electrofishing and seine nets were also used where feasible (e.g. vegetated areas along the shorelines and sandy beaches, respectively). In all cases similar habitat types were sampled in both lakes to avoid effects of differences due to habitat complexity which can interfere in the outcome of interactions between introduced predators (Stuart-Smith et al., 2007). Sampling was conducted during the season of highest precipitation (80 mm per month on average) and when temperatures were highest (10° C average; January to March 2009, May 2010). Both lakes were sampled on four occasions during this period. During May 2010 we also sampled macroinvertabrates using Surber nets in the littoral zone of both lakes. The fishes were measured and weighed in situ, then fixed in 95% ethanol with an injection in the stomach to preserve the

95 stomach contents. The stomachs of S. trutta individuals > 15 cm LT were extracted and fixed directly in the field.

Fish condition

Fish well-being was estimated via the length-weight ratio as W = a Lb, where W is body mass (g), L is LT (cm), and a and b are coefficients of the regression function between W and L. This was done for the native fish species in both lakes. Then the condition factor K as K=100 [W (L 3)-1] (Beckmann, 1948) was estimated. The condition factors of the native species in the salmonid-free lake were compared to those of the same species in the invaded lake using a one- way analysis of variance with the lake as the fixed factor (ANOVA) and Tukey’s post-hoc test of comparisons.

Diet analysis

Stomach contents were identified to the lowest possible taxonomic level. Prey items were classified as being of terrestrial (allochthonous) or aquatic (autochthonous) origin. To characterize the diet by species and lakes, the index of relative importance (IRI) (Pinkas et al., 1971) was estimated. This index combines the numerical, gravimetric, and frequency of occurrence measurements into one value. Intra- and inter-lake dietary overlaps between species using the Pianka index (1973) and the trophic diversity of each fish species per lake were estimated with the Shannon & Weaver index (1949). A non-metric multidimensional scaling analysis (nMDS) based on the IRI values and the Euclidean distance index (PRIMER-E v6) was used to visualize diet clustering patterns among species and lakes.

Isotopic analysis

The isotopic composition of muscle tissue and common prey (Ephemeroptera and Chironomidae larvae) were measured through the proportions δ15N = 15N:14N and δ13C = 13C:12C in each individual, using the ratio δX =1000 [(Rsample - Rstandard) (Rstandard)-1], where X is the proportion in delta (δ) units relative to a standard and Rsample and Rstandard are the absolute isotopic proportions of the sample and the standard, respectively. The standards for carbon and nitrogen were, respectively, fossil calcite (Pee Dee Belemnite of Vienna or VPDB) and atmospheric nitrogen. The results are expressed in parts per thousand (‰), (Ehleringer & Rundel, 1989). The isotopic composition δ15N was analyzed as an indicator of trophic position and δ13C as an indicator of carbon sources (Post, 2002). To normalize the isotope signature to the mean value for a basal consumer within each lake, trophic position was also estimated using the simplest model for a secondary consumer (Post, 2002), as follows: trophic position TP = λ + [(δ15NAplochiton – δ15Nbase) ∆n-1], where λ is the trophic position of the organism used to estimate δ15N base, and ∆n is the enrichment in δ15N per trophic level. Two different prey taxa were used to estimate δ15Nbase, Ephemeroptera and Chironomidae larvae. Both groups were present in the stomach content of all individuals of the three species in both lakes. Based on Vander Zanden et al. (1997), we designed λ= 2.5 for both prey types. As trophic fractionation of

96 nitrogen (∆n) the value proposed by Post (2002), i.e., 3.4‰ was used. For each species and lake, the relationship of isotopic signals with size (LT, cm) was studied using simple linear regression analysis. Finally, the variations of both isotopes and of the ratio δ15N:δ13C between species and lakes were analyzed through ANOVA. Given that the design was incomplete, one-way ANOVA with species-lake as a fixed factor and Tukey as an a posteriori test were used.

RESULTS

In Lake Azul, the fish fauna comprises exclusively three native Galaxiidae: G.platei, A. zebra, A. taeniatus. These three species are also present in Lake Porteño where they coexist with G. maculatus, also a native species, and with the introduced S. trutta . In total, the LT:W ratio of 116 individuals, the stomach contents of 219 individuals, and the isotopic signals of 169 individuals were analysed (see Table I for a breakdown by lake and species). The great majority of the native fishes analyzed were juveniles (except one specimen of A. zebra in Lake Porteño), whereas S. trutta specimens were both juveniles and adults (Table I).

Fish condition

Aplochiton zebra exhibited a higher condition factor in Lake Azul (no salmonids) than in Lake Porteño (with salmonids; ANOVA, F1, 55= 5.54, P < 0.001). Aplochiton taeniatus exhibited instead no difference in condition factor between lakes (ANOVA, F1, 57 = 2.89, P > 0.05; Table I). Length and weight were significantly correlated (P < 0.001) in the two species and growth was generally isometric (allometric coefficient, b, between 2.9 and 3.5) with one exception, A. zebra in Lake Azul which exhibited positive allometry (Table I).

Diet analysis

In Lake Azul (without salmonids) both species exhibited similar diet composition, with A. zebra consuming fewer distinct prey items than A. taeniatus (Table II), but A. taeniatus showed a lower dietary diversity than A. zebra (Table III, Figure 1). Both species showed a high dietary overlap largely due to the consumption of winged insects (adult allochthonous Diptera; Table IV, Figure 1). Both species differed in their diets between lakes (Figure 2); in Lake Porteño (invaded by salmonids), they also differed from each other in their diets. Aplochiton taeniatus exhibited a very narrow trophic niche in Lake Porteño (Table III) with its diet consisting nearly exclusively of winged insects. Aplochiton zebra instead, exhibited a more generalist diet in Lake Porteño than in Lake Azul, reducing the consumption of allochthonous insects (Figure 1).

Salmo trutta presented a diverse diet, consuming prey items similar to Aplochiton sp. in the juvenile stage. The diet of the adults was made up exclusively of the amphipod Hyalella and fishes of the species G. maculatus (Table II, Figure 1).

97

Only the diet of A. zebra broadly overlapped with that of S. trutta in Lake Porteño (Table IV). The diets of these two species were very similar under coexistence but they differed from that of A. taeniatus (Figure 2).

Isotopic analysis

Aplochiton zebra and A. taeniatus from Lake Azul presented significantly higher values of δ15N than those from Lake Porteño (ANOVA, F3, 112= 90.98, P < 0.01, Figure 3), although their δ13C signals did not differ (ANOVA, F3, 112=1.68, P > 0.05). Within Lake Porteño, juveniles of A. taeniatus and A. zebra were enriched in 15N in relation to S. trutta juveniles, and their 13C was also higher than in S. trutta , reflecting a more littoral feeding signal for this group (Figure 3). Only S. trutta presented a significant enrichment of 15N (r = 0.72, P < 0.001) and lower 13C (r = 0.28, P < 0.001) with size, indicating a more authocthonous carbon signal in adults than in juveniles (Figure 4). These results are consistent with the inferences on trophic position as estimated based on both Ephemeroptera and Chironomidae larvae (Table V). Both A. zebra and A. taeniatus reached a lower trophic position in Lake Porteño than in Lake Azul.

DISCUSSION

In this study the body condition, diet and isotopic composition (muscle) between A. zebra and A. taeniatus inhabiting a salmonid invaded ( S. trutta ) and a salmonid-free lake ecosystem were compared. The setting represents a unique natural experiment, facilitating the examination of some of the potential causal mechanisms underpinning the negative effects of salmonids on galaxiids (Macchi et al., 1999; Soto et al., 2006; Habit et al., 2010, 2012). The results indicate that S. trutta exercises important trophic interference over Aplochiton sp ., causing changes in their dietary composition and diet-related behavior and reducing their δ15N signals. This effect is higher in A. zebra than in A. taeniatus , as discussed below, but first a number of caveats are raised and potential limitations of the analyses are discussed. Then the results are discussed in more detail.

Caveats on the comparison of two lakes

It is suggested that the difference between lakes in the trophic ecology and well-being of the two species of Aplochiton results from the interaction with the invasive species S. trutta . Other differences between the two lakes may, however, be influencing their trophic ecology. First, the fish community differs between lakes not only in the presence of brown trout, S. trutta in Lake Porteño but also in that of the native G. maculatus. The results of this study are unlikely to be explained by the presence of G. maculatus in Lake Porteño since this species is generally found in a different habitat. Aplochiton sp. juveniles are commonly found in the littoral zone of lakes, close to tributaries, the same environment inhabited by S. trutta juveniles. Neither G. maculatus nor G. platei are generally found near lake tributaries (Habit et al. 2010). In addition, G. maculatus in lakes feed preferentially on zooplankton (Barriga et al. 2011), not on

98 macroinvertebrates. A significant degree of trophic interference among Galaxias and Aplochiton species is therefore not expected.

A second caveat is that the prey availability was not measured, and diet could differ between lakes due to a difference in food supply. Nevertheless, the prey identification is at high taxonomic level, and significant differences between lakes at this level are not expected. Previous information on the diet composition of A. zebra in Patagonian lake systems is available only from salmonid invaded lakes and in all these lakes A. zebra exhibits a diverse diet based mainly on benthic insect larvae. For example, in Lake Epuyén and Foyén, where A. zebra coexists whit S. trutta , O. mykiss and Salvelinus fontinalis (Mitchill, 1814) (Aigo et al., 2008) it feeds mainly on chironomid larvae and pupae (Lattuca et al., 2007). In Lake Quetro in Chile, also invaded by salmonids, A. zebra was described as a generalized invertebrate predator (McDowall & Nakaya, 1988). Thus, this study describes, for the first time, the diet composition of A. zebra in a salmonid free lake, and it shows that under these conditions A. zebra specializes on winged insects. If the inter-lake difference in diet composition were due to differences in prey availability (i.e., differences in the availability of winged insects), changes in the diet of both A. zebra and A. taeniatus would have been expected, but this was not the case. These results showed that adult insects were available also in the invaded lake since they were present in the stomach content of A. taeniatus in that lake.

Another factor that may influence the relative importance of allochthonous input to a single lake is the shoreline in relation to the area of the lake. Even though Lake Porteño is nearly an order of magnitude larger in surface area than the trout-free lake (Azul), their shoreline to lake area ratio did not differ between lakes (Lake Azul = 0.61, and Lake Porteño = 0.63). Therefore a difference between lakes in the relative importance of allochthonous terrestrial prey input is not expected.

In addition, even though the two species of Aplochiton showed significant differences in the Nitrogenous isotope, they do not differ in the Carbon isotope, which is a good indicator of food source (Post, 2002). This is strong evidence that food sources do not differ between the lakes, suggesting that the detected difference in diet is not primarily explained by differences in prey availability.

Finally, the two lakes also differ in physical characteristics like altitude (Porteño is 31 m.a.s.l. and Azul is 223 m.a.s.l.), distance to the sea (75 km for Porteño and 92 km for Azul) and size (Porteño is 24.4 km2 and Azul is 3 km2). These differences should have a primary effect on lake productivity, but both lakes belong to the same category of deep and oligotrophic lacustrine systems of the Torres del Paine National Park (sensu De Los Ríos & Soto, 2009).

99

Trophic interference

In the absence of S. trutta , the native species A. zebra and A. taeniatus prey heavily on Diptera adults; which likely is possible only if feeding behavior involves jumping out of the water to catch prey. The two species differ, however, in the benthic component of their diet, a division that probably avoids or minimizes interspecific competition. Furthermore, A. zebra exhibits a more generalist diet, while A. taeniatus is a specialist, as indicated by McDowall & Nakaya (1988). The diet of A. taeniatus is largely limited to flying Diptera insects and complemented by only some benthic prey. Even though the diet of A. taeniatus does not overlap with that of S. trutta , the tendency to be more specialists in Lake Porteño suggests a change of predatory habits, probably avoiding benthic preys consumed both by A. zebra and S. trutta .

The diet of A. zebra overlaps that of S. trutta when found in allopatry as is the case in the rivers of the Malvinas-Falkland Islands, where these species are not found sympatrically (Perry, 2007). In allopatry and even more so in sympatry, both A. zebra and S. trutta juveniles consume a broad gamut of benthic macroinvertebrates and winged Diptera, showing a large similarity in their feeding behavior, probably associated with jumping. Moreover, both species occupy very similar habitats in rivers and lakes (Habit et al., 2010). Given the broad overlap in the way they use and exploit resources, these species have a high potential for competition (Pilger et al., 2010). In the rivers of the Malvinas-Falkland Islands, as in low-order rivers in Chiloé Island in Chile, the allopatry between these species has been explained mainly as a result of the competitive exclusion of Aplochiton sp . by S. trutta (McDowall et al., 2001; Young et al., 2010).

In sympatry with S. trutta (Lake Porteño), A. zebra changes its diet, reducing mainly the consumption of winged Diptera. That is, this species presents a more benthophagous diet in the presence of S. trutta that includes items such as Chironomid pupae, Odonata nymphs, and Trichoptera larvae. Therefore, in sympatry with S. trutta, A. zebra feeds less on winged individuals which are probably caught by jumping, thus reducing predation risk (Brown et al., 2006). Paradoxically, the final results is that the diet of A. zebra ends up being more similar to the the diet of S. trutta juveniles, raising the possibility that this shift in diet may reduce predation risk by S. trutta while at the same time it may increase the potential for competition.

The dietary changes of A. zebra and A. taeniatus , and diversification of the diet of A. zebra in sympatry with S. trutta seem to lead to a reduction of their trophic position, as suggested strongly by the impoverishment of their isotopic nitrogen signals (δ15N, Deniro & Epstein, 1981; Cabana & Rasmussen, 1994; Vander Zanden et al., 1999). A similar effect has recently been described by Correa et al. (2012) who showed that the trophic height of large G. platei individuals declined with increasing trout density. In the case of A. zebra , the significant change in its diet and the reduction of δ15N at similar sizes in the presence of S. trutta translate into a significant decrease in body condition factor. This index has been inversely correlated with population abundance for example in salmonids (Arismendi et al., 2011) suggesting that the final result for A. zebra in the presence of S. trutta can be a reduction of its population size. In parallel, in the same system

100 where A. zebra could be declining, S. trutta shows a clear pattern of enrichment in the values of 15N and 13C with size, revealing an ontogenetic change in its diet to more ichthyophagous habits as they grow larger, as reported for S. trutta in other systems (Cucherousset et al., 2007). In this study, we found only G. maculatus and no other native fish in the stomachs of S. trutta adults, though it is possible that this salmonid preys on other galaxiids (McDowall, 2006).

The change with size in the 13C signal in S. trutta reflects a change in the diet from littoral to pelagic and deep-water in larger individuals. In contrast, in both Aplochiton species, the 13C signal indicates a diet associated exclusively with littoral environments (France, 1995; Vander Zanden et al., 1999) in both lakes. These data are consistent with the overlapping diets and habitats of juvenile stages of both native species under study and S. trutta . It is interesting to note that for A. taeniatus , a diet based nearly exclusively on Diptera adults (winged) results in a greater 15N signal than found for the other two species. The more benthofagous diet of S. trutta juveniles is reflected in the lower signal for this isotope (Vander Zanden et al., 1999). Therefore, a reduction in the consumption of winged insects by A. zebra as we found in the invaded lake, can result in a significant change in their trophic position.

Therefore it was shown that S. trutta exerts a strong impact on the trophic ecology of the galaxiid species of the genus Aplochiton, with particular emphasis on that of A. zebra . Such interference could lead to a population decline of this species, even in lacustrine systems, which are recognized as environmental refuges for the galaxiidae in Patagonia (Soto et al., 2006; Habit et al., 2010; 2012). This would explain the current allopatry with S. trutta in fluvial systems (Perry, 2007; Young et al., 2010) dominated by salmonids (Habit et al., 2010).

REFERENCES

Aigo, J., Cussac, V., Peris, S., Ortubay, S., Gómez, S., López, H., Gross, M., Barriga, J. & Battini, M., 2008. Distribution of introduced and native fish in Patagonia (Argentina): patterns and changes in fish assemblages. Reviews in Fish Biology and Fisheries 18, 387-408.

Arismendi, I., Sansana, J. & Soto, D. (2011). Seasonal age distributions and maturity stage in a naturalized rainbow trout (Oncorhynchus mykiss Walbaum) population in southern Chile reveal an ad-fluvial life history. Annales de Limnologie/International Journal of Limnology 47, 133- 140.

Barriga, J. P., Battini, M. Á., García-Asorey, M., Carrea, C., Macchi, P. J. & Cussac, V. E. (2011). Intraspecific variation in diet, growth, and morphology of landlocked Galaxias maculatus during its larval period: the role of food availability and predation risk. Hydrobiologia 679, 27- 41.

Basulto, S. (2003). El Largo Viaje De Los Salmones Una Crónica Olvidada. Editorial Autoedición. Santiago, Chile. 300 pp.

101

Beckmann, W. (1948). The length - weight relationship, factors for conversions between standard and total lengths, and coefficient of condition for seven Michigan fishes. Transactions of the American Fisheries Society 75, 237-256.

Brown, E., Rive, A., Ferrai, M. & Chivers, D. (2006). The dynamic nature of antipredator behavior: prey fish integrate threat-sensitive antipredator responses within background levels of predation risk. Behavioral Ecology and Sociobiology 61, 9–16.

Cabana, G. & Rasmussen, B. (1994). Modelling food chain structure and contaminant bioaccumulation using stable nitrogen isotopes. Nature 372, 255–257.

Correa, C. & Hendry, A. P. (2012). Invasive salmonids and lake order interact in the decline of puye grande Galaxias platei in western Patagonia lakes. Ecological Applications. 22, 828–842

Correa, C., Bravo, A. P. & Hendry, A. P. (2012). Reciprocal trophic niche shifts in native and invasive fish: salmonids and galaxiids in Patagonian lakes. Freshwater Biology. doi: 10.1111/j.1365-2427.2012.02837.x.

Cucherousset, J., Aymes, J., Santoul, F. & Cereghino, R. (2007). Stable isotope evidence of trophic interactions between introduced brook trout Salvelinus fontinalis and native brown trout Salmo trutta in a mountain stream of south-west France. Journal of Fish Biology 71, 210-233.

De Los Ríos, P. & Soto, D. (2009). Estudios limnológicos en lagos y lagunas del Parque Nacional Torres del Paine (51° S, Chile). Anales del Instituto de la Patagonia. 37, 63-71.

Deniro, M. & Epstein, S. (1981). Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta 45, 341–351.

Eby, L., Roach, J. Crowder, L. & Standford, J. (2006). Effects of stocking-up freshwater food webs. Trends in Ecology and Evolution 21, 576 – 584.

Edge, K., Townsend, C. R. & Crowl, T. A. (1993). Investigating anti-predator behavior in three genetically differentiated population of non-migratory galaxiid fishes in a New Zealand river. Journal of marine and freshwater research (New Zealand) 27, 357- 363.

Ehleringer, J.R. & Rundel, P.W. (1989). Stable isotope History, units and instrumentation. Stable isotopes in ecological research. Pringer overlang. New York.

France, R. (1995). Differentiation between littoral and pelagic food webs in lakes using carbon isotopes. Limnology and Oceanography 40, 1310–1313.

García de Leaniz, C., Gajardo, G. & Consuegra, S. (2010). From best to pest: Changing perspectivas on the impact of exotic salmonids in the southern hemisphere. Sistematics and Biodiversity 8, 447–459.

102

Habit, E., Piedra, P., Ruzzante, D., Walde, S., Belk, M., Cussac, V., Gonzalez, J., & Colin, N. (2010). Changes in the distribution of native fishes in response to introduced species and other anthropogenic effects. Global Ecology and Biogeography 19, 697-710.

Habit, E., Gonzalez, J., Ruzzante, D. & Walde, S. (2012). Native and introduced fish species richness in Chilean Patagonian lakes: inferences on invasion mechanisms using salmonid-free lakes. Diversity and Distributions. Published online. DOI: 10.1111/j.1472-4642.2012.00906.x

Ibarra, J., Habit, E., Solis-Lufí, K. & Barra, R. (2011). Juveniles de salmon Chinook (Oncorhynchus tshawytscha) en ríos y lagos de la Patagonia chilena. Gayana 75, 90-99.

Lattuca, M., Ortubay, S., Battini, M.A., Barriga, J.P. & Cussac, V. (2007). Presumptive environmental effects on body shape of Aplochiton zebra (Pisces, Galaxiidae) in northern Patagonian lakes. Journal of Applied Ichthyology 23, 25–36.

Macchi, P., Cussac, V., Alonso, M. & Denegri, M. (1999). Predation relationships between salmonids and the native fish fauna in lakes and reservoirs in northern Patagonia. Ecology of Freshwater Fish 8, 227-236.

McDowall, R.M. & Nakaya, K. (1988). Morphological Divergence in the Two Species of Aplochiton Jenyns (Salmoniformes: Aplochitonidae): A Generalist and A Specialist. Copeia 1, 233-236.

McDowall, R.M., Allibone & Chadderton, W. L. (2001). Issues for the conservation and management of Falkland Islands freshwater fishes. Aquatic Conservation: Marine and Freshwater Ecosystems 11, 473–486.

McDowall, R. M. (2006). Crying wolf, crying foul, or crying shame: alien salmonids and a biodiversity crisis in the southern cool-temperate galaxioid fishes? Reviews in Fish Biology and Fisheries 16, 233-422.

McIntosh, A.R., Townsend, C. R. & Crowl, T. A. (1992). Competition for space between introduced brown trout (Salmo trutta L.) and a native galaxiid (Galaxias vulgaris Stokell) in a New Zealand stream. Journal of Fish Biology 41, 63–81.

McIntosh, A. R., Crowl, T. A. & Townsend, C. R. (1994). Size-related impacts of introduced brown trout on the distribution of native common river galaxias. Journal Marine Freshwater Research (New Zeland) 28, 135-44.

McIntosh, A. R. (2000). Habitat- and size-related variations in exotic trout impacts on native galaxiid fishes in New Zealand streams. Canadian Journal of Fisheries and Aquatic Sciences 57, 2140–2151.

103

Penaluna, B., Arismendi, I. & Soto, D. (2009). Evidence of Interactive Segregation between Introduced Trout and Native Fishes in Northern Patagonian Rivers, Chile. Transactions of the American Fisheries Society 138, 839–845.

Perry, B. J. (2007). Effects of non-native salmonids on native fish fauna in Chile and the Falkland Islands. Submitted in partial fulfilment for the degree of MSc in Aquaculture and the Environment. Department of Biological Sciences School of the Environment & Society University of Wales Swansea. Swansea. United Kingdom. 64 pp

Pianka, E. (1973). The structure of lizard communitues. Annual Review of Ecology & Sistematics 4, 53-74.

Pinkas, L., Oliphant, M.S. & Iverson, I.L.K. (1971). Food habits of albacore, blue fin tuna, and bonito in California waters. California Department of Fish and Game 152, 1-105.

Post, D. M. (2002). Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83, 703–718.

Pilger, T. Gido, K. & Propst, D. (2010). Diet and trophic niche overlap of native and nonnative fishes in the Gila River, USA: implications for native fish conservation. Ecology of fresh water fish 19, 300−321.

Shannon, C. E. & Weaver, W. (1949). The Mathematical Theory of Communication. University of Illinois Press, Urbana.

Simon, K. S. & Townsend, C. (2003). The impacts of freshwater invaders at different levels of ecological organization, with emphasis on salmonids and ecosystem consequences. Freshwater Biology 48, 982-994.

Soto, D., Jara, F. & Moreno, C. (2001). Escaped salmon in the inner seas, southern Chile: facing ecological and social conflicts. Ecological Applications 11, 1750-1762.

Soto, D., Arismendi, I., González, J., Sanzana, J., Jara, F., Jara, C., Guzmán, E. & Lara, A. (2006). Sur de Chile, país de truchas y salmones: patrones de invasión y amenazas para las especies nativas. Revista Chilena de Historia Natural 79, 97-117.

Soto, D., Arismendi, I., Di Prinzio, C. & Jara, F. (2007). Establishment of Chinook salmon (Oncorhynchus tshawytscha) in Pacific basins of southern South America and its potential ecosystem implications. Revista Chilena de Historia Natural 80, 81–98.

Stuart - Smith, R. D., J. F. Stuart - Smith, R.W.G White & L. A Barmuta. (2007). The impact of an introduced predator on a threatened galaxiid fish is reduced by the availability of complex habitats. Freshwater Biology 52, 1555 – 1563.

104

Vanhaecke, D., García de Leaniz, C., Gajardo, G., Young, K., Sanzana, J., Orellana, G., Fowler, D., Howes, P., Monzon-Arguello, C. & Consuegra, S. (2012). DNA barcoding and microsatellites help species delimitation and hybrid identification in endangered galaxiid fishes. PLoS One 7, 1- 10.

Vander Zanden, M. J., Cabana, G. & Rasmussen J. B. (1997). Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (δ15N) and literature dietary data. Canadian Journal of Fisheries and Aquatic Sciences 54, 1142–1158

Vander Zander, M., Casselman, J. & Rasmussen, J. (1999). Stable isotope evidence for the food web consequences of species invasions in lakes. Nature 401, 464–467.

Vigliano, P., Beachamp, D., Milano, D., Macchi, P., Alonso, M., García Asorey, M., Denegri, M., Ciancia, J., Lippolt, G., Rechencq, M. & Barriga, J. (2009). Quantifying predation on galaxiids and other native organisms by introduced rainbow trout in an ultraoligotrophic lake in northern Patagonia, Argentina: a bioenergetics modeling approach. Transactions of the American Fisheries Society 138, 1405–1419.

Young, K., Dunham, J., Sthephenson, J., Terreau, A., Thailly, A., Gajardo, G. & García de Leaniz, C. (2010). A trial of two trouts: comparing the impacts of rainbow and brown trout on a native galaxiid. Animal Conservation 13, 399 - 410.

105

Table I. Number (n) and sizes of individuals studied by lake and type of analysis. Correlation values (r) are given between total length (LT) and weight (W), the parameters of the linear equation between LT and W, and the factor condition (K).

Species Lake n n Size range r a b K ± standard error

Diet Isotopes (LT cm) (LT-W)

A. zebra Azul 19 19 7.2 – 11.1 0.97 -3.35 4.30 0.78 ± 0.04

A. zebra Porteño 52 38 6.0 – 15.5 0.91 -2.11 2.93 0.68 ± 0.02

A. taeniatus Azul 70 49 5.5 – 9.1 0.88 -2.53 3.35 0.58 ± 0.02

A. taeniatus Porteño 12 10 6.2 – 7.0 0.94 -2.67 3.59 0.64 ± 0.01

S. trutta Porteño 66 53 3.8 - 77 0.99 -2.22 3.13 0.82 ± 0.02

1

106

Table II. Dietary composition (percentange of abundance) of the three species (Az: A. zebra ; At: A. taeniatus ; St: S. trutta juveniles and adults) in Lake Porteño (P) and Lake Azul (A).

A.z A.t A.z A.t St juv St adults

Order Family-Species State Habitat (A) (A) (P) (P) (P) (P)

Diptera Undetermined Adult Terrestrial 67.5 76.0 29.9 90.1 24.5

Diptera Chironomidae Larva Aquatic 12.7 1.3 13.2 0.1 15.6

Diptera Chironomidae Pupa Aquatic 0.9 7.9 4.5

Diptera Simulidae Larva Aquatic 0.6 0.9

Plecoptera Austroperlidae Adult Aquatic 0.6

Plecoptera Undetermined Adult Aquatic 0.3 1.7

Plecoptera Gripopterygidae Adult Aquatic 5.7 1.8 0.7 24.5 0.8

Plecoptera Notonemuridae Adult Aquatic 0.9 1.1

Coleoptera Undetermined Adult Terrestrial 2.2 0.7 5.6 3.5 2.2 0.2

Coleoptera Elmidae Larva Aquatic 0.6

Coleoptera Dytiscidae Adult Aquatic 1.3 0.1 0.8

Ephemeroptera Leptophlebiidae Adult Aquatic 4.8 1.6 1.9 0.7 15.3

Trichoptera Undetermined Larva Aquatic 0.9 0.4 8.2 0.3 0.3

Hymenoptera Undetermined Adult Terrestrial 0.9

Hemiptera Undetermined Adult Terrestrial 2.2 5.7 10.6 1.50 0.4

Hemiptera Corixidae Adult Terrestrial 1.3 0.3 1.5

Odonata Undetermined Adult Terrestrial 7.9 8.5 4.2

Amphipoda Hyaella Adult Aquatic 1.7 64.6

Lumbriculida Lumbricidae - Aquatic 4.1

Araneae Undetermined Adult Terrestrial 0.7 1.6 3.5 2.0

Osmeriformes Galaxias maculatus - Aquatic 30.9

107

Table III. Trophic diversity (Shannon & Weaver index, 1949) of the three species in Lake Azul (without salmonids) and Lake Porteño (with S. trutta ).

Dietary breadth Without salmonids (L. Azul) With salmonids (L. Porteño)

A. zebra 1.23 2.28

A. taeniatus 1.02 0.50

S. trutta - 2.44

Table IV. Dietary overlap (Pianka index, 1973) of the species in Lake Porteño and Lake Azul.

Dietary overlap A.taeniatus (L.Azul) S.trutta (L.Porte ño) A.taeniatus (L.Porte ño)

A.zebra (L.Azul) 0.83 - -

A.taeniatus (L.Porte ño) - 0.39 -

A zebra (L.Porte ño) - 0.67 0.81

108

Table V. Trophic position of Aplochiton sp . and Salmo trutta in lakes Azul (non invaded) and Porteño (invaded by S. trutta ).

Trophic level Lake Species Based on Based on Chironomidae Ephemeroptera

Azul A. taeniatus 3.9 3.6

Azul A. zebra 3.6 3.3

Porteño A. taeniatus 3.3 3.2

Porteño A. zebra 2.9 2.8

Porteño S. trutta Adult 3.2 3.1

Porteño S. trutta Juveniles 2.7 2.7

109

Lake Azul Lake Porteño

100

80

60

40

Percentage of abundance abundance of Percentage 20

0 A. zebra A. taeniatus A. zebra A. taeniatus S. trutta S. trutta Juv Adults

Figure 1. Dietary composition of A. zebra . A. taeniatus and S. trutta shown in terms of the prey items with indexes of abundance numeric N > 2%. All individuals are combined by species and age classes to determine the proportion of the relative importance of the items found in the stomachs. Diptera adults ( ), Hemiptera adults ( ), Aracnidae ( ), Coleoptera adults

( ), Trichoptera adults ( ), Ephemeroptera nymphs ( ), Chironomidae larva ( )

Chironomidae pupa ( ), Gripopterygidae nymphs ( ), Lumbricidae ( ), Galaxias maculatus ( ), Odonata nymphs ( ), Hyalella ( ).

110

Figure 2. Visualization of the diet clustering patterns using non metric multi-dimensional scaling analysis (nMDS) of the index of relative importance ( IRI ) based on Euclidean Distance index.

Black triangle: A. zebra ; White circle: A. taeniatus ; Asterisk: S. trutta . Countour lines correspond to Euclidean distance of 0.28 (dotted line) and 0.42 (solid line).

111

Figure 3. δ 13 C:δ15 N ratio for juveniles of A. zebra ( ) , A. taeniatus ( ), and S. trutta ( ) in both lakes.

112

Figure 4. Values of δ 15 N and δ 13 C for S. trutta juveniles and adults in relation to their total length

(cm) in Lake Porteño.

113

DISCUSION.

Patrones de diversidad de la ictiofauna nativa en la Patagonia chilena.

La diversidad de especies en Chile presenta un hotspot en la zona centro-sur del país, esto ha sido evidenciado para flora y fauna, tanto terrestre como acuática (Myers et al. 2000). Esta alta biodiversidad disminuye hacia el sur para la mayoría de los grupos taxonómicos, incluyendo macroinvertebrados (Valdovinos 2008) y peces dulceacuícolas (Soto et al 2006).

Los resultados de esta investigación muestran que la diversidad de la ictiofauna nativa, específicamente, la riqueza de especies de Galáxidos varía con la latitud y es mayor en la zona central de la distribución del grupo, entre 39 ⁰ y 42ºS. Este patrón latitudinal se mostró más fuertemente en las cuencas de origen andino, con especies como G. maculatus dominando en la zona norte, G. platei en la zona sur y las especies del género Aplochiton se presentaron como más comunes en la zona central (Capítulo I). Específicamente en lagos, encontramos que la riqueza de especies nativas, tomando en cuenta todas las especies no solo Galáxidos, se relacionó negativamente con la latitud, mostrando un máximo en lagos de la zona centro-sur del país como el Riñihue (7 especies) y un mínimo en lagos patagónicos, como los de la cuenca del rio Serrano (4 especies) y la isla de Tierra del Fuego, donde solo se ha reportado G. platei (Capítulo II).

Patrones similares, que relacionan negativamente la diversidad de peces dulceacuícolas con la latitud, se han encontrado en cuencas al este de los Andes y en la Patagonia Argentina, donde se explican por la perdida de especies de origen Brasilico hacia el sur (Baigún & Ferriz 2003, Aigo et al. 2008). En general, encontramos que estos cambios de diversidad de especies en cuencas de la Patagonia están dados por la prevalencia de especies australes como los Galáxidos y la desaparición paulatina de especies de otros grupos taxonómicos, como Characidae, Atherinopsidae, Diplomystidae, Trichomycteridae y finalmente Percichthydae (Capítulo II). Estos grupos poseen sus hábitats típicos en cuencas la zona Centro-Sur de Chile con características hidrográficas particulares (Habit 2006) y existen diferentes hipótesis biogeográficas históricas que explican su distribución y diversidad restringida a áreas de endemismo de la denominada Provincia Biogeográfica Chilena (Dyer 2000). Por el contrario, los Galáxidos por características propias del grupo relacionadas a su habilidad de dispersión (Baigún & Ferriz 2003, Barriga et al. 2006) y la adaptación a hábitats fríos habrían colonizado mas al sur (Cussac et al. 2004), estableciéndose en gran parte de la Patagonia.

En este sentido, los patrones de diversidad de la fauna nativa en Chile, estarían fuertemente relacionados con el clima, específicamente la temperatura (Capítulos II y III), la que a través de procesos bioquímicos y relaciones ecológicas afectaría la evolución, características biológicas y la historia de vida de los peces dulceacuícolas, podría explicar gran parte de la distribución y diversidad de las especies (Cussac et al. 2009). Ahora bien, se espera que estos patrones de diversidad y distribución natural de la fauna íctica nativa de Chile estén sufriendo modificaciones

114 a través de la historia producto de las actividades humanas y por la introducción de especies exóticas.

Procesos de invasión de peces introducidos y su efecto sobre las especies nativas.

Las consecuencias que la introducción de especies invasoras puede tener sobre la fauna y los ecosistemas nativos pueden ser observados a distintos niveles de organización biológica (Simon & Townsend 2003, Townsend 2003), sin embargo, determinar cuales efectos son causados por estos procesos de invasión se vuelve una tarea difícil, sobre todo en áreas geográficas donde la intervención humana es intensa y de larga data.

En esta investigación se encontraron cambios a gran escala en la distribución de 5 de las especies más abundantes y ampliamente distribuidas de Galáxidos, reduciéndose en la mayoría de ellas su rango distribucional, producto de los cuales se puede encontrar una disminución de la riqueza de especies en algunas zonas del área de estudio. Esto ocurriría en la zona norte a consecuencia de la pérdida de la especie G. maculatus y en la zona centro-sur (36-41ºS) por la ausencia de otros Galáxidos, A. taeniatus y B. bullocki (Capítulo I). Nuestros resultados sugieren que estos cambios son producto de la interacción negativa entre especies introducidas con las nativas. Sin embargo, constricciones en la distribución de especies nativas podrían estar determinadas por cambios ambientales como variaciones en la temperatura global (Cussac et al. 2004), por otro lado, aumentos en el rango de distribución de especies podrían estar relacionado con los esfuerzos de pesca y mejoramiento de los métodos de captura (Aigo et al. 2008).

Por otro lado, estas interacciones pueden interferir en la abundancia de especies nativas, en este estudio encontramos que las especies nativas se mantienen abundantes donde las abundancias de salmónidos son bajas (Capítulo II), lo que ha sido descrito para Galáxidos en Chile (Arismendi et al. 2009, Soto et al. 2006), Argentina, Tasmania (Ault & White 1994) y Nueva Zelanda (Townsend 1996), como efecto de la interacción negativa con salmónidos. Sin embargo, la presencia de múltiples estresores ambientales (cambios de uso de suelo, degradación de hábitats, construcción de represas entre otros) hacen difícil despejar o individualizar el efecto de esta interacción entre salmónidos y nativos (Flecker & Townsend 1994, McIntosh et al. 2000).

Además, se observó que los efectos serían mayores en ríos que en lagos, probablemente debido a que los ambientes lacustres ofrecen mejores refugios para nativos (Macchi et al. 1999, Pascual et al. 2007) y donde la complejidad de estos hábitats sería un elemento clave para la conservación de la fauna nativa (Pascual et al. 2007, Stuart-Smith et al. 2007).

A una escala menor, para la cuenca del río Serrano, encontramos que a pesar de que la invasión de los salmónidos en las zonas estudiadas no ha generado cambios regionales en la riqueza de especies nativas, estaría impactando sobre algunas de ellas. Encontramos una marcada distribución disjunta entre nativos y truchas lo que ha sido descrito como producto de la depredación o exclusión competitiva (Townsend & Crowl 1991, Towsend, 1996, 2003). Esta interacción estaría desplazando a las especies nativas a hábitats menos productivos y provocando

115 cambios en su conducta alimentaria lo que consecuentemente, se ve reflejado en variaciones en la composición de la dieta, una disminución de su bienestar y posición trófica (Capitulos III y IV).

Encontramos que estos procesos de invasión de salmónidos estarían fuertemente determinados por la conectividad de los sistemas, encontrando la mayor diversidad de estas especies en la zona baja de la cuenca, asociados al ingreso de especies o poblaciones anádromas y una menor diversidad sobre cascadas, donde dominarían las especies nativas (Towsend 1996, 2003).

De acuerdo al aumento de la acuicultura y el fomento a la pesca deportiva, los salmónidos estarían ampliando su distribución y aumentando su diversidad dentro del territorio (Arismendi et al. 2009, Pascual et al. 2009), y es esperable que este tipo de efectos sobre especies nativas aumenten, lo que hace urgente planes y áreas de conservación eficientes.

Efectividad de las áreas silvestres protegidas en la conservación de la ictiofauna nativa.

En Chile un considerable porcentaje del territorio está cubierto por áreas de conservación, sin embargo, éstas no estarían diseñadas en concordancia con la diversidad, distribución y endemismos de la fauna nativa, o no poseen la suficiente información biológica y ecológica para realizar una conservación eficiente sobre todo en un sistema enfocado al resguardo de la fauna terrestre más que la acuática (Januchowski-Hartley et al. 2011, Martínez-Harms & Gajardo 2008, Pauchard & Villarroel 2002, Rodriguez-Cabal et al. 2008, Simonetti & Armesto 1991, Tognelli et al. 2008). Finalmente, dado que los parques y reservas están ubicadas en áreas distintas a las de la mayor diversidad de fauna dualceacuícola nativa, terminan protegiendo a los salmónidos y no las especies nativas (Pascual et al. 2007, Soto et al. 2006).

En esta investigación comprobamos, a una menor escala, que los límites las áreas silvestres protegidas presentes en la cuenca del río Serrano no se ajustan a la distribución actual ni a los patrones de diversidad de peces nativos dentro de este ecosistema, es así como encontramos sitios con la mayor riqueza de especies nativas fuera de estas áreas (Capítulo III). En este sentido estas áreas son más efectivas en el resguardo de las poblaciones de salmónidos, incentivando la pesca deportiva no extractiva, que en la protección de la biodiversidad de Galáxidos.

La implementación de este tipo de áreas protegidas requiere del desarrollo de criterios de gestión sostenible de las comunidades de peces patagónicas con de una perspectiva holística que abarque tanto la conectividad de las cuencas hidrográficas, la presencia de los hábitats críticos, distribución y abundancia de especies nativas y exóticas, así como las demandas y prioridades socio-económicas. Solo así, se podrá lograr un manejo eficiente enfocado en conservar nuestra fauna nativa.

Manejo de la fauna para la conservación de las especies nativas.

Entre las medidas de manejo para la conservación de la fauna nativa, resulta prioritario implementar áreas protegidas que incluyan la mayor diversidad de especies posible. En este

116 estudio se identificaron a gran escala cuatro áreas geográficas de interés para la conservación, donde los efectos de los salmónidos y la actividad humana han sido lo suficientemente bajo como para que las poblaciones de Galáxidos sigan siendo abundantes. Las áreas propuestas son: (1) el río Maullín y la zona costera de la cuenca, (2) la cuenca del río Valdivia, (3) los ríos del norte de la isla de Chiloé y (4) varios lagos en la región de Aysén, la mayoría de las cuales incluye la totalidad de las especies de Galáxidos. Esta ultima incluye los lagos Yulton y Meullín, que todavía se mantienen libres de salmónidos (Soto et al ., 2006), el Lago Thompson, donde G. platei sigue siendo dominante aún cuando se han introducido salmónidos (Habit et al . 2006a), y el lago Riesco, que tiene una gran riqueza de especies considerando su alta latitud (45 º S). Ninguna de estas áreas están protegidas y sólo uno (zona costera de Maullín) ha sido propuesta como un área importante para la conservación de los vertebrados (Tognelli et al ., 2008).

Además, proponemos 15 lagos que identificamos como libres de salmónidos entre los 45 ⁰ y 54 ⁰ de latitud sur, que deberían ser considerados como prioritarios para la conservación de poblaciones únicas de peces nativos. Es decir, los ambientes libres de salmónidos son de vital importancia para la conservación de especies nativas, puesto que representan probablemente la última oportunidad para entender las características ecológicas de las comunidades naturales y servirían como referencia para dimensionar los impactos en ambientes invadidos (Baigún y Ferriz 2003). Así también, los lagos son eficientes refugios para la fauna nativa y sobre todo las zonas litorales de estos, son ambientes relevantes a la hora de establecer medidas para la conservación de la diversidad de especies nativas (Aigo et al . 2008, Macchi et al . 1999).

Adicionalmente a estas medidas de manejo a gran escala, a continuación entregamos algunas propuestas específicas para la cuenca del rio Serrano.

Propuestas de conservación y manejo para la fauna íctica de la cuenca del río Serrano.

Tomando en cuenta la alta vulnerabilidad de las especies de Galáxidos presentes en la cuenca del río Serrano, la conservación de éstas debería ser una prioridad para las áreas silvestres protegidas que se encuentran en este territorio. Por otro lado, siendo la pesca recreativa basada en especies salmonídeas una actividad turística importante para la zona, resulta necesario mantener un “stock” de peces que permita que dicha actividad perdure. Dicho esto y basándonos en los resultados de esta tesis, proponemos medidas generales de conservación de la fauna íctica nativa y manejo de la fauna introducida de la cuenca del rio Serrano,

Los ambientes que consideramos de mayor importancia para la conservación de la fauna íctica nativa son los lagos, puesto que poseen las mayores diversidades de especies nativas y se ha comprobado que son buenos refugios para evitar los efectos de las especies invasoras. Dentro de estos lagos consideramos los criterios de alta diversidad y alta singularidad como los más importantes para definir medidas de conservación.

Lagos con alta diversidad. Específicamente, lagos de baja turbidez y sus ríos tributarios como el L. Porteño y L. Azul, por encontrarse en posición elevada respecto a otros tramos de la cuenca

117 permanecen relativamente protegidos de procesos de invasión y poseen una alta diversidad de especies nativas. Consideramos que estos sistemas son prioritarios para preservación de las poblaciones de especies nativas que aquí habitan, las cuales podrían actuar como poblaciones fuente para la recolonización natural del resto de la cuenca, o en caso de ser necesario, poder implementar planes de traslocación de especies . Por otro lado, para asegurar la mantención de estas poblaciones sería óptimo implementar planes de restauración de los ecosistemas que incluyan mejoramiento de hábitat y erradicación de especies exóticas. El caso del L. Porteño por encontrarse fuera de las áreas protegidas, es un territorio ideal para postular a ampliación de la Reserva Mundial de la Biosfera UNESCO: “Torres del Paine”, y así poder implementar medidas de conservación efectivas.

Lagos de alta singularidad. Estos son lagos de mayor turbidez, baja conectividad y dominancia de G. platei como el L. Pehoe, L. Nordenksjold, L. Paine, L. Dickson, L. Azul, L. Visión del Mundo y L. Mellizas, también deberían considerarse los ríos tributarios y que conectan estos sistemas entre sí. Si bien es cierto, estos ecosistemas no poseen alta diversidad de nativos, se encuentran libres de salmónidos lo que los hace únicos y les otorga un alto valor para la investigación, principalmente en relación al entendimiento de la ecología de estas poblaciones naturales en ecosistemas “prístinos” y a la conservación de poblaciones con alta singularidad genética que han sobrevivido exitosamente a distintos procesos glaciales. Como medida de conservación es fundamental mantener estos ecosistemas libres de salmónidos.

Además de estas medidas de conservación a nivel de ecosistemas, se deberían considerar medidas específicas para las especies en peligro de extinción presentes en la cuenca. Un plan de conservación completo para el grupo Aplochiton debería incluir proyectos de investigación que se enfoquen en determinar, por ejemplo, áreas de reproducción y reclutamiento así como programas de difusión y educación que permitan dar a conocer la importancia de estas especies casi desconocidas por la población local.

Por otro lado, las medidas de manejo de la fauna introducida, principalmente trucha Café y salmón Chinook, deberían asegurar el mantenimiento de actividades productivas en relación con estas especies exóticas en lagos y tramos de la cuenca que se encuentran invadidos, y que históricamente han sido utilizados para ello. Para ordenar el funcionamiento de estas actividades, proponemos que existan áreas donde se erradiquen los salmónidos y se utilicen principalmente para la investigación y la conservación de nativos, y otras áreas donde se manejen los salmónidos y se mantengan los “stock” necesarios para estas actividades, principalmente pesca recreativa.

Lagos prioritarios para erradicación de salmónidos . Estos son lagos invadidos por salmónidos y aislados del resto de la cuenca por barreras naturales (L. Sarmiento y L. Porteño) que ofrecen una excelente oportunidad para implementar planes de erradicación local de truchas que han sido eficientes en otros países (Townsend 2003), y que serían pioneros en Chile. Estos planes deberían incluir el fomento a la pesca extractiva de especies salmonídeas (campeonatos de pesca, letreros y

118 folletos informativos, etc.) y campañas regulares de extracción con redes en el lago y pesca eléctrica en los tributarios.

Lagos y ríos prioritarios para el manejo de salmónidos. Consideramos que ambientes que se encuentran actualmente mas invadidos por salmónidos, como la subcuenca del rio Las Chinas y Tres Pasos, el lago Toro y el rio Serrano, deberían ser consideradas para establecer planes de manejo de la fauna exótica. Dentro de las medidas generales de manejo esta la de sectorizar estos ecosistemas, dejando áreas donde se permita la pesca extractiva en bote o desde las orillas como en el lago Toro y así mantener controladas las poblaciones y la invasión de salmónidos hacia aguas arriba en la cuenca. En otros sectores como el rio Serrano debería incentivarse el uso exclusivo para la pesca deportiva con devolución, protegiendo además los sitios de reproducción y así mantener los tamaños poblacionales adecuados para el desarrollo de esta actividad. Por otro lado, proponemos que se tenga en cuenta la aplicación de medidas de manejo de salmónidos especie especificas, incentivando la captura con extracción de salmón Chinook y la pesca con devolución para la truchas Café y Arcoiris, esta ultima muy poco abundante en la cuenca. Además, creemos que es imperativo establecer medidas de manejo asociadas a la pesca recreativa como las ya señaladas pero que en todo momento se evite la siembra de salmónidos .

Finalmente, esperamos que toda esta información sea un aporte significativo al entendimiento de la interacción entre especies nativas y salmónidas, al manejo de los sistemas acuáticos dulceacuícolas dentro de áreas silvestres protegidas y para la conservación de la fauna nativa de la Patagonia .

119

CONCLUSIONES.

Basados en que la distribución y diversidad de la ictiofauna nativa en la Patagonia fueron determinados por procesos naturales históricos especie específicos, pero que además estarían siendo modificados por procesos de invasión de salmónidos, esperamos encontrar una fuerte relación entre niveles de dominancia de salmónidos y distintos patrones de distribución, comunitarios y poblacionales de la fauna íctica nativa, y de acuerdo a las hipótesis previamente planteadas, las conclusiones de este trabajo son las siguientes:

.- A gran escala, cambios en la distribución y una directa relación entre patrones de diversidad de especies nativas y las dominancia de salmónidos.

Se concluye que a gran escala la diversidad y los cambios encontrados en la distribución de algunos Galáxidos en el centro-sur y Patagonia chilena, estarían actualmente determinados tanto por procesos biogeográficos así como por múltiples cambios ambientales no pudiéndose separar de los supuestos efectos producidos por la invasión de salmónidos.

.- A una escala menor, esperamos encontrar efectos sobre la distribución, riqueza y abundancia locales, así como sobre el estado físico, amplitud y posición trófica de Galáxidos, producto de la dominancia de salmónidos. Por otro lado, esperamos que las áreas silvestres protegidas no muestren efectos significativos sobre la dominancia de salmónidos y abundancia de especies nativas, respecto a las no protegidas.

Se concluye que a menor escala, la distribución disjunta, depredación, interferencias tróficas y disminución del bienestar de poblaciones de Galáxidos en presencia de truchas son una clara evidencia de los efectos que pueden tener los salmónidos sobre la fauna nativa. Además, las áreas silvestres protegidas son poco efectivas en la protección de esta fauna nativa.

En este contexto creemos que es prioritario el establecimiento de tanto planes de conservación a nivel nacional como proyectos de manejo a nivel local que permitan entender en detalle estos procesos de invasión y mitigar sus efectos sobre la fauna nativa.

120

REFERENCIAS.

Aigo J, Cussac V, Peris S et al. 2008. Distribution of introduced and native fish in Patagonia (Argentina): patterns and changes in fish assemblages. Rev Fish Biol Fish. doi: 10.1007/s11160- 007-9080-8

Arismendi, I., Soto, D., Penaluna, B., Jara, C., Leal, C. & León-Muñoz, J. 2009. Aquaculture, non-native salmonid invasions, and associated declines of native fishes in lakes of the northern Chilean Patagonia. Freshwater Biology, 54, 1135-1147.

Arratia, G., G. Rojas & A. Chang. 1981. Géneros de peces de aguas continentales de Chile. Publicaciones ocasionales Museo Nacional de Historia Natural, Santiago de Chile 34: 3-108

Arratia, G. 1997. Brazilian and Austral fresh-water fish faunas of South America. A contrast. En: Tropical Biodiversity and Systematic (Ed. H. Ulrich), pp. 179-186. Proceedings of the international Symposium on Biodiversity and Sistematics in Tropical Ecosystem. Museum Alexander Koenig, Bonn.

Ault, T. R., and R. W. White. 1994. Effects of habitat structure and the presence of brown trout on the population density of Galaxias truttaceus in Tasmania, Australia. Transactions of the American Fisheries Society 123:939–949.

Baigun, C. & R. Ferriz 2003. Distribution patterns of native freshwater fishes in Patagonia (Argentina). Org. Divers. Evol. 3: 151-159.

Barriga J. P., M. A. Battini & V. E. Cussac. 2006 Annual dynamics variation of a landlocked Galaxias maculatus (Jenyns 1842) population in a Northern Patagonian river: occurrence of juvenile upstream migration. J. Appl. Ichthyol. 23, 128–135.

Basulto, S. 2003. El Largo Viaje De Los Salmones Una Crónica Olvidada. Editorial Autoedición. 300 pp.

Battini, M., V. Rocco, M. Lozada, B. Tartarotti & H. Zagarese. 2000. Effects of ultraviolet radiation on the eggs of landlocked Galaxias maculatus (Galaxiidae, Pisces) in northwestern Patagonia. Freshwater Biology 44: 547-552.

Becker L.A., M.A. Pascual & N.G. Basso. 2007. Colonization of the Southern Patagonia Ocean by Exotic Chinook Salmon. Conservation Biology. 21( 5): 1347–1352.

Boy C. C., E. Morriconi and J. Calvo. 2007. Reproduction in puyen, Galaxias maculatus (Pisces: Galaxiidae), in the southernmost extreme of distribution. J. Appl. Ichthyol. 23: 547–554.

Buschmann A. & A. Fortt. 2005. Efectos ambientales de la acuicultura intensiva y alternativas para un desarrollo sustentable. Revista Ambiente y Desarrollo 21(3): 58-64.

121

Campos, H. 1970. Introducción de especies exóticas y su relación con los peces de agua dulce de Chile. Noticiario Mensual del Museo Nacional de Historia Natural, Chile 14 (162) : 6-9.

Chapman A., D. Morgan, S. Beautty & H. Gill. 2006. Variation in life history of land-locked lacustrine and riverine populations of Galaxias maculatus (Jenyns 1842) in Western Australia. Environ Biol Fish. 77:21–37

Correa C. & M. Gross. 2007. Chinook salmon invade southern South America. Biological Invasions. 10(5): 615

Cussac, V., S. Ortubay, G. Inglesias, D. Milano, M. Lattuca, J.P. Barriga, M. Battini & M. Gross. 2004. The distribution of south american galaxiid fishes: the role of biological traits and post- glacial history. Journal of Biogeography 31: 103-121.

Cussac, V.E., D.A. Fernández, S.E. Gómez and H.L. López. 2009. Fishes of Southern South America: a story driven by temperature. Fish Physiology and Biochemistry 35: 29-42. DOI: 10.1007/s10695-008-9217-2.

DGA. 2004. Diagnostico y clasificación de los cursos y cuerpos de agua según objetivos de calidad. Cuenca del río Serrano. Documentos de la Direccion general de aguas. Ministerio de obras publicas. Gobierno de Chile. 95pp.

Dyer, B. 2000. Systematic review and biogeography of the freshwater fishes of Chile. Estudios Oceanológicos (Chile) 19: 77-98.

Dudgeon, D, Arthington, AH, Gessner, MO, Kawabata, ZI, Knowler, DJ, Leveque, C, Naiman, RJ, Prieur-Richard, AH, Soto, D, Stiassny, MLJ & Sullivan, CA. 2006. Freshwater biodiversity: importance, threats, status and conservation challenges, Biological Reviews, 81(2):163-182.

Flecker, A.S. & C.R. Townsend. 1994. Community-wide consequences of trout introduction in New Zealand streams. Ecological Applications, 4, 798–807.

González J. 2005. Evaluación de interferencias tróficas entre truchas y peces nativos en ríos del sur de Chile, a través del análisis de contenidos estomacales. Tesis para optar al grado de Licenciado en Biología Marina. Universidad Austral de Chile.

Habit, E., B. Dyer & I. Vila. 2006a. Estado de conocimiento de los peces dulceacuícolas de Chile. Gayana 70(1): 100-112.

Habit, E. Belk, M., Tuckfield, C. & Parra, O. 2006b. Response of the fish community to human- induced changes in of the Biobío River in Chile. Freshwater Biology, 51, 1-11.

Ibarra J., E. Habit, K. Solis-Lufí & R. Barra. 2011. Juveniles de Salmón Chinook (Oncorhynchus tshawytscha) Ríos y Lagos de la Patagonia Chilena. Gayana, 75(1),90-99.

122

Irz, P., Argillier, C. and Oberdorff, T. 2004. Native and introduced fish species richness in French lakes: local and regional influences. Global Ecology and Biogeography 13, 335–44.

Januchowski-Hartley S., Pearson R., Puschendorf R. & Rayner T. 2011. Freswater and fish diversity: Distribution, Protection and Disturbance in Tropical Australia. Plos One. 6(10:1-11.

Kitano S. 2004. Ecological impacts of rainbow, brown and brook trout in Japanese inland waters. Glob Environ Res 8:41–50

Korsu K., Huusko A., Muotka T. 2008. Ecology of alien species with special reference to stream salmonids. Boreal Environment Research 13 (Suppl. A): 43-52

Korsu K., Huusko A., Muotka T. 2010. Impacts of invasive stream salmonids on native fish: using meta-analysis to summarize four decades of research. Boreal Environ. Res. 15: 491-500

Lattuca, M., M. Battini, and P. Macchi. 2008a. Trophic interactions among native and introduced fishes in a northern Patagonian oligotrophic lake. Journal of Fish Biology 72:1306–1320.

Leprieur, F., Beauchard, O., Blanchet, S., Oberdorff, T. & Brosse, S. 2008. Fish Invasions in the World's River Systems, When Natural Processes Are Blurred by Human Activities. PLoS Biology, 6(2), 404-410.

Leprieur, F., Olden, J.D., Lek, S. & Brosse, S. 2009. Contrasting patterns and mechanisms of spatial turnover for native and exotic freshwater fishes in Europe. Journal of Biogeography, 36, 1899-1912.

Light T, Marchetti MP. 2007. Distinguishing between invasions and habitat changes as drivers of diversity loss among California's freshwater fishes. Conservation biology 21: 434–446,

Macchi, P.J., V. Cussac, M. Alonso and M. Denegri. 1999. Predation relationships between salmonids and the native fish fauna in lakes and reservoirs in northern Patagonia. Ecology of Freshwater Fish 8: 227-236.

Macchi PJ, Pascual MA, Vigliano PH. 2007. Differential piscivory of the native Percichthys trucha and exotic salmonids upon the native forage fish Galaxias maculatus in Patagonian Andean lakes. Limnologica 37:76–87

Mack R., Simberloff D., Lonsdale W., Evans H., Clout M. & Bazzaz F. 2000. Invasiones biológicas: Causas, epidemiología, consecuencias globales y control. Tópicos en ecología. 5: 19pp.

Martínez-Harms M. J. & R. Gajardo. 2008. Ecosystem value in the Western Patagonia protected areas. Journal for Nature Conservation 16: 72—87.

123

McDowall, R.M. (1971a) Fishes of the family Aplochitonidae. Journal of the Royal Society of New Zealand, 1, 31–52.

McDowall, R.M. (1971b) The galaxiid fishes of South America. Zoological Journal of the Linnean Society, 50, 33–73.

McDowall, R.M. 2003. Impacts of Introduced Salmonids on Native Galaxiids in New Zealand Upland Streams: A New Look at an Old Problem. Transactions of the American Fisheries Society, 132, 229–238.

McDowall, R.M. 2006. Crying wolf, crying foul, or crying shame: alien salmonids and a biodiversity crisis in the southern cool-temperate galaxioid fishes? Reviews in Fish Biology and Fisheries, 16(3-4): 233-422.

McDowall R. M. & S. C. Charteris. 2006. The possible adaptive advantages of terrestrial egg deposition in some fluvial diadromous galaxiid fishes (Teleostei: Galaxiidae). Fish and Fisheries. 7: 153–164

McIntosh, A.R. 2000. Habitat- and size-related variations in exotic trout impacts on native galaxiid fishes in New Zealand streams. Canadian Journal of Fisheries and Aquatic Sciences, 57, 2140–2151.,

McIntosh, A.R., C.R. Townsend & T.A. Crowl. 1992. Competition for space between introduced brown trout ( Salmo trutta L.) and a native galaxiid (Galaxias vulgaris Stokell) in a New Zealand stream. Journal of Fish Biology, 41, 63–81.

McIntosh, A. R., McHugh, P., Dunn, N. R., Goodman, J., Howard, S., Jellyman, P.G., O’Brien, L.K., Nystrom, P. & Woodford, D.J. 2010. The impact of trout on galaxiid fishes in New Zealand. New Zealand Journal of Ecology, 34,195-206

Milano, D., V. Cussac, P.J. Macchi, D. Ruzzante, M. Alonso, P. Vigliano & M. Denegri. 2002. Predator associated morphology in Galaxias platei in Patagonian lakes. Journal of Fish Biology 61:138-156.

Myers N, R Mitterneier, C Mitterneier, GB Da Fonseca & J Kents. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853-858.

Naiman RJ, RE Bilby, DE Schindler & JM Hedfield. 2002. Pacific salmon, nutrients and the dynamics of freshwater and riparian ecosystems. Ecosystems 5: 399-417.

MMA. 2012. Lista de especies nativas según estado de conservación. Clasificación de especies, Ministerio del Medio Ambiente de Chile. http://www.mma.gob.cl/clasificacionespecies/

Ortiz-Sandoval J. J. 2007. Estructura trófica de la ictiofauna de la zona limnética y profunda de lagos andino patagónicos de la Región de Aysén: niveles de piscivoría asociados a salmonídeos y

124 especies nativas. Seminario para optar al título de Biólogo. Facultad de Ciencias Naturales y oceanográficas. Universidad de Concepción.

Pascual, M., Macchi, P., Urbanski, J., Marcos, F., Riva Rossi, C., Novara, M. & Dell’Arciprete, P. 2002. Evaluating potential effects of exotic freshwater fish from incomplete species presence– absence data. Biological Invasions, 4, 101–113.

Pascual M. A., Cussac V., Dyer B., Soto D., Vigliano P., Ortubay S. & Macchi P. 2007. Freshwater fishes of Patagonia in the 21st Century after a hundred years of human settlement, species introductions, and environmental change. Aquatic Ecosystem Health & Management. 10:2, 212 – 227.

Pascual, M. A., Lancelotti, J. L., Ernst, B., Ciancio, J. E., Aedo, E. & Garcia-Asorey, M. (2009) The importance of scale, connectivity, and incentives in conservation programs, the case of exotic Salmonids of Patagonia. Frontiers in Ecology and Environment. doi,10.1890/070127.

Pauchard, A. & Villarroel, P. 2002. Protected areas in Chile, history, current status and challenges. Natural Areas Journal, 22, 318-330.

Rodriguez-Cabal, M. A., Nuñez, M. A. & Martinez, A. S. 2008 Quantity versus quality, Endemism and protected areas in the temperate forest of South America. Austral Ecology, 33, 730–736.

Ruzzante, D., S.J. Walde, V.E. Cussac, M.L. Dalebout, J. Seibert, S. Ortubay & E. Habit. 2006. Phylogeography of the Percichthyidae (Pisces) in Patagonia: roles of orogeny, glaciation, and volcanism. Molecular Ecology. 15: 2949-2968.

Ruzzante D. E., S. J. Walde, J. C. Gosse, V. E. Cussac, E. Habit, T. S. Zemlak & E. D. M. Adams. 2008. Climate control on ancestral population dynamics: insight from patagonian fish phylogeography. Molecular Ecology. 17: 2234 - 2244.

Sanderson, B. L., K. A. Barnas, and A. M. Wargo Rub. 2009. Nonindigenous species of the Pacific Northwest: an overlooked risk to endangered salmon? BioScience 59:245-256.

Saunders, D. L., Meeuwig, J. J. & Vincent, A. C. J. 2002. Freshwater protected areas, strategies for conservation. Conservation Biology, 16(1), 30-41.

Simon, K.S., & C.R. Townsend. 2003. The impacts of freshwater invaders at different levels of ecological organisation, with emphasis on salmonids and ecosystem consequences. Freshwater Biology 48:982-994.

Simonetti, J. & Armesto, J. J. 1991. Conservation of Temperate Ecosystems in Chile, Coarse vs. Fine Filter Approaches. Revista Chilena de Historia Natural, 64, 615-626.

125

Soto D., F. Jara & C. Moreno. 2001. Escaped Salmon in the Inner Seas, Southern Chile: Facing Ecological and Social Conflicts. Ecological Applications. 11(6):1750-1762.

Soto, D., I. Arismendi, J. Gonzalez, J. Sanzana, F. Jara, C. Jara, E. Guzman & A. Lara. 2006. Sur de Chile, país de truchas y salmones: patrones de invasión y amenazas para las especies nativas. Revista Chilena de Historia Natural. 79(1): 97-117.

Soto, D. Arismendi, I. Prinzio, C. Di et al. 2007. Establishment of Chinook salmon (Oncorhynchus tshawytscha) in Pacific basins of southern South America and its potential ecosystem implications. Revista Chilena de Historia Natural 80(1): 81-98.

Stuart-Smith RD, Stuart-Smith JF, White RWG, Barmuta LA. 2007. The impact of an introduced predator on a threatened galaxiid fish is reduced by the availability of complex habitats. Freshwater Biology 52: 1555-1563

Tognelli M. F., P. I. Ramirez de Arellano & P. A. Marquet. 2008. How well do the existing and proposed reserve networks represent vertebrate species in Chile?. Diversity and Distributions. 14:148–158.

Townsend, C. R. 1996. Invasion biology and ecological impacts of brown trout Salmo trutta in New Zealand. Biol. Conserv. 78: 13-22.

Townsend, C. 2003. Individual, population, community, and ecosystem consequences of a fish invader in New Zealand streams. Conservation Biology 17: 38 – 47.

Townsend C.R. & Crowl T.A. 1991. Fragmented population structure in a native New Zealand fish: an effect of introduced brown trout? Oikos, 61, 348–354.

Valdovinos C. 2008. Diversidad de Especies: Invertebrados dulceacuícolas. En: Conama (ed) Biodiversidad de Chile: Patrimonio y desafíos. Capítulo II. Nuestra Diversidad Biológica: 202- 223. Tercera edición, Ocho Libros Editores, Santiago, Chile. Vigliano P. H., D. A. Beauchamp, D. Milano, P. J. Macchi, M. F. Alonso, M. I. García Asorey, M. A. Denegri, J. E. Ciancio, G. Lippolt, M. Rechencq & J. P. Barriga. 2009. Quantifying Predation on Galaxiids and Other Native Organisms by Introduced Rainbow Trout in an Ultraoligotrophic Lake in Northern Patagonia, Argentina: A Bioenergetics Modeling Approach, Transactions of the American Fisheries Society, 138(6):1405-1419

Vila I, L Fuentes & M Contreras. 1999. Peces Límnicos de Chile. Boletín Museo Historia Natural, Chile, 48: 61 - 75.

Zattara E & AC Premoli. 2005. Genetic structuring in Andean landlocked populations of Galaxias maculatus: effects of biogeographic history. Journal of Biogeography 32, 5–14.

126

Zama A. 1986. Some biological observation on Galaxiid fishes, Galaxias platei, collected in Southern Chile. Japanese Journal of Ichthyology. 32(4):450-453.

Zemlak T.S., E.M. Habit, S.J. Walde, M. A. Battini, E. D.M. Adams & D. E. Ruzzante. 2008. Across the southern Andes on fins: Glacial refugia, drainage reversals and a secondary contact zone revealed by the phylogeographic signal of Galaxias platei in Patagonia. Molecular Ecology.

127