UNIVERSIDADE FEDERAL DO PARANÁ RAUL RENNÓ BRAGA FUSÃO INVASORA: HIERARQUIZAÇÃO DA HIPÓTESE, AVALIAÇÃO EXPERIMENTAL E “FRAMEWORK” PARA TESTE E SÍNTESE CURITIBA 2016 FUSÃO INVASORA: HIERARQUIZAÇÃO DA HIPÓTESE, AVALIAÇÃO EXPERIMENTAL E “FRAMEWORK” PARA TESTE E SÍNTESE Tese apresentada ao Programa de Pós- Graduação em Ecologia e Conservação – PPGECO da Universidade Federal do Paraná – UFPR, como requisito parcial para obtenção do título de Doutor. Aluno: Raul Rennó Braga Orientador: Prof. Dr. Jean Ricardo Simões Vitule Co-Orientador: Prof. Dr. André Andrian Padial Curitiba 2016 DEDICATÓRIA Ao meu pai Mozart Rocha Braga e a minha mãe Maria Aparecida Rennó Braga, que sempre me incentivaram a estudar, aos quais devo, em grande parte, o que hoje sou. Aos meus irmãos, Marcelo Rennó Braga e Raquel Rennó Braga, pela compreensão e ajuda. Ao meu amigo e orientador Jean Ricardo Simões Vitule que ao longo de muitos anos acreditou no meu trabalho e mostrou o caminho para chegar onde estou. Enfim, a todos os colegas e educadores que fizeram parte da minha formação pessoal e intelectual. AGRADECIMENTOS Agradeço ao professor orientador Doutor Jean Ricardo Simões Vitule pelo acompanhamento e revisão do estudo. Agradeço aos meus co-orientadores André Andrian Padial e Jonathan Jeschke e aos professores, Sidinei Magela Thomaz, Roger Paulo Mormul pelo apoio e ajuda nos momentos necessários. Meu especial agradecimento à minha família, aos membros do Laboratório de Ecologia e Conservação da Universidade Federal do Paraná, do Grupo de Pesquisa em Ictiofauna do Museu de História Natural do Capão da Imbuia, do “Ecological Novelty Group” da Freie Universität Berlin e a todas as pessoas que colaboraram direta ou indiretamente para a realização deste trabalho dentre elas: Maria Aparecida Rennó Braga, Mozart Rocha Braga, Marcelo Rennó Braga, Raquel Rennó Braga, Bruno Kazuo Nakagawa, Fabrício de Andrade Frehse, Marcos Ostrowski Valduga, Vanessa Maria Ribeiro, Vanessa Salete Daga, Hugo Bornatowski, Igor Kintopp Ribeiro, Matheus Oliveira Freitas, Vinícius Abilhoa, Florian Ruland, Stefan Linzmaier, Wolf Saul, Igor De Paiva Affonso, Luiz Guilherme Ribas, Patrícia Dall’ Agnol, Juliana Wojciechowski, Suelen Pereto, Eduardo Ribeiro da Cunha, Emanuel Silva, Fernanda Ceschin, Sebastião Rodrigues, Alfredo Soares da Silva, Valdenir Ferreira de Souza. Agradeço também à Instituição Universidade Federal do Paraná e ao Curso de Pós-graduação em Ecologia e Conservação pela excelência em estudo oferecida e a Capes pelo apoio financeiro em forma de bolsa de estudo. LISTA DE FIGURAS Prefácio Figura 1. Tendência temporal do número cumulativo de espécies invasoras de acordo com (a) modelo de resistência biótica e (b) modelo de fusão invasora (adaptado de Ricciardi 2001). Capítulo I Figure 1. Weighted data on level of empirical support for the different types of interactions. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 2. Weighted data on level of empirical support for the ecological level of evidences. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 3. Weighted data on level of empirical support for non-native species coevolution. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 4. Weighted data on level of empirical support for the different types of effects. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 5. Weighted data on level of empirical support for the different habitats. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 6. Weighted data on level of empirical support for the different taxonomic groups. Letters above bars indicate significant differences (U-tests, p < 0.05). Figure 7. Schematic illustration of the hierarchy of hypotheses for the IMH. Empirical tests are classified according to three criteria: (1) type of interaction; (2) ecological level of evidence; and (3) outcome of interaction. Color codes represent sub-hypotheses with more than 50% of support considering weighted data (green boxes), and more than 50% of questioning studies for red boxes. White boxes represent sub-hypothesis with less than five studies and so no comparisons were made. Figure A1. Number of studies according to weights. Figure A2. Unweighted data on level of empirical support for the different analysis. Figure A3. Evidence on the IMH acording to research method. Letters S represent studies with and NS without statistical analysis. Capítulo II Figure1. Experimental design and different mesocosms (boxes) replicated five times. Center box represents the simplified native biota and the others represent each treatment with added single or combined non-native species. Delta was calculated to account for alterations in measured parameters related to single, additive or synergistic effect of the chosen non-native species. Figure 2. Water turbidity for each invasion scenario in relation to the native (control) mesocosms after 20 days. Dashed line indicates the control treatment. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis(A). Data are mean, standard error (box) and 95% confidence interval (whiskers). Figure 3. Total dissolved solids for each invasion scenario in relation to the native (control) mesocosms after 20 days. Dashed line indicates the control treatment. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis (A). Data are mean, standard error (box) and 95% confidence interval (whiskers). Figure 4. Dissolved oxygen concentrationsfor each invasion scenario in relation to the native (control) mesocosms after 20 days. Dashed line indicates the control treatment. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis (A). Data are mean, standard error (box) and 95% confidence interval (whiskers). Figure 5. Native (E. najas) and non-native (H. verticillata) macrophytes abundance of size class 1 (< 15 cm) fragments after 20 days. Letters below bars represent which non- native species were present. Hydrilla verticillata (H), Limnoperna fortunei(L) and Astronotus crassipinnis (A).Only invasion scenarios with both macrophyte species present are shown. Data are mean, standard error (box) and 95% confidence interval (whiskers). Figure 6. Native prey survival after 20 days. Dashed line indicates the control treatment. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis (A). Data are mean, standard error (box) and 95% confidence interval (whiskers). Figure 7. Network of positive interactions between three non-native species inhabiting the Paraná River floodplain leading to a negative effect upon a native fish species. Lines ending in arrows indicate positive effects and lines ending in solid circles indicate negative effects. Solid lines indicate evidence provided by our present study, dashed lines indicate evidence provided by published evidence and dotted lines indicate partially supported evidence by our study and published evidence. Figure A8. Experiment location and layout. Figure A9. Curtains closed during direct sunlight hours and mesocosms protected with 2.5 cm mesh. Figure A10. Evaluation of how many times it was necessary to agitate macrophytes in order to remove water and standardize weights added to mesocosms. Figure A4. Mesocosm layout of a trial containing all three non-native species. Figure A5. Flexible square used for mussel counting. Figure A6. Delta of phytoplankton densities compared through Sorensen dissimilarity index for each invasion scenario in relation to the native (control) mesocosms. Solid line represent the average Sorensen dissimilarity index within control mesocosms. Dashed lines represent control’s minimum and maximum values of Sorensen dissimilarity index. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis(A). Figure A7. Delta of phytoplankton densities compared through Bray-curtis dissimilarity index for each invasion scenario in relation to the native (control) mesocosms. Solid line represent the average Bray-curtis dissimilarity index within control mesocosms. Dashed lines represent control’s minimum and maximum values of Bray-curtis dissimilarity index. Letters below bars represent which non-native species were present. Hydrilla verticillata (H), Limnoperna fortunei (L) and Astronotus crassipinnis(A). Figure A8. Influence of A. crassipinnis presence on native (E. najas) and non-native (H. verticillata) macrophytes abundance of size class 1 (< 15 cm). Data are mean, standard error (box) and 95% confidence interval (whiskers). Capítulo III Figure 1. Each circle represents an ecological level. P is population, C is community and E is ecosystem. Circle sizes are proportional to the ease of data collection, with the largest being the easiest and the smallest the hardest. Shading represents the likely severity of the total impact of invasional meltdown, with darker the most severe and lighter the least severe. Shading in circle junctions represents the power of hypothesis- testing for each combination of ecological level studies. Studies combining the community level and any other level are stronger than the combination of population