Cascading effects in mutualistic networks

Author(s)

Marilia Palumbo Gaiarsa

• Document type: Doctoral Thesis
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Instituto de Biociências (IBIOC/SB)
• Defense date: 2017-10-19

Examining board members

Paulo Roberto Guimaraes Junior; Marina Correa Cortes; Esther Sebastián Gonzalez; Carolina Reigada Montoya; Leonor Patrícia Cerdeira Morellato

• Advisor: Paulo Roberto Guimaraes Junior

Abstract

Species in ecological communities are linked through interactions. Perturbations, such as fluctuations in abundance, can flow from a species to another through ecological interactions. As a consequence, perturbations can ripple across species assemblages resulting in cascading effects that can potentially affect all species in the community. Ecological assemblages differ both in terms of species composition and in the way in which interactions are organized. As a result, different ecological communities form interaction networks that differ both in their structures as well in the interaction strengths connecting pairs of species within networks. Given that species and interactions are being lost at alarming rates, it is imperative to comprehend how robust communities are to extinction drivers. Moreover, if we are to prevent communities\' collapse and restore lost interactions, we have to understand how communities are assembled, as well as if and how robustness change through time. Despite continued effort by ecologists, it remains unclear how community structure is related to cascading effects and whether interaction strength affects network robustness by enhancing or dampening cascading effects due to multiple extinction drivers. In this thesis, I combine empirical data on weighted mutualistic networks, numerical simulations, and theoretical networks to explore how robust different network structures are to different extinction drivers, and how robustness change as networks assemble. First, I investigate how the structure of mutualistic networks affects perturbation spreading time--a proxy of network robustness to cascading effects. I found that networks are more robust to cascading effects when I incorporate interaction strengths, since simulations in which interaction strength was included had higher perturbation spreading times. Species richness, modularity, and nestedness had a strong, positive effect in perturbation spreading time regardless of the interaction strengths. Then, using theoretical networks with a fixed number of species and number of interactions, I was able to disentangle the effects nestedness and modularity have on robustness. I explore how robustness to different extinction drivers, in addition to cascading effects, is related to nestedness and modularity. Networks with greater nestedness and modularity were more robust to cascading effects, whereas networks with intermediate nestedness levels were the most robust to species removal. Modularity had no effect on robustness to species removal. Most importantly, I show that robustness depends not only on the type of extinction driver assessed, but also on the measure being used to quantify robustness. Finally, I use an eight-year dataset of plant-pollination networks following habitat restoration to explore how the assembly of plant-pollinator communities, and their robustness, changes as community assembles. I found that species occupy highly dynamic network positions through time, causing the assembly process to be punctuated by major network reorganizations. There was no relationship between years since restoration and robustness to perturbation spreading and to species removal. Altogether, these results contribute to broaden our understanding of the mechanisms behind biodiversity maintenance. If we are to protect and restore ecological communities, it is essential to consider not only the species per se, but also all species that depend directly and indirectly of that particular species, to unravel the relationship between network structure and community robustness to extinction drivers, and to understand how network structure and robustness changes as communities assemble (AU)

• FAPESP's process: 13/13319-5 - Ecological and evolutionary cascades in mutualistic networks
• Grantee: Marília Palumbo Gaiarsa
• Support Opportunities: Scholarships in Brazil - Doctorate