2018 ESA Annual Meeting (August 5 -- 10)

COS 123-9 - Predicting the collapse of plant-frugivore networks due to habitat loss

Thursday, August 9, 2018: 4:20 PM
252, New Orleans Ernest N. Morial Convention Center
Mariana Morais Vidal, Department of Ecology, University of São Paulo, São Paulo, Brazil, Paulo R. Guimarães Jr, Department of Ecology, Universidade de São Paulo, Sao Paulo, Brazil and Jean Paul Metzger, Institute of Biosciences, University of São Paulo, São Paulo, Brazil
Background/Question/Methods

Habitat loss and fragmentation lead to nonrandom extinction patterns. If the sequence of species extinctions is nonrandom, we may be able to predict the implications of species loss to the organization and functioning of ecological communities. In this study we used empirical data to estimate a realistic sequence of bird species extinctions resulting from habitat loss in the endangered and highly diverse Brazilian Atlantic forest. We used data on bird species occurrence from 97 sampling sites. As a proxy for bird species sensitivity to habitat loss we used the minimum percentage of habitat cover in an 800 m radius around the sampling sites where each species was detected. Then, we simulated realistic and random sequential extinctions in plant-frugivore networks from a pristine area, considering increasing scenarios of habitat loss. We investigated the consequences of habitat loss on the organization of plant-frugivore interactions and we evaluated how robust pristine networks were to the realistic extinction sequence.

Results/Conclusions

Simulated bird extinctions led to secondary extinctions, decreasing the number of interactions in the networks, increasing connectance and decreasing relative nestedness. The consequences of realistic sequence of extinctions on network structure were quite similar to those of random extinctions. Accordingly, the models in which we randomly removed species reproduced the robustness of frugivory networks to realistic species deletions. In addition to these results, we identified an apparent threshold in the response of network structure to habitat loss. When the percentage of habitat was reduced to less than 30% of the landscape, the organization of the frugivory networks changed dramatically, possibly due to massive bird extinctions. We conclude that the consequences of realistic species extinctions on the organization of plant-frugivore assemblages in the studied system are similar to random extinctions, suggesting that the number of species being lost, rather than the order of species extinctions, is the key determinant of its impacts on the organization of frugivory networks.