Exploring Species Persistence And Community Robustness Based On Metapopulation Theory

Due to habitat destruction,species invasion,climate change and environment pollution,habitats for animals and plants have been severely destroyed,which causes habitat loss and patch fragmentation,thereby leading to species extinction and biodiversity loss.As in nature,there exist complex trophic interactions between plants and animals,specific species extinctions might affect the survival of other trophically-linked species.In particular,because of trophic cascading effect,those species at higher trophic levels go extinct more easily.Therefore,how to assess metapopulation persistence and community robustness has become one of the central issues in current ecology.In this study,we mainly focus on investigating both population and community maintenance,specifically:(i)effects of habitat patch loss on metapopulation dynamics and persistence;(ii)effects of selective species extinctions on the robustness of mutualistic and antagonistic networks.Firstly,combining metapopulation and network theories,we study the responses of species survival to different modes of patch removal in patch networks with contrasting dispersal heterogeneities,including regular networks,random networks,exponential networks and scale-free networks.Among them,different patch removal modes depend on patch linking degree,including removing most connected to least connected patches,random removal,and removing least connected to most connected patches.Without patch loss,the interaction between patch network heterogeneity and species life-history traits determines species viability.In other words,species with low relative extinction rates can survive more easily in regular networks,while those with high relative extinction rates can perform best in scale-free networks.Under random patch removal,species with higher dispersal network heterogeneity can tolerate much more patch loss,theoretically demonstrating that previous studies based on lattice-or randomly-structured models,might have seriously underestimated species extinction thresholds.However,if we remove those highly linked patches firstly,then these species are quite vulnerable,and suffer very high extinction risks.To explore the underlying mechanism of these results via network analysis,we find that there exists a strong positive correlation between overall metapopulation size and the largest cluster size.Next,we make a systematic comparative analysis of the relationship between network architecture(including connectance,nestedness and modularity)and community robustness to species loss in mutualistic versus antagonistic networks.We collect and compile a large data set of 186 mutualistic and 82 antagonistic bipartite networks,covering a wide range of network size and structures.Then we apply the topological approach(i.e.repeated species sequential deletions)to explore the coevolution in the architecture-robustness relationship.Generally,both mutualistic and antagonistic networks show similar structural responses to species deletion,yet how network architecture changes with species loss is greatly determined by the deletion order of species identified with linking degree.Compared to antagonistic networks,mutualistic networks are more vulnerable to species loss,especially when more connected species are removed firstly.However,the architecture in both types of network plays a similar role in mediating community stability,unlike previous views that network structure favoring stability fundamentally differs between mutualism and antagonism.More importantly,different sequential deletions result in diverse(positive and negative)architecture-robustness relationships.Overall,both patch loss and species extinctions seriously threaten the maintenance of ecosystem functioning.This study theoretical demonstrates that those key patches(patches with high linking degree)and key species(those species with high degree)play an important role in metapopulation persistence and community maintenance,therefore providing the theoretical basis and decision support for biodiversity conservation and management.