Dataset Information

A Bayesian network approach to trophic metacommunities shows that habitat loss accelerates top species extinctions.

ABSTRACT: We develop a novel approach to analyse trophic metacommunities, which allows us to explore how progressive habitat loss affects food webs. Our method combines classic metapopulation models on fragmented landscapes with a Bayesian network representation of trophic interactions for calculating local extinction rates. This means that we can repurpose known results from classic metapopulation theory for trophic metacommunities, such as ranking the habitat patches of the landscape with respect to their importance to the persistence of the metacommunity as a whole. We use this to study the effects of habitat loss, both on model communities and the plant-mammal Serengeti food web dataset as a case study. Combining straightforward parameterisability with computational efficiency, our method permits the analysis of species-rich food webs over large landscapes, with hundreds or even thousands of species and habitat patches, while still retaining much of the flexibility of explicit dynamical models.

SUBMITTER: Haussler J

PROVIDER: S-EPMC7702078 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

A Bayesian network approach to trophic metacommunities shows that habitat loss accelerates top species extinctions.

Häussler Johanna J Barabás György G Eklöf Anna A

Ecology letters 20200927 12

We develop a novel approach to analyse trophic metacommunities, which allows us to explore how progressive habitat loss affects food webs. Our method combines classic metapopulation models on fragmented landscapes with a Bayesian network representation of trophic interactions for calculating local extinction rates. This means that we can repurpose known results from classic metapopulation theory for trophic metacommunities, such as ranking the habitat patches of the landscape with respect to the ...[more]

PMID: 32981202

Similar Datasets

Project description:BACKGROUND:Although a number of recent studies have uncovered remarkable diversity of microbes associated with plants, understanding and managing dynamics of plant microbiomes remain major scientific challenges. In this respect, network analytical methods have provided a basis for exploring "hub" microbial species, which potentially organize community-scale processes of plant-microbe interactions. METHODS:By compiling Illumina sequencing data of root-associated fungi in eight forest ecosystems across the Japanese Archipelago, we explored hubs within "metacommunity-scale" networks of plant-fungus associations. In total, the metadata included 8080 fungal operational taxonomic units (OTUs) detected from 227 local populations of 150 plant species/taxa. RESULTS:Few fungal OTUs were common across all the eight forests. However, in each of the metacommunity-scale networks representing northern four localities or southern four localities, diverse mycorrhizal, endophytic, and pathogenic fungi were classified as "metacommunity hubs," which were detected from diverse host plant taxa throughout a climatic region. Specifically, Mortierella (Mortierellales), Cladophialophora (Chaetothyriales), Ilyonectria (Hypocreales), Pezicula (Helotiales), and Cadophora (incertae sedis) had broad geographic and host ranges across the northern (cool-temperate) region, while Saitozyma/Cryptococcus (Tremellales/Trichosporonales) and Mortierella as well as some arbuscular mycorrhizal fungi were placed at the central positions of the metacommunity-scale network representing warm-temperate and subtropical forests in southern Japan. CONCLUSIONS:The network theoretical framework presented in this study will help us explore prospective fungi and bacteria, which have high potentials for agricultural application to diverse plant species within each climatic region. As some of those fungal taxa with broad geographic and host ranges have been known to promote the survival and growth of host plants, further studies elucidating their functional roles are awaited.

Project description:Empirical patterns that emerge from an examination of food webs over gradients of environmental variation can help to predict the implications of anthropogenic disturbance on ecosystems. This "dynamic food web approach" is rarely applied at the coastal margin where aquatic and terrestrial systems are coupled and human development activities are often concentrated. We propose a simple model of ghost crab (Ocypode quadrata) feeding that predicts changing dominant prey (Emerita talpoida, Talorchestia sp., Donax variablis) along a gradient of beach morphology and test this model using a suite of 16 beaches along the Florida, USA coast. Assessment of beaches included quantification of morphological features (width, sediments, slope), macrophyte wrack, macro-invertebrate prey and active ghost crab burrows. Stable isotope analysis of carbon ((13)C/(12)C) and nitrogen ((15)N/(14)N) and the SIAR mixing model were used to determine dietary composition of ghost crabs at each beach. The variation in habitat conditions displayed with increasing beach width was accompanied by quantifiable shifts in ghost crab diet and trophic position. Patterns of ghost crab diet were consistent with differences recorded across the beach width gradient with respect to the availability of preferred micro-habitats of principal macro-invertebrate prey. Values obtained for trophic position also suggests that the generalist ghost crab assembles and augments its diet in fundamentally different ways as habitat morphology varies across a highly dynamic ecosystem. Our results offer support for a functional response in the trophic architecture of a common food web compartment (ghost crabs, macro-invertebrate prey) across well-known beach morphologies. More importantly, our "dynamic food web approach" serves as a basis for evaluating how globally wide-spread sandy beach ecosystems should respond to a variety of anthropogenic impacts including beach grooming, beach re-nourishment, introduction of non-native or feral predators and human traffic on beaches.

Dataset Information

A Bayesian network approach to trophic metacommunities shows that habitat loss accelerates top species extinctions.

Publications

A Bayesian network approach to trophic metacommunities shows that habitat loss accelerates top species extinctions.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets