A working guide to harnessing generalized dissimilarity modelling for biodiversity analysis and conservation assessment

Mokany, Karel; Ware, Chris; Woolley, Skipton N. C.; Ferrier, Simon; Fitzpatrick, Matthew C.

doi:10.1111/geb.13459

Cited by 73 publications

(82 citation statements)

References 110 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Backward elimination (Mokany et al, 2021) was used to account for potentially high correlations between pH and Na, K and Cl. Only factors that explained a proportion (>0%) of nematode compositional dissimilarity models were retained in the final models.…”

Section: Statistical Analysesmentioning

confidence: 99%

Context‐dependent role of abiotic and biotic factors structuring nematode communities along two environmental gradients

et al. 2022

View full text Add to dashboard Cite

Although abiotic environmental factors have been historically regarded as the dominant deterministic process in microbial community assembly, recent studies indicate that biotic interactions may be equally significant. However, the extent to which both processes are important in assembly of belowground communities is unknown. Along two environmental gradients: alkalinity (ranging from pH ~7 to ~11) and habitat type (lakes, shorelines, and prairies around lakes) present in the Western Nebraska Sandhills, we used 18S rRNA gene marker metabarcoding and statistical analyses, including generalized dissimilarity modelling (GDM), to evaluate the dynamics between abiotic and biotic factors that might play a role in nematode community assembly. Lakes supported the least diverse and prairies the most diverse communities with completely distinct compositions. We also observed a potential role of alkalinity in shaping these communities but only in lakes. Generally, GDMs indicated the influence of both abiotic and biotic factors. However, their relative importance in explaining community variability was dependent on the habitat. Biotic factors influenced the lake communities most, followed by shorelines and prairies, explaining ~47%, 27% and 8% of the variation, respectively. In contrast, the role of abiotic factors was relatively similar in lakes, shorelines and prairies (~15%, 18% and 14% of the variation, respectively). Most variation in the shorelines (62%) and prairies (82%) remained unexplained, suggesting the potential importance of factors associated with specific traits or a stronger role of stochastic processes. Nevertheless, our findings suggest both deterministic processes are important in nematode community assembly, but their specific contributions are context‐dependent.

show abstract

Section: Statistical Analysesmentioning

confidence: 99%

Context‐dependent role of abiotic and biotic factors structuring nematode communities along two environmental gradients

et al. 2022

View full text Add to dashboard Cite

show abstract

“…It models the spatial turnover of composition between all samples as a function of the environmental differences between the locations (Ferrier et al, 2007). We used geographical distances (Latitude and Longitude) and environmental gradients (PCA axes) as predictors, and Bray–Curtis community distance matrix as response variables (Mokany et al, 2022b; Woolley et al, 2017). To prepare the data for GDM analysis, the “formatsitepair function” in the package “gdm” was used (Ferrier et al, 2007; Mokany et al, 2022a).…”

Section: Methodsmentioning

confidence: 99%

“…We used geographical distances (Latitude and Longitude) and environmental gradients (PCA axes) as predictors, and Bray-Curtis community distance matrix as response variables (Mokany et al, 2022b;Woolley et al, 2017). To prepare the data for GDM analysis, the "formatsitepair function" in the package "gdm" was used (Ferrier et al, 2007;Mokany et al, 2022a)…”

Section: Beta Diversity (Turnover and Nestedness)mentioning

confidence: 99%

Patterns of variation in plant diversity vary over different spatial levels in seasonal coastal wetlands

Dwyer

Millett

Jones

et al. 2022

Diversity and Distributions

View full text Add to dashboard Cite

Aim: To quantify the responses of alpha and beta diversity to multivariate gradients, incorporating variation in environmental and management variability in coastal dune slacks.Location: United Kingdom dune slacks.Methods: Plant community composition, plant nutrient status and soil characteristics were measured for 164 quadrats in 41 dune slacks across 12 coastal sand dune systems. Data were collated on climate and atmospheric deposition. Hydrological regimes at daily resolution were modelled and calibrated using daily-to-monthly site measurements, from which we calculated quadrat-level hydrological metrics. Alpha diversity (richness, Shannon diversity and Pielou's evenness) metrics and beta diversity (turnover and nestedness) for species and genera were calculated across three spatial levels from sand dune system (highest) to dune slack to quadrat (lowest).Results: Diversity patterns depended on the spatial and taxonomic level considered.At smaller spatial levels (between dune slacks and between quadrats), alpha and beta diversity varied along gradients driven by soil characteristics, water table depth and atmospheric deposition. At larger spatial levels (between sand dune systems), patterns of beta diversity were a consequence of plant nutrient status. There was little variability in alpha diversity along this same gradient, with only small changes in Pielou's species evenness. Patterns at a coarser taxonomic level (genus) mirrored those at the species level. Main conclusion:We show that patterns of variation in plant diversity are dependent on the spatial level considered, but taxonomic level made little difference in understanding these patterns. Therefore, if we do not consider patterns across different spatial levels, important environmental and management drivers could be missed. The high biodiversity value and degree of threat to these European protected habitats makes such understanding invaluable for their conservation.

show abstract

“…The performance of the compositional GDM used to make predictions was assessed using cross-validation, we observed how well the model predicted the dissimilarity within a randomly withheld 20% of sites. This process was repeated a thousand times and the model performance metrics, including mean absolute error, where averaged (Mokany et al, 2022).…”

Section: Generalized Dissimilarity Modellingmentioning

confidence: 99%

Chilean bee diversity: Contrasting patterns of species and phylogenetic turnover along a large-scale ecological gradient

Marshall

Ascher

Villagra

et al. 2022

Preprint

View full text Add to dashboard Cite

Chile's isolation and varied climates have driven the evolution of a unique biodiversity with a high degree of endemism. The Mediterranean-type biome of Central Chile is one of 35 global biodiversity hotspots and has been highlighted as one of Chile's most endangered areas. It is threatened by anthropogenic land use change impacting the integrity of local biomes and associated species. This area holds the most extensive collections of the country with high endemicity regarding native bee species. Characterising habitat requirements of bees is a pressing priority to safeguard them and the ecosystem services they provide.We investigated broad-scale patterns of bee diversity using newly accessible expert-validated datasets comprising digitized specimen records from both Chilean and US collections and novel, expertvalidated type specimen data for the bees of Chile. We used a generalised dissimilarity modelling (GDM) approach to explore both compositional and phylogenetic β-diversity patterns across latitudinal, altitudinal, climate and habitat gradients in well-sampled bee assemblages in Central Chile. Using the GDM measures of increasing compositional and environmental dissimilarity we categorised and compared the most important drivers of these patterns and used them to classify 'wild bee ecoregions' (WBE) representing unique assemblages.Turnover of bee assemblages was explained primarily by latitudinal variation (proxy for climate) along Chile. However, temperature variations, precipitation and the presence of bare soil also significantly explained the observed patterns. In comparison, we observed less turnover in phylogenetic biodiversity corresponding to spatial gradients. We were able to develop six de novo ecoregions (WBE) all with distinct taxa, endemic lineages, and representative species. The WBE represent distinct spatial classifications but have similarities to existing biogeographical classifications, ecosystems and bioclimatic zones.

show abstract

A working guide to harnessing generalized dissimilarity modelling for biodiversity analysis and conservation assessment

Cited by 73 publications

References 110 publications

Context‐dependent role of abiotic and biotic factors structuring nematode communities along two environmental gradients

Context‐dependent role of abiotic and biotic factors structuring nematode communities along two environmental gradients

Patterns of variation in plant diversity vary over different spatial levels in seasonal coastal wetlands

Chilean bee diversity: Contrasting patterns of species and phylogenetic turnover along a large-scale ecological gradient

Contact Info

Product

Resources

About