Abstract:To support coastal planning through improved understanding of patterns of biotic and abiotic surrogacy at broad scales, we used gradient forest modeling (GFM) to analyze and predict spatial patterns of compositional turnover of demersal fishes, macroinvertebrates, and macroalgae on shallow, temperate Australian reefs. Predictive models were first developed using environmental surrogates with estimates of prediction uncertainty, and then the efficacy of the three assemblages as biosurrogates for each other was assessed. Data from underwater visual surveys of subtidal rocky reefs were collected from the southeastern coastline of continental Australia (including South Australia and Victoria) and the northern coastline of Tasmania. These data were combined with 0.01 degree-resolution gridded environmental variables to develop statistical models of compositional turnover (beta diversity) using GFM. GFM extends the machine learning, ensemble tree-based method of random forests (RF), to allow the simultaneous modeling of multiple taxa. The models were used to generate predictions of compositional turnover for each of the three assemblages within unsurveyed areas across the 6600 km of coastline in the region of interest. The most important predictor for all three assemblages was variability in sea surface temperature (measured as standard deviation from measures taken interannually). Spatial predictions of compositional turnover within unsurveyed areas across the region of interest were remarkably congruent across the three taxa. However, the greatest uncertainty in these predictions varied in location among the different assemblages. Pairwise congruency comparisons of observed and predicted turnover among the three assemblages showed that invertebrate and macroalgal biodiversity were most similar, followed by fishes and macroalgae, and lastly fishes and invertebrate biodiversity, suggesting that of the three assemblages, macroalgae would make the best biosurrogate for both invertebrate and fish compositional turnover.

A Bayesian latent allocation model for clustering compositional data with application to the Great Barrier Reef

Bayesian clustering of spatially distributed compositional data with application to the Great Barrier Reef

A Bayesian Contiguous Partitioning Method for Learning Clustered Latent Variables.

Bayesian clustering of high-dimensional data via latent repulsive mixtures

Congruence in demersal fish, macroinvertebrate, and macroalgal community turnover on shallow temperate reefs

Bayesian spatial clustering of extremal behaviour for hydrological variables

Exploring coral reef communities in Puerto Rico using Bayesian networks

Mathematical modelling and uncertainty quantification for analysis of biphasic coral reef recovery patterns

Monitoring through many eyes: Integrating disparate datasets to improve monitoring of the Great Barrier Reef

Among-site variability in the stochastic dynamics of East African coral reefs

Spatial non-parametric Bayesian clustered coefficients

A Bayesian Approach to Restricted Latent Class Models for Scientifically-Structured Clustering of Multivariate Binary Outcomes

Bayesreef: A Bayesian inference framework for modelling reef growth in response to environmental change and biological dynamics

Extending the Latent Dirichlet Allocation model to presence/absence data: A case study on North American breeding birds and biogeographical shifts expected from climate change

Bayesian model selection for the latent position cluster model for Social Networks

Bayesian compositional regression with microbiome features via variational inference

Data-driven determination of zooplankton bioregions and robustness analysis

Bayesian Model Averaging Over Tree-based Dependence Structures for Multivariate Extremes

Optimal Bayesian estimators for latent variable cluster models

A Bayesian hierarchical model for disease mapping that accounts for scaling and heavy-tailed latent effects

A Bayesian Hierarchical Time Series Model for Reconstructing Hydroclimate from Multiple Proxies