So Many Variables: Joint Modeling in Community Ecology.

By David I. Warton, F. Guillaume Blanchet, R. O’Hara, Otso Ovaskainen, Sara Taskinen, Steve C. Walker, Francis K. C. Hui. Trends in Ecology & Evolution, 2015

802 citations recorded by the source corpus.

Subfields: Ecological Modeling, Environmental Chemistry, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/e590e01b15a5cff22984681b93771f776ff8d62f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecological Modeling

In Environmental Chemistry

In Nature and Landscape Conservation

Citations in the corpus

Joint species distribution modelling with the r‐package Hmsc (2020, methodology): “In a comparison among a large number of single-species and JSDMs, HMSC ranked first in terms of predictive performance (Norberg et al., 2019).”
Modelling Palaeoecological Time Series Using Generalised Additive Models (2018, background): “Formal multivariate versions of GLM or GAMs are currently an important area of research within ecology (see (Warton et al., 2015) for a recent review), where they go by the name joint species distribution models (JSDMs).”
Selecting critical features for data classification based on machine learning methods (2020, background): “Climate, Ecology, and Environ‑ mental The analysis of noisy ecological data [25], variables in ecology modelling [26], number of counts termites [27], community ecology and integrating species, traits, environmental, space [28], parameter in rainfall forecasting [29, 30], global climate zone [31],…”
Modelling of species distributions, range dynamics and communities under imperfect detection: advances, challenges and opportunities (2017, background): “Newly proposed “latent variable models” provide a flexible way to incorporate correlation in analyses of many species (Warton et al. 2015, Ovaskainen et al. 2016a, Ovaskainen et al. 2016b).”
Using joint species distribution models for evaluating how species‐to‐species associations depend on the environmental context (2017, methodology): “However, as pointed out by Warton et al. (2015), an unre- solved issue of current JSDMs is that they assume that the species associations are constant over environmental gradi- ents, as well as more generally over space and time.”
Biotic interactions in species distribution modelling: 10 questions to guide interpretation and avoid false conclusions (2018, background): “To overcome this problem, one can make the entries of P a function of a few random auxiliary (latent) covariates (Warton et al., 2015).”
Parsimoniously Fitting Large Multivariate Random Effects in glmmTMB (2024, neutral): “…effect is required to account for correlation in species responses within a sample – we expect correlation across species due to inter-specific interactions, or response to unobserved environmental conditions (Warton et al. 2015) and we wish to be able to estimate positive or negative correlations.”
Joint dynamic species distribution models: a tool for community ordination and spatio‐temporal monitoring (2016, background): “Latent variable models are seeing increased use in community ecology as a sparse method for analysing correlations among multiple species (Ovaskainen et al ., 2015; Thor-son et al ., 2015a; Warton et al ., 2015).”
Defining and evaluating predictions of joint species distribution models (2020, background): “This latent variable is not to be confused with the latent variable/latent factor concept of latent factor models (Warton et al., 2015; Wilkinson et al., 2019).”
How to make more out of community data? A conceptual framework and its implementation as models and software. (2017, methodology): “If we exclude the environmental covariates X from the analysis, then the latent variables behind an association matrix can be viewed as a model-based ordination (Warton et al. 2015b).”
MAcroecological Framework for Invasive Aliens (MAFIA): disentangling large-scale context dependence in biological invasions (2020, methodology): “With the development of Community Assembly by Trait Selection (CATS; Shipley et al 2006; Warton et al. 2015b) the classical fourth-corner problem (Legendre et al. 1997) and its implementation in joint models for abundance (Warton et al. 2015a), the analysis of trait × environment interactions, have…”
Habitat modelling of tracking data from multiple marine predators identifies important areas in the Southern Indian Ocean (2018, background): “Further effort should also be directed towards exploring alternative approaches for modelling the distribution and overlap of this set of marine top predators, including multivariate models (e.g., Elith & Leathwick, 2007; Warton et al., 2015).”
Co-occurrence is not evidence of ecological interactions. (2020, methodology): “Although this argument may suggest that using models that account for environmental filtering is appropriate (e.g. JSDMs, Ovaskainen et al., 2010; Warton et al., 2015; D’Amen et al., 2018), it should not be interpreted this way.”
A comprehensive evaluation of predictive performance of 33 species distribution models at species and community levels (2019, background): “By comparison, the more recently developed joint species distribution models (JSDMs) represent the response of entire species assemblages to environmental variation, assuming, for example, that species with similar traits have similar responses (Warton et al. 2015, Ovaskainen et al. 2017).”
Hierarchical Species Distribution Models (2016, background): “Although this review is limited to hierarchical models for a single species, the hierarchical species distribution modeling approach can be readily extended to include multiple, and possibly interacting, species [56, 79, 97].”