Analysing chemical-induced changes in macroinvertebrate communities in aquatic mesocosm experiments: a comparison of methods

Szöcs, Eduard; Brink, Paul J. Van den; Lagadic, Laurent; Caquet, Thierry; Roucaute, Marc; Auber, Arnaud; Bayona, Yannick; Liess, Matthias; Ebke, Peter; Ippolito, Alessio; Braak, Cajo J. F. ter; Brock, T.C.M.; Schäfer, Ralf B.

doi:10.1007/s10646-015-1421-0

Cited by 25 publications

(22 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The effect of these three low-dose mixtures was evaluated in a series of community-level metabolic end points (see Materials and Methods) ( 47 ). Multivariate generalized linear models (GLM MV ) ( 48 ) revealed a statistically significant interaction between the factors treatment (control, Mix 16, Mix 16-4, and Mix 16/10) and time (model 1 in Table 1; ANODEV, Wald P = 0.03) but no effects from treatment or time separately. Changes in model communities were expressed mainly on Y eff and β-Glu activity (section 4 in S1), indicating relevant effects on both the autotrophic and heterotrophic components of the microbial community.…”

Section: Resultsmentioning

confidence: 99%

“…Statistical significance of individual or nested GLM MV model fits was evaluated by ANODEV ( 70 ), and MV test statistics were constructed using a Wald statistic ( 69 ), where P values were approximated by resampling rows using residual permutation as described by Szöcs et al . ( 48 ) (see section 4 of S1 for details). Power was, in general, >0.76 for relevant effect sizes.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Hidden drivers of low-dose pharmaceutical pollutant mixtures revealed by the novel GSA-QHTS screening method

et al. 2016

View full text Add to dashboard Cite

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Hidden drivers of low-dose pharmaceutical pollutant mixtures revealed by the novel GSA-QHTS screening method

et al. 2016

View full text Add to dashboard Cite

show abstract

“…For community analyses, GLM for multivariate data (Warton et al 2012) have been proposed as alternative to Principal Response Curves (PRC) and yielded similar inferences, but better indication of responsive taxa (Szöcs et al 2015). For community analyses, GLM for multivariate data (Warton et al 2012) have been proposed as alternative to Principal Response Curves (PRC) and yielded similar inferences, but better indication of responsive taxa (Szöcs et al 2015).…”

Section: Simulationsmentioning

confidence: 99%

“…For example, extremely low samples sizes (n < 5) are common in many mesocosm studies (Sanderson 2002;Szöcs et al 2015). For example, extremely low samples sizes (n < 5) are common in many mesocosm studies (Sanderson 2002;Szöcs et al 2015).…”

Section: Introductionmentioning

confidence: 99%

Ecotoxicology is not normal

Szöcs

Schäfer

2015

Environ Sci Pollut Res

Self Cite

View full text Add to dashboard Cite

Ecotoxicologists often encounter count and proportion data that are rarely normally distributed. To meet the assumptions of the linear model, such data are usually transformed or non-parametric methods are used if the transformed data still violate the assumptions. Generalized linear models (GLMs) allow to directly model such data, without the need for transformation. Here, we compare the performance of two parametric methods, i.e., (1) the linear model (assuming normality of transformed data), (2) GLMs (assuming a Poisson, negative binomial, or binomially distributed response), and (3) non-parametric methods. We simulated typical data mimicking low replicated ecotoxicological experiments of two common data types (counts and proportions from counts). We compared the performance of the different methods in terms of statistical power and Type I error for detecting a general treatment effect and determining the lowest observed effect concentration (LOEC). In addition, we outlined differences on a real-world mesocosm data set. For count data, we found that the quasi-Poisson model yielded the highest power. The negative binomial GLM resulted in increased Type I errors, which could be fixed using the parametric bootstrap. For proportions, binomial GLMs performed better than the linear model, except to determine LOEC at extremely low sample sizes. The compared non-parametric methods had generally lower power. We recommend that counts in one-factorial experiments should be analyzed using quasi-Poisson models and proportions from counts by binomial GLMs. These methods should become standard in ecotoxicology.

show abstract

“…Warton et al () showed that MvGLMs, in contrast to distance‐based methods, can differentiate between location (difference in mean) and dispersion (difference in mean–variance relationship) effects. Szöcs et al () found that the statistical power of MvGLMs was higher or at least equal to that of principal response curves (a form of redundancy analysis) when used for the analysis of ecotoxicological semifield studies. However, systematic studies of data sets with known properties are lacking and this paucity of studies hampers our capacity to make informed decisions on the selection of methods for multivariate data analysis.…”

Section: Introductionmentioning

confidence: 99%

Should ecologists prefer model‐ over distance‐based multivariate methods?

Jupke

Schäfer

2020

Ecology and Evolution

Self Cite

View full text Add to dashboard Cite

Ecological data sets often record the abundance of species, together with a set of explanatory variables. Multivariate statistical methods are optimal to analyze such data and are thus frequently used in ecology for exploration, visualization, and inference. Most approaches are based on pairwise distance matrices instead of the sites‐by‐species matrix, which stands in stark contrast to univariate statistics, where data models, assuming specific distributions, are the norm. However, through advances in statistical theory and computational power, models for multivariate data have gained traction. Systematic simulation‐based performance evaluations of these methods are important as guides for practitioners but still lacking. Here, we compare two model‐based methods, multivariate generalized linear models (MvGLMs) and constrained quadratic ordination (CQO), with two distance‐based methods, distance‐based redundancy analysis (dbRDA) and canonical correspondence analysis (CCA). We studied the performance of the methods to discriminate between causal variables and noise variables for 190 simulated data sets covering different sample sizes and data distributions. MvGLM and dbRDA differentiated accurately between causal and noise variables. The former had the lowest false‐positive rate (0.008), while the latter had the lowest false‐negative rate (0.027). CQO and CCA had the highest false‐negative rate (0.291) and false‐positive rate (0.256), respectively, where these error rates were typically high for data sets with linear responses. Our study shows that both model‐ and distance‐based methods have their place in the ecologist's statistical toolbox. MvGLM and dbRDA are reliable for analyzing species–environment relations, whereas both CQO and CCA exhibited considerable flaws, especially with linear environmental gradients.

show abstract

Analysing chemical-induced changes in macroinvertebrate communities in aquatic mesocosm experiments: a comparison of methods

Cited by 25 publications

References 41 publications

Hidden drivers of low-dose pharmaceutical pollutant mixtures revealed by the novel GSA-QHTS screening method

Hidden drivers of low-dose pharmaceutical pollutant mixtures revealed by the novel GSA-QHTS screening method

Ecotoxicology is not normal

Should ecologists prefer model‐ over distance‐based multivariate methods?

Contact Info

Product

Resources

About