Changes in the analysis of temporal community dynamics data: a 29-year literature review

Abstract

Background Understanding how biological communities change over time is of increasing importance as Earth moves into the Anthropocene. A wide variety of methods are used for multivariate community analysis and are variously applied to research that aims to characterise temporal dynamics in community composition. Understanding these methods and how they are applied is useful for determining best practice in community ecology. Methodology We reviewed the ecological literature from 1990 to 2018 that used multivariate methods to address questions of temporal community dynamics. For each paper that fulfilled our search criteria, we recorded the types of multivariate analysis used to characterise temporal community dynamics in addition to the research aim, habitat type, location, taxon and the experimental design. Results Most studies had relatively few temporal replicates; the median number was seven time points. Nearly 70% of studies applied more than one analysis method; descriptive methods such as bar graphs and ordination were the most commonly applied methods. Surprisingly, the types of analyses used were only related to the number of temporal replicates, but not to research aim or any other aspects of experimental design such as taxon, or habitat or year of study. Conclusions This review reveals that most studies interested in understanding community dynamics use relatively short time series meaning that several, more sophisticated, temporal analyses are not widely applicable. However, newer methods using multivariate dissimilarities are growing in popularity and many can be applied to time series of any length.

The dual nature of metacommunity variability

“…Approaches to studying temporal taxonomic beta diversity have proliferated in recent decades. While descriptive and ordination-based approaches dominate the literature, new metrics are uncovering novel insights into community dynamics as multispecies time series increase in length and availability (Buckley et al 2021). For example, studies have quantified the turnover in community composition between time points or relative to a baseline using dissimilarity metrics (Dornelas et al 2014), shifts in species ranks in relative abundance (Avolio et al 2019), and by partitioning compositional change into its turnover and nestedness components (Baselga 2010, Podani et al 2013, Magurran et al 2019.…”

Local compositional change and regional stability across 3000 years of coral reef development

“…The variety of statistical tools to calculate and model trends in temporal turnover has expanded substantially (Buckley, Day, Case, et al., 2021; Legendre & Gauthier, 2014). Most studies, however, still estimate turnover using raw dissimilarities, as we did here, potentially because ecological time series are rarely long enough for more sophisticated methods (Buckley, Day, Lear, et al., 2021). Our results highlight that, within the relative simplicity of measuring turnover via dissimilarities, the meaning and resulting β diversity trends depend on the reference frame.…”

Trends in ultra-processed food availability and its association with diet-related non-communicable disease health indicators in the Portuguese population

“…As an ecological study, results have to be cautiously interpreted due to the possibility of ecological fallacy. For this reason, associations identified at group level data should not be assumed to be true for individuals (47,51,52) . Even tough, results have the capacity to highlight for the relevance of the subject and might pose hypothesis for further analysis.…”