Systematic biodiversity change, not loss

A recent paper by the Biodiversity and Behaviour Group sheds new light on the biodiversity crisis. It shows there has NOT been a consistent drop in numbers of species found locally around the world. Instead, our investigation of 100 communities and over 35,000 species in terrestrial, freshwater, and marine habitats from the poles to the equator, reveals a consistent change in which species are found in any one place.

The paper – Assemblage time series reveal biodiversity change but not systematic loss, published in Science, is based on work by Maria Dornelas, Anne Magurran and Hideyasu Shimadzu (together with Faye Moyes and Caya Sievers) in collaboration with Nick Gotelli (University of Vermont, USA) and Brian McGill (University of Maine, USA).

There are two key results of our study. First, there was neither a systematic loss nor a systematic gain in the number of species recorded through time. 59 communities showed an increase in species richness through time and 41 communities showed a decrease. The measured rate of change (species/time) was small in most studies.

The second result is that 79 of the 100 communities showed substantial changes in species composition, measured relative to the baseline of the first available survey of the community.


Although global extinctions and declines of many species have been well-documented, these results suggest that simple counts of species richness in a small area may not show consistent downward trends (depending on the kind of species and the sampling method, the definition of “small area” ranges from one square yard to several square miles). However, the set of species living in these same small areas has changed substantially over relatively short time scales of years to decades.

In a nutshell, there has been a change through time in the identity of species, but not, on average, in the number of species recorded in these monitoring studies.


The management and policy implications of our study are that conservation scientists will need to shift from just talking about how many species are found in a place to talking about which species are found in a place. Since species composition changed more often than species number, these kinds of changes should be a focus for future study.

Brian also has a regular blog which this week discusses the paper and surrounding issues, it can be seen here.

The figures and statistics from our paper can be reproduced from the material provided in databases s1, s2 and Table S1. For other analyses, the underlying abundance data are available at the websites listed in Table S1 and the reference list. Please note that, in some cases, access to these data is restricted and permission must be obtained from the original authors.