Various biodiversity indicators, such as species richness, total abundance and species diversity indices, have been developed to capture the state of ecological community over space and time. As biodiversity is a multifaceted concept, it is important to understand the dimension of biodiversity reflected by each indicator for successful biodiversity conservation and management. Here we present a method to characterize and classify biodiversity indicators according to their dynamical similarity and apply the methodology to the monitoring data of a marine fish community under the intermittent anthropogenic warm water discharge. Our analysis reveals that ten biodiversity indicators can be classified into three super-groups that possibly reflect the same dimension of biodiversity. Group I (species richness) with the most robust dynamics to the temperature changes; Group II (species diversity and total abundance) showing an abrupt change in the middle of the monitoring period, presumably due to a change in temperature; Group III (species evenness) with the highest sensitivity to temperature changes. The finding suggests that selecting indicators from different super-groups might help to fully assess biodiversity response to environmental change.

Species richness, evenness, and taxonomic diversity have been proposed and used as indicators of biodiversity. However, the governing equations for the dynamics of these indicators are often unknown and an understanding of their mechanism is limited. By employing a nonlinear forecasting-based method to the time series of 10 biodiversity indicators of a marine fish community, we aimed to detect changes in their dynamics and classify the indicators according to the level and timing of dynamic changes. The 10 biodiversity indicators were classified into three super-groups: Group I (species richness and community center of distribution) with the most unchanged dynamics; Group II (species diversity and total abundance) with dynamics that had an abrupt change in the middle, presumably due to an increase in local temperature; and Group III (species evenness) with highly variable dynamics. Finally, a hypothesis regarding the mechanistic basis of the dynamic similarities between the biodiversity indicators was proposed.