How can assemblage structure indices improve monitoring of change in bird communities using ongoing survey data?

Traditional monitoring of biodiversity has focused on the abundance or distribution of individual species, or on simple summary indices of average trends or diversity or species richness. Supplementing these approaches with an index system that measures subtle changes in assemblage structure is highly desirable and appears not to have been previously attempted. Subject to validation, such indices could inform sensitively about ecosystem health (defined as proximity to a stable or desired state) and the natural capital status or potential ecosystem service delivery.

The available approaches for monitoring biological assemblages and the data to which they might be applied are reviewed, including consideration of sources of error. A particularly promising method is the calculation of ‘diversity profiles’ (Leinster & Cobbold, 2012), which present a ‘signature’ describing the range of biological variation, considering both species’ identities and their relative abundances.

The diversity profile approach was tested using a long-term bird monitoring data set for woodland birds in the UK, applying species similarities based on phylogeny, habitat selection and ecological function (i.e. diet). Tests were conducted using three nested sets of species with different levels of specialization with respect to woodland. The influence of quantifiable sources of error was evaluated using bootstrapping.

Profile shape and temporal changes in diversity varied in magnitude and pattern with the choice of similarity measure and species set, with the pattern of temporal change in diversity showing either an increase or a decline, depending on the choice of analysis. This shows that the profile approach has the potential to reveal subtle variations in assemblage structure, although careful application and interpretation of outputs are required. The bootstrapping results revealed that uncertainty in profile values was high, but that this was due chiefly to uncertainty in interspecific similarities, as opposed to sampled abundance values, but this relates to the specific application to the datasets tested, rather than to the method in general.

Different approaches to quantifying differences between species (phylogenetic, habitat preference and dietary) revealed different aspects of assemblage variability.

Combined with the calculation of reference levels of diversity for known assemblages or sets of conditions such that observed variation can be interpreted quantitatively, assemblage monitoring via diversity profiles could add valuable, quantitative inference to ongoing monitoring, and hence the assessment of ecosystem health, potentially bridging the gap between conservation and a purely economic valuation of the status of natural capital.