Agriculture, Ecosystems and Environment 98 (2003) 87–98 Biodiversity indicators: the choice of values and measures Peter Duelli * , Martin K. Obrist Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf-Zürich, Switzerland Abstract Ideally, an indicator for biodiversity is a linear correlate to the entity or aspect of biodiversity under evaluation. Different motivations for assessing entities or aspects of biodiversity lead to different value systems; their indicators may not correlate at all. For biodiversity evaluation in agricultural landscapes, three indices are proposed, each consisting of a basket of concordant indicators. They represent the three value systems “conservation” (protection and enhancement of rare and threatened species), “ecology” (ecological resilience, ecosystem functioning, based on species diversity), and “biological control” (diversity of antagonists of potential pest organisms). The quality and reliability of commonly used indicators could and should be tested with a three-step approach. First, the motivations and value systems and their corresponding biodiversity aspects or entities have to be defined. In a time consuming second step, a number of habitats have to be sampled as thoroughly as possible with regard to one or several of the three value systems or motivations. The third step is to test the linear correlations of a choice of easily measurable indicators with the entities quantified in the second step. Some examples of good and bad correlations are discussed. © 2003 Elsevier Science B.V. All rights reserved. Keywords: Biodiversity; Indicator; Arthropods; Correlate 1. Who needs biodiversity indicators? National and regional agencies for nature conserva- tion, agriculture, and forestry have to monitor species diversity or other aspects of biodiversity, both before and after they spend tax money on subsidies or eco- logical compensation management, with the aim of enhancing biodiversity (European Community, 1997; Ovenden et al., 1998; Wascher, 2000; Kleijn et al., 2001). Similarly, international, national or regional non-governmental organisations (NGOs) may want to monitor aspects of biodiversity at different levels and scales (Reid et al., 1993; IUCN, 1994; Cohen and Burgiel, 1997). In scientific research biodiversity * Corresponding author. Tel.: +41-1-739-2376; fax: +41-1-739-2215. E-mail address: peter.duelli@wsl.ch (P. Duelli). indicators can be used as quantifiable environmen- tal factors. Since the biodiversity of even a small area is far too complex to be comprehensively mea- sured and quantified, suitable indicators have to be found. Those who are responsible for comparing and eval- uating biodiversity have a strong incentive to choose a scientifically reliable and repeatable indicator, which inevitably increases costs. The financing agencies usu- ally opt for a financially “reasonable” approach, which often results in programmes addressing only essential work. The resulting compromises make optimisation of the choice of biodiversity indicators and methods of fundamental importance. A recent international electronic conference on bio- diversity indicators (http://www.gencat.es/mediamb/ bioind, 2000) has revealed widely differing views on why and what to measure and quantify. 0167-8809/$ – see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0167-8809(03)00072-0