Figure 1.
Overview structure of SpecSel, the species' selection algorithm, outlining the process to identify the optimal indicator set for each set size.
SpecSel has been implemented in Java and the program, including detailed coding for the search tree component, can be freely downloaded from https://www.uea.ac.uk/computing/specsel.
Figure 2.
Relationship between the number of species in the indicator and the average sensitivity score of constituent species in the most sensitive combination for that set size for the pan-European and alternative indicators drawn from all possible species.
Average sensitivity scores calculated as average of niche breadth*reliance across constituent species, with higher scores associated with less sensitive indicators. See Figure S1 for the equivalent figure for pan-European and alternative indicators drawn only from species currently covered by PECBMS.
Table 1.
Species included in the MINIMAL sets for the main pan-European indicator (Main), the equivalent indicator drawn only from species currently covered by PECBMS (Main-PECBMS) and the forest-type specific and regional indicators.
Table 2.
Species included in the BREAKPOINT sets for the main pan-European indicator (Main), the equivalent indicator drawn only from species currently covered by PECBMS (Main-PECBMS) and the forest-type specific and regional indicators.
Figure 3.
Temporal dynamics of pan-European woodland bird indicator, drawn from species currently covered by PECBMS, between 1980 and 2011.
Lines show index values, based on the geometric mean of constituent species' population trends, for MINIMAL, BREAKPOINT, SENSITIVE, the existing pan-European woodland bird index (CURRENT) and COMMUNITY sets. Equivalent figures for the regional and woodland type indicators are provided in Figures S2, S3, S4, S5, S6, S7.
Table 3.
Summary of comparisons between the temporal dynamics of alternative index sets (MINIMAL, BREAKPOINT, SENSITIVE and, for the pan-European and regional indicators, existing indicator sets CURRENT) for each indicator type and that of an index based on the population dynamics of all species in the candidate pool from which the sets had been drawn (COMMUNITY).