Estimating species richness: calibrating a large avian monitoring programme

摘要: Summary 1 Species richness is the most widely used measure for diversity of a biological community. Unfortunately, number species counted usually biased measure, as not all present may be detected. Use counts proxy true requires assumption constant (over space and time) detectability. This index hardly ever tested and, if violated, comparisons over time, or other dimensions, example different habitats, will distorted. In monitoring programmes one therefore needs to know proportion that are detected how this affected by external factors. 2 We capture–recapture techniques calibrate Swiss breeding bird survey, where recorded annually in c. 270 1-km2 quadrats during two three visits interest focused on annual trends regional comparisons. Hitherto, analysis has been restricted counts, while detectability its determinants known. We interpolated jackknife estimator compute mean 268 2001–03 related space, observer, survey effort biology. 3 Mean averaged 0·89 (SD 0·06, range 0·72–1·00), with no significant difference among years significant, but small, differences. Observers differed, surprisingly relation their experience quadrat. Detectability was positively visit duration. Larger communities had lower A slight violation population closure because staggered arrival migrants did introduce any measurable bias into our results. 4 Synthesis applications. Species programme high varied little recognized sources heterogeneity. Nevertheless, increased standardization should considered While these results pleasing show using indices need always induce serious bias, conditions programmes, future programme, quite different. Both ecological studies, way risk minimization, ought rigorously estimated whenever possible avoid detection spurious effects changes