BIJLSMA RG (1982) Census problems with Bullfinches Pyrrhula pyrrhula as breeding bird. LIMOSA 55 (1): 9-16.

During 1975 and 1976 the SW. Veluwe (c. 200 km2, of which 110 km2 suitable for Bullfinches) was surveyed in order to obtain information on distribution, abundance and reproduction of the breeding population. The main objective of this study was to point out the impact of several factors on the reliability of the mapping method.
      Methodological research was carried out in a number of sampling plots of 5-350 ha (658 ha in sum). These were surveyed by helpers and controlled by the author who was intimately known with each plot. Mapping was performed according to standard rules, using 1 : 10 000 maps. Each plot was visited 4-10 times during March through August, spending on average 5-350 min/ha (tab. 1). To compare these results with the exact population size, nests were searched in each plot. Most nests were found; information was gathered on the approximate location of other nests.
      Song activity reached a peak at the end of April and during May. The maximum percentage of territories recorded during three years varied from 43 to 60 but average percentages were much lower (12.8-21) (fig. 1). Singing males were heard from sunrise to sunset but the overall pattern was inconsistent (fig. 2). Cloudy weather with rain, low temperatures and wind (> 4 Beaufort) had an adverse effect on singing and behavioural activities. On the contrary Bullfinches were highly activated after thunderstorms when the sun bursts through, on calm mornings in May and in late summer. Individual differences in song frequencies were found. Both sexes regularly made long-distance flights (up to 400 m from the nest) during which singing occurred (by both sexes) on arbitrary points along these flight lines without eliciting territorial behaviour of neighbouring pairs. Territorial behaviour was missing and most pairs tended to breed in loose groups of 3-4 pairs in suitable breeding places, especially areas with young conifers. Nests were only 10-30 m apart in those groups. Polyandry was found in an intensively studied area during 1976, 1979 and 1980 in respectively one out of 31 pairs, two out of 27 pairs and one out of 28 pairs. A female with two males was observed thrice and a female with three males once.
      Survey efficiency of eleven observers using the standard mapping method averaged 70%. Only one survey resulted in a registration of 100%. The remaining surveys recorded 29-78% of the pairs present, which variation was probably due to differences in experience of the observers (fig. 3, cf. tab. 1). Survey reliability also depended on the time spent per area. In woodland plots of 20, 40, 100 and 350 ha 12, 9, 6 and 5 min/ha respectively were devoted to the census. After eight visits 89% of the territories in the 20 ha plot were recorded and only 39% in the 350 ha plot (fig. 4).
      These results together make the mapping method unreliable for censusing Bullfinches during the breeding season. The only reliable census method consists of systematic nest searching by experienced observers in well-known plots of 100 ha or less. A comparison of the mapping method and nest searching is given in tab. 2. In two small plots the results of nest searching agreed with the number of territories mapped; in the other plots underestimates of 22-71% were found (x = 37%). In areas of 100-1000 ha a modified mapping method is suggested in which every seemingly suitable breeding place was thoroughly inspected in order to locate the nests or to elicit pairs to betray their approximate nest site. Using this method in a census plot of 20 ha woodland, 100% of the pairs was recorded after five visits, compared with 89% after eight visits when the mapping method was used (fig. 4). Besides greater efficiency and accuracy the modified mapping method had an extra advantage in that it could be used irrespectively of day-time, time during the breeding season and weather. In areas larger than 1000 ha woodland it is advisable to select sampling plots (of 10-40 ha) in which all nests must be located by systematic searching. These figures should be extrapolated. Care must be taken in selecting plots because of the tendency of Bullfinches to cluster their nests in suitable breeding places which are restricted both in time and space.
      Censuses in 1975 and 1976 on the SW. Veluwe revealed 850-900 pairs on 110 km2 During 1977-79 some samples of breeding densities on other parts of the Veluwe were taken and on account of these data (and in combination with the SW. Veluwe results) the breeding population of the Veluwe (? 1200 km2) was estimated to be 5000-6000 pairs. Therefore, the estimate of 4000-5000 pairs for the Netherlands (van Dijk 1979) is too low and 10 000 pairs will be a better figure.. Eurasian Bullfinch Pyrrhula pyrrhula

[free pdf] [dutch summary]