Porovnanie a priori a a posteriori prístupov pri analýze potravných gíld vo vtáčích spoločenstvách: modelový príklad

Abstract

Two fundamental approaches of guild structure analyses – a apriori and a posteriori have not been empirically compared within the framework of a model study at one site up to now. This is the first attempt to describe, compare, and analyze the main differences between the two approaches using the example of bird assemblage. The study was conducted in a primeval beech-fir forest in the Šrámková National Nature Reserve. Population densities of breeding birds were estimated using an improved version of the mapping method in a 27.5 ha (500 × 550 m) study plot. An a apriori foraging guild analysis was based on the foraging characteristics of birds from previous studies. In total, 10 types of guilds were defined for a apriori assemblage analysis: stream foragers, ground foragers, foliage gleaners, bark foragers, airspace foragers, and plant eaters. An a apriori data matrices (25 × 10, 57 × 10) was prepared for comparative analyses. Bird species were divided into guild categories before the field observations of foraging birds started. The a posteriori classification was based on extensive field survey of the study plot. Foraging observations were collected from the middle of May until the end of July in 1997-99. In total, 2 921 random point observations of foraging birds were recorded on the data sheets with a standardized set of variables indicating species, sex, day time, duration, foraging height, direction of foraging movement, substrate type, and foraging strategy. Only, 25 bird species were subjected to further statistical analyses based on the criterion ≥ 40 observations per species. Multivaried statistical procedures such as hierarchical cluster analysis and correspondence analysis were applied to the classification of guild patterns. The emerged pattern consists of 7 types of guilds: ground foragers, stream foragers, bark peckers, bark gleaners, foliage gleaners, airspace foragers, and sky foragers. Comparing the same set 25 bird species, a priori approach detected only 6 guild types and incorrectly classified 5 species (20 %). Only 5 types of identical a apriori guilds (71.43 %) were detected in the emerged a posteriori pattern. Incorrect classification was mainly cause by the lack of previous information on foraging ecology and niche requirement of species. It is summarized that when correctly applied and based on reliable information from past studies, a apriori approach is a fast, cheap, and powerful tool for basic functional analyses in wildlife management. In contrast, a posteriori approach is very time consuming and requires large sample sizes of foraging observations. Only, 40-50 % of species meets satisfactory sample size for further statistical analyses. In addition to practical application for wildlife management and landscape planning, such study may be used for testing hypothesis in theoretical ecology. Further recommendations and improvements based on the experience in the field, with data analyses, and general methodology are given.