A review and evaluation of territory mapping methods in bird censuses

Mapping methods have played a crucial role in bird monitoring and for testing ecological hypotheses, theories and accuracy of other count techniques. The purpose of this review is to describe bird territory mapping methods, their development, principles, types and accuracy, and to compare their estimates with those of other quantitative methods. The mapping method was developed by a US ornithologist, A.B. Williams, in 1936. Since then, these methods have been used in large‐scale monitoring schemes in the United States, Canada, Great Britain, Sweden, and currently in the Netherlands and Switzerland. Modification, improvements and standardization of the original Williams’ mapping method led to the development of six different types of the mapping method: Consecutive Flush, International Standard, Combined Version, Total Mapping, Circular Mapping and Digital Mapping methods. Clustering (amalgamation) algorithms have been used to reduce human subjectivity in interpreting species maps. Imperfect bird detection during surveys was identified as a major source of bias in bird monitoring. Various mixture models were developed to solve this problem and helped to improve the accuracy of estimating population abundance, trends and species richness. Inter‐observer and inter‐analyst variability was identified as a major source of bias in mapping method estimates. Mapping methods vary widely in accuracy, and results differ among species and regions. Total Mapping, usually applied to single or few populations, can produce the most realistic population estimates. The Combined Version is considered one of the most accurate methods for territorial species and may be the most useful for more precise assemblage‐wide studies. Future development of these techniques should involve testing mapping efficiency and accuracy for individual species and determining the variability of estimates among species, habitats, seasons and observers, as well as the development of new algorithms for computerized evaluation of species maps and statistical models for correcting imperfect detection in monitoring surveys.