Population size and spatial distribution of the white stork Ciconia ciconia in Poland in 1958 with insights into long-term trends in regional and global population

Although white stork Ciconia ciconia counts go back to the XIX century, making the species a model in population investigations, substantial gaps remain in the knowledge of its large-scale population dynamics. In particular, incomplete past estimates from the core breeding areas leave the long-term changes uncertain. In this paper, we provide the earliest estimation of population size and spatial distribution of the white stork in its main stronghold (Poland) to reconstruct the species long-term trends. Based on original survey data collected during the 2(nd) (1958, incomplete in Poland) and 3(rd) (1974, successful) International White Stork Census (IWSC) we compared stork numbers in a random sample of villages surveyed on both occasions. We applied linear models to estimate the population growth rate between 1958 and 1974 and to assess spatial variation in population change across the country. Finally, we collated worldwide stork numbers obtained from all IWSCs and discuss the long-term population trends in light of new data. The stork population in Poland in 1958 was estimated at 46,100 breeding pairs and the nationwide density at 14.7 pairs/100 km(2). A strong decline (by 30.2%, 1.88% per year) was noted between 1958 and 1974 across Poland with prominent spatial variation reflecting differences in local densities. The strongest declines in absolute terms affected the most abundant populations in Eastern Poland. Our data show that in the mid-20th century, the stork population in Poland was close to the current level, but it experienced a massive decline during the 1960s-1980s. This decline was consistent with trends in the worldwide and regional European populations, contrary to earlier statements indicating limited, if any, changes in the core European area. Overall, our data expand knowledge on the long-term dynamics in the white stork numbers and show that even massive changes may easily go undetected if based on non-solid data.