Natural Regulation of the Baltic Population of the Great Cormorant Phalacrocorax carbo sinensis: The Interplay between Winter Severity and Density Dependence

After a period of continuous increase and range expansion, the Baltic Great Cormorant population has stabilised in large parts of its range in recent years. Ringing recoveries reveal that considerable proportions of the population winter in areas that can be affected by prolonged frost periods. There is evidence that winter severity is an important density-dependent regulation factor: if the population is large, ice cover of coastal and inland water surfaces during harsh winters affects the population by reducing the availability of food resources. As long as the population remained small, however, it was not affected even by very cold winters, since the remaining accessible food resources were presumably still sufficient. The analysis presented here uses the average winter temperature in Germany as a proxy for winter severity in the frost-affected parts of the wintering areas of Baltic Cormorants. The Baltic Cormorant population in 1980–2016 is estimated from annual counts in Denmark, Schleswig-Holstein, Mecklenburg-Western Pomerania, Estonia, Finland and Gotland, which account for about 50% of the total population. The interplay between winter severity and density dependence is analysed using a linear and a non-linear regression model approach. The non-linear model gives a better description of the relationship between the size of the Baltic breeding population during the year (n), the winter temperature Tn, and the population size during the previous year (n–1). According to the model, a population of less than 41,400 breeding pairs would not suffer declines during even the coldest winters recorded since 1882. In 1989, the Baltic Cormorant population exceeded for the first time the threshold value for density-dependent regulation caused by severe winters. The winter 1995/96 was then the first one cold enough to cause a population decline. According to the model, during the years 2002/2003, 2005/06, 2008/09, 2009/10 and 2010/11 the winters have been cold enough to reduce population numbers. Furthermore, the model shows that the regulative winter effect is restricted to the low temperature range.