Summary: Population trends of 100 winter bird species in Finland in 1957-1996

Risto A. Väisänen & Tapio Solonen 1996: Suomen talvilinnuston 40-vuotismuutokset, Linnut-vuosikirja 1996, pp. 70-97.

• The whole article in Finnish (with tables and figures)

Long-term population changes of one hundred bird species wintering in Finland have been monitored during the last 40 years (for 20- and 30-year reports, see Sammalisto 1977 and Hildén 1988). The data has been collected in the national winter bird censuses organized by the Zoological Museum of the Finnish Museum of Natural History. Each winter censuses were conducted along the same 300-500 routes, the average length of which is about 10 kms. The three census periods last two weeks each: "early winter" from 1 to 14 November, "mid-winter" from 25 December to 7 January and "late winter" from 21 February to 6 March. For detailed instructions of the method, see Hildén et al. (1991).

The total data represents 3360 routes, over 40 000 censuses and about 15 million bird observations. Each row of Table 1 presents the censuses of one winter. The Table also shows eight five-winter periods, within which the data have, for most analyses, been combined in the early, mid- and late winter censuses. There are no data for the late winter census in the first two periods, and no data for the early winter census in the first four periods (the data of the single autumn 1975 is not used to represent the fourth period). Censuses have been made by several thousand voluntary ornithologists, mostly in southern and western Finland (Fig. 1). The geographical distribution of the censuses has remained rather stable (Table 2). Water bird censuses were, however, poorly represented in the first two periods.

For the eight five-winter study periods there was significant variation in certain factors that affect population levels of birds: winter chill (Fig. 2) and the cone crop of spruce (Picea abies) and pine (Pinus sylvestris) (Fig. 3).

Some properties of the population indices calculated for the five-year periods are compared in Fig. 4. The index to the left corresponds to that used in the species diagrams. Numbers that follow the species name in each diagram indicate of the size of the data. For example, in the Willow Tit the first figure tells that 14.5 individuals per 10 kms were observed in a median period in early winter censuses. Multiplying this unit by 2000 (since the average route length is about 20 000 km; cf. Table 2), converts to the absolute median number of observations for the period of a unit of five early winter censuses (in this case 29 000 Willow Tits).

The statistical significance of population change has been tested crudely by calculating it for the slope of regression, where x = mid-year of a five-winter period and y = the population index converted to a logarithmic scale. We assume that the species has steadily increased or decreased between the periods by a multiplicative constant that deviates from unity. The test does not reveal curved population changes. The results of the tests are shown in the three columns at furthest right of the Appendix, respectively for the early, mid- and late winter censuses. For waterfowl the data of the first two five-year periods were excluded from the test.

An index of general change of abundance of 100 winter bird species was calculated by a principal component analysis on the population indices of the last six five-year periods (Fig. 5; periods 1 and 2, meaning winters 1957-1966, were not used in this calculation, because their waterfowl data are not representative). The score value of the first principal component ordinates the species from those that have increased the most to those that have decreased most (columns "Population change: Winter" and "Rank" in the Appendix). For each species, the 40-year change of the wintering population is compared with the corresponding change in its Finnish breeding population during 1970-1990 (here we follow the system of Tucker & Heath (1994): +2 = large increase of at least 50%; +1 = small increase of 20-49%; 0 = stable, or change of less that 20%; -1 = small decrease of 20-49%; -2 = large decrease of at least 50%; f = fluctuating with changes of at least 20%, but no clear trend).

There is notable correspondence between species' winter and summer indices. Exceptions arise from the following points: (1) in several species winter observations mostly concern birds that do not breed in Finland, but have moved here from more northern or eastern areas; (2) the winter indices (1967-1996) cover a period that is ten years longer than that of the summer indices (1970-1990); (3) in some species a change in the winter population can be explained by technical reasons such as improved observer identification skills, increased use of telescopes, or improved sampling of the species' preferred habitats.

Of the 100 winter bird species studied here, about half have increased and one third decreased during the last 30 years - rather few have remained stable. Corresponding change has been noticed among the most abundant birds breeding in Finland (cf. Järvinen 1984). The diagrams of the species in ecological or systematic groups have been arranged on the basis of the winter index from the most increased to most decreased. Further, annual changes of wintering and breeding populations are compared in about 20 species that represent sedentary birds or partial migrants. The latter indices come from repeated land bird censuses and start, depending of the size of the data, between summers 1978 and 1984.

Waterfowl numbers noted in the winter bird censuses depend on of the extent of ice. This has been found to vary in a complex manner in the Aland Islands, where most winter waterfowl observations are made (Hario et al. 1993). Observations of the Canada Goose (Fig. 6: Bracan), Mute Swan (Cygolo) and Whooper Swan (Fig. 7: Cygcyg) have increased in tandem with the growth of their Finnish breeding populations. The reasons for the population growth of the following species is more obscure: Cormorant (Fig. 6: Phacar), Tufted Duck (Aytful), Scaup (Aytmar), Long-tailed Duck (Clahye), Goosander (Fig. 7: Mermer) and Goldeneye (Buccla). The origin of the censused winter populations of these species is not known - the birds most probably do not breed in Finland. Further, starting from the 1970s, telescopes came into use, which considerably increased the observations of eg. Tufted Ducks and Scaups.

Other population fluctuations of waterfowl presented in Figs. 7-8 have not shown clear trends, excluding the Mallard, which has strongly declined - especially in the late winter census (Fig. 8: Anapla). Gulls have been most abundant in those five-winter periods with mild weather (Fig. 9), but the general growth of their numbers can be explained by the increased use of telescopes.

No permanent changes has been noted in the winter numbers of the Sparrowhawk (Fig. 10: Accnis), Goshawk (Accgen) or Rough-legged Buzzard (Butlag). Efficient winter feeding has progressively increased the numbers of Golden Eagle (Aquchr) and White-tailed Eagle (Halalb). However, the winter indices exaggerate the population growth, because new census routes have been founded near the feeding sites and the improvement of optical equipment has made the detection of large raptors easier. The downward trend of the Merlin (Falcol) has also been noted in breeding populations (Haapala et al. 1997).

Although owl observations are scarce in the winter bird census data, the results coincide fairly well with those of the Finnish raptor monitoring programme (Haapala et al. 1997) and reveal, for example, the increase of the Eagle Owl during the last 30 years (Fig. 11: Bubbub) and the lack of long-term trends in the Hawk Owl (Surulu) and Pygmy Owl (Glapas). The Tengmalm's Owl (Aegfun) has been scarce in recent winters, because several years have passed since the latest good breeding summer for the species in Southern Finland (1991). The Tawny Owl (Stralu) increased during the first 15 years of winter monitoring, but has steeply declined later.

Numbers of wintering Great Grey Shrikes were notable in the mid-1970s (Fig. 11: Lanexc), when autumnal observations of the species were also frequent (Hildén & Hildén 1996). A new peak density has been noticed in the winters of the 1990s.

Pheasant numbers increased in the early years of winter bird censuses, remained stable for decades, but have declined by half in the 1990s (Fig 12: Phacol). The Partridge (Perper) has decreased more steeply with the modernization of agriculture, although it has slightly recovered recently. Populations of these gallinaceous birds of fields are dependent on winter feeding and perhaps also reintroductions. The town of Helsinki stopped winter feeding of pheasants in 1991-1992 and the population crashed rapidly.

There is an alarmingly steep decrease in the Hazel Grouse (Fig. 12: Bonbon), Black Grouse (Tetrix), Willow Grouse (Laglag) and Capercaillie (Teturo), which have suffered from environmental changes of Finnish forests and peatlands, and apparently also overhunting. The decline of the Willow Grouse and Capercaillie has been especially heavy in Southern Finland, where the bulk of the winter bird censuses have been made.

The scant data of the Collared Dove shows the growth phase of the Finnish population (Fig. 13: Strdec). No special trends can be seen in the diagrams of the Stock Dove (Coloen) and Wood Pigeon (Colpal). Feral Pigeons has decreased - perhaps control measures against this species has affected it in towns (Colliv).

The large fluctuation during the two first five-winter periods makes the indicated increase of the Great Spotted Woodpecker less convincing (Fig. 14: Denmaj). The upward trend becomes more apparent, when minimum, median and maximum winters are shown separately for each five-year period (Fig. 15). Two good woodpecker winters (1957 and 1959) were noticed during the first period, but the other three years were only modest. The spruce cone crop was poor during the second period and woodpecker numbers correspondingly low. Since then their abundance has grown 2-3-fold both in minimum, median and maximum winters. In the same period breeding bird censuses (unpubl. line transect data) indicated a population increase of only about 60 % between the 1940s-1950s and 1970s-1980s. This is explained by winter feeding which has attracted more and more woodpeckers to settled areas, where they have been noticed in the winter censuses. Although the annual wintering population varies more than the breeding population (Fig.16), in most years the winter numbers can be used to predict the size of the breeding population.

The Grey-headed Woodpecker has increased in the winter census data (Fig. 14: Piccan), but apparently the birds have only accumulated more effectively on winter feeding sites by the census routes. The Black Woodpecker has been most numerous during periods of mild winters in the beginning of the 1970s and in the turn of the 1980s-1990s (Fig. 14: Drymar). The other three species in Fig. 14, Lesser Spotted Woodpecker (Denmin), Three-toed Woodpecker (Pictri) and White-backed Woodpecker (Denleu) have declined strongly due to changes in the structure of Finnish forests. Upward trends during the latest period were caused by a large-scale woodpecker irruption from Russia in the autumn of 1993. It did not result in a permanent growth of breeding populations in Finland.

Starting from the 1970s, invasions of the Siberian subspecies of Nutcracker (Nucifraga caryocatactes macrorhynchos) have become stronger and more regular, and a few have settled to breed in Finland, but the subspecies breeding in SW Finland (N. c. caryocatactes) has also extended its breeding area. Flocks of the Siberian birds have apparently had the stronger effect on the increase in winter observations (Fig. 17: Nuccar).

The Magpie strongly increased in the 1970s and early 1980s due to a decrease in persecution (Figs. 17-18: Picpic). It has started to breed in settled areas, including towns. A decrease in persecution apparently has also helped on the growth of the Raven population from the 1950s-1960s (Fig. 17: Corrax). The tiny population of Rooks that tries to overwinter in Finland has fluctuated without any clear trend (Fig. 17: Corfru). Irregular autumnal invasions make Jay data problematic because the invasions increase the variance of the winter and summer population indices. This factor probably explains the uninformative appearance of Jay diagrams (Figs. 17-18: Gargla). The species has increased in the late winter census. Early springs have, however, become more common and thus improved detectability of the Jay. This may explain the trend.

The Jackdaw declined in winter bird censuses especially in the 1950s-1960s, but it may be that flocks were larger and recorded more effectively in the early census years (Fig. 17: Cormon). Winter flocks make the winter population indices much more variable compared with summer population indices (Fig. 18). Hooded Crow numbers did not change much during the first 20 years of winter monitoring, but decreased by half during the last 20 years (Fig. 17: Cornix). Wintering and breeding populations have declined similarly (Fig. 18). A downward trend is apparent in the small winter census data of the Siberian Jay (Fig. 17: Perinf).

Weather causes large annual variation in the seed crop of trees (Fig. 3), which in turn causes strong variation in the abundance of species specialized in their use. Single good winters notably affect the population index of a five-year period. For example the increasing trend of the Two-barred Crossbill mainly results from two large invasions to northern Finland and Sweden in the autumns of 1986 and 1995 (Fig. 19: Loxleu) when the spruce cone crop was abundant (Fig. 3; data in winters 1987 and 1996). The birds bred in the following spring and summer and disappeared the next autumn. Invasions have possibly been intensified in this species, as comparable immigration has not been reported before.

Increasing trends in the diagrams of the Arctic Redpoll (Fig. 19: Carhor) and Parrot Crossbill (Loxpyt) have been partly or totally caused by improved skills in identifying these species. Crossbill (Loxsp.) data include both Loxia curvirostra and unidentified Crossbills, but represents mainly L. curvirostra, because it is clearly more abundant in South Finland, where most of the data is gathered, than L. pytyopsittacus. The area and volume of the spruce forests have increased in South Finland in recent decades when excellent Crossbill years (1957, 1968, 1979, 1984, 1990 and 1993) became more common in the winter bird census data.

The seeds of birch (Betula) and alder (Alnus) are basic winter food of the Siskin (Fig. 19: Carspi) and Redpoll (Carmea). One period, when the population was low in five successive winters, notably affects the diagrams of both species.

Among birds feeding on rowanberries (Sorbus), the Waxwing (Fig. 19: Bomgar), Blackbird (Turmer) and Bullfinch (Pyrpyr) have become more numerous. (The winter index of the Fieldfare is unreasonably high mainly for technical reasons (see Appendix); its diagram (Fig. 20: Turpil) has been moved below the above three species).

Seven great thrush winters correspond to abundant rowanberry crops in 1957, 1965, 1970, 1984, 1990, 1993 and 1996 (Fig. 21). In these years Fieldfares were abundant in the early and mid-winter censuses, and tiny remains of flocks appeared even in the late winter census. The Waxwing was also numerous in great thrush winters, but common in mid- and late summer censuses also in 1974, 1977, 1978, 1979, 1981, 1986 and 1992. The winter abundance of the Fieldfare has not been related with annual or long-term changes of breeding numbers, but Waxwings may have increased in northern boreal forests.

In the early years of monitoring both species were rather scarce in ordinary winters (no abundant food supply, Fig. 22). In recent decades differences between winters in the abundance of Fieldfares have notably levelled off. Good Waxwing winters have become more common, and very poor ones rare.

The rowanberry crop also causes strong fluctuation in winter numbers of the Blackbird (Fig. 23). Winter invasions cause fluctuations in the numbers of the Bullfinch, but winter feeding of birds may be taken as the main reason for the long-term increase. No special trend have occurred in the generally scarce winter numbers of Redwing (Fig. 20: Turili). The abundance of the Pine Grosbeak (Pinenu) has fluctuated, but the trend may be declining.

Winter long-term changes have been similar in three basic members of the coniferous forest tit guild, the Coal Tit (Fig. 24: Parate), Willow Tit (Parmon) and Crested Tit (Parcri) (cf. Hildén & Väisänen 1991). This has also been evident in the annual variation of the wintering and breeding populations. In recent years the species' trends have not followed each other (Fig. 25). Willow Tit numbers have declined by half in five years. The Siberian Tit has slightly recovered in the 1980s both in winter (Fig. 24: Parcin) and summer censuses from the low level of the 1970s, but the long-term trend is to a strong decline during this century. Fluctuations of the Goldcrest (Figs. 3 and 25: Regreg) and the Treecreeper (Cerfam) have been regulated by the occurrence of harsh winters. In the former species the crash was especially strong in the period of severe winters in the mid-1980s and in the later species in the late 1960s. The Blue Tit (Fig. 24: Parcae) and Great Tit (Parmaj) have increased mainly due to winter feeding. Winter numbers of the Blue Tit have grown about 20-fold in 40 years. This has been aided by the eutrophication of waters and the subsequent expansion of reedbeds, an important winter food source for the Blue Tit (Hildén 1988). Numbers of Great Tits stabilised at a high level already in the 1970s. In certain years notable discrepancies have occurred in this species between winter and summer census results (Fig. 25).

Several other passerines commonly visiting feeding sites have increased their winter populations. The growth of the Tree Sparrow population has been exponential (Fig. 26: Pasmon). Nuthatch invasions from Siberia have become more numerous in this century (Väisänen 1996a). Of 14 invasions the largest ones in the autumns 1962, 1976, 1983 and 1995 especially affect the diagram (Fig. 26: Siteur). After the latest invasion a small population has started to breed in Finland in 1996-1997. Greenfinch (Carchl) numbers have grown about 30-fold in the 40 years, although the growth of the breeding population of this partial migrant has been more moderate (Fig. 27). A larger proportion of the population has started to winter in Finland than previously, attracted by abundant food at feeding sites.

The numbers of Yellowhammer doubled in the winter bird censuses during the first 25 years (Fig. 26: Embcit). Since then the breeding population has decreased by about 20 % (Fig. 26). In the winter data this trend has been masked by strong annual fluctuations and more efficient accumulation of the birds on feeding sites beside the census routes. The same factor has partly masked the decline of the House Sparrow that has continued two decades (Figs. 26-27: Pasdom): the breeding population has decreased by half in about ten years. The decline of the Starling that has continued from the mid-1970s, has been apparent also in winter censuses (Fig. 26: Stuvul). Starling flocks strongly affect the census results of early winter compared with the breeding numbers (Fig. 27).

Among more rare winter visitors, the results of the Reed Bunting are highly dependent on the exceptional winter of 1993 (Fig. 28: Embsch), when an abundant reed seed crop possibly attracted the species to winter fairly abundantly in southern Finland (Hildén & Väisänen 1993). The high numbers of the Hawfinch (Fig. 28: Coccoc) in the late 1970s were probably caused by the rich elm seed crop of 1978 (Sammalisto 1979, Tiainen 1983). The 40-year diagram of the Linnet (Fig. 28: Carcan) is the most peculiar one among the 100 species presented here; it is dominated by one large flock of some 1 600 birds noted in the mid-winter census of 1985.

Increasing winter population trends of the Wren (Fig. 28: Trotro), Dunnock (Prumod) and Robin (Erirub) can be explained by improved skills to seek these birds beside the census routes in the most important wintering area, the Aland Islands. The diagram of the Dipper reflects changes in censusing practices that earlier took in many places used by the mainly Norwegian birds wintering in Finland (Fig. 29: Cincin). Goldfinch (Fig. 29: Carcar) numbers depict a real decline that started in the 1960s, and recent recovery. Abundance in the Long-tailed Tit (Aegcau) is regulated by autumnal invasions from Russia and high mortality in periods of harsh winters. No special trend emerges from Twite data (Fig. 29: Carris). The breeding population of Snow Buntings has declined, but the same trend is not apparent in winter census data (Fig. 29: Pleniv), which hardly concerns the Finnish breeding population. The mid-winter indices of the Skylark are based on scarce data, but broadly reflect the general decline of this species as a breeder following wider use of modern agricultural practices (Fig. 29: Alaarv). The early and late winter indices of the Skylark mostly reflect weather variation. In mild autumns and very early springs notable numbers of migratory Skylarks have been observed in the censuses.