Tropical migrants form a much larger proportion of the avifauna of deciduous forest in eastern North America than in Europe. In North America, each square kilometre of this forest can support up to 600 pairs of long-distance migrants, which can form 50-75% of the entire bird assemblage (Lynch 1987). Given the immense size of this forest, which in pre-settlement times extended from southern Canada to the Gulf Coast and from the Atlantic to the Great Plains, it could then have accommodated nearly two billion migrants (Terborgh 1989). With the deforestation that followed the arrival of Europeans, this number surely diminished, probably to less than half its original value. At one time all these migrants were presumably accommodated in winter mainly within the tropical forests of Central and South America to which the remainder still migrate.
Deforestation in the Neotropics has been much more recent than in North America, occurring mainly since 1950.1 So for many years after the destruction of large areas of breeding habitat, the reduced migrant numbers were probably not occupying their more intact wintering habitat to full capacity. But with the more recent and rapidly accelerating deforestation of the Neotropics, declines in migrant populations from this cause become increasingly likely. These statements may hold as a generalisation, but whether particular species are affected probably depends on the extent of forest removal within their particular wintering range, and on whether they can live in the secondary habitats that replace the forest.
In North America, most recent information on trends in bird populations comes from the Breeding Bird Survey (BBS), managed by the United States Fish & Wildlife Service, in cooperation with the Canadian Wildlife Service. Since 1966, this scheme has provided annual data on bird populations, collected mainly by amateur observers, from randomly chosen sites in various habitats across the
1In even earlier times, at the height of the Mayan, Aztec and Inca civilisations, deforestation of Central and South America was more extensive than it is now, most of the recent forest being re-growth, less than 1000 years old.
continent. The scheme has produced a geographically much wider, and statistically more valid, data-set than was available for Europe over the same period.
For 133 species of Nearctic-Neotropical migrants (excluding waterbirds and shorebirds), the data were deemed adequate for analysis of long-term trends. Over the whole 36-year period, 23 species showed significant overall continent-wide increases, while 18 showed significant overall continent-wide declines, the remaining species showing no significant overall trends. However, patterns changed between 1966-1979 and 1980-1991. In the first period, 34 species showed significant continent-wide increases, while 19 showed significant declines (Table 25.1; Peterjohn et al. 1995, largely confirming Sauer & Droege 1992). However, in the later 1980-1991 period this pattern was reversed, when 15 species showed significant continent-wide increases, while 35 species showed significant decreases. Only the Upland Sandpiper Bartramia longicauda, House Wren Troglodytes aedon, Solitary Vireo Vireo solitarius, Warbling Vireo Vireo gilvus, Red-eyed Vireo Vireo olivaceus and Blue Grosbeak Guiraca caerulea showed significant increases through both periods, while the Chimney Swift Chaetura pelagica, Eastern Wood Pewee Contopus virens, Lark Sparrow Chondestes grammacus and Grasshopper Sparrow Ammodramus savan-narum showed declines through both periods. Most other species showed both temporal and regional variation in trends. The overall pattern indicated substantial regional and temporal variation in trends, but perhaps with more eastern forest species than expected declining in the 1980s. Many of these findings at national level were confirmed by local counts over shorter or longer periods, either in breeding areas or at migration sites (Terborgh 1989, Askins et al. 1990, Peterjohn et al. 1995).
From spring and autumn migration counts at Long Point on Lake Eyrie, Ontario, during 1961-1988, only four out of 33 species of Neotropical migrants increased in numbers seen per day, while 29 decreased. These figures differed significantly from those of short-distance temperate migrants, in which 11 out of 23 species increased and 12 decreased (Hussell et al. 1992). Despite these net trends, most species fluctuated substantially from year to year. Tropical migrants tended to decrease in the 1960s, increase in the 1970s and decrease in the 1980s, while many intra-continental temperate-zone migrants showed the opposite pattern. Nevertheless, nine (27%) of the 33 tropical migrants showed consistent declines over the 28 years, compared with only three (13%) of the 23 temperate migrants (statistically, a non-significant difference between groups). In general, the same patterns were revealed by both spring and autumn counts, which in turn correlated with BBS indices for Ontario, where many of the migrants were presumed to breed. Counts from some other migration stations showed some similar and
Was this article helpful?