Density Trends and Range Boundary Constraints of Forest Birds Along a Latitudinal Gradient

We plotted the density distributions of 41 land-bird species along a 1,200-km transect spanning 7°28′ (865 km) of latitude through relatively uniform bottomland deciduous forest in middle North America. Standardized counts and observations at 12 survey stations, closely matched in habitat structure and widely distributed along the route, provided population data for all species and indices of total avian foraging pressure (consuming biomass) on each of six major foraging substrates. Density curves for species fluctuated considerably from station to station but tended to be level across range centers and slope peripherally to north and south boundaries at rates of 3-30% per degree of latitude. Substrate foraging pressures declined northward on the aerial and midfoliage substrates and southward on the low-foliage substrate. Summed community densities showed no significant latitudinal trends. We used the distinctive distribution patterns of climate (smooth latitudinal gradients), habitat structure (irregular mosaics of vegetation patches), and competition (reciprocally sloping density gradients) to identify and evaluate the role of these three constraints along the transect. Progressive latitudinal trends in species abundance thus were attributed to climatic factors, irregular station-to-station fluctuations to habitat factors, and inversely sloping density trends in paired profiles to competition. On this basis all species apparently responded to both climatic and habitat factors, and about half of the species showed suggestions of competition. In a correlation analysis across the 12 stations, latitude per se most closely matched density distribution in 12 species, one or another of the habitat parameters in 25 species. We proposed that season length (days available for breeding activity) was the principal constraining attribute of latitude at northern range boundaries, day length (hours available for feeding and provisioning young) at southern boundaries. Boundaries have been essentially stable during the past 50-100 yr in most species, but the northern boundary expanded northward in one species following human-induced habitat enhancement, and temporarily receded southward in another following a winter of severe stress. We attribute this general stability of range boundaries over time to within-population gene flow and the associated peripherally declining mean fitness of phenotypes adapted to central range conditions along radially diverging environmental gradients. We suggest that two boundary lines should be recognized for each species, an inner functional boundary at the line where birth rates drop below death rates, and an outer empirical boundary at the limit of recorded occurrences.