THE SELECTION AND USE OF FRUIT BY BIRDS IN AN EASTERN FOREST

In order to gain insight into the above questions, I studied the feeding habits of birds in a 600 ha tract of woods belonging to the Institute for Advanced Studies in Princeton, Mercer Co., New Jersey in the fall and winter of 1974-75. I found 23 species of fruit-bearing plants in the woods and along its margins. The species composition varied in different successional stages and with different moisture and soil conditions. To take into account the variety of stand types represented in the forest, I chose study sites in 3 different, relatively homogenous areas. The first study area comprised 235 m2 in an area grown in young red maple (Acer rubrum) and white oak (Quercus alba) with American beech (Fagus gran(difolia) and some flowering dogwoods (Cornus fiorida); the diverse understory included mapleleaf viburnum (Viburnum acerifolium), false Solomon's seal (Smilacina racemosa), spicebush (Lindera bensoin) and Virginia creeper (Parthenocissus quinquefolia). In this plot, only the mapleleaf viburnum and false Solomon's seal produced fruit. The second plot was 470 m2 of white oak, tuliptree (Liriodendron tulipifera) and shagbark hickory (Carva ovata), with subcanopy of flowering dogwood, blackgum (Nyssa sylvatica) and American beech. Understory included mapleleaf viburnum, New York fern (Thelypteris noveboracencis), greenbrier (Smilax rotundifolia), false Solomon's seal and spicebush. The dogwood and blackgum trees produced fruit, but most of the blackgum berries had already dropped when the study began. The viburnum, false Solomon's seal and spicebush had fruit. The third plot was 470 m2 of floodplain forest near the Stony Brook River. An open canopy of 2 large white oaks with box elders (Acer negundo) beneath allowed much sunlight to pass through. Understory was mostly tall, woody spicebush and vine poison ivy (Rhus ra(licans) densely entwined the trunks of the oaks. Both the spicebush and poison ivy produced large crops of berries. All fruit on each plant in the 3 plots was counted at the outset of the study, as were the