Bird feeders and their effects on bird-window collisions at residential houses

Feeding wild birds creates an important link between homeowners and conservation. The effects of bird feeders and year-round feeding on birds have not been well studied, however, particularly in relationship to bird-window collisions. We determined effects of bird feeder presence and placement on bird-window collisions at residential homes. Paired month-long trials in which a feeder was either present or absent for one month and then removed or added for the second month were completed at 55 windows at 43 houses. In each trial, homeowners were asked to search their study window daily for evidence of a bird-window collision. During the study there were 51 collisions when there was no bird feeder and 94 when the feeder was present. The season when each trial was set up was the best individual predictor of bird-window collisions. The largest number of collisions was observed during fall migration and the lowest during the winter months. There were no collisions at 26 of the study windows. High variance was observed in the number of collisions at different houses, indicating that effects of bird feeders are context dependent. Changing the occurrence, timing, and placement of feeders can alter collision rates but is only one of many factors that influence whether a residential house is likely to have a bird window-collision or not.Original Abstract: Le fait de nourrir les oiseaux sauvages cree un lien important entre les proprietaires de maisons et la conservation. Toutefois. les effets des mangeoires sur les oiseaux et les effets de nourrir les oiseaux a l'annee n'ont pas ete approfondis. surtout en ce qui a trait aux collisions d'oiseaux avec les fenetres. Nous avons determine les effets de la presence et de l"emplacement de mangeoires sur les collisions d'oiseaux avec les fenetres de maisons. Des essais apparies d'une duree d'un mois, au cours desquels une mangeoire etait soit presente soit absente pour le mois, puis etait soit desinstallee soit installee pour le mois suivant. ont ete realises a 55 fenetres de 43 maisons. Lors de chaque essai, nous avons demande aux proprietaires de verifier quotidiennement la presence d'un indice de collision d'un oiseau avec la fenetre a l'etude. Dans le cadre de cette recherche, 51 collisions avec une fenetre sont survenues lorsqu'il n'y avait pas de mangeoires et 94 lorsqu'une mangeoire etait installee. La saison de l'essai s'est revelee etre la meilleure variable explicative des collisions d'oiseaux avec les fenetres. L