Single large versus several small: The SLOSS debate in the context of bird responses to a variable retention logging experiment