Thus I am very sympathetic to efforts to argue for a evolutionary or biological basis for these feelings, which appear to be common to most human cultures. E.O. Wilson, the father of "Sociobiology" and evolutionary psychology, has written a book "Biophilia" that essentially argues that our preference for natural scenes is innate, the product of a psychology that evolved in paleolithic times. I would like this to be a correct view, but alas, it is, like most of evolutionary psychology, more like Rudyard Kipling's "Just so Stories" than hard science.
It is one thing to simply note trees as a metaphor for shelter, and thus to find it natural that architectural designs (such as the Metropol Parasol in Seville shown in the picture) that incorporate the tree metaphor would be pleasing to us. It is quite another hang this all on the supposed cognitive neuroscience of embodied cognition, as Sarah Williams Goldhagen, the architecture critic for The New Republic, has done in a rather confused piece. A recent post by Voytek, and the discussion following, point out a number of reservations and relevant points.