I only have the suggestion that I learned a long time ago, but unfortunately I don't have any scientific measurements to confirm it. It very well may be my superstition but here it is. It's simple enough to try but exhausting to try to measure and confirm. 

N1EA Antenna Rule #1:

No part of the antenna shall touch or be within one foot (305 mm) of any object conducting or non-conducting including leaves or tree branches including antenna supports. For antenna supports I made a one foot separation from the end of the antenna insulator at the end of the wire antenna and the antenna support (building, tree, tower or PVC pipe) made from Kevlar rope. Kevlar rope is my choice as it's thin diameter but quite strong. 

My friend and inventer Michael Laronda, WA1OMI measured using instruments the impedance of the antenna while we moved the antenna away from objects when the impedance of the antenna no longer changed, we noted that when separation distance in all cases for conductors and most surprisingly non-conductors was equal or greater than one foot all interactions with the antenna stopped. Antennas that previously only performed poorly then performed much better. As I said, only the measurement of how much separation was needed to have the antenna no longer be effected by mutual impedance was repeated and carefully measured. The efficiency improvements were very significant but our notes have been lost long ago, but we just made it a design parameter that no conductor or non conductor be within one foot (305 mm) of any antenna except non-conducting support ropes which were tied to the antenna end insulators. Branches and vegetation were just cut away. 

73

David Ring N1EA