Planarity and size of orb-webs built by Araneus Diadematus (Araneae: Araneidae) under natural and experimental conditions

Orb-weaving spiders build more or less planar webs in a complex, three dimensional environment. How do they achieve this? Do they explore all twigs and branches in their surroundings and store the information in some form of mental map? Or do they at first just build a cheap (i.e. few loops, possibly non-planar) web to test the site and - if this first web is successful (i.e. the web site is good) - later build subsequent improved and enlarged webs, by re-using some of the anchor points and moving other anchor points? The second hypothesis is supported by the fact that the garden cross spider Araneus diadematus CLERCK (Araneidae) usually builds several webs at the same site, re-using structural parts of one web for subsequent webs. To further test the second hypothesis, we measured and assessed the planarity of first and of subsequent webs built in the field and in the laboratory.