Are results of artificial nest experiments a valid indicator of success of natural nests?

Studies using artificial nests to assess rates or patterns of nest predation have increased during recent decades. However, the critical assumption that temporal or spatial patterns of relative predation rates of artificial nests parallel those of natural nests has begun to be questioned. We compared the daily survival rate for 365 artificial nests with that estimated for 295 open nests of passerines, and analyzed the trends in survival rates for both nest types among three breeding seasons and three species of nest plants in the central Monte Desert, Argentina. Daily survival rate for artificial nests (0.737) was significantly lower than that for natural nests (0.925). Trends in survival rates of artificial nests over years and among species of nest plants only partially reflected those of natural nests. Our results support the well established finding that artificial nests may not accurately estimate actual rates of nest predation. Artificial nests may provide an additional source of data to natural nests when testing ecological hypotheses, but artificial nest experiments should be carefully designed (i.e., realistic) and should attempt to identify predators of artificial and natural nests to validate the experimental results.