Differences in predators of artificial and real songbird nests: Evidence of bias in artificial nest studies

In the past two decades, many researchers have used artificial nests to measure relative rates of nest predation. Recent comparisons show that real and artificial nests may not-be depredated at the same rates, but no one has examined the mechanisms underlying these patterns. We determined differences in predator-specific predation rates of real and artificial nests. We used video cameras to monitor artificial nests baited with quail and plasticine eggs and Field Sparrow (Spizella pusilla) and Indigo Bunting (Passerina cyanea) nests infield habitats in central Missouri (U.S.A.). Although daily predation estimates (all predators pooled) were similar between artificial and real nests, predators differed substantially in their depredation of artificial versus real nests. Snakes were the major predator at real nests, and raccoons (Procyon lotor) were the major predator at artificial nests. We found strong support for models that distinguished predation between two or among three predator groups and between artificial and real nests. There was no snake predation of artificial nests, and the odds of predation of artificial nests was 115-551% (95% confidence interval) and 2-154% of the odds of predation of real nests by mammals and birds, respectively. Artificial nests with plasticine eggs could not be used reliably to identify predators. In several cases plasticine eggs were marked by mice, and raccoons were recorded on video removing the quail egg. Because biases for artificial nests were positive for some predators and negative for other predators (and could be compensating), and potentially existed for all predator groups, conclusions based on artificial nest studies should be suspect even when there is evidence that overall predation rates are similar among real and artificial nests.