Spatial variation in shorebird nest success: Implications for inference

Estimates of nest success are widely applied in order to evaluate a multitude of theoretical and practical issues. Frequently, however, researchers fail to limit their inferences to the appropriate spatial scale. We evaluated small-scale variation in nest success of Western Sandpipers Calidris mauri during a four-year study on the Yukon-Kuskokwim Delta in western Alaska. We use these data to demonstrate that small-scale variation in nest success can significantly alter a researcher's interpretation of the factors affecting that reproductive parameter. In the absence of a statistically valid sampling design, researchers must be very careful about making inferences for areas beyond their actual study site. Properly designed studies allow for broader inferential power, but the logistical and financial hurdles involved in designing and implementing such a study are daunting. Metareplication can enhance one's confidence in the interpretation of local results, but should not be seen as a substitute for well-designed sampling schemes implemented across broad geographic scales.