Please use this identifier to cite or link to this item:
Measurement errors in Hawaiian forest bird surveys and their effect on density estimation.
|Title:||Measurement errors in Hawaiian forest bird surveys and their effect on density estimation.|
|Authors:||Camp, Richard J.|
|Issue Date:||Mar 2007|
|Abstract:||Reliable count data are necessary for valid density estimation. Before each Hawaiian bird survey, observers go through a training and calibration exercise where they record measurements from a station center point to flagging placed about the station. The true distances are measured, and when an observer's measurements are within 10% of truth the observer is considered calibrated and ready for surveying. Observers tend to underestimate distances, especially for distant measures (e.g., true distance > 50 m). All proposed empirical distribution functions failed to adequately identify the function form of the calibration data. The effect of measurement errors were assessed with populations of known density in a simulation study. By simulations, using the true distances, the conventional estimator seems unbiased; however, in the presence of measurement errors
the estimator is biased upward, resulting in overestimated population sizes. More emphasis should be made to minimize measurement errors. Observers’ measurement errors should be small with deviances less than 10%, for example, and observers should recalibrate frequently during surveys. Truncation is not a surrogate for increased accuracy. When there are relatively large amounts of measurement error estimators to correct the errors should be developed and used. Measurement errors to birds heard but not seen needs to be calibrated, and adjustment parameters included in measurement error correction models.
|Appears in Collections:||Hawaii Cooperative Studies Unit (HCSU)|
This item is licensed under a Creative Commons License