Effects of Vegetation and Background Noise on the Detection Process in Auditory Avian Point Count Surveys.

maximum detection distance;point count surveys;detection probability

Pacifici, Jamian Krishna ; Dr. Theodore R. Simons, Committee Co-Chair,Dr. Thomas R. Wentworth, Committee Member,Dr. Kenneth H. Pollock, Committee Co-Chair,Pacifici, Jamian Krishna ; Dr. Theodore R. Simons ; Committee Co-Chair ; Dr. Thomas R. Wentworth ; Committee Member ; Dr. Kenneth H. Pollock ; Committee Co-Chair

University:North Carolina State University

关键词: maximum detection distance;    point count surveys;    detection probability;   

Others  :  https://repository.lib.ncsu.edu/bitstream/handle/1840.16/159/etd.pdf.pdf?sequence=1&isAllowed=y

美国|英语

来源: null

【 摘 要 】

We used a system capable of simulating avian census conditions when birds are detected aurally to evaluate environmental effects on the detection process.We were interested in quantifying the effects of observers, species, and background noise on detection probability and maximum detection distance in two habitats (mixed pine/hardwood forest and deciduous forest) and two leaf conditions (leaves on and leaves off).We found significant effects of background noise, habitat, and leaf conditions on maximum detection distance.Maximum detection distance decreased on average by 36.61 m +- 5.44 with the addition of background noise, 29.52 m +- 6.25 with leaves on trees, and 35.68 m +- 5.58 between mixed pine/hardwood forest and deciduous forest.Average estimated detection probabilities varied greatly by factor combination and species, but in general they decreased with the addition of background noise, under leaf-on conditions, and in deciduous forest sites compared to mixed pine/hardwood sites.Average detection probabilities at 100 m ranged from 0 to 1 among species and under different environmental conditions.Average detection probability estimates never exceeded 0.19 for the Black-and-white Warbler (Mniotilta varia) at 100 m in deciduous forest.We found that a relative hearing index explained a maximum of 37% of the variation in observer detection probabilities.Interacting factors differed among species and were therefore difficult to predict.Environmental conditions can impart substantial bias in auditory point count data.We suggest estimating detection probability directly and incorporating this estimate in abundance estimates.