ACCURACY TESTS FOR SONIC TELEMETRY STUDIES IN AN ESTUARINE ENVIRONMENT

We evaluated accuracy and precision of a directional sonic telemetry system and 2 positioning systems to study sea turtle (Cheloniidae) use of estuarine habitat in Core Sound, North Carolina. Accuracy and precision of location estimates affect the power of statistical tests for use of habitat studies and define the amount of movement that can be reliably measured. Angle errors associated with the sonic system averaged -2.5 degrees +/- 5.67 (SD) for a 95% error are of +/-11.34 degrees(range - 17-12 degrees). We obtained 45 location estimates after correcting 90 bearings for bias. Location errors (E), obtained from stationary positions at 400-1,200 m, ranged from 14.6 to 281.0 m with a median of 75.6 m. The 90 and 95% confidence areas for these data were 11.1 and 21.2 ha, respectively. Location error varied (P < 0.01) with geometric mean distance (D-g) between receivers and transmitters. Areal measures of confidence obtained at the D-g 500-600 m were the smallest (2.1-2.8 ha) among 3 distance intervals within 500-1,200 m. Attained levels of accuracy and precision were adequate to determine turtle movement and distribution in relation to selected fisheries activities, but of limited value for use of habitat studies. Inaccurate position estimates of monitoring platforms (e.g., boats) also affect location estimates. Precision of position estimates of a stationary boat anchored at a known location (i.e., channel marker) were poor, averaging 62 m, when obtained from Long Range Navigation System (LORAN). In contrast, positions obtained from Differential Global Positioning System (DGPS) varied by 3 m. The DGPS did not affect (P = 0.94) location estimation. Average difference between estimates using known location coordinates and those obtained from DGPS was 0.56 with a 95% confidence interval of +/-1.29 m. We recommend that DGPS be used when evaluating sea turtle use of habitat. The DGPS was more accurate than LORAN and was unaffected by geography.