Epperly et al.: Aerial surveys for sea turtles 



257 



and to validate methodology. Two planes, each car- 

 rying two observers positioned on the same side, con- 

 ducted 12 flights over an area where painted ply- 

 wood "turtles" were deployed. Turtle models repre- 

 senting loggerhead turtles of 30, 60, and 90 cm 

 length, were attached to three anchored ground lines. 

 Within an overflight pass, turtles of one size were 

 grouped on a single line. All three lines could con- 

 tain turtle models during any one pass. The number 

 of turtle models of one size and the line on which 

 they were placed during a single pass were chosen 

 at random, but the experiment was constrained such 

 that a total of six turtle models of each size were 

 displayed within every three passes; the actual num- 

 ber of a given size displayed during a pass ranged 

 from to 4. The number, location, and size of the 

 turtle models were unknown to the observers. Alti- 

 tude and speed were identical to that used in the 

 general survey ( 152 m and 128 km/h). The airplane 

 flew on a line 0.10-0.30 km from the models. Analy- 

 sis of variance techniques were used to examine the 

 contribution of observers, turtle model size, and the 

 interaction of observer and model size to the error in 

 the counts. 



Results and discussion 



Under the ideal conditions under which the aerial 

 survey experiment was performed, no significant dif- 

 ferences were detected among observers (AN OVA, df=3, 

 P=0.89). Within the range of sizes tested, turtle size 

 was not a significant factor in the observers' ability to 

 sight turtles ( ANOVA, df=2, P=0.24). On average, 97.2% 

 of the actual number of "turtles" were sighted during a 

 pass (range 50-100%). We concluded that interobserver 

 variability was not a major factor and that turtles could 

 be sighted accurately in relatively turbid waters. The 

 experiment did not test for the effect of fatigue on an 

 observer's ability to sight turtles. 



The inshore waters of temperate latitudes are sea- 

 sonally repopulated with sea turtles. Nearly all sea 

 turtles in Pamlico and Core Sounds, North Carolina, 

 are immature individuals (Epperly et al., in press, 

 a). Based on public reports, there is evidence that 

 turtles immigrate into Core and Pamlico Sounds in 

 the spring, disperse throughout the sounds in the 

 summer, and emigrate from the sounds in the late 

 fall and early winter (Epperly et al., in press, a). 

 Results of the aerial surveys confirm this general 



Figure 3 



Seasonal sea turtle sightings in aerial surveys of Core and Pamlico Sounds, 1989-91. There 

 were no fall surveys of southern Pamlico Sound, and Core Sound was the only area flown 

 during the winter. The Core Sound area is enlarged to the right of each figure. (A) March- 

 May; (B) June-August; (C) September-November; (D) December-February. 



