November surveys. Green turtles were only noted during August surveys of SFLA but 

 were possibly included in the unidentified turtle category during other surveys. 



Turtles were rarely sighted in Texas surveys. Of those recorded, less than half as 

 many were seen in November as in August. Kemp's ridley turtle was identified only in 

 STEX but was present in both seasons sampled. Loggerhead turtles were sighted in 

 Texas, but with much lower frequencies than in Florida. 



The total number of turtle sightings and the relative frequency of the component 

 species varied within predictable limits. As an example, daily counts of loggerhead 

 turtles, unidentified turtles and all turtles in NFLA during August had ranges of 13 to 31, 

 1 to 13, and 16 to 44 respectively (Table 6). August surveys in SFLA resulted in ranges for 

 the same groupings as follows: 12 to 23, 7 to 15, and 19 to 40. 



DAILY VARIATION 



The daily variation in sightings of individual species within a single season and 

 geographic area can significantly effect estimates of faunal composition and 

 abundance. The magnitude of variation observed was in part dependent upon factors such 

 as abundance, observability, group size, and mobility. Empirically, taxa that were 

 relatively common on any one day were observed on subsequent days. The number of 

 groups on subsequent days was usually less than 50% above or below the initial 

 frequency. Taxa that were observed one to four times in a single day's flight were less 

 likely to be detected during every day of subsequent flights. Examples of both commonly 

 and sparsely sighted taxa are detailed in Tables 17 and 18. In the case of the unidentified 

 groups such as dark terns, daily variation may have been the result of being able to 

 identify them in certain lighting conditions or group sizes but not in others. The daily 

 totals of all tern groups sighted in August-SFLA surveys are 12, 15, and 20. This pattern 

 shows a greater predictability than the individual taxa in part because of increased 

 sample size and in part because when more groups are specifically identified, a decrease 

 in unidentified taxa is expected. Such variation would be reduced even more when group 

 sizes are considered. Five groups on one day averaging five birds/group may exactly 

 correspond to one group of 25 birds seen on a subsequent day during a different feeding 

 situation, etc. Both group size and group abundance have been analyzed for estimations 

 of relative abundance. 



On the basis of the example data in Tables 17 and 18 it is apparent that daily 

 replicates add to the understanding of faunal composition. In SFLA during August (Table 

 18), the first day's flight resulted in the observation of three mammal taxa and nine bird 

 taxa. On the subsequent day, three mammal taxa and eight bird taxa were noted. 

 However, of the three mammals observed on day 2, one was new to the data set, raising 

 the known number of mammal fauna to four. Of the eight birds seen on day 2, two were 

 new on the data set, resulting in a known bird fauna of 11. On day 3, 1 mammal and 11 bird 

 taxa were observed. Although no new mammal taxa were added, two additional bird taxa 

 were added to the data set even though two other taxa known to occur in the area were 

 not seen. Comparing the actual number of taxa observed with those known on the basis 

 of previous observations, a similar pattern can be seen in August data from. NFLA (Table 

 17). For mammals on days 1, 2, and 3, respectively, the ratios of observed/known taxa 

 are 2/2, 5/6, and 5/7. For birds the ratios are 7/7, 6/7, and 8/9. As of day 3, it was 



39 



