suggest that terns, which are more conspicuous because of their coloration and ability to 

 fly, are n*>re detectable than bottlenose dolphins. Terns can be seen from greater 

 distances, although at extreme distances they are often not identified to species. Thus, 

 the area covered with respect to terns is significantly greater than the area covered for 

 bottlenose dolphins. The percent aerial coverage for August-NFLA was: bottlenose 

 dolphins, 4.5%; royal terns, 14%; and unidentified terns, 21.5%. 



A comparison of the total number of individuals, the maximum observation 

 distance, and the estimated density for the example taxa in Table 30 suggests several 

 relationships relevant to evaluating the data. First, the taxa for which the most 

 sightings were made are not necessarily judged to be the most abundant. Density 

 estimates necessarily compensate for the observability of the animal away from the 

 transect line. Those animals which are viewed at large distances must be seen more 

 often than those with short observation distances. For example, the larger number of 

 royal terns (281) seen in STEX during November resulted in a density estimate of 0.36 

 birds/km , whereas the smaller number (31) seen during August yielded an estimate of 

 0.20 birds/km . The 186 royal terns seen in NTEX-August surveys resulted in a density 

 estimate only twice as large as the density estimate resulting from the 31 royal terns 

 sighted in NTEX during November. In the first case, the aggregation of birds during 

 November allowed detection at greater distances, and thus facilitated the greater area 

 to be surveyed. Part of the increase in the total number of birds sighted was a result of 

 increased area and part was the result of increased abundance. It may be significant that 

 the decrease in royal terns in NTEX during November was accompanied by an increase in 

 the species in STEX. A similar increase is seen in the density of royal terns in SFLA 

 during November (Table 31); the change in NFLA is less pronounced (Table 32). This 

 possibly suggests that the new birds in SFLA during November were arriving from areas 

 other than NFLA. 



Comparison of seasonal and geographic variation in other density estimates for 

 Florida subunits (Tables 31 and 32) allows several trends to be defined. Laughing gulls 

 were twice as abundant in SFLA as in NFLA during November. Loggerhead turtles were 

 only slightly more abundant in SFLA than in NFLA during August. However, loggerhead 

 density in NFLA during November was below the number needed for density calculations, 

 whereas they were only slightly less abundant during November. 



The brown pelican, a bird without pronounced seasonal movements, had similar 

 density estimates in NFLA during November and August. Density estimates for royal 

 terns were similar during the two surveys in NFLA (Table 31). In SFLA, higher November 

 concentrations of royal terns are suggested by both absolute numbers and density 

 estimates (Table 32). 



The estimates of densities available from the Pilot Study remain to be confirmed 

 by repeated measurements and increased sample sizes. Confidence limits for most 

 estimates are wide; however, these limits will be narrower when estimates spanning an 

 increased number of observations are available. As the ecological zonation of each 

 species is defined it will be important to calculate densities in appropriate portions of 

 the subunits rather than considering a generalized distribution throughout the study area. 



Future studies will also allow the determination of seasonal trends by allowing 

 comparison of an entire sequence of estimates instead of merely two points in time. 



59 



