; ARHAM ET AL AERIAL CENSUS OF THE BOTTLENOSE DOLPHIN 



bid water. On such days the only clear water was 

 in the lee of barrier islands and headlands where 

 the fetch was limited. 



Increased turbidity limits the observer's 

 chances of sighting underwater animals and not- 

 ing mud boils and trails. For underwater animals, 

 however, the overall effect on the number of sight- 

 ings was tempered because submerged dolphins 

 will frequently be spotted when they eventually 

 surface. More important was the negative 

 influence of high turbidity on the observer's abil- 

 ity to note surface signs. For example, on the two 

 low-wind days 12 out of 68 ( 17. 7<^ ) sightings were 

 cued by surface perturbations. In contrast, on the 

 three medium to high-wind days only 8 of 83 

 (9.6%) of sightings were signaled by this cue. The 

 effect was probably more important than those 

 data indicate, for frequently the observer's atten- 

 tion was drawn to an area by subtle surface signs 

 and then, if a dolphin's body showed at the surface, 

 it was usually the second rather than the first cue 

 that was logged. As stated earlier, we have re- 

 duced the effects of weather on the population 

 estimates by excluding the results of survey 5, 

 when the wind effects were extreme, from the den- 

 sity computations. 



Regarding our inability to sight all herds, the 

 supposition that all target animals will be seen is 

 basic to the strip transect method (Eberhardt 

 1978). However, in terms of line transect theory 

 ( Seber 1973), which assumes that the herds will be 

 randomly distributed, the frequency histogram of 

 the estimated perpendicular sighting distances 

 (Figure 5) gives strong evidence that one of these 

 assumptions was incorrect, probably the former, 

 as follows. First, only 3 of 144 sightings were made 

 at under 50 m range. The aircraft's configuration 

 which severly limits searching the water directly 

 under and adjacent to the flight path was the 

 major cause of this discrepancy. (A secondary fac- 

 tor was discomfort to the observer's necli caused by 

 attempting to look down at a steep angle.) The 

 only sightings made directly under or close to the 

 track were when the aircraft was in a steep turn, 

 and frequently herds were noted at moderate 

 ranges when we were circling on a previous sight- 

 ing. Secondly, the systematic decrease of the sight- 

 ing frequencies from 50 to 200 m, suggesting a 

 negative exponential curve, and the "tail" out to 

 400 m must at least in part reflect the inherent 

 inefficiency of the observers to see beneath the 

 water's surface at low angles or to detect relatively 



small, low-contrast objects at even moderate dis- 

 tances. 



Three factors, dolphin movement, nature of the 

 terrain, and observer experience, may have had 

 mixed effects on the estimates, as follows. 



Regarding effects of dolphin movement between 

 the open Gulf of Mexico and the bay behind the 

 barrier islands, it was originally planned that vol- 

 unteer observers stationed adjacent to the passes 

 would note the numbers and directional move- 

 ments of bottlenose dolphins during the hours of 

 the survey. However, a week's delay in starting 

 the field work and the subsequent resumption of 

 college classes following Easter vacation made it 

 necessary to cancel that observational phase. At 

 the termination of survey 4, however, we flew 

 homeward just outside Matagorda Peninsula and 

 Island. Outside Pass Cavallo at least 50 T. trun- 

 catiis were seen lolling in small herds in and just 

 outside the surf zone. These dolphins may have 

 either been moving in from the Gulf or out of the 

 bays, but their proximity to the beach and the pass 

 indicates that there was frequent movement of 

 dolphins between the two environments. 



Factors of bathymetry of the bays and the na- 

 ture of the terrain were not considered by the 

 analysis. While T. truncatus were occasionally 

 noted in shallow water just inside the barrier is- 

 lands, extensive regions in the middle of the bays 

 and in the shoreward areas were covered with a 

 thin layer of water over sand and mud flats and 

 there are numerous reefs and islands. Thus, 

 within most of the 800 m swaths used to compute 

 the density estimates there was territory that was 

 not available to the dolphins that could legiti- 

 mately be subtracted from the area searched. On 

 the other hand, by multiplying the searched area 

 by 5.76 (Equation (3)) to obtain an estimate of the 

 total number of dolphins we were sometimes at- 

 tributing dolphin habitat to dry land. This is par- 

 ticularly true for the Port Aransas ship channels 

 that were limited to about 600 m width and were 

 surrounded by large land areas. 



We feel that observer experience possibly also 

 biased the accounts. Tursiops triincati/s herds ap- 

 pear to occupy a home range (Caldwell 1955; 

 Shane 1977) and we frequently sighted herds that 

 were of similar size and in the same approximate 

 location of herds noted on previous surveys. The 

 observers tended to concentrate their attention on 

 these areas and thus searched them more 

 efficiently in the latter surveys. 



593 



