to the total number of animals counted (9.3 ±2.0%) 

 is about the same as previously reported (Leath- 

 erwood 1979). 



Distribution and Behavior 



As can be seen from Figures 1-3 the distribution 

 of dolphin herds in the area was hardly homogen- 

 ous. Some 28 herds i'Zn ) of the total were sighted 

 in the narrow Aransas Pass ship channels (mainly 

 transect 1) and 211 (2.3'-; i of the animals counted 

 were in these herds. (This marked difference in 

 densities is discussed below.) Transect 18 across 

 Matagorda Bay was another area of high dolphin 

 density. While we noted only eight herds on this 

 line they were relatively large and accounted for 

 I49c of the dolphins sighted.. In general, aside from 

 the ship channels, the shoreward side of the bar- 

 rier islands and locations close to the beach ap- 

 peared to be favorable situations for T. truncatiis, 

 whereas sightings were rare in the middle of large 

 bays. 



When possible, the apparent behavior of the 

 herds was coded as either traveling, playing, feed- 

 ing, or resting. Of the 97 herds classified, about 

 half i48..5'^r ) were considered to be feeding. Side or 

 upside down swimming by dolphins actively pur- 

 suing prey as reported by Leatherwood ( 197.5) was 

 frequently observed. This was particularly true in 

 the shallow regions inside the barrier islands 

 where Gunter ( 1954) reported that bottlenose dol- 

 phins frequently chase mullet, Mitgil cephalus. 

 Feeding appeared to be associated with herd size, 

 for of the 17 herds composed of 15 or more indi- 

 viduals, 13 (76.59; ) were considered to be feeding. 

 The next most common behavior was "traveling," 

 and 36 herds (37.1'*) were assigned to that be- 

 havioral mode. 



Perpendicular Sighting Distances 

 and Sighting Cues 



As previously indicated, in most cases we esti- 

 mated the perpendicular distance from the plane's 

 track to the sighted herd. In addition we also log- 

 ged the nature of the observation which first 

 alerted us to the presence of a dolphin herd, the 

 "sighting cue" (Figure 5). During the field work. 

 11 different codes were used but these could be 

 reduced to four classes: 1) surface perturbations 

 such as mud trails or boils, scars, and splashes; 2) 

 an animal's body seen below the water (most eas- 

 ily noted when the dolphins are rolling or swim- 



jH suRFiCe PERTunemiONS. 12 5% 



I I ANIMAL BELOW SU«f«E. Z I 5% 



I 1 ANIMAL OR SLOW AflOvE SLIREACE, 58 5*4 



I I UNCERTAIN OR UNNOTED, 76% 



'-tuMm-i^-S 



ESTIMATED PERPENDICULAR DISTANCE: i i 

 SIGHTING FROM TRANSECT LINE (ml 



Fir.lfRp: 5 — Frequency distribution of estimated perpendicular 

 distances of bottlenose dolphin herd sightings from transect 

 lines on surveys 1-5, Port Aransas to Matagorda. Texas. Histo- 

 grams are divided into the relative ratios of sighting cue classes. 



ming upside down and their contrasting light ven- 

 tral surfaces are showing); 3) an animal's body, or 

 part of it, or its condensed respiratory exhalation 

 "blow" noted above the water surface; and 4) "cue 

 uncertain or unnoted." 



The "animal above surface" cue was effective at 

 all ranges and was the predominant sighting cue, 

 accounting for 58.3'r of all sightings (Figure 5). 

 The "animal below surface" instigated 21.,5'i of 

 the sightings, but was more important at ranges 

 under 200 m, contributing 28 of the 96 sightings 

 (29.2''7f ) at these ranges, whereas, at ranges -200 

 m, only 3 of 48 sightings (6.29r ) were signaled by 

 this cue. As will be discussed later, the effective- 

 ness of both underwater sightings and surface per- 

 turbations appeared to be vulnerable to weather 

 conditions. Most questionable or unrecorded 

 sighting cues occurred on the initial survey. 



DISCUSSION 



Possible Biases to Population Estimates 



Several factors, both operational and analytical, 

 influenced the results, in some cases prejudicing 

 the counts upward and in others to lowering them. 

 We first discuss two factors, effects of weather and 

 inability to sight all herds, that tended to cause 

 underestimates. 



Relatively strong southwest winds (22-41 km/h) 

 blew constantly for several days during the field 

 operations. The wind's major effect on searching 

 efficiency was not sea state, as is the case in the 

 open ocean, for splashes were seldom the sighting 

 cue, but rather the stirring of bottom materials 

 into suspension creating large areas of highly tur- 



592 



