significant difference in stranding rates was detected among years; however, the relatively 

 low power of the ANOVA test suggested some undetected differences among years. Using 

 the upper 95% confidence interval on the 1986-1990 monthly stranding means as a criterion 

 for detecting significant differences, the number of strandings during January-March 1990 

 was significantly greater than during the preceding four years. 



The slope of monthly stranding rates regressed against Texas coastal monthly mean 

 sea surface temperature was significant; however, a low correlation coefficient suggested that 

 there was no simple linear relationship. A negative exponential relationship between first 

 semester stranding rates and the preceding December-January mean sea surface 

 temperature was detected; the lowest January-December mean sea surface temperatures 

 preceded the highest January- June stranding incidence. A weak relationship between 

 dolphin strandings and air temperature for the same period was not significant. Sea surface 

 temperature anomaly data from NOAA's Oceanographic Monthly Summary for December 

 1989- January 1990 ranged from -04 to -2.0 °C The persistence of the negative anomaly 

 throughout the winter of 1989-90 suggested abnormally low sea surface temperatures. Mean 

 salinity varied significantly among months, among years, and among months within years; 

 however, there was no significant relationship between bottlenose dolphin stranding and 

 salinity. Monthly mean offshore transport varied significantly among years, but there was 

 no apparent relationship between monthly mean offishore transport and monthly mean 

 bottlenose dolphin stranding. Examination of seasonal stranding and offishore transport 

 suggested a weak, but significant, inverse relationship during the spring. Other than an 

 inverse relationship between winter sea surface temperatures and spring dolphin mortality 

 rate, these analyses detected no strong significant relationships between bottlenose dolphin 

 strandings and other enviromnental variables. 



The association of low winter sea surface temperatures with an increased spring-time, 

 dolphin stranding rate suggests t ■- possibility of thermaUy-induced stress. Alternatively, the 

 observed association between v, .er sea surface temperature and spring bottlenose dolphin 

 stranding rates may be less dirt -. An estimated 2.7 million fish, of which approximately 

 2.6 million were striped mullet . Jugil cephalus), died in East Matagorda Bay after a severe 

 cold spell in December 1989. ^nd smaller kills occurred in Texas Bays from Sabine to 

 Laguna Madre Bay. It is possible that bottlenose dolphins were forced to switch to inferior 

 prey items because of unusual weather-related fish migration patterns or fish mortalities. 

 However, the available food habits study results do not indicate M. cephalus as a major prey 

 item in stranded dolphin stomachs. 



The inverse relationship between spring bottlenose dolphin stranding and offshore 

 currents (Ekman transport), although not strong, may contribute to an increase in beach-cast 

 mortalities in the spring and thus, an apparent increase in mortality rate. An increased 

 nearshore occurrence of bottlenose dolphins, with normal mortality rates, during the spring 

 season could also contribute to an apparent increase in mortality. 



