260 



Fishery Bulletin 100(2) 



SCDNR personnel were trained to identify charac- 

 teristic signs of human interaction, such as attached 

 gear, rope or line marks, net marks, propeller cuts, 

 straight-edge knife cuts, puncture wounds, etc. Ani- 

 mals with any of these characteristics were consid- 

 ered to be positive for human interaction. An as- 

 sessment of stomach contents (full or empty), and 

 histopathology (when available) to determine ante- 

 or post-mortem injuries were used to corroborate 

 findings. Photographs, and both necropsy and pa- 

 thology reports from the United States Ai-med Forc- 

 es Institute of Pathology (AFIP, Washington, D.C.) 

 also were used to confirm these reports. Those ani- 

 mals that did not show any human interaction char- 

 acteristics determined with photographs, necropsy, 

 or AFIP pathology reports were considered to have 

 no human interaction. For animals that were too 

 decomposed, not fully examined, examined by un- 

 trained personnel, for whom there were inconclu- 

 sive findings from necropsy or pathology reports, 

 human interactions could not be determined. There- 

 fore, each stranded bottlenose dolphin was classified 

 into one of three categories to determine if human 

 interaction was a possible cause, or contributing fac- 

 tor, in the death of the animals: 1 ) positive human 

 interaction, 2) no human interaction, and 3) human inter- 

 action could not be determined (CBD). 



Trends in the stranding data related to year, month, sea- 

 son, gender, and two specific age classes: neonate and fe- 

 males >220 cm, were investigated by using chi-square (X" > 

 goodness-of-fit and analysis-of-variance (AN OVA) meth- 

 ods. A chi-square test for equal proportions was used to de- 

 termine if there were differences in the number of strand- 

 ings between years. An ANOVA was used to determine 

 if there were differences in the number of strandings be- 

 tween months. To determine if there were seasonal trends 

 in the strandings, the data were stratified into gi'oups of 

 three months representing four seasons: January-March 

 (winter). April-June (spring). July-September (summer), 

 and October-December (fall) (Fig. 2). Expected number 

 of strandings for each season was determined by aver- 

 aging over the 5-year period. To determine whether the 

 stranding pattern for any given year deviated significantly 

 from the "norm," we compared each years seasonal num- 

 ber of strandings with the expected seasonal number by 

 using a chi-square goodness-of-fit test. An ANOVA was 

 performed to determine if there was a difference in the 

 number of strandings between seasons. A chi-square good- 

 ness-of-fit test was used to determine seasonal trends be- 

 tween zones. 



A chi-square goodness-of-fit test was used to determine 

 if there was any difference in the proportion of male and 

 female bottlenose dolphin strandings. A chi-square test for 

 trend was used to test the hypothesis that there would be 

 a downward trend in the number of animals that were of 

 unknown sex due to increased training of stranding net- 

 work volunteers in determining the sex of bottlenose dol- 

 phins. A chi-square goodness-of-fit test was used to deter- 

 mine if there was a difference in the number of strandings 

 of females >220 cm between seasons. 



25 



20 



22 



ro 15 



E 10 



E 



12 



Jan Feb Mar 

 Winter 



Apr May Jun 

 Spring 



Months 



Jul Aug Sep Oct Nov Dec 

 Summer Fall 



Figure 2 



Number of bottlenose dolphin strandings in South Carolina for 

 each month and season of the year from 1992 to 1996 (n = 153). 

 The numbers over each bar represent the number of strandings 

 reported for that month. 



Results 



Yearly trends 



From 1992 to 1996, 153 bottlenose dolphin strandings were 

 reported along the coast of South Carolina. The number of 

 strandings each year ranged from a low of 28 in 1992 to 33 

 in 1993 (.v=30.6) (Table 1); there was no significant differ- 

 ence among years (X" test for equal proportions, P=0.968). 

 Prior to 1992 the highest number of bottlenose dolphin 

 reported stranded for one year was 17 in 1991 (the year 

 the network was formed), excluding the unusual mortality 

 eventof 1987 1/! =60). 



Monthly trends 



Over the five-year period, the greatest number of reports ( 22, 

 or 1AA'''< ) of bottlenose dolphin strandings occuiTed during 

 July and the least in January (/;=2) and October (/!=3) (Fig. 

 2). Tliere was no yearly differences in the total number of 

 strandings by month fi-om 1992 to 1996 (ANOVA, P=0.172). 



Seasonal trends 



The highest number of strandings occurred in spring 

 («=53, ZA.&^ I and the lowest number of strandings were 

 recorded in winter (/;=26, 17.0'^^f ). Strandings during the 

 years 1992, 1994. 1995. and 1996 did not deviate from 

 the expected pattern (x^ goodness-of-fit; P=0.994. 0452, 

 0.379, and 0.062, respectively), but the seasonal pattern in 

 1993 was significantly different (P=0.016). This was due in 

 large part to the high number of strandings (/( = 14) in the 

 fall, when we expected the number to be less than seven. 



The ANOVA analysis indicated that the mean number of 

 strandings differs significantly between seasons (P=0.021 ). 



