IRVINE ET AL: AERIAL SURVEYS FOR MANATEES AND DOLPHINS 



Multiple Range Test (Steel an d Torrie 1960). A 

 square root transformation (\J herdsize + 0.5) 

 was applied to the counts to make them suitable 

 for parametric analysis (Steel and Torrie 1960). 

 Computations were performed with programs of 

 the Statistical Analysis System (Helwig and 

 Council 1979) at the University of Florida, 

 Gainesville, Fla. 



RESULTS 



Manatees 



Two hundred and ninety-seven groups of 

 manatees, totaling 554 individuals, were ob- 

 served during 121.8 survey hours (Fig. 1). Num- 

 bers sighted (Table 1) and average number of 

 individuals per groups (Table 2) varied by 

 county and month. Total numbers of manatees 

 sighted increased from September to November, 

 but the total per county consistently increased 

 only in Monroe County. Total counts were not 

 statistically compared among counties because 

 habitat type, weather, and amount of survey area 

 were not equivalent. 



Ninety-four percent of the groups sighted con- 

 sisted of one to four animals (Fig. 2). Group sizes 

 were not observed with equal frequency, and 

 more than half of the 297 sightings were of single 

 animals (P<0.005; chi-square). However, 367 

 (66.2%) of the 554 manatees sighted were in 

 groups. Pooled samples of all counties indicated 

 that group size-frequency distributions did not 

 vary significantly between months (F>0.80; chi- 

 square). 



Mean group size for the pooled sample of all 

 sightings was 1.9 (SE = 0.12). A subset of data, 

 including only those counties with sightings in 



each month (Monroe, Lee, and Sarasota Coun- 

 ties ), was analyzed as a two-way ANOVA. This 

 analysis provided no evidence of a month by 

 county interaction (P>0.85), indicating that any 

 pattern of monthly variation in group size was 

 comparable for those three counties. Monthly 

 variation in average group size, analyzed as a 

 separate one-way ANOVA for each county, was 

 significant (P<0.05) only in Hillsborough County, 

 due to high December counts at warmwater 

 effluents. 



Numbers of manatees sighted were not pro- 

 portional to the amount of survey time in each 

 habitat type or salinity (F<0.005; chi-square). 

 Pooled samples from all counties indicated that 

 numbers of manatees sighted per month varied 

 significantly by salinity and habitat (P<0.0005; 

 chi-square). Except in December, substantially 

 more manatees were sighted in marsh-river hab- 

 itats than in other habitat types, and most were 

 in brackish water (Table 3). 



From 51 to 100 manatees, representing 54.3 to 

 75.7% of those sighted on July to November sur- 

 veys and 58.5% overall, were observed in Monroe 

 and Collier Counties (Table 1). Manatees were 

 consistently sighted in Whitewater Bay, Cheva- 

 lier Bay, and in the Lopez River (ENP, Monroe 

 County), but the largest concentrations were 

 found in Collier County from Marco Island to 

 Chokoloskee. Manatees may have been over- 

 looked if they were not near the surface or 

 creating surface wakes or mud trails because of 

 water turbidity (estimated visibility 0-0.5 m). 



A maximum of two manatees per survey was 

 sighted in Charlotte Harbor (Charlotte County; 

 Fig. 1 ) and small numbers of manatees were con- 

 sistently sighted in Pine Island Sound, Matlacha 

 Pass, San Carlos Bay, in the lower reaches of the 



Table 1.— Numbers of manatees and bottlenose dolphins observed, by county, during aerial surveys in 

 western peninsular Florida from July to December 1979. C = calves. 



'Incomplete survey 

 2 Not surveyed 



623 



