Though the increase does not represent an interannual doubling of the 

 population, the change was significant, based on comparisons of 95% confidence 

 limits (Figure 8). The increase was evident through all four abundance estimation 

 methods, and it ran counter to the patterns of consistency across years demonstrated 

 for Tampa Bay and Sarasota (Wells et al. 1995; Wells and Scott 1990). Our 

 evaluation approach was to first examine corroborative indicators of the change, 

 and then to test hypotheses about the possible biological or methodological source(s) 

 of the increase. 



The apparent increase in numbers of dolphins during 1993-1994 was 

 corroborated by changes in the number of dolphins sighted per unit of sighting 

 effort. For this analysis, we divided the sum of the final best point estimates of 

 numbers of dolphins for each sighting for each year by the number of kilometers of 

 survey transects for that year. This density indicator should be less prone to 

 potential biases that might have resulted from violations of mark-recapture 

 assumptions. The number of dolphins per km increased by 14% from 1991 through 

 1993 and 1994 (Table 7). This measure provided additional supportive evidence of 

 an increase in the numbers of dolphins in Charlotte Harbor. We hypothesized three 

 potential biological sources of dolphins to account for the increase: (1) through 

 recruitment of young, (2) through an influx of new dolphins, and /or (3) from the 

 return of previously identified individuals. 



If the increase was due to recruitment of young, then several expectations 

 follow. If we assume that Charlotte Harbor is a relatively closed population unit, 

 and the entire increase resulted from reproduction, then the number of young-of- 

 the-year during a given year should be greater than or equal to the change in 

 abundance from the previous year. As can be seen from Table 5, production of 

 young was nearly 2.5 times greater in 1993 than in 1990. At no time, however, does 

 reproduction during one year entirely account for abundance increases in the next 

 year. 



If recruitment of young accounted for some, but not necessarily all, of the 

 apparent abundance increase, then the proportion of marked animals (m/n for 

 Method 2, Table 3) should decline over the years, since identifying marks tend to be 

 acquired with age, and calves tend to be less marked than older animals. The 

 accumulation of young-of-the-year from several years of increased reproductive 

 output should be reflected in increased numbers of unmarked calves and juveniles 

 in later years. The proportion m/n did in fact decline, from 0.80 in 1990, to 0.58 in 

 1994, suggesting a dilution of the pool of marked animals by young, as-yet 

 unmarked individuals. 



Any increase indicated from mark-recapture analyses that is due to 

 recruitment of young, should be expected to be reflected by other indicators that are 

 not based on marked animals. Increases in numbers of young-of-the-year should 

 result in subsequent increases in calves. The number of young-of-the-year per 

 kilometer of survey transect tripled from 1990 through 1991, 1992, and 1993 (Table 7). 



