station 1, the first experimental sample to equal or surpass control values of 

 H' and J' was collected in October, about 2 months after dredging. From that 

 time until November of the next year, only 5 of 14 experimental samples showed 

 evidence of faunal recovery. Recovery was demonstrated somewhat better by H' 

 and J' data from the six stations sampled in July 1977. At four borrow pit 

 stations, experimental samples had the same or higher diversity and equitability 

 values than control samples. Also, average H' for experimental samples was 

 higher than that for control samples, and averages of J' were the same inside and 

 outside borrow pits. 



A review of diversity and equitability results suggests the following: 

 (1) the benthos off Panama City Beach exhibited an annual cycle in which species 

 diversity and abundance were greater in warm water months than in winter; (2) 

 faunal recovery in the borrow pit at station 1 was evident to a considerable 

 degree within 2 to 3 months after dredging, and became nearly complete by the 

 end of sampling in November 1977; and (3) faunal recovery also occurred within 

 1 year of dredging in at least half of the six borrow pits sampled. To further 

 test these inferences, sets of biotic data from control and experimental samples 

 were evaluated using Morisita's index of faunal similarity and stability analyses. 

 Morisita's index was first used to develop similarity matrices (App, D) , and then 

 to perform a classification analysis that arranged control and experimental 

 samples in the form of a dendrogram according to their various degrees of likeness 

 (App. E) . Two stability analyses were made (App. F) . The first shows the amount 

 of sample variation among the control and experimental samples when compared to 

 the centroid of the statistical faunal cluster calculated from all base-line and 

 control data. The second shows time to faunal recovery by plotting experimental' 

 sample data against the nearest mathematical edge of the same statistical cluster. 



d, Morisita's Index . Similarity matrices were calculated and displayed for 

 time-sequence samples from station 1, and for one-time collections at stations 1 

 to 6 (App. D) . A regular pattern of light cells (no similarity) and dark cells 

 (high similarity) was not evident because 45 percent or more of station-to- 

 station comparisons in both values had faunal overlap of at least 50 percent. 

 For additional clarification, the same data were used to generate a classification 

 analysis for presentation as a cluster diagram (App. E) . In performing the 

 necessary calculations, a Q-mode (normal) analysis was made to show faunal 

 relationships on a station-to-station basis; no data transformations were made 

 because doing so would obscure the dominant ranking of any faunal elements in 

 the samples; and group averaging was selected as the sorting strategy. 



For time-sequence samples, the first five (1 September 1977-experimental to 

 10 August 1976-experimental) show very little similarity to any other samples 

 and were therefore considered unrelated, or outliers. These outliers include 

 two summer-fall experimental samples taken 1 year after dredging, two similar 

 winter collections taken about 6 months after dredging, and the first experimental 

 sample taken a few days after the dredging. The interpretation here is that the 

 two experimental samples 1 year after dredging are as unrelated to other samples 



25 



