control data. This graph shows a large variation occurring in control and 

 experimental samples, and at corresponding times, both appear about equally 

 distant from the centroid — distance to maximum community stability. In other 

 words, control samples did not show close connections to the centroid, nor 

 did they follow a seasonal or any other discernible pattern in relation to 

 that point. Likewise, experimental samples showed no definite postconstruction 

 deviation from the centroid, and followed no subsequent trend that might have 

 indicated recovery. In fact, when respective sample distances from the centroid 

 were compared in a Mann-Whitney U-Test, it was found that variations among 

 control and experimental samples were statistically indistinguishable. The point 

 emphasized by this analysis is that faunal variation was a major feature of both 

 control and experimental samples. 



In the second graph, the y-axis scale (labeled distance to cluster edge) 

 refers to the edge of the statistical community (to a 95-percent confidence 

 level) that has the centroid as its midpoint (App. F) . The zero point on the 

 scale represents the nearest edge of the community, higher positive values are 

 increasing distances from the edge, and negative values show that the experimental 

 sample falls inside the cluster about the centroid and cannot be statistically 

 separated from it. Experimental samples along the x-axis are arranged by day 

 number in postdredging sequence. The x-y plots show that an experimental sample 

 first touched the edge of the centroid cluster on day 332 (5 July 1977) , about 

 11 months after dredging was completed at station 1. This intersection of an 

 experimental sample with the zero line represents time to faunal recovery. 

 However, in several later samples, the plot again falls outside the cluster edge, 

 and does not return until October, 14 months after dredging and 1 month before 

 sampling ended. This situation may be due to normal sample variation. 



VI. CONCLUSIONS AND DISCUSSION 



Study results indicate several general conclusions related to hydrology, 

 sediments, and benthic fauna of borrow pits and undredged adjacent bottom. 

 Hydrological measurements included temperature and salinity, recorded quarterly 

 at stations A and B in 1974 and 1975, and monthly at station 1 during a 20-month 

 period between April 1976 and November 1977. Temperature data showed that 

 regular seasonal changes are subject to rather wide year-to-year variations. 

 Summer temperature was the most consistent, but in spring, fall, and winter, 

 observed yearly differences were on the order of 10° Celsius. In part, 

 fluctuations of this magnitude could conceivably mediate events responsible 

 for changes in benthic diversity and abundance recorded in base-line, control, 

 and experimental samples. 



Salinity was characteristically high (above 32 parts per thousand); however, 

 a low value of 26 parts per thousand, recorded in August 1975, showed that the 

 study area may at times be influenced by estuarine water masses from St. Andrew 

 Bay and perhaps other areas as well (Salsman and Ciesluk, 1978) . As with 

 temperature, such periodic change could be translated into adjustments in 

 community structure. In the case of salinity, however, the effects might be 



27 



