sampling periods which were similar 

 in species composition and abundance. 

 These analyses also identified spe- 

 cies groups which were similar in 

 distribution temporally. 



Preliminary analysis of the ben- 

 thic macroinvertebrate data indicated 

 that several of the dominant species 

 were relatively ubiquitous. There- 

 fore, a classification technique 

 which was unbiased toward species 

 dominance and yet included both 

 quantitative and qualitative infor- 

 mation seemed desirable to employ. 

 The Canberra-Metric similarity mea- 

 sure of Lance and Williams (1967) 

 was employed to determine similari- 

 ty between the entities of sampling 

 period and infaunal species. 



When choosing species to be in- 

 cluded in the analysis, we arbitrar- 

 ily chose those whose total abundance 

 were greater than 30 individuals; 

 this criterion alleviated the very 

 infrequently taken species whose low 

 abundances would have contributed 

 very little to the overall analyses. 

 This follows Day et al. (1971) who 

 found that results of similar analy- 

 ses were not reliable using only rare 

 species and that similar results were 

 obtained using data with and without 

 the rare species. Data standardiza- 

 tion needs were considered by check- 

 ing the normality of density distri- 

 butions for a number of the dominant 

 species. The skewness and kurtosis 

 were found to be relatively constant 

 among samples for these species, thus 

 allowing for the use of raw data 

 scores in the cluster analyses 

 (Scheefe 1959; Downing 1979). 



RESULTS 



PRE-EVENT TRENDS 



The most dominant recurring pat- 

 tern observed during the pre-event 

 study of benthic infauna was that 



there was a major difference in com- 

 munity structure between the channel 

 stations, 1 to 3, and the shallow wa- 

 ter stations, 4 to 6 (Figure 1). 

 These differences were primarily re- 

 lated to variations in sediment 

 structure plus the perturbations of 

 periodic dredging and constant 

 shrimping and shipping activities in 

 the channel which were not present in 

 the shallower waters ( < 5m) . There- 

 fore, for our continued observations 

 of the benthos following the fresh- 

 water inflow event of September 1979, 

 we chose to sample both a channel and 

 shallow water station (1 and 4 res- 

 pectively) to adequately document any 

 changes in benthic infauna related to 

 the event. 



There were several trends in the 

 historical benthic data that sug- 

 gested we might expect changes to oc- 

 cur in the ecosystem as a result of 

 a period of intensive freshwater in- 

 flow. Several significant correla- 

 tions existed between both general 

 community variables as well as in- 

 dividual species densities and the 

 environmental variable of salinity. 

 For example, total infaunal density 

 for the historical data base was 

 negatively correlated with salinity 

 (r = -0.34, p< 0.001, n = 159), sug- 

 gesting that community densities in- 

 creased when salinity decreased. A 

 polychaete population, Mediomastus 

 californiensis , displayed a negative 

 correlation (r = -0.25, p<0.002, n = 

 159) with salinity. In addition, 

 three bivalves, Mysella planulata , 

 Mulinia lateralis , and Abra aequalis , 

 all showed significant negative cor- 

 relations of -0.43, -0.28, -0.32, re- 

 spectively, with salinity (p < 0.001, 

 n = 159). 



Besides the significant corre- 

 lations with salinity cited above 

 the benthic community variables of 

 infaunal species number and total 

 density showed some interesting 

 trends in relation to salinity 



492 





