more frequently disturbed communities indicates greater disturbance frequency maintains the community at a 

 earlier stage of succession (Pearson and Rosenberg 1976; 1978; Dauer 1993; Trueblood et al. 1994; Weisberg et 

 al. 1997). If a commumt)' is disturbed every two weeks, it cannot proceed beyond the coloiuzation stage. Thus, 

 the length of time since the last disturbance regulates the communit)' (Connell 1978). In addition, communities 

 subjected to altered flow disturbance (i.e., increased or decreased) had higher abundance and biomass than 

 natural (control) communities (Fig. 7) explaining the frequency*flow interaction (Fig. 6). These results indicate 

 disturbance may play an important role in regulating macrobenthic community dynamics and increase 

 secondary' production as suggested by Rhoads et al. (1978). 



.\11 disturbance frequenc)' levels, including undisturbed ambient samples, were dominated by the polychaete 

 Streblospio benedicti . Streblospio benedicti is an opportunistic species (Grassle and Grassle 1974). 

 Dominance by opportunists is a key characteristic mdicating early succession, or highly disturbed, estuanne 

 macrobenthic communities (Pearson and Rosenberg 1976; 1978; Thistle 1981; Dauer 1993; McCook 1994; 

 Weisberg et al. 1997). Dominance of S. benedicti in the current study may indicate the community of station C 

 is highly disturbed by natural environmental variation, e.g., as the broad salinity ranges and flow conditions 

 foimd during this study (Fig. 4). 



Community abundance, biomass, and diversity of undisturbed sediments were lower than that of all disturbance 

 treatments after the first (7 May) biweekly samples (Fig. ). The undisturbed sediment communit)' is the 

 ambient, reference community against which coloni2ation of defaunated (disturbed) sediment was compared. 

 Increasing abundance, biomass, and diversity were expected in defaunated sediment until disturbed and 

 undisturbed communities were similar. In contrast, abundance of the bimonthly disttirbance frequency level 

 was 8 times that of the undisturbed community; biomass of bimonthly level was twice that of undisturbed 

 sediment; and diversity of bimonthly level was higher than that of undisturbed sediment. 



There are three possible explanations for the differences between undisturbed and defaunated communities. 

 The structure of experimental trays may attract organisms as a refuge or alter water flow affecting deposition 

 and recruitment (Butman 1986b; Snelgrove et al. 1993). The defaunated sediment may attract macrobenthos or 

 promote macrobenthic reproduction. Macrobenthic succession of defaunated sediment may be mitiated by a 

 poptJation burst of opportunistic species. The population burst may exceed late succession commumt)' 

 abundance and biomass, but return to normal levels after a period longer than 8 weeks. One, two, or all three 

 explanations may be responsible for the observed differences. 



Temporality 



Temporality, "the quality or state of being temporal (Mish 1985, p. 121 4)", is a property of communities that 

 arises in all studies from the complexity of ecological interactions in the natural environment. Communit)' 

 variation through time is not in itself a new finding. Temporalit)', however, is the unexplamed temporal 

 component of communit)' variation that may exceed that of experimental treatments. Examples of temporality 

 can be found in benthic communities of the Savin Hill mudflat, Boston Harbor (Trueblood et al. 1994), 

 microbial communities of the Parker River salt marsh, Rowley, MA (Montagna and Ruber 1980), and epifaunal 

 communities at Beaufort, NC (Sutherland and Karlson 1977; Holm et al. 1997). In all these cases, natural 

 commumt)' variation over rime (i.e., temporality) exceeded the effects of experimental mampulations. 



In the present study, collection date had the strongest effect on macrobenthic abundance (Figs. 7 and 8; 

 Tables 2 - 4), biomass (Figs. 7 and 8; Tables 1 - 4), diversity, and community structure (Fig. 9). Thus, natural 

 temporal variation of salinity, temperature, and dissolved oxygen (Fig. 4), which may be associated with 

 seasonality and flooding, played a greater role in determining community structure than flow or disturbance 

 frequency treatments (Tables 2-4). There were three community states through time (Fig. 9). The first group, 

 representing 23 April - 4 June 1 997, was dominant when average salinity varied between 1 1 and 1 8 %o, average 

 dissolved oxygen declined from 11.22 mg 1 ' to 3.89 mg 1 ', and temperature varied between 24.5 and 26.5 °C 

 (Fig. 4). The low abundance and biomass estimates of 4 June, compared with previous dates (Figs. 7 and 8) is 

 probably related to overnight hypoxia that likely occurred under low-flow conditions observed that day (Fig. 4). 

 After 4 June, the community state shifted because of increased abundances of chironomid larvae, Streblospio 

 benedicti and Laeoneris culveri (Fig. 9b). Streblospio benedicti appears to have had a recruitment event 

 between 4 June and 20 June when total average S. benedicti abundance increased from 1 1,000 to 94,000 m^. 

 The recruitment event appeared to be greatest under increased flow conditions. This indicates a possible flow 

 treatment and sampling date interaction whereby flow velocity may have affected active substrate selection, 



F-8 V Effects of Temporality, Disturbance Frequency and Water Flom 

 on an Upper Estuarine Macroinjauna Community 



