The flow tieatment tests are generally valid because of non-significant interactions with date, except for the 

 undisturbed samples (Table 3). No sigmficant abundance differences were found among flow treatments. The 

 biomass differences among flow treatments for biweekly and monthly levels (Fig. 6.) were significant (Table 3). 

 At the biweekly and monthly frequency levels, decreased flow yielded significandy greater biomass than control 

 and increased flow levels (Tukey test). Abundance changed over collection dates for all frequency levels. 

 Biomass changed over the long-term, but over short-term (biweekly) time intervals. 



The flow*date interaction was significant for abundance and biomass for the undisturbed treatment (Table 3). 

 This interaction is especially interesting because it represents how undisturbed sediment changed over time as a 

 result of the flow treatments alone (Fig. 7). In undisturbed sediment, highest abundance and biomass values 

 were found within structures that altered flow. Except for the first collection date, the abundance and biomass 

 in the control plots were always lower than the altered flow plots indicating a possible structure effect. So, the 

 interaction was due to the first collection date (when all flow levels were similar) and the similarity' of flow 

 alteration structures (which alternated between highest and second highest values). 



Frequency Effects 



Disturbance frequency tests were affected by significant frequency*date interaction effects, except in decreased 

 flow structures ^able 4). At the decreased flow level, average biomass of monthly (1.20 g m "), bimonthly 

 (1.15 g m^, and weekly (0.90 g m^ frequencies were similar to each other, but were significandy different from 

 undisturbed average biomass (0.69 g m ■^. 



Effects of the disturbance frequency experiment on the natural communit)' is represented by changes in the 

 undisturbed sediment communit)' in control level of flow treatments. There was a significant frequenc)'*date 

 interaction for control flow levels for both abundance and biomass (Table 4). Average abundance and biomass 

 was lower in undisturbed sediment than in all disturbance treatment levels except for biweekly samples 

 collected on 07 May (Fig. 8). The lower abundance and biomass in undisturbed samples may indicate 

 disturbance enhances community productivity'. On most dates, average abundance and biomass in bimonthly- 

 frequency, control-flow, samples were greater than other disturbance frequency samples. In general, there was 

 a trend of increasing abundance and biomass from biweekly, to monthly, to bimonthly distixrbance frequencies, 

 indicating growth or recruitment with time. 



Community Structure 



Ten taxa were found during the experiment. Streblospio benedicti and Laeoneris culveri were present on each 

 sampling date. Chironomid larvae were observed 23 April and 7 May, but appeared in abundance 20 June, 

 persisting through 1 August. Polydora comuta Bosc was present between 4 June and 2 July, but was most 

 abundant on 20 June. Polydora comuta was observed only once in 31 observations at the control- flow plot 

 indicating a possible positive structure effect. Mysidopsis aknyra Bowman was present between 23 April and 

 22 May, and was found only in undisturbed samples. Nermerteans were found on 20 June and 02 July in 

 undisturbed and monthly samples. Mulinia laterahs (Say) was foimd infrequendy between 23 April and 20 June 

 in undisturbed sediment of decreased and control flow treatments. Mediomastus ambiseta (Hartman) was 

 fotmd infrequendy between 4 June and 1 August. Amphipods and oligochaetes were each present in a single 

 sample during the course of the study. 



Average abundance and biomass were greatest for samples collected on 20 June 1997 (Figs. 7 and 8). Average 

 abundance was 422,000 m'^ and average biomass was 5.62 g m'^. Streblospio benedicti accounted for 96 % of 

 abundance and 76 % of biomass in these samples on average. The bimonthly- frequenc}', increased-flow plots 

 contained the greatest average abundance and biomass of all samples in the study. For example, one sample 

 contained 4,730 organisms (1,342,000 m ■^, of which 4,667 (1,324,000 m "') were S. benedicti . Many S. benedicti 

 were very small, presumably lecithotrophic larvae or post larval juveniles, but sizes were not measured. 



Principal component analysis (PCA) yielded three distinct collection date groups (Fig. 9a), but no flow or 

 disturbance frequency effects. PCI separated samples collected 20 June and 02 July from those collected 

 17 July and 01 August. PC2 separated samples collected 23 April - 04 June from those collected 20 June - 

 01 August. Chironomid larvae (CL) loaded 0.97 on PC2 (Fig. 9b) indicating their importance to characterizing 

 communities samipled 20 June and later. Streblospio benedicti and Laeoneris culveri loaded positively for both 



F-6 V Effects of Temporality, Disturbance Fnquengi and WaterFbw 

 on an Upper Estuarine Macroinfauna Community 



