Laserfiche WebLink
bags and sediment deposited on dte forest floor above the <br />diversion dam. The procedure ensured collection of neazly <br />100% of all sediment passing through treatments. The <br />treatments were then compazed on the basis of the total <br />exported sediment. <br />Inflow and outflow runoff concentration data were <br />analyzed in a paired sample design. Mean reductions <br />between inflow and outflow concentrations were determined <br />and used in analysis of vaziance (ANOVA). The hypothesis <br />for this comparison was that treatments were equal in <br />capacity [o fdter sediment-laden runoff. Mean reductions in <br />runoff wncen[mtion and sediment yield were used as <br />dependent variables for investigation. Independent variables <br />considered in the ANOVA were treatment method, road <br />grade, and mnoff volume. Where ANOVA indicated signifi- <br />cant differences, individual treatment means were compazed <br />using Duncan's Multiple Range Test. <br />RESULTS <br />Fifty-three sampling events were monitored during this <br />initial phase of the study, which covered a period from <br />28 August 1997 to 27 December 1998. A sampling event <br />consisted of an individual storm. Storm intensities ranged <br />from 1.0 to 31.0 mm/hour with an average intensity of <br />4.4 mrn/hour during the lCrmonth study period (fig. 3). <br />Inflow and outflow mnoff concentrations for each associated <br />erosion control treatment were collected for each rain event. <br />Sedunent deposifion was measured on four occasions during <br />the study period by collecting sediment bags and measuring <br />sediment deposited on the forest floor. <br />140 <br />120 <br />~ 100 <br />~ 80 <br />60 <br />.Q <br />.d 40 <br />a <br />20 <br />RUNOFF CONCENTRATION REDUCTIONS <br />The reductions in runoff concentration deviated from <br />normality and were normalized with a logarithmic trans- <br />formation. ANOVA detected significant treatment effects <br />(p> 0.001) in the concentration reductions. No signif-scant <br />differences were detected in concentration reductions be- <br />tween vegetation, sediment fence, and settling basin treat- <br />ments (table 2). Settling basin and rip-rap treatments were <br />statistically similaz in capacity to filter sediment laden <br />mnoff. <br />The settling basin and rip-rap treatments were less <br />effective than other treatments a[ filtering sediment during <br />large storm events (>40 tnm). Runoff during these storms <br />exceeded the design capacity for the settling basins and <br />flushed trapped sediment from the rip-rap treatment. The <br />settling basin was 100% effective on all but 10 events, but <br />yielded a mean reduction of only 85 over the study period. <br />TOTAL ERPORTED SEDA~NT <br />Sediment export was expected to be least below treat- <br />ments with the highest reductions in runoff concentrations. <br />No significant treatment differences in mean sedunent export <br />or runoff volume were found during ttte 16-month study <br />period (table 3). Sediment export for vegetation, sediment <br />fence, rip-rap, and settling basin treatments were 43.4, 43.0, <br />30.1, and 23.0 kg, respectively. <br />In an attempt to obtain a better understanding of the <br />sediment export data, a detailed look was taken at the two <br />components measured: sediment deposited on the forest <br />floor, and sediment collected in sediment bags. On average <br />for the four treatments, sediment deposited on the forest floor <br />accounted for more than 93% of total sediment export for all <br />four treatments. Sediment collected in sediment bags <br />^ 5 4, ^^ ^e ^a ,lry tih ti0 ,~^ ~°' ~~ eP e'? e~ a9 bti <br />Storm Event # <br />~ Precipitation ®Stonn Intensity <br />Figure 3. Oluerved predpifadon and intensity at 1Lskegee site during study period. <br />0 <br />5 ~ <br />t <br />10 £ <br />15 ~ <br />20 M <br />C <br />25 <br />30 E <br />35 y <br />40 <br />TRANSACRONS OP'ni8 ASAE <br />