Water

Publications

Importance of Flood Flows to the Productivity of Dryland Rivers and Their Floodplains

Final Report
Prof P.M. Davies, Prof S. E. Bunn and Ms F. Balcombe
Environment Australia, 2003

Controlled Experiment Simulating Floodplain Inundation

A controlled experiment was conducted to determine the rates of primary productivity and ecosystem respiration associated with inundation of floodplain soils. The experiment was conducted on the banks of Cooper Creek and designed to simulate the effect of a flood. Previously-dry floodplain soils were collected in ten replicate large plastic containers and inundated with water from the creek in a flow-through system (Figures 7 and 8).

Figure 7

Figure 7: The equipment used prior to deployment. The Nally bins contained floodplain soils and water was pumped to swimming pools and trickled through the bins. The perspex domes enclosed the substrate within the bins. The data loggers (blue next to the bins) recorded dissolved oxygen at 10 minute intervals, pumps re-circulated water within the domes and voltage controllers attached to 6v batteries ensured flow saturation across sensor membranes without unduly disturbing the substrate.

Figure 8a Figure 8b

Figure 8: The production rate experiment in progress

Measurements of benthic metabolism were made on these soils over three periods; two days after inundation (Time 1), five days after inundation (Time 2) and after eight days (Time 3). Rates of gross primary production (GPP) were initially low at about 250 mg.C.m2.day-1, at the end of the experiment, mean rates had doubled to over 510 mg.C.m2.day-1 (Figure 9). Rates of ecosystem respiration (R24) showed a similar pattern to GPP, indicating substantial photo-respiration (Figure 10).]

Figure 9: Mean (SEM) rates of benthic GPP after the three measurement periods (n=7 each time period).

Figure 9

Statistical differences were T3>T2>T1 (ANOVA on log10 transformed data, p<0.05, all cases).

Figure 10: Mean (SEM) rates of benthic ecosystem respiration (R24) after the three measurement periods (n=7 each time period).

Figure 10

Statistical differences were T3>T2>T1 (ANOVA on log10 transformed data, p<0.05, all cases).

During the flood of March/April 2000, rates of GPP were highly patchy; changing substantially in space and time. After over two weeks of inundation, mean rates on the floodplain were about double that measured in the experiment. This suggests the temporal component of the flooding is an important determinant of overall GPP and probably overrides the effects of turbidity associated with depth.

Key

   Links to another web site
   Opens a pop-up window