|
<br />FLUCTUATING FLOWS A.T\lD BACKWATER HABITATS
<br />
<br />1131
<br />
<br />weighted by the previous backwater volume and the inflow volume (the volume of water flowing into the backwater
<br />during the current time step):
<br />
<br />T.b"i = [(VI] ' Tb,I-I) + T,I1,1 ' (VI - V{w dJ
<br />VI
<br />
<br />(2)
<br />
<br />(If VI :::; VI_ J, this step was skipped and we proceeded to the next step, a consideration of the effects of weather on
<br />backwater temperature.)
<br />Next, the water temperature model calculated the hourly net heat flux from Tb,i and the current hour's weather
<br />data, Then the updated backwater temperature was determined by adding the temperature change resulting from the
<br />net heat flux to Tb,i' To avoid the complexity of modelling ice, we assumed that backwater temperature never falls
<br />to 0, This assumption is reasonable for the time period of interest (July through 31 December 1994), Daily mean,
<br />minimum and maximum temperatures were calculated at midnight fTOm the day's hourly values, These daily
<br />temperature variables were used to model invertebrate production.
<br />
<br />Invertebrate production and .food availability
<br />
<br />The food availability model was developed as a very simple representation of how mainstem flow fluctuations
<br />affect the availability of food for fish inhabiting backwaters. We first modelled production of invertebrate food, then
<br />the loss of food resulting from water exchange with the mainstem and, finally, the density of invertebrate food.
<br />Given the absence of site-specific information on invertebrate communities and production, we combined the
<br />empirical approach of Morin and Dumont (1994) and the experimental data of Morin and Dumont (1994) and Maier
<br />et aI, (990) to simulate daily invertebrate production, For simplicity, we used chironomid larvae (Order Diptera),
<br />the most abundant member of the Green River benthic invertebrate community (Grabowski and Hiebert, 1989) to
<br />represent all invertebrate prey, both benthic and planktonic, This simplification is justified by the observations that
<br />(1) the stomach contents of juvenile pikeminnow (20-40 mm TL) are typically composed of 70-97% benthic
<br />invertebrates by volume (Grabowski and Hiebert, 1989; Muth and Snyder, 1995), (2) chironomids constituted a
<br />large proportion of those benthic invertebrates found in juvenile pikeminnow stomachs (68-84% by volume;
<br />calculated from Muth and Snyder, 1995) and (3) only the very smallest pikeminnow (<20mm TL) consume
<br />significant quantities of zooplankton (0-40% of stomach contents by volume; Grabowski and Hiebert, 1989). Our
<br />approach combined (1) an instantaneous growth model with (2) the simple assumption that mean invertebrate
<br />biomass density (D, g . m2) remained constant and equal to the mean annual invertebrate biomass for the study
<br />reach (i,e. backwaters and mainstem habitats).
<br />We modelled the daily instantaneous specific growth rate of an individual chironomid larva (G, g' g-l) as a
<br />power function of the average daily temperature in the backwater (T, oc) and mean invertebrate size (M, g):
<br />
<br />1 6E 6 1({),26(),f). (,{)'(X)6,f~)j
<br />G= ' - ,e'\ "
<br />M
<br />
<br />(3)
<br />
<br />This equation was created by fitting the combined data of Morin and Dumont (1994) and Maier et aI, (1990)
<br />using nonlinear regression, With this equation, invertebrate growth rate is positive at temperatures greater than or
<br />equal to O('C, increasing to a maximum at 230C and decreasing as water temperature increases further.
<br />Daily food production (P, g' m-:'.) was then calculated as by Morin and Dumont (1994):
<br />
<br />D.G
<br />P=-
<br />2
<br />
<br />(4)
<br />
<br />In our simulations, we estimated D by multiplying the average density of invertebrates observed in Green River
<br />backwaters throughout the year (15, prey, mm2) by the average mass of a chironomid larva (M, g).
<br />We assumed that all of the daily invertebrate production was available for consumption by fish unless
<br />invertebrates are removed as the backwater drains into the mainstem, We did not model inflow ofinvertebrates from
<br />the main stem because invertebrate biomass and production are typically higher in backwaters than in the main
<br />channel of the Green River (Grabowski and Hiebelt, 1989), We assumed that the proportion of inveltebrates
<br />
<br />Copyright (C, 2006 John Wiley & Sons, Ltd,
<br />
<br />Ri vcr Res, Applic. 22: 1125-1142 (2006)
<br />DOl: IO,1002lrra
<br />
|