Laserfiche WebLink
<br />LR <br /> <br />Fish food organism and water auality evaluation <br />Weekly sampling will be implemented to establish the quantity (densities and biomass) and diversity of <br />macro invertebrates occurring in natural and application floodplain wetlands and adjacent river reaches. <br />Macroinvertebrate production will be measured using surber samples, kich samples, and benthic cores. <br />During periods of water flow between inundated areas and the river, drift nets. will be used to estimate <br />transfers of particulate organic matter and macroinvertebrates. In addit,on, water quality parameters that <br />affect these organisms will be measured from weekly water samples using an autoanalyzer, and other water <br />quality parameters will be measured using Hydrolabs. Water samples will be collected at outflow areas of <br />floodplains and analyzed for nutrient concentrations to determine influx concentrations. Response <br />variables to be measured for fish food organisms and water quality evaluation are as follows: quantity of <br />organisms, specics and size composition, and water quality variables (dissolved oxygen, etc.) <br /> <br />Rinarian reSDonses and effects <br />This component will be combined with the fish food and water quality crew. Their duties include: <br />physical mapping, including vegetation. Response variables to be measure for riparian responses and <br />effects are as follows: vegetation composition, density, plant material input, and subsequent changes <br />following levee removal. <br /> <br />Understanding how vegetation within flooded Dottomlands progresses from tamarisk to cottonwoodJwillow <br />and vise versa is important. Because riparian and aquatic vegetation are known to playa critical role in <br />macroinvertebrate production and in providing refuges for small fish from predators, vegetation <br />composition and subsequent changes following levee removal will be enumerated using transect samples <br />obtained during invertebrate sampling. <br /> <br />Manipulated sites <br />Old Charley Wash was used as a manipulated site in FY1996. Response variables measured were as <br />follows: gear type effectiveness, relative abundance, cohort analysis between years, reproduction, relative <br />abundance and composition offish food organisms, food habits, and growth rates for fishes within the river <br />reaches associated with manipulated sites. No manipulated sites ~.~ scheduled for experimentation in <br />FY1997. If levee removal evaluation creates more questions that can be addressed in outyears (e.g., <br />FY 1998) with manipulated sites, we envision their inclusion once again. <br /> <br />Wj.YS..!J!1~tS~g~~<!.<\ilill~'jjIt.w~;a:desc~'sl\!l'..2l~~!.~,@.!Im~e4.19n:s!l<;:;adai9~i9~Q~~~ <br />!pp~!iCjii!Site~~),l[~~$2lliP.e.e;seas-@l!i~~.s.@"n!@.1:<W!eCiSio~<!f.i\be E,~~~]{~~Y"iLffii!ioJjl!,3~uru <br /> <br />Slatistical Desi~n and Data Analvsis: <br /> <br />Data will be dealt with in a series of a priori comparisons. First, linear combinations of variables will be <br />produced from those collected. These will ultimately become the response variables using canonical <br />correlation analyses or factor analyses, depending on the linearity of the system. Water quality and <br />macroinvertebrates will be put into 2-3 linear combinations to reflect productivity. Nonnative fish <br />responses will also be placed into 1-3 components depending on the densities, the number of species <br />collected and the range of age classes. Native fishes will probably be left in their raw fonn since the <br />numbers collected will be sman (at least at first). Once components have.been generated, a series of binary <br />comparisons will be made. First, productivity changes obselVed in inundated versus control areas will be <br />compared using repeated measures ANOV A's. Productivity will probably be represented by at least four <br />axes, including primary and secondary (invertebrates) productivity and native and nonnative fish density or <br />biomass estimates. Because samples will be collected weekly for 4-6 weeks, short-term time series will be <br />able to be generated. Because the sampling protocol calls for at least three months of monthly sampling <br />following the weekly sampling, we will have about five months of 'monthly' data to put into a longer <br />time-series (again, using repeated measures). <br /> <br />11 <br />