Laserfiche WebLink
<br />. <br /> <br />. <br /> <br />accurately reflect habitat conditions actually used by the fish. Further, <br />because gear capabilities limit our ability to effectively sample all habitats <br />(particularly deep, swift areas), utilization data used to construct SI curves <br />can be biased toward shallow, low velocity habitats that are more easily <br />sampled. For more detailed information on curve development, see Bovee <br />(1986). <br /> <br />Depth, velocity and substrate are weighted equally and treated independently <br />in the PHABSIM model, however, the model is most sensitive to the velocity <br />criteria. As a result, velocity SI curves which peak at zero, (Valdez et al. <br />1987, Attachment 1) account for predictions of optimal WUA at close to zero <br />flow for adults. For adult Colorado squawfish the velocity curve has a peak <br />utilization value of 1.00 at 0.00 ft/sec; which is not surprising considering <br />that 76.4 percent of the data used to construct the curve came from pools <br />and/or eddies. Similarly, the adult humpback chub curve has a peak value of <br />1.00 at 0.23 feet per second, with 49.6 percent of the data from pools and/or <br />eddy habitats. <br /> <br />Winter requirements of Colorado squawfish and humpback chub in the Yampa River <br />are not completely understood, and were not included in this analysis. <br />Continued research and SI curve development for the Yampa River are in <br />progress, however, there is some question that the PHABSIM models may not work <br />well under ice. Winter habitat requirements for endangered fish may be quite <br />different from summer habitat requirements. <br /> <br />Conclusions <br /> <br />Ideally a species model must include important ecosystem variables that allow <br />for a direct connection to a response by the population, i.e. growth, <br />recruitment, survival, standing crop or some other quantifiable species <br />response, otherwise, results will not be biologically meaningful. In lieu of <br />such a model, critical limiting habitats and/or other quantifiable variables <br />need to be identified, and related to species habitat utilizations. In <br />addition, detailed field data specific to those habitats and/or variables <br />shQuld be collected to develop more definitive habitat versus flow <br />relationships. <br /> <br />In a "natural" system like the Yampa River (no major dams or diversions), the <br />availability of physical habitats above a certain unknown minimum threshold, <br />mayor may not be limiting to endangered fishes. It must be recognized, <br />however, that there are many more environmental factors potentially <br />influencing population size than just the availability of gross physical <br />habitat. The physical microhabitat component of IFIM requires the assumption <br />that fish population response has some correlation to changes in WUA. This <br />assumption has never been validated for large, turbid rivers. It has not been <br />demonstrated that WUA, as defined by depth, velocity, and substrate has any <br />relationship to endangered fish populations whatsoever. Determining the <br />degree of accuracy of existing endangered fish SI curves and the corresponding <br />predictive reliability of PHABSIM, and what it means to the overall health of <br />fish populations will require many years of extensive field validation and <br />hypotheses testing. Until species criteria can be validated, they are of <br />limited value. Model outputs based on existing SI curves for the Yampa River <br />do not, in most instances, seem to accurately predict habitat versus flow <br /> <br />17 <br />