7758 - Laserfiche WebLink

Home Browse Search

28 BIOLOGICAL REPORT 24 Life stages of target biota are sampled or otherwise monitored (fish preferences are often determined from animals fitted with radio transmitters) across the range of the hydraulic variables to derive "habi- tat suitability curves." Intuitively, this is a logical approach, but it is often biased by sampling error, especially in large, deep, and often turbid rivers, where the biota are difficult to capture or see; (3) the net suitability of use of a given locality (transect cell) is quantified by a variable called weighted usable area (WUA), which is a derived relation between plan area of the transect cell (area avail- able) and the habitat preference indices (from suit- ability curves) for velocity, depth, and substratum. The WUA is calculated cell by cell and summed for the entire reach and over a range of discharges. Hence, increments of WUA for a stream become a continuous function of discharge. Easy to read and more detailed descriptions of the IFIM are given by Gore and Nestler (1988) and Nestler et al. (1989). This procedure has been widely used to justify flow provisions in regulated streams throughout North America, in some cases leading to state statutes to guarantee protection of aquatic biota (Reiser et al. 1989a). Even though the IFIM has become an industry standard (Reiser et al. 1989a), it has a number of faults that are not widely recognized or under- stood within management circles. Concern exists regarding use of suitability curves as probability functions (Patten 1979; Mathur et al. 1985; Moyle and Baltz 1985); the assumption of independence of depth, velocity, and substratum (Patten 1979; Malthur et al. 1985); the lack of a demonstrated relation between WUA and a meaningful measure of productivity or biomass (Mathur et al. 1985; Bowlby and Roff 1986; Conder and Annear 1987; Scott and Shirvell 1987); and lack of any relation- ship with regard to many other ecosystem proc- esses, such as predation and other density-de- pendent relationships, which clearly influence population structure (Moyle and Baltz 1985; Bowlby and Roff 1986; Orth 1987; Stanford and Ward 1992a). To my knowledge none of these criticisms has been resolved, nor is it likely they will be. However, these criticisms have been placed in perspective with respect to the rationale and intent of the IFIM, which is often misunder- stood, misrepresented, and misused (Gore and Nestler 1988). For example, the model was not intended to predict biomass; it is a physical habi- tat simulator. Even when the model is ap- plied properly, a variety of problems may emerge depending on input choices, which necessitates a clear understanding of how the model works. The simulator can use a variety of hydraulic predictors (e.g., the HEC-2 flow model of the U.S. Army Corps of Engineers), each of which has biases and therefore will result in different WTJA calculations (Gan and McMahon 1990). Suitability curves not derived on site (i.e., curves given in the literature) are often used, which can also bias output (Gore and Nestler 1988). The IFIM was used in an attempt to derive flow recommendations for specific river segments of the Upper Colorado River Basin for the endan- gered fishes. However, in the analysis WUA often was maximized for various life history stages of squawfish and humpback chub at very low flows that in the historical record were exceeded most or all of the time (Rose and Hann 1989). Such output is nonsense because the ecological data for these fishes clearly shows the importance of back- waters and eddies that occur at much higher flows. The problem here is that the IFIM probably should never have been used in the big river reaches of the Upper Colorado River Basin. When low velocity habitats are abundant, as they are throughout the potamon of the Colorado River system, the simulator underestimates the WUA; in fact, the model cannot deal with zero-flow habi- tats. This explains why the IFIM works well only in small streams where the channel is charac- terized by uniformly varying flow (e.g., the low velocity profile reflects steady, uniform flow, which is also an assumption of the HEC-2 hydrol- ogy simulator that is often used in IFIM; my observations). Also, habitat suitability curves were probably biased because the fish were diffi- cult to observe or collect in the usually turbid, deep water of the Yampa and Green rivers (Rose and Hann 1989), which is precisely why the adult fish monitoring program (U.S. Fish and Wildlife Service 1987b) emphasizes shallow, shoreline habitats that can be sampled effectively by elec- trofishing. However, the fishes routinely use deep- water habitats (e.g., Tyus and McAda 1984; McAda and Kaeding 1991), and movement be- tween habitats (e.g., channel, backwaters) on a diel basis cannot be accounted for in the method. The utility of the IFIM evolved a great deal during the period that data were being gathered in the Upper Colorado River Basin studies, and deficien- cies in the method with regard to the Colorado River were probably not apparent at the time much of the data were gathered.