The greatest amount of research and development activity relevant to instream 

 flow assessments during the 1970s was in this microhabitat description area and more 

 specifically the physical microhabitat models used to evaluate usability under 

 different streamflow regimes. Most models have been criticized because: (1) they 

 were not supported by a rigorous mathematical development; (2) they lacked clear 

 definition of the significance of the usability index; and (3) they were limited by the 

 statistical techniques used to estimate weighting functions. 



Mathematically, fish microhabitat models may be presented in the "USA" form:" 



U = S . A 



where: 



(1) 



U = usability-which is a relative index value of the environment as habitat for the 

 target organism, 



S - suitability-which is the organism's voluntary or involuntary preference for 

 combinations of environmental attribute values (i.e., depth, velocity, sub- 

 strate), and 



A - availability-which is the distribution of the values of the environmental 

 attributes in a stream segment. 



Recent work by the IFG has refined microhabitat analysis by developing improved 

 hydraulic simulation models, weighted criteria for the life stages of target fish species, 

 and the introduction of stochasitc or time-series streamflow data so that the habitat 

 usability can be displayed over time for each species-life stage. 



The first task undertaken by the IFG was the modification of both the conceptual 

 view of the stream reach and the available hydraulic simulation models. Rather than 

 viewing the stream reach as a series of depth, velocity, and substrate contours, the 

 stream reach was modeled as a series of small cells or elements. The length of a cell is 

 the distance halfway upstream and downstream from a transect to adjacent transects. 

 Each transect is subdivided into a number of subsections, the width of each being 

 translated as the width of the cell. This is illustrated in Figure 3. 



State-of-the-art hydraulic models were then ungraded, so that instead of one 

 average depth and velocity for a cross section, the depths and velocities of all the cells 

 could be predicted. This was accomplished through improvements in the single 



Figure 3. Conceptualization of simulated stream reach. Shaded subsections have 

 similar depth and velocity ranges. 



134 



