8208 - Laserfiche WebLink

Home Browse Search

percent abundance or availability of each type was then averaged across all pertinent fish. River- wide habitat use data (Appendix tables III and IV) indicated that none of the habitat types are com- pletely avoided. Thus, we assumed that all habitat types present within a study site at a given flow were available for the fish to use, and we assumed that relative abundance (percent total water area) of a given habitat type is a measure of the relative availability of that type. Based on the stable channel configuration between and during the two studies, radiotelemetry and mapping, we also assumed that habitat availability when the fish use data were collected was the same as when the habitat mapping data were collected (see Appendix IV for a discussion of this). To determine if adult squawfish prefer particular types of habitat, we compared usage with avail- ability (Williams and Marshall 1938, Hess and Rainwater 1939, Jacobs 1974, Swanson et al. 1974, Chesson 1978, Gilmer et al. 1975, Johnson 1980, Osmundson 1990). The degree of preference, or lack thereof, for a particular habitat type is estimated by the average difference between the percent that that type contributes to the total water area available to an individual fish and the percent fre- quency of use of that type by the individual fish. If there is no preference, fish should be located in the various habitat types in the same frequency as the occurrence or availability of those types. For example, if 20% of the total water area is comprised of pool habitat, one would expect 20% of the fish locations to be in pools if habitat selection was random, i.e., no preference. If the fish exhibit a preference for certain habitat types, i.e., more use than availability would predict, we assume that those types are important in fulfilling some biological need. Maximizing the quantity and quality of such habitats is viewed as benefiting the fish and is therefore a goal of flow management. To determine preference, we compared habitat usage with habitat availability for each fish that had one or more locations within the four study sites. Locations falling outside the study sites were not used. Percent availability of each habitat type within a given fishes range (one or more study sites) was subtracted from the percent use of that type by that fish. Differences were then averaged across all fish. The mean difference was then used as a measure of the degree of preference for that habitat type. Those types with positive values (>0) were considered to be preferred; the higher the value, the more preferred. Negative values were interpreted simply as a lack of preference for a type rather than an active avoidance of it (see Johnson 1980). Determining Flow Levels That Maximize Preferred Habitat As flow level vanes, the area of each habitat type also changes. The goal was to identify those flows that provide the maximum amount of the preferred habitat. For each flow level, all mapped areas of a given habitat type were summed for each site and then totaled for all four sites. This provided the total area of each habitat type provided at each of 11 flow levels. To select a level that provided the most of a preferred habitat we looked only at the range of flows that were less than or equal to the historic range for the particular season of interest. In doing so, we assumed the species or population was adapted to virgin conditions but that historic conditions were also apparently adequate. The decline of these fish populations coincided with major water withdrawals from the system during the mid part of the century. Average historic summer flows in the 15-mile reach are based on the 1902-1942 period of record (closure of the first dam, Green Mountain, was in fall 1942). Virgin flows were higher because diversions for the Grand Valley were put on line prior to the historic period (the historic period starts when gages were built and records were first kept) . These first local diversions largely impacted the summer flows in the 15- mile reach. However, anecdotal reports indicate that Colorado squawfish and razorback suckers 14