9665 - Laserfiche WebLink

Home Browse Search

11/14/01 draft report, Schmidt and Box One value of our effort is in providing a template with which to organize available information, to assess information gaps and needs, and to evaluate the priorities in new data collection efforts. Walters et al. (2000) have assembled an extensive array of physical and ecological information about the riverine ecosystem of the Colorado River in Grand Canyon and developed a conceptual ecosystem model that highlights linkages among operations of Glen Canyon Dam, fluvial processes, ecological processes, and economic impacts of various decisions. Despite the simplicity of many aspects of this model, it now serves as a template for organizing scientific studies and identifying research priorities whose outcome will add precision and accuracy to the model. In the same way, we hope that our model, despite its errors, will inspire subsequent field measurements to evaluate our assumptions and improve the predictive capability of the model. As discussed above, discrepancies between model predictions and field sampling data may not entirely be due to inadequacies of the model. Pikeminnow populations are low in the Colorado River system, and the sampled larval population in late summer is a very small proportion of the early season drift. Models can serve as valuable tools with which to interpret limited count data, because limited sampling data may not adequately represent larval densities (e.g., Link and Sauer 1998). The variability in field-based population estimates due to sparse sampling efforts masks most of the longitudinal differences in distribution of larval fish that are accepted as the paradigm of larval fish management in the middle Green River. A vailability of a model that predicts the longitudinal distribution of fish may lead to revisions of temporal and spatial aspects of present sampling protocols to better characterize population dynamics at the time transport into backwaters initially occurs. The model also provides a set of testable hypotheses about larval drift, backwater populations, and the control on those attributes by river hydrology and geomorphology. Modeling predictions suggest that the present sampling program inadequately characterizeslongitudinal patterns of drift during early summer and fails to account for losses tothe largest proportion of each year's drifting population. By the time sampling is conducted in late summer, a very small proportion of larval fish can be found in the backwaters-of the middle Green River. Reaches near the Yampa River are not presently sampled, butthe model predicts that larval populations in those reaches may be significant. Historically, larvalJish were found in these reaches (Muth et al. 2000). Conclusions The fall population of larval Colorado pikeminnow in the middle Green River is a small proportion of the total number that initially drift downstream in early summer. The existing fall sampling program is either extremely inefficient or there are very few fish, because more than half of the backwaters seined between 1990 and 1995 had no larvae. Previously published estimates of larval fish populations in late summer (McAda 1993) are less than 1 percent of the number of larvae that drifted in those years. The few numbers of backwaters actually sampled and the large number 17