 |
After more than a century of intense salmon fishing, today's catches are estimated at less than half of the 1890's. Salmon catch has been declining and the State of Oregon wanted to know why. CSE was contracted to research and prepare a report on the status of salmon in western Oregon and northern California. Under the direction of CSE President Dan Botkin, a blue ribbon panel of scientists was assembled to tackle the salmon issue. NEW! One product of the CSE project for the state of Oregon on salmon and forest practices is a paper introducing a new way to forecast the number of returning salmon on a river. The abstract of that scientific paper appears below, along with a link to the complete paper. What CSE found was that there is a set of common beliefs that form the background to the debate about salmon in the Pacific Northwest. These beliefs include the ideas that prior to European settlement: there was a super-abundance of salmon; the forests were mainly composed of large ancient trees; this extensive old-growth was essential to the abundance of salmon; and the forests and salmon were at a "steady state".
Initially, CSE researchers encountered a major obstacle. Available data were inadequate to provide definitive answers to the questions posed by Oregon Senate Bill 1125. For example, statistically valid counts of salmon exist for only two of the 26 rivers that flow into the Pacific Ocean in the study area. The most widely used method for counting salmon (peak-count method) is statistically unreliable. Data sets identifying the number and locations of dams are contradictory. These are only a few examples of the lack of reliable data available to the team.
CSE responded to the Oregon Senate's questions with a broad range of recommendations designed both to increase the state of knowledge about salmon habitats and to serve as a basis for a comprehensive plan. Among CSE's recommendations were the following:
The overall response to Dr. Botkin's presentation of the CSE report, Status and Future of Salmon in Western Oregon and Northern California, at seminars on May 15-17, 1995 in Oregon was positive. Representatives from the timber industry and salmon conservationists were seen charting amicably following the presentation. State Senator Bill Bradley (D-24) said, "It (the report) doesn't relieve anyone from responsibility. We've got to point the finger at all of us." Bill Maxon, the west coast fisheries coordinator for the U.S. Department of Fish and Wildlife stated, "We need to get these results out to the public and look at how to implement the recommendations. It is now up to the citizens and their elected officials to transform the report's recommendations into constructive policy initiative that may hold the key to the destiny of this potent symbol of the Pacific Northwest.
The abstract of the newly published paper is here. The complete text of the scientific paper is now available at the following website: http://www.naturestudy.org/pdf/Salmon_Paper_12$31$99.pdf . ENVIRONMENTAL VARIATION AND POPULATION CHANGE: FORECASTING SPRING CHINOOK RUNS IN TWO OREGON COASTAL RIVERS Daniel B. Botkin1, Matthew J. Sobel2, Lloyd G. Simpson3, Kenneth Cummins4, and Lee M. Talbot5 Copyright © The Center for the Study of the Environment 2007 Abstract. In this paper, we analyze correlations between adult spring chinook salmon (Oncorhynchus tshawytscha Walbaum) returns and river flow, hatchery releases and ocean troll catches in the Rogue and Umpqua Rivers and their tributaries in southern Oregon. Adult returns and water flow vary widely from year to year. Results suggest that variation in stream flow and number of smolts released by a hatchery can be used to forecast numbers of returning adults three and four years in the future. In contrast, present methods restrict forecast to a few months prior to a season. Water flow correlates positively with adult returns in 173 of the 174 cases. Hatchery releases are negatively correlated in 17 out of 17 four years before and 16 of 17 releases three years before. This suggests that hatchery releases may be ineffectual or actually diminish wild escapements. The statistical procedure known as jackknife models is used because the data set is small (1975 to 1992). With this method, 174 jackknifed models have good statistical properties, including R2 values from 0.60 to 0.98 and from 0.50 to 0.96 for the three- and four-year models, respectively. The analysis suggests that environmental variation is useful as part of forecasting methods, providing much longer lead times than available with current methods. The complete paper may be downloaded in PDF format by clicking on this link. 1 University of California, Santa Barbara, and Center for the Study of the Environment, Santa Barbara, CA 93160.
|
||
| CSE: Home // Who Is? // Approach // Membership // Services // Map | |||
Copyright © 1997-2006 The
Center for the Study of the Environment |
SALMON:
MYTHS VERSUS KNOWLEDGE
Evolving
Future Policy 
