WSP11169 - Laserfiche WebLink

Home Browse Search

. ()-1159 enough to tag, unknown proportions of fish attempting to rear in mainstem vs LCR, relatively low numbers of fish captured and recaptured). If such noisy early estimates were our only recruitment "indices", we would have serious doubts interpreting the results of any experiment aimed at improving recruitment (e.g. exotic fish removal). But with stock assessment models, we can integrate these early estimates with data collected in subsequent years as the fish grow and become fully vulnerable to tagging (and other indexing methods). This integration still requires assumptions about stability of survival rates (otherwise when we first see some of the recruits from a given cohort as 3-yr olds, 4-yr olds, etc. we would have no way to estimate how many additional young fish must have been present earlier in order to have produced these survi vors). Stock assessment data analysis should be viewed to some degree as a problem in risk management, where we must tradeoff between using noisy point sample (short-term mark- recapture and catch per effort index) information, versus using more complex methods built around assumptions (particularly about stability of survival rates over time) that cannot be fully tested with the available historical data. We can demonstrate that assessments of population trend (but not current abundance) are highly robust to such assumptions (we get about the same downward trend pattern for every survival assumption that we have thought to test so far). Furthermore this downward trend suggested by the stock assessment models is also indicated by independent catch rate data (a measure of relative abundance) in the LCR. However, this does not mean that we have obtained the "correct" answer to date. In short, there is no fundamentally "right" or "wrong" methodology, and no single "best" estimate of stock status and trend. There has been some demand by Orand Canyon stakeholders to "give us a number" representing scientific consensus about the best assessment methodology and best point estimate of current chub stock size. Such demands are common in fisheries assessment and management situations in general, and represent a fundamental misunderstanding (or deliberate misrepresentation) about what scientists can and should provide. What we can provide is a set of probability distributions for stock size and trend, based on alternative assumptions about the data. Scientists cannot, and should not, be expected to agree upon how to deal with the risk management problem of which assumptions to "trust", and for us to pretend such consensus might exist would be dishonest and misleading. Moreover, it is not a requirement or even a real need for effective policy design that we do produce a particular number or estimate. Perfectly reasonable judgments about management can be made on the basis of probabilistic assessments and statements about relative likelihood of various outcomes, just as humans must do in practically all decision situations that involve substantial public and private investments. To demand a single number from scientists is as unrealistic as it would be for a stock market investor to demand a single earnings number from a stockbroker. However, your stockbroker may be able to give sound advice about how your portfolio is trending and whether or not strategic changes in your investments are wise. It is this type of information that we are able to provide relative to status and trends of humpback population dynamics. Although there is considerable uncertainty in the absolute abundance of humpback chub, particularly with regard to most recent abundance, the suite of models including competing assumptions and formulations all depict a "down-turn in the market" (Figure 6). Draft - April 21,2003 5