ScholarWorks@UA

Length-based models and population analyses for northern shrimp Pandalus borealis Krøyer

DSpace/Manakin Repository

Show simple item record

dc.contributor.author Fu, Caihong
dc.date.accessioned 2015-03-16T19:38:31Z
dc.date.available 2015-03-16T19:38:31Z
dc.date.issued 2000-08
dc.identifier.uri http://hdl.handle.net/11122/5120
dc.description Thesis (Ph.D.) University of Alaska Fairbanks, 2000 en_US
dc.description.abstract The lack of basic knowledge on stock dynamics o f northern shrimp Pandalus borealis, a protandric hermaphrodite, has caused difficulty in regulating fishing effort on a scientific basis and in understanding potential causes behind population fluctuations and collapses. Previous length-based population models (LBMs), developed for other species, are undesirable primarily for two reasons: (1) individual cohort dynamics are masked; (2) variations in annual natural mortality (M) are ignored. This research was primarily aimed at developing a more advanced LBM that provides estimates of parameters such as recruitment (R), fishing mortality (F) and especially annual M. Simulation-estimation experiments were conducted to evaluate model performance. Despite model complexity, annual M can be well estimated provided measurement errors in survey biomass estimates are low. The common assumption of constant M created biased parameter estimates. Estimated M of P. borealis in Kachemak Bay, Alaska increased steadily in the 1980s. Retrospective projections showed that the increasing trend in M in the 1980s resulted in the population collapse. The ultimate goal o f stock assessment is to develop sound harvest strategies. With the widely observed abundance fluctuations in shrimp populations, it is impossible to manage solely based on conventional methods, such as maximum sustainable yield (MSY). Thus, harvest strategies were compared under various situations o f M and R. With M increasing over time, it is important to execute threshold management, i.e., closing the fishery at population levels below a threshold value. Simulations indicated that overfishing caused by underestimated M or overestimated R can be greatly alleviated if the population is sampled once every year. Life history aspects of sex change, growth, M, and their seasonal variations were also incorporated into the LBM. Populations with protandrous animals are likely to be subject to recruitment overfishing; merely protecting older females while allowing high exploitation on younger males can lead to population collapse. Fishing after spring egg hatching is superior to fishing after mating and egg extrusion in fall when F is high. In summary, the length-based model developed here provided a convenient framework for understanding population processes and harvest strategies and should be useful for a variety of hard-to-age species. en_US
dc.description.tableofcontents 1. Introduction and overview -- 1.1. Distribution and fisheries worldwide -- 1.2. Review of stock assessment -- 1.3. Overview of biological data pertinent to assessment of P. borealis -- 2. Estimability of natural mortality and other population parameters in a length-based model: Pandalus borealis Krøyer in Kachemak Bay, Alaska -- 3. Retrospective projection using Monte Carlo simulation: an application of a length-based model to Kachemak Bay northern shrimp -- 4. Analyses of harvest strategies for Pandalus shrimp populations -- 5. The role of sex change, growth, and mortality in Pandalus population dynamics and management -- Summary -- Bibliography -- Appendix 1. Detailed descriptions of the length-based population model for chapter 2 -- Appendix 2. Details of the length-based model for chapter 3. en_US
dc.language.iso en_US en_US
dc.title Length-based models and population analyses for northern shrimp Pandalus borealis Krøyer en_US
dc.type Thesis en_US
dc.type.degree phd en_US
dc.identifier.department Fisheries Division en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search ScholarWorks@UA


Advanced Search

Browse

My Account

Statistics