Investigating ecosystem-level effects of gillnet bycatch in Lake Erie: implications for commercial fisheries management

Simple item page
dc.contributor.authorLi, Yanen
dc.contributor.committeechairJiao, Yanen
dc.contributor.committeememberGuo, Fengen
dc.contributor.committeememberHallerman, Eric M.en
dc.contributor.departmentFisheries and Wildlife Sciencesen
dc.date.accessioned2017-04-04T19:49:36Zen
dc.date.adate2010-08-17en
dc.date.available2017-04-04T19:49:36Zen
dc.date.issued2010-07-22en
dc.date.rdate2016-10-07en
dc.date.sdate2010-08-03en
dc.description.abstractLake Erie supports one of the world's largest freshwater commercial fisheries. Bycatch has become a concern in current fisheries management. This study focused on four species in Lake Erie that include two major commercial and recreational species, walleye (Sander vitreus) and yellow perch (Perca flavescens); an invasive species, white perch (Morone americana); and an endangered species, lake sturgeon (Acipenser fulvescens). The analyses were based on two datasets, the Partnership Index Fishing Survey (PIS) data and the commercial gillnet logbook data. The bycatch of walleye, yellow perch and white perch was predicted by a delta model developed on the PIS data. Discards were estimated as the difference between predicted bycatch and landed bycatch. Results highlighted bycatch and discard hotspots for these three species that have great management implications. Three classification tree models, a conditional inference tree and two exhaustive search-based trees, were constructed using the PIS data to estimate the probability of obtaining lake sturgeon bycach under specific environmental and gillnet fishing conditions. Lake sturgeon bycatch was most likely to be observed in the west basin of Lake Erie. The AdaBoost algorithm was applied in conjunction with the generalized linear/additive models to analyze catch rates of walleye, yellow perch and white perch. Three- and five-fold cross-validations were conducted to evaluate the performance of each candidate model. Results indicated that the Delta-AdaBoost model yielded the smallest training error and test error on average. I recommend the Delta-AdaBoost model for catch and bycatch analyses when data contain a high percentage of zeros.en
dc.description.degreeMaster of Scienceen
dc.identifier.otheretd-08032010-125745en
dc.identifier.sourceurlhttp://scholar.lib.vt.edu/theses/available/etd-08032010-125745/en
dc.identifier.urihttp://hdl.handle.net/10919/76836en
dc.language.isoen_USen
dc.publisherVirginia Techen
dc.rightsIn Copyrighten
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subjectmodel analysisen
dc.subjectgillneten
dc.subjectLake Erieen
dc.subjectcommercial fisheriesen
dc.subjectbycatchen
dc.titleInvestigating ecosystem-level effects of gillnet bycatch in Lake Erie: implications for commercial fisheries managementen
dc.typeThesisen
dc.type.dcmitypeTexten
thesis.degree.disciplineFisheries and Wildlife Sciencesen
thesis.degree.grantorVirginia Polytechnic Institute and State Universityen
thesis.degree.levelmastersen
thesis.degree.nameMaster of Scienceen

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
etd-08032010-125745_Li_Y_T_2010.pdf
Size:
9.05 MB
Format:
Adobe Portable Document Format
Download

Collections

Masters Theses