Multiscale habitat suitability modeling for canary rockfish (Sebastes pinniger) along the Northern California coast

Author

Portia N. Saucedo, Humboldt State UniversityFollow

Graduation Date

Summer 2017

Document Type

Dissertation/Thesis

Program

Master of Science degree with a major in Natural Resources: option Environmental Science and Management

Committee Chair Name

Dr. Jim Graham

Committee Chair Affiliation

HSU Faculty or Staff

Second Committee Member Name

Dr. Brian Tissot

Second Committee Member Affiliation

HSU Faculty or Staff

Third Committee Member Name

Dr. Joe Tyburczy

Third Committee Member Affiliation

HSU Faculty or Staff

Keywords

uncertainty, MaxENT, ENMTools, modeling, geospatial, marine, predictive, interface, BlueSpray, ROC, AIC, AUC

Subject Categories

Natural Resources

Abstract

Detailed spatially-explicit data of the potential habitat of commercially important rockfish species are a critical component for the purposes of marine conservation, evaluation, and planning. Predictive habitat modeling techniques are widely used to identify suitable habitat in un-surveyed regions. This study elucidates the predicted distribution of canary rockfish (Sebastes pinniger) along the largely un-surveyed northern California coast using data from visual underwater surveys and predictive terrain complexity covariates. I used Maximum Entropy (MaxEnt) modelling software to identify regions of suitable habitat for S. pinniger greater than nine cm in total length at two spatial scales. The results of this study indicate the most important environmental covariate was proximity to the interface between hard and soft substrate. I also examined the predicted probability of presence for each model run. MaxEnt spatial predictions varied in predicted probability for broad-scale and each of the fine-scale regions. Uncertainty in predictions was considered at several levels and spatial uncertainty was quantified and mapped. The predictive modeling efforts allowed spatial predictions outside the sampled area at both the broad- and fine-scales accessed. This approach demonstrates that single-species suitable habitat can be defined with species-specific covariates. Further, this approach may be applicable to other rockfish species to aid fisheries management in the delineation of essential fish habitat as well as in conservation efforts in marine spatial planning.

Comments

Original degree title: Master of Science degree with a major in Natural Resources, option Environmental & Natural Resource Sciences

Citation Style

APA

Recommended Citation

Saucedo, Portia N., "Multiscale habitat suitability modeling for canary rockfish (Sebastes pinniger) along the Northern California coast" (2017). Cal Poly Humboldt theses and projects. 71.
https://digitalcommons.humboldt.edu/etd/71