Aquaculture conference Technical Day, Nelson 2015

Transcript

1. BLUE MUSSEL OVER-SETTLEMENT PREDICTIVE MODEL JAVIER ATALAH CAWTHRON INSTITUTE, NELSON Fisheries & Aquaculture research series 16th of September 2015, Nelson
2. NEW ZEALAND'S GREEN MUSSEL AQUACULTURE • The most important aquaculture sector $US218M pa (73% of NZ aquaculture revenue) • The Marlborough Sounds the largest farming area • Reliant on wild caught juveniles (spat) Photo: Kevin Heasman
3. BIOFOULING AND THE MUSSEL INDUSTRY • Significant threat to the industry • High biomass and density of biofouling • Strong competitors for food and space • Blue mussel (Mytilus galloprovincialis) major problem
4. BLUE MUSSEL OVER-SETTLEMENT IMPATCS • Spat lines heavily over-settled are dumped • Blues carried through 1-2 grow-out stage to “final seeding” • Removed by size grading or optical sorting • Crop loss when lines lifted from water • Increased weight and extra floatation • Increased factory processing costs • ~NZ$20 million p.a. (ca. 12% of regional revenue)
5. IMPACTS OF BLUES ON ANNUAL PERNA YIELD
6. RATIONALE AND OBJECTIVES • Few management options available • Considerable operational problems • How can be avoid? Long term and seasonal patterns Model response to drivers Forecast and inform farmers Blue = Bad Green = Good
7. SPAT MONITORING HISTORICAL MFA SPAT DATASET Blue and green spat monitoring ~40 yr Collector rope deployed for 2 wk 2, 4, 10 and 15 m depth Overall >50 sites
8. SAMPLING SITES
9. La niña El niño SOUTHERN OSCILLATION INDEX
10. BMPO historical dataset Environmental data BMPO + environmental data data Predictive model Web app to forecast blue and green spat settlement
11. LONG-TERM TEMPORAL TREND PELORUS Blues Greens
12. PELORUS SEASONAL RECRUITMENT PATTERNS Blues Greens
13. RECRUITMENT AND TEMPERATURE Blues Greens Temperature (° C)
14. DEPTH PATTERNS Greens Blues
15. PREDICTIVE MODEL • Spatio-Temporal Bayesian Models • Latent Gaussian models using integrated nested Laplace approximations • INLA – R package
16. http://www.r-inla.org/ SPATIOTEMPORAL STRUCTURE
17. THE WEB APP https://cawthron.shinyapps.io/BMOP/
19. THE BLUES RADAR
20. FUTURE WORK • Refine web app and launch it • Select and validate predictive mode • Input live data into model and app • Spat counts using imagery analysis • Feedback from end-users shiny@cawthron.org.nz • Mobile devices app
21. THANKS Marine Farming Association Hayden Rabel, Biosecurity team, Modelling team Sustainable Farming Fund (MBIE) Culture Shellfish Programme (MBIE) NIWA, MSQP