Statistical Survey Design for Assessing Response to MPA Zoning: Application to Reef Fishes in South Florida

Transcription

1 Statistical Survey Design for Assessing Response to MPA Zoning: Application to Reef Fishes in South Florida Jerald S. Ault & Steven G. Smith Professor of Marine Biology & Fisheries Schooling permit in Tortugas Bank Ecological Reserve, June 2004.

2 An average day of grouper fishing in the Keys in the 1930 s!

3

4 Fishery Systems Science (FSS) Data Acquisition Population Dynamics Habitats Bioeconomics Coral Reef Ecosystem Model Building Physical Biological Humans Habitats Human-Fishery Sector Fishery Resource Risk Assessment

5 A Large Marine Ecosystem? HYCOM-MICOM Forecast for September 27, 2006

6 GeoSpatial Model of Biophysical Dynamics Human Impacts Layer (fishing, water quality, zoning) Predator Cohort Patch layer Demographics Natural Mortality Age, number, length, weight, biomass, position (x,y), velocity. Bioenergetic Growth Prey Cohort Patch layer Recruitment Hydrodynamics layer (currents, salinity, temperature, DO) Movements & Migrations Habitat layer (bathymetry, benthic substrates) y x Ault et al Can. J. Fish. Aquatic Science 56(S1) Ault et al N. Amer. J. Fish. Mgmt. 19 Ault, Luo, & Wang CRC Press Wang, Luo & Ault Bull Marine Science 72(3) Humston, Olson and Ault Trans. Amer. Fish. Soc. 133

7

8 Observable Assessment Indicator Variables Frequency Average Size (TL, mm) Minimum Legal Size (24 inches) Unexploited Length (mm) Frequency , 42.8 lbs SPR = 100% MSY Exploitation Length (mm) L( t) Frequency Fishing Mortality Rate = F( t) a a λ c N ( a, t) F(t) a λ a c 36.4, 30.1 lbs SPR = 35.1% N(a,t) da Current Exploitation Total length (mm) L( a, t) da 27.9, 12.6 lbs SPR = < 6%

9 Model Cross-Validations Fishery-Dependent Observables Fishery-Independent Observables Catch (kg) Comm Observed Rec Observed Rec Predicted Average Length (mm) RVC Headboats Crec Obs C comm Obs CpredRec CpredComm Index value (fish/177m2) FI Recruitment Observables Year year RVC juv pred RVC juv RVC exploited pred RVC exploited

10 1.20 Estimated Fishing Mortality Components Fishing Mortality Rate Commercial Recreational 0.00 F estimates Year aspic ss FI Lbar FD Lbar Fmsy

11 Bigger is (Biologically) Better! Individual Length (cm) Length Weight Weight (kg) Population Black Grouper Biomass (mt) Biomass Fecundity Eggs (millions) Age (years) 0

12 But, BIGGER is Better!!

13 Fishery Sustainability Decision Metrics MSY Most likely estimate for F msy (F)

14 Management Benchmarks: FL Keys Reef Fish Community GROUPERS SNAPPERS GRUNTS Overfishing (30% SPR) White Tomtate Sailors French Cottonwick Bluestriped Margate Yellowtail Vermillion Silk Schoolmaster Red Mutton Lane Hogfish Gray Dog Cubera Blackfin Black Yellowmouth Yellowfin Yellowedge Speckled Hind Scamp Rock Hind Red Hind Red Nassau Graysby Goliath Gag Coney Black Exploited Coral Reef Fishes 10 0 Ault, J.S,, Bohnsack, J.A., and G.A. Meester Fishery Bulletin 96: (Best Publication Award 1998, NOAA NMFS Scientific Publication Office) Ault, J.S., Smith, S.G., and J.A. Bohnsack ICES Journal of Marine Science 62: % Spawning Potential Ratio

15 Florida Keys Reef Fish Community Limit Control Rule 8 Hogfish Scamp F/Fmsy Yellowfin Nassau Overfished Black Gray Graysby Yellowtail Red Dog Cottonwick Mutton Margate Schoolmaster Bluestriped White Red Hind Stock overfished, no overfishing (recovering) MSY Limit Lane Rock Hind Overfishing, but stock not overfished No overfishing, not overfished Goliath French Tomtate Unexploited Grouper Snapper Grunt B/Bmsy Ault, Bohnsack, and Meester Fishery Bulletin 96(3): (Best Publication Award) Ault, Bohnsack, Smith, and Luo Bulletin of Marine Science 78(3): Ault, Smith, and Bohnsack ICES J. Marine Science 62:

16 Yield-per-Recruit Traditional Fishery Management Benchmarks & Controls Spawning Potential Ratio SPR-per-Recruit Yw/R: > 9.14 kg 92% increase SPR: > % increase

17 Yield in Numbers-per-Recruit Transitional Productivity Dynamics of Florida Keys Black Grouper Stock Yield in Numbers Yield in Weight(kg)-per-Recruit Yield in Weight Years Post Management Intervention

18 Designing the Tortugas No-Take Marine Reserves Meester, Ault, Smith, Mehrotra Sarsia 86: Meester, Mehrotra, Ault, Baker Management Science 50:

19 The Managed Florida Keys Coral Reef Ecosystem FLORIDA Miami Florida Bay Gulf of Mexico Florida Straits

20 Pilot & Pre-Survey Analyses Habitat Characterization & Mapping Species Lifestage-Habitat Associations Model-based Habitat Assessment Community Dynamics Analyses Adaptive Precision or Iterative Learning Design Analysis & Sampling Allocation Conduct 2-Stage StRS Survey Data Assimilation Post-Survey Analysis Design-based Estimates Multispecies Stock Assessments Spatial Ecosystem Models for Resource Risk Assessment of Management Alternatives

21 Sampling Survey Design Goal: Estimate population & community metrics Accurate, precise, low-cost Proportions (presence-absence, cover) Means (density, richness) Totals (abundance) Species and Size composition metrics by life-stage (juvenile,adult,exploited) Key Properties Finite population within finite area Distribution-free

22 Domain Mapped Holocene Reefs (live coral habitat) Broward-Palm Beach: 195 km 2 Florida Keys: 500 km 2 Tortugas: 325 km 2 Design Variable Density (representative suite of species, life stages) Stratified Random Design Heterogeneous Spatial Distribution Stratification Variables Reef Habitat Class Depth Geographical Subregion Spatial Management Zone (e.g, no-take reserves)

23 (reef tract) Stratified Random Sampling Design Single Stage Design (Habitat Class, Depth, Zone) (200 x 200 m) Two Stage Design (177 m 2 )

24 200 m Fish: 177 m 2 per cylinder Lobster: 500 m 2 belt Coral: 10 m 2 belt

25 Precise, Cost-Effective Survey Designs Stratification Scheme: Partitions survey domain into subregions of low, moderate, and high variance Allocation Scheme: Allocation based on stratum size and variance More samples in larger strata More samples in higher variance strata

26 Linking Reef Fish Spatial Abundance & Benthic Habitats Patchy hard-bottom Patch reefs Pinnacles Low Degree of Patchiness High Low-relief spur & groove Low-relief hard-bottom Rocky outcrops Medium-profile reef High-relief spur & groove Reef terrace Low Vertical Relief High

27 Reef Terrace Pinnacles Patchy Hardbottom in Sand Medium Profile Reef Reef Terrace Patch Reefs Rocky Outcrops High-Relief Spur & Groove Low-Relief Spur & Groove Low-Relief Hardbottom Figure 2

28 Tortugas Bank NTMR DTNP RNA FKNMS Boundary Tortugas Bank Fished Dry Tortugas National Park Boundary Florida Dry Tortugas Miami

29 Standard deviation Mean density

30 Randomization Guards Against Bias Stratum-Level Estimates Second-Stage Units Within Primary Units Primary Units Within a Stratum Equal probability of selection for each sample unit Domain-Wide Estimates Stratum Weighting Factor Further guard against bias in domain-wide estimates for spatially heterogeneous populations

31 Performance Measures [ ] SE[ D ] D CV = D Ability to detect differences = 95% CI for avg density = approx. 2 SEs or twice the CV

32 [ ] + + = h h h h st h h uh h h h h h uh h uh h N s w D V s w m s w s w s w n * * = h uh h uh h h s w s w n n * Optimal (Neyman) Sample Allocation Primary units to achieve specified CV

33 CV 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Florida Keys Reef Fish: Sampling Allocations Neyman Optimal Sampling Allocation n*

34 Survey Design Estimation of MPA Effects Accounts for: Different mix of habitats inside & outside MPAs Disparity in survey area inside & outside MPAs Enables analysis of: Population-level impacts Inside vs. outside comparisons

35 Percentage Population Size Change in 2004 relative to Baseline Estimates Bank Fished Black Grouper + 84% Red Grouper - 43% Hogfish - 27% Mutton Snapper - 45% Yellowtail Snapper - 19% Gray Snapper - 96% White Grunt + 7% Bluestriped Grunt + 50% Spotted Goatfish + 133% Redband Parrotfish + 121% Foureye Butterfly + 86% Purple Reeffish + 31% Bank MPA + 120% + 38% + 6% + 303% + 367% - 51% + 24% + 13% + 326% + 26% - 18% + 42% National Park + 128% - 9% + 50% + 142% + 132% + 270% + 2% + 242% + 175% + 26% + 32% + 263% Total Domain + 124% 2% - 19% + 109% + 181% + 39% + 4% + 159% + 198% + 56% + 13% + 76% Signif. *** ns ns *** * ns ns ns *** ns ns ***

36 Baseline N = 209 Abundance = 277, % Frequency N = 558 Abundance = 621, % Length (cm)

37 Tortugas Bank, Fished n = % 2004 Frequency Tortugas Bank, NTMR n = % Dry Tortugas NP n = % Length (cm)

38 Protected Open Baseline 94-97

39 Summary Strive for strategic balance in Florida reef resource management. Necessary tactics involve a combination of reserves and traditional management, while mitigating environment and chance. With more people is our management strategy sufficient? Optimum management strategy involves multiple control methods with closed areas to buffer uncertainty. Maps & consistent sampling resources to detect changes. Good for fish, ecosystem, fishermen and Florida s economy!