Effects of changes in biotoxin closures on recreational shellfish harvest demand

Transcription

1 Effects of changes in biotoxin closures on recreational shellfish harvest demand Leif Anderson and Stephanie Moore Northwest Fisheries Science Center NOAA Fisheries

2 Recreational clam and oyster harvesting is a popular activity in Puget Sound 133,688 estimated harvester days on managed beaches (WDFW, 2013) Beaches are often subject to health-related environmental closures based on ongoing water quality assessments Local resource management agencies (e.g., Puget Sound Partnership) have set a priority to reduce the risks of biotoxin and pollution closures on human health

3

4 Harvest closures Health-related closures set by Washington Dept. of Health Health-related closures may be due to: Biotoxins diarrhetic shellfish toxins, paralytic shellfish toxins, domoic acid Pollution (bacteria, viruses, or other) Closures are posted online, in other media, and (often) at public beaches

5

6

7 What are the effects of likely climate scenarios on harmful algal bloom growth rates? biotoxin-related closures? What are the effects of biotoxin closures on effort? recreational use value?

8 Data for economic model Puget Sound Recreational Shellfishing Survey 3 focus groups and 3 sets of one-on-one interviews : Seattle, Bellingham, Silverdale Up to 6 contacts: phone, prenotice, 1st mailing, postcard, 2nd mailing, 3rd mailing Fielded in 2013 Estimated response rate ~50% Data include location of most often used beach, contingent behavior questions

9 Experimental design 4 contingent behavior questions on each of 25 survey versions Attributes type of closure (biotoxin, pollution), length of closure (1, 2, 3, 12 months) species affected (all clams and oysters, butter clams only), additional distance to a nearby beach that is fully open (5, 10, 20, 30 miles) Computerized search algorithm to max D-efficiency How many trips would you take if the beach you most often use for harvesting clams and oysters was closed?

10

11 Economic model Estimate demand for 3 trip types together in order to capture substitution between types of trips incomplete demand system 3 trip types: harvest, alternate beach, non-harvest Demand for trips = f (travel cost, demographics, closures) x i = α i z exp k β ik P k + γ i y, negative binomial functional form Standard restrictions for integrability (γ i = γ k, β ik = 0, i k, β ii < 0, α i (z) > 0) x i = exp α i + μusealt + φpriv + δ ni Closed n + β i P i + γy n Closed is measured in months Heterogeneous baseline demand for trip types (α i ~N) Anderson and Plummer Marine Resource Economics. 32(1):

12 Variable Coefficient Mean Coefficient SD Trip types (α i ) Harvest ***.83728*** Alternate *** *** Nonharvest *** *** Price (β i ) Harvest *** Alternate *** Nonharvest *** Closed (δ ni ) Harvest Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Alternate Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Nonharvest Biotoxin, All Species Biotoxin, Butter Only Pollution Income (γ) *** Priv(φ) *** Usealt (μ) *** Scale *** Respondents 120 Sample size 2640 Log-likelihood at zero Log-likelihood at convergence * p<0.10, ** p<0.05 *** p<0.01

13 Variable Coefficient Mean Coefficient SD Trip types (α i ) Harvest ***.83728*** Alternate *** *** Nonharvest *** *** Price (β i ) Harvest *** Alternate *** Nonharvest *** Closed (δ ni ) Harvest Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Alternate Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Nonharvest Biotoxin, All Species Biotoxin, Butter Only Pollution Income (γ) *** Priv(φ) *** Usealt (μ) *** Scale *** Respondents 120 Sample size 2640 Log-likelihood at zero Log-likelihood at convergence * p<0.10, ** p<0.05 *** p<0.01

14 Variable Coefficient Mean Coefficient SD Trip types (α i ) Harvest ***.83728*** Alternate *** *** Nonharvest *** *** Price (β i ) Harvest *** Alternate *** Nonharvest *** Closed (δ ni ) Harvest Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Alternate Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Nonharvest Biotoxin, All Species Biotoxin, Butter Only Pollution Income (γ) *** Priv(φ) *** Usealt (μ) *** Scale *** Respondents 120 Sample size 2640 Log-likelihood at zero Log-likelihood at convergence * p<0.10, ** p<0.05 *** p<0.01

15 Variable Coefficient Mean Coefficient SD Trip types (α i ) Harvest ***.83728*** Alternate *** *** Nonharvest *** *** Price (β i ) Harvest *** Alternate *** Nonharvest *** Closed (δ ni ) Harvest Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Alternate Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Nonharvest Biotoxin, All Species Biotoxin, Butter Only Pollution Income (γ) *** Priv(φ) *** Usealt (μ) *** Scale *** Respondents 120 Sample size 2640 Log-likelihood at zero Log-likelihood at convergence * p<0.10, ** p<0.05 *** p<0.01

16 Variable Coefficient Mean Coefficient SD Trip types (α i ) Harvest ***.83728*** Alternate *** *** Nonharvest *** *** Price (β i ) Harvest *** Alternate *** Nonharvest *** Closed (δ ni ) Harvest Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Alternate Biotoxin, All Species *** Biotoxin, Butter Only *** Pollution *** Nonharvest Biotoxin, All Species Biotoxin, Butter Only Pollution Income (γ) *** Priv(φ) *** Usealt (μ) *** Scale *** Respondents 120 Sample size 2640 Log-likelihood at zero Log-likelihood at convergence * p<0.10, ** p<0.05 *** p<0.01

17 Willingness to pay Willingness to pay for a harvest trip = $144 harvest trip to an alternate beach = $117 non-harvest trip = $98

18 Climate model Regional atmosphere and ocean model simulations derived from global climate models ( km grid downscaled to 12 km) Downscaled simulations are inputs to 3-D model of the Puget Sound Future projections of temperature and salinity (2050) Experimentally-derived growth function (temp, salinity) Resulting output is a window in space and time favorable to an Alexandrium bloom Moore et al Harmful Algae 48: 1-11

19 Additional days favorable for Alexandrium growth Moore et al Harmful Algae 48: 1-11

20 Linked model example Beach: Birch Bay State Park, Washington State, USA Increase in favorable bloom conditions: 30 days (163 vs 133) Trips: Harvest Alternate Non-harvest Total Current Future Resulting loss in value: t i δ t exp α β i + μusealt + φpriv + δ ni Closed nt + β i P i + γy i n x it where Closed nt = f(temp t, salinity t )

21 Linked model example Beach: Birch Bay State Park, Washington State, USA Increase in favorable bloom conditions: 30 days (163 vs 133) Trips: Harvest Alternate Non-harvest Total Current Future $663 $522 Resulting loss in value: t i δ t exp α β i + μusealt + φpriv + δ ni Closed nt + β i P i + γy i n x it where Closed nt = f(temp t, salinity t )

22 Next steps Improve link between growth-favorable conditions and biotoxin closures Develop full time series of climate scenarios, rather than only endpoints Discounting hyperbolic? Expand estimates to Puget Sound level using effort estimates from Washington Department of Fish and Wildlife

23 Questions?