Dave Secor U Mass Dartmouth SMAST, 2017 Atlantic Offshore Renewable Energy Development and Fisheries Workshop Priorities and Challenges of Monitoring Designs Ecological Impacts BACI (before-after-control-impact) v. Gradient Designs Examples: WE Impacts on Sedentary and Migratory Fishes in MAB Effect Size and Scale of Impact Helen Bailey et al. 2014
The first BACI design? Does stocking fjords with cod larvae result in increased juvenile numbers? Captain Gunder Dannevig Hatchery - protagonist Director JohanHjort Hjort Hatchery-antagonist Year Stocked Fjord-1 Stocked Fjord-2 Control Fjord-1 Control Fjord-2 Control Fjord-3 meanimpact meancontrol 1 11.4 1.9 1.5 4.1 2.7 22.5 2.8 2 33.7 10.8 10.9 49 112 11.4 57.3 No before treatment Impact and effect size confounded by year Response not suitably scaled to question Dahl, K. 1909. The problem of sea fish hatching. Special Part B, No. 5. In Rapport sur les Travaux de la Commission A, dans la Période 1902 1907, 39 pp. Ed. by J. Hjort. Rapports et Procès-Verbaux des Réunions du Conseil International pour l Exploration de la Mer, 10. 844 pp. Secor, D.H. 2015. Migration Ecology of Marine Fishes. JHU Press 305 p.
BACI Design BAG (Gradient) Design - = before Impact Impact Gradient β before - = after before - after = effect size β before - β after = effect size β after Design elements BACI BAG Site selection Control site issues Informed by effect size Independence Pseudo-replication? Non-independence assumed Confounding environmental signal Predictions Problem Discrete (Y/N) bounded effect; difficult to model Some can be incorporated Effects gradient; supports models
2016 Black Sea Bass Noise impact study Before After BAG Design: Attenuated Noise, informed by literature, constrained by reef habitat 15 black sea bass implanted with transmitters at each site, surrounded by receivers Effects: Loss rate; activity rate
Fish remaining in array Hurricane Hermine (early September, 2016) 15 Hermine 10 Outer 5 Inner 0 Middle Jun Jul Aug Sep Oct Nov Dec Jan Date
Oceanographic models can inform impact scale, effect size, BAG Design Middle Site, Cold pool destruction S. MAB, Cold pool destruction Ming Li, Horn Point Laboratory: Finite-Volume Community Ocean Model: FVCOM, configured for MAB
Experimental Design: Before, After, Gradient Baseline Impacts on migratory fishes off DelMarVa: Flyway Concept Impact 1 Impact 2
Baseline: Depth strata (incorporation of confounding environ signal) constrained by coverage Impact 1: Attenuated noise Impact 2: Deviation from baseline, adjusted for bias owing to depth strata BOEM Award: Movement and Habitat Selection by Migratory Fishes within the Maryland Wind Energy Area and Adjacent Reference Sites (PIs Secor and Bailey)
Fall, Spring Flyway: Atlantic sturgeon Unique Atlantic Sturgeon 38.5 Nov Dec Jan Feb 38.4 38.3 38.2 38.5 Mar Apr May Jun Latitude 38.4 38.3 38.2 38.5 Jul Aug -75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6 Number 38.4 5 10 38.3 15 20 38.2-75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6 Predictive Responses: Depth, Season Longitude
Unique Striped Bass 38.5 Nov Dec Jan Feb 38.4 38.3 Latitude 38.2 38.5 Mar Apr May -75.2-75.0-74.8-74.6 38.4 38.3 38.2-75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6 Longitude Winter, Spring Flyway: Striped bass: Nov-Aug; Summer Desert Number 20 40 60
Winter Flyway, Predicted Temperature Window: Striped bass Cumulative detections Dec Jan Feb Rate of shelf cooling Inner Sites Jan Apr Feb May -75.2-75.0-74.8-74.6 Number 20 40 60 WEA Sites -75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6 Feb -75.2-75.0-74.8-74.6 May Longitude Number 20 Outer Sites 40-75.2-75.0-74.8-74.6-75.2-75.0-74.8-74.6 60
Effect Size Summary: Advantages of Gradient Design Design elements BAG Examples Site selection Independence Confounding environmental signal Predictions Informed by effect size Non-independence assumed Some can be incorporated Effects gradient; supports models BSB: Literature, Experts, Experiments Migr: Concept, Literature BSB: Noise Migr: Depth strata, temperature, noise BSB: None considered Migr: Depth, temperature, season BSB: Possible, unrealized Migr: Depth x season; temperature window Remaining Issue: Effect size and scale of impact before - after = effect size β before - β after = effect size Population Survival Recruitment Reproduction Migration Growth Activity Metabolism Impact Scale (amplitude, duration, extent, exposure)
Tagged Species PI, Institution N Tagging Location Period of Tag Activity St Bass D. Secor, UMCES 71 Potomac River Spring 2014-Winter 2016 St Bass D. Secor, UMCES 40 Potomac River Spring 2017-Winter 2019 St Bass D. Secor, UMCES A. Higgs, NYSDEC 100 Hudson River Spring 2016-Winter 2018 St Bass B. Gahagan, MADMF 224 Coastal MA Summer 2015-Winter 2017 At Sturg C. Stence, MD DNR 30 Nanticoke Fall 2015-Fall 2025 At Sturg D. Fox, DE State Univ. >500 Delaware 2010-2025 At Sturg M. Balazik, VCU 155 James River 2010-2025 At Sturg K. Dunton/M. Frisk, SUNY Stony Brook 495 NY Bight 2010-2025
Hare Tortoise MIDDLE SOUTH SOUTH Non WEA NORTH Non WEA MIDDLE NORTH
Activity Index 15 15 10 10 Receivers visited hr -1 5 5 0 Before Hermine After Hermine 0 Evacuated Remained
Fish remaining in array -12-10 -8-6 -4-2 0 0 2 4 6 8 10 12 0 2 4 6 8 10 12 14 0 5 10 15 ARIMA Intervention Analysis: Both temporary and permanent losses associated with passage of Hermine Northern Middle Southern 0 2 4 6 8 10 12 14 Observed time series Fitted time series Fitted interventions Jul Aug Sep Oct Nov Dec Jan Jul Aug Sep Oct Nov Dec Jan Jul Aug Sep Oct Nov Dec Jan