Nephrops in KattegatSkagerrak SDU FOI University of Southern

Nephrops in Kattegat/Skagerrak SDU & FOI (University of Southern Denmark & Institute of Food and Resource Economics)

Presentation • • • Data sources Collected data Challenges Suggestions/Discussion Hypotheses Questions

Data sources 1. DFAD-database – – – Based on log-book data Contains individual fishing vessel data & catches Used for general/descriptive section 2. FOI account statistics − − − Sample of commercial Danish fishing vessels Account statistics i. e. business economic data Used in socioeconomic section

Data sources (continued) 3. Danish Directorate of Fisheries − Log-book-, inspection- and violation databases − Due to confidentiality reasons data is anonymised − Used for section on control & enforcement

Collected data − Unique opportunity to have access to Danish control/enforcement data − Danish Directorate of Fisheries very cooperative and supportive − Time consuming to establish contact and access to control-data causing a delay in the upload of data

Collected data (cont. ) − Status: Data is collected and will be uploaded ASAP − Enforcement data is sensitive and can only be uploaded on an aggregated level

Challenges 1) Since enforcement data is anonymised and economic data does not include all vessels the three data sources have not been linked → Average numbers and possible loss of individual characteristics driving “violation behaviour” (Defining groups fishermen could be divided in)

Challenges (cont. ) 2) Difficulties in how to define enforcement effort. We have: − Total number of different types of inspections 20052006 (boardings- and dock-side inspections, administrative- and paper “inspections”) − − Number of inspections at individual fishing vessel level Hours at sea for inspection vessels (for total Danish sea area) → Average inspections in Kattegat/Skagerrak can be calculated

Challenges (cont. ) 3) Only average cost data of boarding and dock-side inspections → By this no idea of the shape of the cost curve (i. e. increasing/decreasing marginal costs) and different assumptions/scenarios may be needed)

Challenges (cont. ) 4) 5) Violations leading to a sanction are only included and not ongoing cases or violations without a sanction. May cause too low a number of violations and consequently violation-fraction No biological data. TACs are based on LPUE (observed and experiments), estimated fishing effort and TV-monitoring. However, Nephrop stocks appear stable (1991 -) and we are considering using a length-based model in the description of this.

Challenges (cont. ) 5) (cont. )

Challenges (cont. ) 5) (cont. )

Challenges (cont. ) 6) How to take discards of undersized Nephrops into account 7) Observed control efforts are already prioritized/focused on areas where violations have severe ecological consequences → random sample is not obtained (characteristics of violators are missing. See slide 6)

Challenges (cont. ) 8) Difficult to separate violations in fish species i. e. to identify “Nephrops-violations” from “Codviolations” if both have been landed. 9) Danish Directorate of Fisheries: “There seems no clear and unambiguous relation between the number of inspections and the detected violations” (!) (Due to changes in the control strategy? Change in behaviour of fishermen? )

9) (cont. ) Challenges (cont. )

Suggestions/Discussion a) To derive effort (in COBECOS software effort is restricted to btw. 0. 001 and 0. 999) 1. Effort = #inspections per fishing vessel Max # inspections per fish. Vessel 2. Effort = #inspections per fishing vessel Total # inspections

Suggestions/Discussion (cont. ) b) → P(inspection) = #inspections for individual vessel #fishing trips (measured as #landings) P(violation) = #violations for individual vessel #inspections for individual vessel P(fine) = #sanctions for individual vessel = 1 #violations for individual vessel P(detection) = P(insp. )*P(violation)*P(sanction) = #sanctions for individual vessel #fishing trips for individual vessel

Suggestions/Discussion (cont. ) b) (cont. ) P(detection) depends on the number of fishing trips → more fishing trips = less chance of detection. (Is this in accordance with theoretical model? ) Actual- and expected probability of being sanctioned may vary a great deal Different probabilities for first-time and previous violators? (Previous violators are – to a certain extent – more often inspected)

Hypotheses I. II. Previous violators face a higher P(inspection) – given their P(violation) - P(detection) is also higher → but P(violation) can change over time Increased control efforts increases P(detection) – given the P(violation).

Questions 1) 2) 3) 4) 5) A number of violations are “unintended” due to complicated regulation. These violations are “irrational”, how do we deal with that? Calculation of P(detection)? Expected vs. observed probabilities. Calculation? Identification of previous violators? Hypothesis I-II? (Previous slide)

Questions (cont. ) 6) 7) Distribution of supporting enforcement activities (e. g. pooling of data) on different types of enforcement effort? (Suggestion: ABC-costing? ) Can COBECOS software handle more species?

Thank for your interest as well as valuable comments, suggestions and input.