OSPAR CEMP Guidelines

____________________________________________________________________________________

OSPAR Commission OSPAR Agreement 2017-09

1

OSPAR CEMP Guidelines

Common Indicator: BH3 Extent of Physical damage to predominant and special habitats1

(OSPAR Agreement 2017-09)

This OSPAR biodiversity indicator is still in the early stages of implementation and as a result of iteration

and learning, it is anticipated that there will be evolution of the methods and approaches documented in

the CEMP guidelines. Version updates will be clearly indicated and be managed in a phased approach via

ICG-COBAM through its expert groups and with the oversight and steer of BDC.

Contents

1. Introduction 2

1.1. General introduction to the indicator 2

1.2. Components 3

2. Monitoring 4

3. Data specifications 4

3.1. Data acquisition and preparation 4

3.2. List of data sources 4

3.3. Data reporting, handling and management 4

4. Assessment method 5

4.1. Parameters and metrics 5

4.2. Assessment criteria 5

4.3. Spatial Analysis and trend analysis 6

4.4. List of Annexes 10

5. Change Management 12

6. References 13

1 Annex 5 and Annex 8 are to be completed

____________________________________________________________________________________


2

CEMP Guidelines:

Common Indicator Extent of Physical Damage to predominant and special habitats (BH3)

1. Introduction

This document provides a documentation of the methodology of the indicator ‘Extent of Physical Damage

to Predominant and Special Habitats (BH3)’. The document is an OSPAR Guideline for the Coordinated

Environmental Monitoring and Assessment Programme (CEMP). The CEMP guideline is published as OSPAR

Other Agreement 2017-09.

1.1. General introduction to the indicator

The aim of this indicator is to evaluate to what extent the sea floor and its associated ecology, species and

habitats are being damaged by human activities. The indicator is designed to assess all subtidal habitat

types at a sub-regional level i.e. predominant habitats and MSFD special habitats, including OSPAR

Threatened and/or Declining habitats (OSPAR Agreement 2008-6). It uses a combination of spatial analyses

to extrapolate data and knowledge from local studies to larger areas, and therefore it is regarded as

particularly useful for assessing large sea areas where currently only limited data are available.

Physical disturbance of the seabed by human activities such as fishing, sand extraction or offshore construction especially endangers habitats with larger and fragile species and species attached to the sea floor. In many regions of the OSPAR marine area, a shift in community composition has been reported where large and long-lived species have been replaced by small and fast-growing opportunistic species and scavengers that profit from disturbance and the availability of dead organisms (OSPAR, 2010; EEA, 2015). The impact of bottom trawling on the seafloor is considered to be the most widespread, as other activities are equally or more intense but spatially more limited.

The indicator will build upon two types of underlying information, i) the distribution and sensitivity of

habitats (resilience and resistance), and ii) the distribution and intensity of human activities and pressures

that cause physical damage, such as mobile bottom gear fisheries, sediment extraction and offshore

constructions. These two sources of information (pressure and sensitivity) are combined to calculate the

potential damage to a given seafloor habitat, and the trends across the six-year period.

At present the focus of the work is on predominant habitats according to the EU Marine Strategy

Framework Directive (MSFD) and how to incorporate special consideration of those habitats listed under

the OSPAR threatening and declining list. The EUNIS (European nature Information System)2 level 3

classification has been used as a proxy for the MSFD predominant habitats. Biogeography has been taking

into account for the development of this indicator in order to assess variations of sensitivity and

disturbance within a sub-regions containing similar physical and biological characteristics.

2 Eunis classification: http://eunis.eea.europa.eu/habitats-code-browser.jsp

____________________________________________________________________________________


3

Considerations on all special habitats according to the MSFD however should also feed into the indicator

assessment process where feasible.

The methodology used have been thoroughly tested and reviewed by national and OSPAR experts and

through focus workshops, and it represent a realistic approach to assess the distribution of impacts across

the regions based on current knowledge and using all evidence available. However, it is important to note

that the strength of any assessment is dependent on the quality of the data, and this will in turn dictate the

power and utility of the resultant information

The indicator is still under development. The following limitations should be noted:

Distribution and proportionality of partial indicator pressure data used at this stage. Using data

from >12m vessels, limits the dataset to waters largely beyond 6/12nm and therefore will

underestimate impact on those geographical areas where inshore fleets are based;

Pressure type, limited to sea bed abrasion from fishing and not including the other pressures which

result in physical damage Impacts from small vessels and information from other activities causing

physical damage will be added at a later stage.;

Parts of the indicator calculations are based on categorical approaches. Development of a

quantitate approach to assess sensitivity and disturbance for large scale assessment is currently

under development;

The indicator is not able to assess historical damage, which had caused the deterioration and

modification of habitats in the past;

Calculation of a final physical damage index per habitat type and sub-region.

1.2. Components

Biodiversity component: Benthic habitats

MSFD criterion & indicators (COM Dec 2010): 1.6 (habitat condition) and 6.1 (Physical damage, having

regard to substrate characteristics)

OSPAR Threatened and Declining Species and Habitat List:

Table 1: Habitats from the OSPAR Threatened and Declining Species and Habitat list (OSPAR Agreement 2008-6) which

could be assessed using this indicator

HABITATS

Coral Gardens

Cymodocea meadows

Deep-sea sponge aggregations

Littoral chalk communities

Lophelia pertusa reefs

Maerl beds

Modiolus modiolus beds

Oceanic ridges with hydrothermal vents/fields

Flat oyster (Ostrea edulis) beds

Sabellaria spinulosa reefs

____________________________________________________________________________________


4

Seamounts

Sea-pen and burrowing megafauna

Zostera beds

2. Monitoring

There are no specific monitoring requirements associated with this indicator, although the results on levels

of disturbance and associated confidence can be used to target monitoring programmes or one-off surveys.

It is expected that data from monitoring programmes, in particular those associated with BH1 –Condition of

Typical Species, and BH2 – Condition of Benthic habitat communities (multimetric indices) will be used to

improve the evidence base and algorithms underpinning the metrics and concepts, and to calibrate and

ground-truth the results.

3. Data specifications

3.1. Data acquisition and preparation

Data are used from pre-exsisting sources (outlined below).

3.2. List of data sources

Pressure data: Aggregated gridded VMS data for surface abrasion and subsurface abarsion; (eg.

Fishing Data based on data from ICES 2016 report:

http://www.ices.dk/sites/pub/Publication%20Reports/Advice/2016/Special_Requests/OSP

AR_further_development_of_fishing_intensity_and_pressure_mapping.pdf)

Habitat data: aggregated to broadscale habitat; (e.g.

Combined map partly produced using EMODnet outputs:

EMODnet Seafloor Habitats interactive map: http://www.emodnet-

seafloorhabitats.eu/default.aspx?page=1974&LAYERS=HabitatsCeltNorth,Region&zoom=3

&Y=51.759999999887654&X=2.269999999995032)

Benthic Species Data: (eg. Marine Recorder snapshot v442 (taken 28/08/2016)

http:jncc.defra.gov.uk/page-1599)

3.3. Data reporting, handling and management

Pressure data and Habitat data have undergone QA/QC as part of the processing undertaken in their

creation but are also subject to QA checks throughout the data processing steps of the indicator. Metadata

http://www.ices.dk/sites/pub/Publication%20Reports/Advice/2016/Special_Requests/OSPAR_further_development_of_fishing_intensity_and_pressure_mapping.pdf

http://www.ices.dk/sites/pub/Publication%20Reports/Advice/2016/Special_Requests/OSPAR_further_development_of_fishing_intensity_and_pressure_mapping.pdf

http://www.emodnet-seafloorhabitats.eu/default.aspx?page=1974&LAYERS=HabitatsCeltNorth,Region&zoom=3&Y=51.759999999887654&X=2.269999999995032



http://jncc.defra.gov.uk/page-1599

____________________________________________________________________________________


5

is also completed throughout the process to document steps accurately. Bethic species data are restricted

to those collected from 2000 onwards.

4. Assessment method

4.1. Parameters and metrics

The final parameter/metric of this indicator is the index of physical damage for a given habitat across a region.

The components of the analysis are:

A composite habitat map showing the extent and distribution of habitats (based on observational and modelled data), including the mapped extent of any relevant features (e.g. records and distribution of particular species and biotopes like EUNIS Level 5 habitats or other biological characteristics). For the purpose of this assessment a biotope is defined as ‘the combination of an abiotic habitat and its associated community of species (Connor et al. 2004). All habitat data were combined at EUNIS level 33;

Tables relating benthic habitat types to habitat sensitivity scores based on their resistance and resilience (recoverability) (Tillin et al. 2010; BioConsult, 2013; Tilling & Tyler-Walters, 2014). The sensitivity is assessed at species, biotope and level 3 level, depending on what habitat mapping information is available within each grid cell (c-square);

Distribution and intensity of pressures causing physical damage. This analysis focussed on surface

and sub-surface abrasion caused by bottom trawling (for fishing from vessels greater than 12m

only) within 0.05° grid cells (c-squares) (JNCC, 2011; ICES, 2015; Church et al, 2016);

Distribution of levels of disturbance per habitat type per year: Calculation of disturbance based on

the intensity and duration of pressures and habitat sensitivity per pressure type. Please note that

the pressures of abrasion (non-fisheries), siltation and selective extraction are not currently

included in the assessment, but will be incorporated in future developments of the indicator.

Data generated by all the elements above are combined using a step-wise approach to calculate the total area of different levels of disturbance, and a combination of those, across the region, per habitat type. The results are used to calculate the levels of variability of fishing intensity and trends in disturbance per year and across a six-year period.

4.2. Assessment criteria

Assessment unit/scale (Temporal and spatial)

The spatial assessment of this indicator is presented at EUNIS level 3, and has been prepared by combining

the sensitivity and pressure4 data from habitats, biotopes and species within the EUNIS level 3 habitat

polygons. For this assessment the OSPAR regions have been subdivided following biogeographic boundaries

(OSPAR, 2016) into: Southern North Sea (SNS) and Northern North Sea (NNS); Southern Celtic Sea (SCS) and

Northern Celtic Sea (NCS), English Channel (CH) and Bay of Biscay/Iberian Peninsula. Please note that for

3 Eunis classification is currently being revised, but it has not been singed off. For the purposes of this assessment we have used the

EUNIS classification version 2007-11

____________________________________________________________________________________


6

Region IV, only a partial assessment has been possible in this assessment, as at present insufficient habitat

and sensitivity data on the deep seafloor areas are available.

Two temporal scales are used:

1. Annually to calculate the distribution of disturbance within a year and,

2. Within an MSFD cycle (6 years) to calculate the total aggregated values for a whole cycle.

The temporal aggregation across a cycle of 5 to 6 years is calculated using the aggregation of values from

the pressure and disturbance. The method used to assess habitat and sensitivity data does not have a

temporal scale associated with the spatial layers, although within the sensitivity results the resilience values

are based mainly of the longevity of habitats and species as it is one of the key elements to assess their

recoverability

Please note the aggregation method within the MSFD cycle is currently under the development

Baseline/ reference level (To be developed and discussed)

Environmental target (To be developed and discussed)

4.3. Spatial Analysis and trend analysis

The indicator method is based on a series of analytical steps to combine the distribution and intensity of

physical damage pressures with the distribution and range of habitats and their sensitivities. The indicator

will use an additive approach for future inclusion of multiple other pressures.

Figure 1. Conceptual overview of the indicator showing the different components of the indicator.

____________________________________________________________________________________


7

An overview of the concept is showing in Figure 1, illustrating the main results produced under each of

steps of the analysis. A detailed description of each of the steps is described below:

Step 1: Extent and distribution of habitats

An important component of this indicator is the production of a composite habitat map showing the extent

and distribution of predominant and special habitats and their associated sensitivities. This map is

produced using a combination of benthic survey data and modelled habitat maps. As a basis for the

assessment, a full coverage EUNIS level 3 habitat map has been produced for the OSPAR Maritime Area,

integrating maps from surveys and broad-scale models. Please note that at present, the new EUNIS

classification has not been taken into consideration as it has not been published.

The EUNIS habitats are mapped at different levels of detail, from level 3 physical habitats to level 6

biological communities and then aggregated to EUNIS level 3 where more detail is provided. The majority

of the habitat maps were obtained from the European Marine Observation and Data Network (EMODnet)

Seafloor Habitats portal5, including the broad-scale physical habitat map, EUSeaMap (EMODnet, 2010 &

2016), and more detailed habitat maps created from survey data available through EMODnet, or as part of

OSPAR habitat data calls. The information on the coverage and type of data have been taken into account

for the calculation of confidence maps (see step 6). The specification for a habitat map for the assessment

of BH3 included the following conditions:

1. To contain information on the relevant EUNIS habitat/biotope type at any level between levels 3

and 6;

2. To refer data on biotopes to Level 3 of the EUNIS habitat classification system;

3. To use the broad-scale modelled map, EUSeaMap at EUNIS level 3 when higher resolution maps

from surveys are not available

4. To use the best available evidence on habitat data;

5. To cover the greatest possible area of the OSPAR North-East Atlantic region;

6. To contain no overlaps.

Mapping rules were established in order to decide objectively which of the overlapping datasets would be

the sole occupant in the overlapping area. Where a EUNIS habitat map developed from survey data

overlapped with a broad-scale predictive habitat map, a threshold confidence score of 58% was used as a

simple rule for deciding whether or not to favour the habitat map from survey. This threshold was based on

the MESH protocol (Mapping European Seabed Habitats) (EMODnet, 2010). Within the MESH scoring

system, for any map to have a score greater than 58%, the survey techniques must have used a

combination of remote sensing and ground-truthing to derive the habitat types, hence physical and

biological elements are included for its production. Therefore, 58% was deemed to be the lower threshold

at which an overlapping survey map is considered to be of higher quality than the broad-scale predictive

map. Pre-processing conditions and rules for the combining of data are outlined in Annex 1.

Step 2: The assessment of habitat sensitivity

The sensitivity of benthic habitats is determined based on a combination of the resistance (tolerance) and

resilience (recoverability) of key structural, functional and characterising species of the habitat in relation

to a defined intensity of each pressure Tilling et al, 2010; BioConsult, 2013; Tillin and Tyler Walters, 2014).

Due to data limitations, the sensitivity scores are defined using a categorical scoring approach (Tilling et al, 5 EMODnet Seafloor Habitats portal: www.emodnet-seafloorhabitats.eu/webgis

http://www.emodnet-seabedhabitats.eu/webgis

____________________________________________________________________________________


8

2010). Sensitivity assessments for ecological groups have also been undertaken using Bray-Curtis cluster

similarity analysis and Multidimensional Scaling, where resistance and resilience scores are assigned to

groups of species with similar biological traits (e.g. burrowers) (Tillin and Tyler Walters, 2014).

The sensitivity map is created with three steps, using best available evidence where present:

Species records from survey data, that match a list of species assigned to a specific ecological group, are mapped using their maximum sensitivity value (based on the combination of resilience and resistance). Data are plotted as the intersection between the habitat polygon and a 0.05° grid;

If there is a high enough density of species recorded and agreement between the species and the underlying habitat they are within, the sensitivity from the same species records used in step one are used to assign a modal sensitivity to the surrounding habitat polygon. For example, if data coverage is sufficient, and the substrate and habitat type in the surrounding polygons are the same, then the same sensitivies are applied to these areas;

Finally, in order to act as a background map, and to fill in areas not covered by the first two steps, the habitat map outlined in Step 1 is used to assign EUNIS level 3 benthic habitat sensitivities to the whole area. The sensitivities used in this step are often a range (from very low sensitivity to very high), in which case the maximum sensitivity is selected.

The maps are then combined geographically to show the highest confidence information across all regions.

Step 3: The assessment of the extent and distribution of physical damage pressures

For annual assessments: The first step is to determine the relevant human activities causing physical

pressures and their spatial and temporal extent. Bottom trawling is known to be affecting a large area of

the seafloor (Dinmore et al 2003; Eastwood et al, 2007; Foden et al, 2010, 2011; JNCC 2011; Jennings et al

2012) so the assessment method currently focuses on the corresponding pressures, surface abrasion

(damage to seafloor surface features) and subsurface abrasion (penetration and/or disturbance of the

substrate below the surface of the seafloor). Pre-processed vessel monitoring system (VMS) fishing data

are used to calculate the ‘swept area’ of a specific group of fishing gears (métiers6 if this level of

information is available). The swept area is calculated using the parts of the fishing gear in contact with the

seabed and it is calculated on the width of fishing gear (in metres) multiplied by the average vessel speed

(in knots) and the time fished. This calculation is undertaken on a cell-by-cell (grids or c-squares) basis per

gear and per year, using data covering six years from 2010 to 2015, with data collection and processing co-

ordintaed by ICES (See ICES Working Group Spatial Fisheries Data and Annex 4 for the detailed method).

Only the part of gear in contact with the seafloor is used for the analysis, and as a result all the gears have

been classified according to type and/or metier group (Eigaard et al, 2015; Church et al, 2016). The swept

area ratio (proportion of cell area swept per year; SAR) is then calculated by dividing the swept area by the

grid cell area. The trawling effort is classified with an intensity scale ranging from ‘none’ to ‘very high’ (cell

area swept more than 300 %). Separate Geographical Information System (GIS) layers are produced for

surface abrasion and subsurface abrasion.

For assessments across a cycle (6 years): Several statistical approaches were explored in order to analyse

the variability on the level of fishing across years, including regression analysis and the use of percentiles,

e.g. 95-percentile. However, these approaches were not deemed suitable due to not only the limited

number of years available, but also the high level of variable on the SAR values.

6 Metier is a group of fishing operations targeting a similar (assemblage of) species, using similar gear, during the same period of the year and/or within the same area and which are characterised by a similar exploitation pattern (Definition from the European Union Data Collection Framework

6).

____________________________________________________________________________________


9

A simple analysis of variance is used with the classified surface and sub-surface yearly results, in order to

differentiate between cells with low and high SAR variability. This analysis allows the distinction between

areas where fishing intensity seems to be consistent or at similar levels across years, from those where

fishing intensity levels fluctuates. A grid cell was considered to be ‘variable’ when the variance analysis

showed a change in three or more categories of the classified SAR across all years.

In order to produce a layer showing the aggregated surface and subsurface pressures that took into

account the variations on fishing pressures across years, a generic rule was used:

for cells with low variability (consistent fishing: i.e. constantly under a similar fishing pressure) the

mean of SAR across all years is calculated,

for cells with high variability (i.e. fishing pressure variable) the highest SAR value across all years is

selected to define the pressure category as it represent the maximum level of exposure within the

cycle.

Step 4: The combination of pressure intensity and habitat sensitivity

The degree of disturbance of a habitat is a prediction based on the spatial and temporal overlap of its sensitivity and exposure to a specific pressure. As a first approach to set up a disturbance matrix for the pressure ‘abrasion’, the modelling results of Schroeder et al. (2008) using fishery-induced mortality rates of selected benthic species with different ecotypes (r- and K-selected species) for the fishing gears beam and otter trawl were used as a basis. The decrease in abundance was averaged over the different species and gears to obtain a logarithmic curve for the physical impact of bottom trawling. The values derived from the function were applied to create a disturbance matrix combining sensitivity and extent of pressure (BioConsult, 2013).

The results from the sensitivity and pressure spatial layers are combined via this matrix, producing ten categories of disturbance (0-9, where 0 is no disturbance, and 9 is the greatest amount of disturbance possible). The matrix is used to calculate the disturbance for each surface and sub-surface abrasion per year. The highest value from both disturbance categories is then selected to calculate the combined disturbance values.

At present other activities which could cause physical damage pressures are not included. It is anticipated

that due to the different nature of the pressures ‘selective extraction’, ‘abrasion’ and ‘changes in siltation’,

separate disturbance matrices or algorithms will be required which will take into account the spatial

distribution of pressures and the temporal effects. This information is not currently available and will be

included on the next round of assessments for an overall calculation of disturbance caused by physical

damage.

Step 5: Disturbance aggregation method and trend analysis

Disturbance values across years are combined using the aggregated fishing pressure spatial layers,

developed in step 3 above. Results are used to calculate trend between years in those grid cells orc-

squares identified as variable. This allows the variation of disturbance across years per habitat type to be

assessed. The trend analyses are simple plots over the six-year period, rather than a linear regression which

was not possible due to the small number of years assessed. For the purpose of the aggregation a simple

rule was choosen based on expert judgemnt using using the moderate disturbance values as the middle

point for the split. Using this rule the disturbance categories were aggregated into two groups:

Disturbance categories 0 to 4, representing lower levels of disturbance;

____________________________________________________________________________________


10

Disturbance categories 5-9, representing higher levels of disturbance.

Step 6: Confidence assessments

In order to spatially represent confidence in the available data, a numeric method of calculating confidence

was adapted from OSPAR (2015). The method multiplies relative measures of confidence on a scale of 0 to

1, where there is a difference in confidence between categories or classes used in a data layer.

A numerical score (0.33, 0.66 or 1) was manually assigned by the assessor to each of the different attributes

used to create the sensitivity layer. A high confidence score was given a numeric value of 1, medium 0.66

and low 0.33. The different methods used to create the sensitivity layer were taken in turn and a numeric

confidence score was assigned to each of the attributes: confidence based on underlying data; confidence

within data source (such as MESH confidence for habitats); and confidence in the sensitivity of the habitat

to a pressure.

Further technical guidance on each of the spatial analysis and analytical steps can be found in the set of

annexes accompanying this document.

4.4. List of Annexes

Annex 1 – Development of composite habitat maps

Annex 2 – Sensitivity Assessment of Species and Habitats

Annex 3 – Development of sensitivity maps

Annex 4 – Development of surface and sub-surface abrasion layers from bottom gear fishing

Annex 5 - Development of physical damage pressure layers for non-fishing activities (To be completed)

Annex 6 – Calculation of disturbance values and trends

Annex 7 – Confidence assessments

Annex 8 – Quality control (QC) and quality assurance (QA) procedures (To be completed)

4.5. Presentation of Assessment results

Some examples of the presentation of the results of this indicator:

____________________________________________________________________________________


11

Figure 2. Aggregated Surface abrasion pressure using 2010-2015 data series. Pressure unit is swept area ratio (the

proportion of grid cell swept by fishing gear). The hatched area around the UK showing the areas where inshore

fisheries activity from vessels < 12m is higher than those >12m. Non-OSPAR Maritime area (Mediterranean and Baltic

Seas) not included.

Figure 3. Spatial distribution of aggregated disturbance using 2010-2015 data series across OSPAR sub- regions.

Disturbance categories 0-9, with 0= no disturbance and 9= highest disturbance. Plots show percentage area of OSPAR

sub-regions in disturbance categories 0-4 (none or low disturbance) and 5-9 (high disturbance) across reporting cycle

(2010-2015). The percentage was not included for the Bay of Biscay and Iberian coast due to the lack of complete

data.

____________________________________________________________________________________


12

Figure 4. Spatial distribution of confidence assessments for the physical damage indicator.

Figure 5. Percentage of total area of each habitat in disturbance category 5-9 in each of the OSPAR sub-regions for

coarse, sand, mud and mixed sediments.

5. Change Management

Responsibility for this CEMP guideline and follow up of indicator assessments falls under the OSPAR

Biodiversity Committee, the work is undertaken by the expert group for benthic habitats which provides

input to ICG-COBAM.

____________________________________________________________________________________


13

The indicator as described in this document remains under methodolical development to improve the

outputs, specifically in a number of areas:

• the lack of data from small vessel/inshore fisheries and other activities causing physical damage

across regions (e.g. offshore construction);

• review of the biogeographical assessment units and environmental data;

• review of the sensitivity and disturbance methods

• development of reference conditions;

• a better understanding of the spatial and temporal impacts of different fishing gear types.

6. References

BioConsult (2013): Seafloor integrity - Physical damage, having regard to substrate characteristics

(Descriptor 6). A conceptual approach for the assessment of indicator 6.1.2: ‘Extent of the seafloor

significantly affected by human activities for the different substrate types’. Report within the R & D project

‘Compilation and assessment of selected anthropogenic pressures in the context of the Marine Strategy

Framework Directive’, UFOPLAN 3710 25 206.

Connor, D, W., Allen, J, H., Golding, N., Howell, K, L., Lieberknecht, L, M., Northern, K, O. & Reker, J, B.

(2004) The Marine Habitat Classifcation for Britain and Ireland Version 04.05. JNCC, Peterborough ISBN 1

861 07561 8 (internet version) www.jncc.gov.uk/MarineHabitatClassifcation

Church N.J., Carter A.J., Tobin D., Edwards D., Eassom A., Cameron A., Johnson G.E., Robson, L.M. &

Webb K.E. (2016) JNCC Recommended Pressure Mapping Methodology 1. Abrasion: Methods paper

for creating a geo-data layer for the pressure ‘Physical Damage (Reversible Change) - Penetration

and/or disturbance of the substrate below the surface of the seabed, including abrasion’. JNCC report

No. 515, JNCC, Peterborough.

Dinmore, T., Duplisea, D. E., Rackham, B. D., Maxwell, D. L. & Jennings, S. (2003). Impact of a large-scale

area closure on patterns of fishing disturbance and the consequences for benthic communities. ICES Journal

of Marine Science, 60, 371-380.

Eastwood, P. D., Mills, C. M., Aldridge, J. N., Houghton, C. A. & Rogers, S. I. (2007). Human activities in UK

offshore waters: an assessment of direct, physical pressure on the seafloor. ICES Journal of Marine Science,

64, 453-463.).

Eigaard Ole R., Bastardie F., Breen M., Dinesen G. E. 1, Hintzen N. T., Laffargue P., Mortensen L. O., Nielsen

J. R., Nilsson H. C., O’Neill F. G., Polet H., Reid D. G., Sala A., Skold M., Smith C., Sørensen T. K., Tully O.,

Zengin M., and Rijnsdorp A. D. (2015). Estimating seafloor pressure from demersal trawls, seines, and

dredges based on gear design and dimensions. ICES Journal of Marine Science, xxxxx

EMODnet (2010) MESH Confidence Assessment Available at: http://www.emodnet-seafloorhabitats.eu/default.aspx?page=1635

EMODnet (2016). "EUSeaMap 2016 Broadscale predictive habitat maps - Draft Interim outputs:

http://www.emodnet-seabedhabitats.eu"

Foden, J., Rogers, S. I. & Jones, A. P. (2010). Recovery of UK seafloor habitats from benthic fishing and

aggregate extraction - towards a cumulative impact assessment. Marine Ecology Progress Series, 411, 259–

270.

http://www.jncc.gov.uk/MarineHabitatClassifcation

http://www.emodnet-seafloorhabitats.eu/default.aspx?page=1635

http://www.emodnet-seafloorhabitats.eu/default.aspx?page=1635

____________________________________________________________________________________


14

Foden, J., Rogers, S. I. & Jones, A. P. (2011). Human pressures on UK seafloor habitats: a cumulative impact

assessment. Marine Ecology Progress Series, 428, 33–47.

ICES (2015). Report of the Working Group on Spatial Fisheries Data (WGSFD).

http://www.ices.dk/sites/pub/Publication%20Reports/Expert%20Group%20Report/SSGEPI/2015/01%20W

GSFD%20-%20Report%20of%20the%20Working%20Group%20on%20Spatial%20Fisheries%20Data.pdf

Jennings, S., Lee, J., & Hiddink, J. G. (2012). Assessing fishery footprints and trade-offs between landings

value, habtat sensitivity, and fishing impacts to inform marine spatial planning and an ecosystem approach.

ICES Journal of Marine Science, 1-11.

JNCC (2011). Review of methods for mapping anthropogenic pressures in UK waters in support of the

Marine Biodiversity Monitoring R&D Programme. Briefing paper to UKMMAS evdience groups. Presented

06/10/2011.

OSPAR (2010). Quality Status Report 2010. OSPAR Commission. London.

OSPAR (2015). Confidence and Uncertainty in assessing cumulative effects. Presented at the OSPAR

Meeting of the Intercessional Correspondence Group on Cumulative Effects, ljmuiden (NL): 26-27 Feb 2015

"Bringing it all together”

OSPAR (2016) Intermediate Assessment 2017 Resources, http://www.ospar.org/work-areas/cross-cutting-

issues/intermediate-assessment-2017-resources, Accessed online: 01/06/2016

Schroeder, A., L. Gutow & M. Gusky (2008): FishPact. Auswirkungen von Grund-schleppnetzfischereien

sowie von Sand- und Kiesabbauvorhaben auf die Meeres-bodenstruktur und das Benthos in den

Schutzgebieten der deutschen AWZ der Nordsee (MAR 36032/15). Report for the Federal Agency for

Nature Conservation.

EEA (2015). State of Europe’s Seas, European Environment Agency, Report N2/2015

Tillin, H.M., S.C. Hull & H. Tyler-Walters (2010): Development of a Sensitivity Matrix (pressures-MCZ/MPA

features). Defra Contract No. MB0102 Task 3A, Report No. 22.

http://jncc.defra.gov.uk/pdf/MB0102_Sensitivity_Assessment%5B1%5D.pdf

Tillin, H. & Tyler-Walters, H., (2014): Assessing the sensitivity of subtidal sedimentary habitats to pressures

associated with marine activities - Phase 1 Report, JNCC Report 512A. http://jncc.defra.gov.uk/page-6790


associated with marine activities - Phase 2 Report, JNCC Report 512B. http://jncc.defra.gov.uk/page-6929


associated with marine activities - Phase 3 Sensitivity Proformas (not published on the JNCC website but

can be supplied upon request).






OSPAR CEMP Guidelines - Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications: Annex 1: Development of a composite habitat map for OSPAR Regions

1

Annex 1: Development of a composite habitat map for OSPAR Regions

Introduction

The specification for a habitat map for the assessment of BH3 included the following conditions:

1. To contain biotope data or smaller units of EUNIS classification (e.g. EUNIS level 6) 2. To refer data on biotopes to level 3 of the EUNIS habitat classification system; 3. To use broad-scale predictive EUNIS level 3 data when high resolution data are not

available 4. To use the best available evidence; 5. To cover the greatest possible area of the OSPAR North-East Atlantic region; 6. To contain no overlaps.

Whilst the UK’s full-coverage EUNIS level 3 layer integrating maps from surveys and broad-scale

models1 provides a map to meet these conditions for the UK continental shelf, further work was

required to extend to the OSPAR area.

The datasets listed below were combined together following the method in a way that ensured that

only one data source was present at any one location.

Input datasets used

The following complete datasets were used in the combination method, the blue text in brackets

describing the short name used in the method text.

UK full coverage EUNIS level 3 layer integrating maps from surveys and the EUSeaMap model2 (UK combined map);

Broad-scale predictive physical habitat map for the Atlantic, Greater North Sea and Celtic Seas3, interim 2016 version (EUSeaMap broad-scale map);

EUNIS habitat maps from surveys4 (EUNIS survey maps).

Other habitat maps submitted for the assessment (Biotope survey maps)

1 UK full-coverage EUNIS level 3 layer integrating maps from surveys and broad-scale models method can be found here:

http://jncc.defra.gov.uk/page-6655#EUNIScombined

2 Method available from: http://jncc.defra.gov.uk/page-6655#EUNIScombined

3 Made available by EMODnet Seabed Habitats : http://www.emodnet-seabedhabitats.eu

4 Available on the EMODnet Seabed Habitats interactive map: http://www.emodnet-

seabedhabitats.eu/default.aspx?page=1974&LAYERS=Eunis,Region&zoom=3&Y=51.759999999887654&X=2.269999999995

032



http://www.emodnet-seabedhabitats.eu/

http://www.emodnet-seabedhabitats.eu/default.aspx?page=1974&LAYERS=Eunis,Region&zoom=3&Y=51.759999999887654&X=2.269999999995032




2

Method

Summary of method

The input datasets intersect in certain geographic areas. Therefore, to comply with the “no overlaps”

rule, other rules were established in order to objectively decide which of the overlapping datasets

would be the sole occupant the overlapping area.

EUNIS habitat maps from surveys from the EMODnet Seabed Habitats portal are supplied with an

accompanying MESH confidence score5, which describes the quality of the processes that were used

to create the map in question.

Where the OSPAR combined map overlapped anything, it was favoured.

Where EUNIS habitat maps from surveys overlapped, the map with the highest MESH confidence score was favoured.

Where a EUNIS survey map overlapped with the EUSeaMap broad-scale map a threshold MESH confidence score of 58 % was used as a simple rule for deciding whether or not to favour the survey map. Within the MESH scoring system, for any map to have a score greater than 58 %, the survey techniques must have used a combination of remote sensing and ground-truthing to derive the habitat types. Therefore, 58 % was deemed to be the lower threshold at which an overlapping survey map “wins” against a broad-scale map.

Once these rules were set, a method of Erase and Merge was used to ensure that the differing

datasets would not intersect when compiled together. The process creates a ‘cookie cutter’ hole of

the dataset to be input, before merging the two datasets together (FigureA1.1).

Pre-processing the input datasets

Before the datasets could be combined, their attribute tables were altered in the following way to

ensure conformity in the conflation process:

5 MESH confidence assessment method available from: http://www.emodnet-seabedhabitats.eu/default.aspx?page=1635

FigureA1.1. Example of the Erase/merge process. From an initial layer (1) the area covered by a "superior" map is cut out with the erase tool (2) and then the secondary map inserted in place with the merge tool (3). This ensures that no overlaps exist between the two layers once merged.

http://www.emodnet-seabedhabitats.eu/default.aspx?page=1635


3

String field “GUI6” was added to the EUSeaMap broad-scale map for auditing purposes. and attributed a single GUI with the value ‘EUSM2016’.

Integer field Polygon was added to the EUSeaMap broad-scale map for auditing purposes.The value was equal to the “OBJECTID” value.

Field “EUNIS_L3” was added to all datasets where it was not already present.

Combining the datasets

1. “Select By Attributes” was used to select all features from the EUNIS survey maps or Biotope survey maps where the MESH confidence scores were greater than 58 and these were exported as a new layer.

2. The ArcGIS ERASE tool was used with the layer created in step 1 as the erase layer and EUSeaMap broad-scale map as the input layer.

3. The ArcGIS MERGE tool was used to merge the layer created in step 1 with the layer created in step 2.

4. The ArcGIS ERASE tool was used with the UK combined map as the erase layer and the layer created in step 3 as the input layer.

5. The ArcGIS MERGE tool was used to merge the UK combined map with the layer created in step 4.

6. For the final layer created in step 5, values were entered for “EUNIS_L3” via the field calculator.

a. Where the values in “HAB_TYPE” were valid EUNIS codes, the first 4 characters from “HAB_TYPE” were copied to “EUNIS_L3” via the field calculator7.

b. Where the values in “HAB_TYPE” were not valid EUNIS codes (for example ‘DSea CSed’ or ‘Deep circalittoral seabed’), the entire field of “HAB_TYPE” was copied to “EUNIS_L3”.

Final Product

The output is a single feature class containing no overlaps, standardised so that the field “EUNIS_L3”

can be used across the dataset for habitat information and the “POLYGON” and “GUI” fields can be

used as an audit trail to the original source.

The coverage of the final product and its respective sources can be seen in Figure A1.2, and a first

draft of the combined map in Figure A1.3.

To find metadata and original study boundaries associated with each original habitat map from

survey, you may type the 8-digit code from the GUI field into the metadata search page on the

EMODnet Seabed Habitats portal: http://www.emodnet-seabedhabitats.eu/search.

6 Globally Unique Identifier, a field from the EMODnet Seabed Habitats data exchange formats to refer to a specific dataset,

taking the form of a 2 letter ISO 3166-1 alpha-2 country code followed by six digits (e.g. PT001001).

7 Parser: VB Script. Code: EUNIS_L3 = Left( [HAB_TYPE] , 4 )

http://www.emodnet-seabedhabitats.eu/search

http://www.emodnet-seabedhabitats.eu/def


4

Figure A1.2. Coverage of the final habitat map showing the data sourced from each input dataset.


5

Figure A1.3. Combined EUNIS Level 3 Habitat Map for OSPAR Regions produced by JNCC for BH3-

physical damage indicator

OSPAR CEMP guidelines – Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications: Annex 2: Sensitivity Assessment of species and habitats

1

Annex 2: Sensitivity Assessment of Species and Habitats

Introduction

Sensitivity of ecosystem components are determined by two aspects: the ability to

withstand disturbance or stress (resistance or tolerance) and the ability and time needed to

recover from a perturbation and return to the previous state (resilience or recoverability).

(Tillin et al. 2010). Both elements are essential for a sensitivity analysis, in particular when

the data are going to be evaluated to assess not just impact at a particular location at a

point in time, but also the temporal effects of pressures over time. A species or habitat that

is considered sensitive is therefore one that has both low levels of resistance and resilience,

whereas a species that is considered to be not sensitive or has low sensitivity is one with

both high levels of resistance and resilience.

Sensitivity assessments of an ecosystem

The sensitivity of an ecosystem can be assessed in a number of different ways. More

traditional methods have assessed the sensitivity of a broad scale habitats on a categorical

scale of none to high based on expert judgement (e.g. Tillin et al., 2010), such as the

information used in the UK sensitivity assessment carried out for MCZ designation. More

recent methods however, analyse the functional groups within the habitat based on

physical traits of key and characterising species, to assign sensitivities to these groups (e.g.

Tillin & Tyler-Walters, H. 2014 a). The use of ecological groups allows the use of a wide set

of data for the analysis of sensitivity, instead of only using biotopes or other habitat polygon

data.

Characteristic species should be those that significantly influence the ecology of habitat

(Bildstein et al. 2014; BioConsult 2013; Rachor & Nehmer 2003; Rachor 2007; Tillin & Tyler-

Walters, H 2014a). They could be species which provide a distinct habitat that supports an

associated community, or one that is important for community functioning through

interactions with other species, or species which are used for the definition of a habitat. The

loss or degradation of one of these species would severely affect the viability, structure and

function of the habitat and may result in the loss of the habitat or a changed classification.

Ecological groups should not be species specific, but rather consist of groups of ecologically

similar species, e.g. fragile erect epifauna on cobbles and boulders (Tillin & Tyler-Walters, H

2014a, 2014b, 2014c).

Examples of this approach include 16 ecological groups identified within subtidal sediments

within UK waters, comprising 96 characterising species (Tillin et al. 2014) and identification


2

of 45 characterising species within the German EEZ (Bildstein et al. 2014). Both of these

studies use multivariate ordination (nMDS) and Bray-Curtis similarity cluster analyses to

identify the ecologically important groups and species. For a full method description, please

refer to the reports referenced.

As this indicator is carried out at a regional scale, it is acknowledged that species may react

differently in different marine regions due to variable environmental factors, the

characteristics of local populations and their role in the benthic assemblage. We therefore

combine sensitivity assessments specific to regions in order to provide the most

representative result. Despite this flexibility, it has to be ensured that sensitivity is based on

the same underlying factors; resistance and resilience in relation to a specific pressure and

that the same scales for resistance and resilience are used as well as the same combining

matrix.

Assessment of resistance

Resistance (tolerance) of a species or habitat reflects the susceptibility to damage or loss as

a result of a pressure on the seabed (Holling, 1973). The likely resistance is estimated with

respect to a specified magnitude and duration of change in order to provide a standard level

against which to assess. Resistance of a species or habitat is assessed with the following

scale:

Table A2.1. Assessment scale used for determining resistance of a species or habitat

Resistance Description

None Severe decline and/or physical-chemical parameters also affected e.g. removal of habitat that could cause a change of habitat type. A severe decline/ reduction relates to the loss of more than 75 % of the extent, density or abundance of the selected species or habitat element.

Low Significant mortality of species with some effects on physical-chemical character of habitat. A significant decline/reduction relates to the loss of 25%-75% of the extent, density or abundance of the selected species or habitat element.

Medium Some mortality of species without change to habitat type. ‘Some mortality’ relates to the loss of up to 25% of the extent, density or abundance of the selected species or habitat element.

High No significant effects to the physical-chemical character of habitat and no effect on population viability of species but potential effects to biological processes like feeding, respiration and reproductive rates.

Assessment of resilience


3

Resilience (recoverability) describes the ability of a habitat or species population to restore

from damage sustained as a result of a physical impact on the seabed (Holling, 1973).

Resilience of organisms is especially dependent on the ability of the species to regenerate,

regrow, recruit or recolonize and the extent of damage incurred. Recovery is only possible

when the impact has stopped or has been removed. Resilience of characteristic species is

assessed with the following scale:

Table A2.2. Assessment scale for resilience

Resilience Description

Very Low At least 25 years to recover structure and function

Low Full recovery within 10-25 years

Medium Full recovery within 2-10 years

High Full recovery within 1-2 years

Very high Full recovery within 1 year

Sensitivity matrix (combination of resistance and resilience)

The resistance and resilience scores are combined to produce an overall sensitivity score for

a species or habitat. A matrix is used to automate this combination and results in a category

of sensitivity, ranging from 1 to 5 (with 5 being the most sensitive). This matrix can be

applied to different sensitivity assessments.

Table A2.3. Sensitivity matrix combing resistance and resilience scores to produce a

sensitivity score ranging from 1 to 5, where 5 is the most sensitive.


4

Sensitivity

Resilience

very low (>25 yr.)

low (>10-25

yr.)

medium (>2-10 yr.)

high (1-2 yr.)

very high (<1 yr.)

Res

ista

nce

none 5 4 4 3 2

low 4 4 3 3 2

medium 4 3 3 2 1

high 3 3 2 2 1

References

Bildstein, T., D. Fiorentino, C.-P. Günther, R. Pesch, P. Rückert, W. Schröder, B. Schuchardt (2014): Cluster 6 Biotopkartierung: Endberichtsentwurf - Teil Nordsee. Report on behalf of the Bundesamt für Naturschutz. BioConsult (2013): Seafloor integrity - Physical damage, having regard to substrate characteristics (Descriptor 6). A conceptual approach for the assessment of indicator 6.1.2: ‘Extent of the seabed significantly affected by human activities for the different substrate types’. Report within the R & D project ‘Compilation and assessment of selected anthropogenic pressures in the context of the Marine Strategy Framework Directive’, UFOPLAN 3710 25 206. Holling, C, S. (1973) Resilience and Stability of Ecological Systems, Annual Review of Ecology and Systems, 4, 1-23. Rachor, E.& Nehmer, P. (2003): Erfassung und Bewertung ökologisch wertvoller Lebensräume in der Nordsee. Rachor, E.; Reiss, H.; Degraer, S.; Duineveld, GCA.; van Hoey, G.; Lavaleye, M.; Willems, W.; Rees, H.L. (2007): Structure, distribution, and characterizing species of North Sea macrozoobenthos communities in 2000. In: Rees HL, Eggleton JD, Rachor E, Berghe E van der (eds): Structure and dynamics of the North Sea benthos. Copenhagen, p 46–59. Salzwedel, H.; Rachor, E.; Gerdes, D. (1985): Benthic macrofauna communities in the German Bight. -Veröffentlichungen des Institutes für Meeresforschung Bremerhaven 20: 199-267. Tillin, H.M., S.C. Hull & H. Tyler-Walters (2010): Development of a Sensitivity Matrix (pressures-MCZ/MPA features). Defra Contract No. MB0102 Task 3A, Report No. 22. http://jncc.defra.gov.uk/pdf/MB0102_Sensitivity_Assessment%5B1%5D.pdf



5

Tillin, H. & Tyler-Walters, H., (2014a): Assessing the sensitivity of subtidal sedimentary habitats to pressures associated with marine activities - Phase 1 Report, JNCC Report 512A. http://jncc.defra.gov.uk/page-6790 Tillin, H. & Tyler-Walters, H., (2014b): Assessing the sensitivity of subtidal sedimentary habitats to pressures associated with marine activities - Phase 2 Report, JNCC Report 512B. http://jncc.defra.gov.uk/page-6929 Tillin, H. & Tyler-Walters, H., (2014c): Assessing the sensitivity of subtidal sedimentary habitats to pressures associated with marine activities - Phase 3 Sensitivity Proformas (not published on the JNCC website but can be supplied upon request).



OSPAR CEMP guidelines – Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications: Annex 3: Development of sensitivity map

1

Annex 3: Development of Sensitivity maps General Approach In order to undertake the most reliable sensitivity assessments of habitats, the best available evidence should be used where it is available. We are therefore proposing to take a layered approach when creating sensitivity maps, placing areas with higher confidences of sensitivity on top of areas of lower confidence. For the UK, the best available evidence of sensitivity comprises of the species records held in the database ‘Marine Recorder’ combined with the sensitivities assigned via Tillin & Tyler-Walters (2014) for subtidal sediments and Maher et al. (2016) for rocky habitats following the approach to sensitivity outlined in Annex 4. Species records held in Marine Recorder that match the list of characterising species outlined in Tillin & Tyler Walters and () can be mapped with their associated sensitivities. This will be done by matching biotope records from marine recorder to their associated sensitivities from Tillin et al. (2010). Due to the large spatial scale of the indicator, we propose to use the sensitivity information described in the step above and to extrapolate this to habitat polygons provided using the combined habitat map outlined in Annex 3. This step involves validation measures and the use of confidence to ensure that wrong assumptions are not being made. Finally, where there is no sensitivity information provided by either of the two steps mentioned above, we propose to use sensitivities assigned to EUNIS level 3 from Tillin et al. (2010) to the combined map outlined in Annex 1.


2

Step 1 Step 2 Step 3

Confidence in data

Survey Database

Species record, Lat long location

Speciessensitivity value

Sensitivity assessment

Map with highest sensitivity per grid cell from survey records

Species record, lat long location + Grid

Combined survey and modelled map

Modelled map

Polygons from the combined map with species + sensitivities +Grid

Polygons from the combined map with species + sensitivities

Sensitivity assessment

BSHpolygon records + sensitivity

BSH sensitivityvalue

BSHpolygon records + sensitivityGrid

Map with highest sensitivity per BSH polygon frommodelled map

Map with modal sensitivity per habitat polygon extrapolated from survey records

Final sensitivity map

Figure A3.1. Outline of the proposal for the three step process used in creation of sensitivity map of the BH3 indicator


3

Step 1- Using survey data

In this step, survey point data is assigned to a sensitivity score. This includes species records (UK species data are taken from the database Marine Recorder; available to download from http://jncc.defra.gov.uk/page-1599) corresponding to the list of subtidal sediment characterising species outlined in Tillin & Tyler- Walters (2014) or rocky characterising species outlined in (). For German waters a list of characterising species was also included in this step. An exercise was carried out to align the sensitivities between the countries. Species data from survey maps submitted by other countries could also be used as part of this step of the analysis if similar alignment methods were undertaken.

Once the points have been assigned a relevant sensitivity score, they are joined to an intersect of the habitat map outlined in Annex 1 and a grid with the dimensions 0.05 o x 0.05 o, enabling the user to assign a sensitivity score to the immediate area surrounding the points. The grid used has the same dimensions used in the VMS fishing pressure squares, which will help to facilitate the combination of pressure and sensitivity information later in the process. By using the intersect of the habitat map and the grid, detail of small sensitive areas is not over represented. In the cases where a sensitivity range was found due to multiple survey records per grid cell, the maximum sensitivity values were selected. Selecting the maximum sensitivity per grid cell reduces the risk of missing any highly sensitive species.

The steps taken in order to achieve this are captured in Table A3.1 and the output from this stage can be seen in Figure A3.2.

Table A3.1. Outline of steps taken to produce a sensitivity map using survey point data

Step Programme used

Product

Query survey point data and assign sensitivity scores (including numeric representations of sensitivity scores, eg. High=3, medium=2, low=1, not sensitive=0)

Access 1.1- Table with survey point (species or biotope) record, location, sensitivity

Import table into Arc and use ArcGIS tool CLIP to restrict to required extent (eg. UK EEZ, OSPAR boundary)

Arc GIS 1.2- Spatial display of points with associated sensitivities in required area

Use ArcGIS tool INTERSECT to intersect habitat map and vector grid layer1.

Arc GIS 1.3- Intersect of habitat map and grid.

Use ArcGIS tool SPATIAL JOIN to join the points (product 1.2) to

Arc GIS 1.4- Grid vector layer with maximum sensitivity per cell

1 The grid polygon layer used in this step is the same 0.05 x 0.05 minute grid used in the UK VMS pressure layers


4

habitat map- grid intersect (product 1.3) using ‘one to one’ feature and setting merge rule for sensitivity value to maximum.

(See Figure 5-2).

Figure A3.2. Example output of Step 1, using survey data to map sensitivity to a pressure (for this example, surface abrasion). For the UK, we have used species records that have sensitivities assigned via Tillin & Tyler-Walters (2014) ‘Ecological Groups’ sensitivity assessment along with Tillin et al. (2010).

Step 2- Extrapolating sensitivity information from Step 1

In step 2, the point data from step 1 is joined to the habitat map and the sensitivity information assigned to the habitat polygon that it falls inside. Within each polygon, the modal sensitivity is selected to provide the best representation of the sensitivity taking into account the risk of assigning a misrepresentative sensitivity to a large area. This step is carried out separately for rock and sediment habitats to ensure correlation between the points and habitat polygons. A threshold density of points per area of habitat polygon is


5

applied at this step to prevent large areas being assigned a sensitivity from a single point. The steps taken in order to achieve these are outlined in table A3.2 and the output from Step 2 can be seen below in Figure A3.3.

Table A3.2. Outline of steps taken to extrapolate survey point sensitivity values to habitat polygons

Step Programme used

Product

Create a unique ID number for each polygon in habitat map

Arc GIS 2.1Habitat map with unique ID for each polygon

Select ArcGIS tool SPATIAL JOIN between Product 1.1 (spatial distribution of survey data points with associated sensitivities) and Product 2.1. using ‘one to one’ feature and setting merge rule for sensitivity value to mode.

Arc GIS 2.2- Habitat map with modal sensitivity per polygon from survey records in Step 1.

Calculate area of each polygon and ‘count’ of number of points per polygon. Divide area by no. of points to calculate point density and exclude any polygons with less than 1 point per 20km2.

2.2- Habitat map with modal sensitivity per polygon from survey records in Step 1.


6

Figure 3.3. Output of Step 2- extrapolating the sensitivity information from survey point records to habitat polygons. This example from the UK shows the modal sensitivity of species records assigned sensitivities via Tillin & Tyler-Walters (2014) ‘Ecological Groups’ sensitivity assessment per habitat polygon.

Step 3- Using modelled habitat maps

This step provides full area coverage of contracting parties’ marine area, utilising the combined habitat map outlined in Annex 1, mapped to EUNIS level 3 sensitivity scores taken from Tillin et al. (2010). Where a range of sensitivities is given for a habitat, we select the maximum. The output of this step can be seen in Figure A3.4 and the steps are outlined in greater detail in Table A3.6.

Table A3.-1: Outline of steps taken to map EUNIS level 3 sensitivities to combined habitat map.

Step Programme used Product

Join EUNIS L3 sensitivities to combined habitat map

Arc GIS 3.1- Combined habitat map with associated sensitivity


7

Figure A3.4: . Output of step 3: Combined habitat map (Annex 3) with maximum EUNIS level 3 surface abrasion sensitivities displayed, taken from Tillin et al. 2008 “MB0102”.

Combining Steps 1, 2 and 3

We combine the maps outlined in Steps 1, 2 and 3 above by ‘placing’ them on top of one another showing the information with the better confidence on top of the maps with lower confidences. This uses the same ‘Erase and Merge’ method used in the production of the combined habitat map, which is described in Figure A1.1. The process for combining the sources of sensitivity information can be seen in Figure A3.5 and the steps taken are outlined in table _ below.

Step Programme used

Product

Use the Arc tool ERASE to erase the area Arc GIS 4.1- 2.3 with the area of 1.4


8

covered by the output of Step 1 from the output of Step 2

removed.

Use the Arc tool ERASE to erase the outputs of Steps 1 (1.4) and 2 (2.3) from Step 3 (3.1)

Arc GIS 4.2 3.1 with the areas of 1.4 and 2.3 removed.

Use the Arc tool MERGE to Merge the output of Step 1 (1.3), product 4.1 and product 4.2

Arc GIS 4.3- Single layer sensitivity map with information combined from steps 1, 2 and 3.


9

Figure A3.5: Diagram displaying the combining sensitivity maps produced in Steps 1, 2 and 3 showing the layers with higher confidence above layers with lower confidence.

OSPAR CEMP guidelines – Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications Annex 4: Development of surface and sub-surface abrasion layers from fishing with bottom-contact gear

1

Annex 4: Development of surface and sub-surface abrasion layers from fishing with bottom-contact gear

Introduction A review was conducted by JNCC on the current methods used to map fishing activity and the abrasive pressure it can exert on seabed habitats (JNCC 2011;). Based on this review, a series of recommendations were made to inform a common approach to mapping abrasion. The study area was limited to the UK continental shelf designated area within the British fisheries limit. The study specifically focused on turning fishing activity information for vessels greater than 12 m in length into a meaningful abrasion pressure score; small vessel activity was not considered (Church et al, 2016). The proposed method has been further tested and modified by the ICES WG SFD (2015, 2016) to validate the recommendations and produce the final method for the whole of OSPAR.

Methodology:

Data Two types of data may be used for this method:

Anonymised VMS ’ping’ data; or Aggregated VMS data grids.

Anonymised VMS ‘ping’ data are not currently available to all data users, due to their commercially sensitive nature. Data are sometimes aggregated to a grid before release to anonymise the movements of individual fishing vessels. Both ‘ping’ and gridded data are considered to make the approach relevant to most users. To create a swept area abrasion data layer at the OSPAR regional scales, the method needs to be applied to all gridded VMS data submitted by participating countries. Please note that for the years 2009–2011 VMS was mandatory for fishing vessels larger than 15 m and during the years 2012/2013 VMS was mandatory for fishing vessels larger than 12 m. For a number of reasons, not all vessels in the 12–15 m category were VMS enabled by 2013 so it is likely that more vessels will transmit data in this size class over the coming years.

Pre-processing VMS ping data were linked to vessel logbook information in order to determine location and the type of fishing gear being deployed. A speed rule filter is used to distinguish between fishing and non-fishing events. Decisions on the most appropriate filter to be applied were taken by the submitting body, although boats travelling between 1-6 knots is considered indicative of active fishing (Mills et al 2007; Lee et al. 2010). As a result, all data received by ICES was assumed to represent fishing activity only (see associated caveats 4.1.6 of ICES report). Point data representing VMS pings were assigned to c-square -grids at a resolution of 0.05° x 0.05° decimal degrees (Geographical Coordinate System World Geodetic System


2

1984 (GCS WGS84)) i.e. unprojected. As such, the base unit for calculating swept area in the methodology is the c-square grid cell (c-squares provide a simplified system that allows for spatial indexing of geographic data e.g. fishing effort). Within each grid cell the number of pings was summed based on the time interval between pings, which produced an estimate of total fishing time for each cell. The grid was aligned to, and was a subdivision of, larger ICES rectangles to allow comparison with catch data that are collected at this scale (Lee et al. 2010). Te sum of the interval between all the fishings pings occurring in each grid cell is used to produce the estimated total fishing time. The ICES WG SFD has developed a template for the ongoing submission of VMS and Log-book data in collaboration with OSPAR to standardise the data used for these calculations.

Gear types A list of fishing gears for inclusion within this method was developed from the available gear types listed within the Defra MB0106 project (Lee et al. 2010). The predominant activities associated with physical abrasion, in terms of extent and intensity, are ‘fishing: demersal trawling’ and ‘fishing: dredging’. Gear coding in logbooks is not typically suited for quantitative estimation of seafloor pressure (swept area and impact severity). The EU FP7 “BENTHIS” project developed a method to overcome this information deficiency. The approach is to use the relationships between gear dimensions and vessel size (e.g. trawl door spread and vessel engine power (kW)) for different métiers to assign quantitative information on bottom contact (e.g. width of gear). Métier is a group of fishing operations targeting a similar (assemblage of) species, using similar gear, during the same period of the year and/or within the same area and which are characterised by a similar exploitation pattern (Definition from the Data Collection Framework1). As part of this project a list of 14 different functional gear categories was created and then populated with information on vessel size (m), power (kW) and gear specifications for each métier collected in a pan-European industry-based questionnaire survey (Eigaard et al, 2015). This study enabled statistical modelling of the vessel size/ engine power and gear size relationships for different métiers. The estimates of width of fishing gear causing abrasion (surface and subsurface) and speed for each of the BENTHIS métier and JNCC gear group can be found in Table A4.1

Gear widths and speeds Different gear types interact with the seabed in different ways and subsequently exert different levels of abrasive pressure both in terms of the area of substrate affected and the penetration depth. These considerations are central to the recommended method as gear width (determined by gear type) is a key component of the swept area calculation and can contribute differently to the area estimates of surface and subsurface abrasion respectively (Church et al, 2016).

1 http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2008:346:0037:0088:EN:PDF

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2008:346:0037:0088:EN:PDF


3

Due to differences in the characteristics of target species, otter trawls, beam trawls and scallop dredges vary in their physical interaction with the sea bed. Generally speaking, demersal otter trawls are designed to target fish and invertebrates close to the seabed while beam trawls and scallop dredges target species that live on the seabed or are partially buried in the sediment (Løkkeborg, 2004). Different components of the gear can interact with the sea bed in different ways. For instance, for otter trawls, the towed otter doors represent a relatively small spatial footprint but penetrate deeply into the seabed, as opposed to the ground ropes (foot-rope) between the doors which have a relatively much larger spatial footprint, but do not typically penetrate the seabed as deeply. The sweeps/bridles have the largest contact area with the seabed, however, the degree of impact from these components is still poorly understood (Valdemarsen et al. 2007). Subsequently, otter doors have the potential to disturb both subsurface infaunal and epifaunal communities, whereas ground ropes are likely to only disturb surface epifaunal communities. In linking the abrasion pressure to benthic response and habitat vulnerability, splitting the pressure into surface and subsurface components allows better discrimination of the potential ecological effects. Beam trawling involves the use of a rigid, typically metal, beam to keep the trawl net open instead of hydrodynamic otter boards, and so generally sweeps a smaller total area. Several components on the trawl (shoes, tickler chains, chain mat) are potentially capable of penetrating the sea bed across the width of the gear (Bergman and van Santbrink, 2000). The penetration depth of a beam trawl depends on the weight of the gear and the towing speed, but also on the type of substrate (Paschen et al., 2000). Similarly scallop dredges are specifically designed to disturb the sea bed surface and penetrate the upper few centimetres of the sediment with dredge teeth mounted along the whole width of the gear (Løkkeborg, 2004). The lists of gear groups (based on the level 4 classification), including the range of values for width and speed used can be found at Church et al (2016), and for the metiers in Eigaard et al, (2015)

Swept area ‘Swept area’ is generally considered to be an estimate of the area of seabed in contact with the fishing gear. It is a function of gear width, vessel speed and fishing effort (JNCC 2011). The area of seabed swept by a vessel was calculated per gear type per annum. The swept area method can be applied to both aggregated VMS and VMS pings. This calculation was carried out for each demersal and dredging fishing gear type:

1. For the VMS pings, ‘Swept area’, SA, (m2) can be calculated per ping, multiplying the ‘width of fishing gear’, w, (m) by the ‘recorded speed’, v, (m.min-1) and the ‘time fished’ (each ping representing the area swept since the last recorded ping), e, (min) to get an estimate of area covered per gear, per ping (Equation 1). The pings were then aggregated by summing on a grid at 0.05 0 (which aligned with the aggregated VMS data). The resulting swept area was calculated as m2, per cell, per annum (m2.cell-1.yr-1)


4

2. For aggregated data the area of seabed swept by a vessel was calculated per Benthis

métier group (gear type in absence of metier, see Table A1-4) per annum and based

on the abrasion methodology proposed by ICES WG SFD (2015) and Church et al

2016. The fishing area swept (Swept Area) was calculated per grid cell (m2.cell-1). It

was calculated by multiplying the values of bottom contact, w, (m) by the ‘average

vessel speed’ v, (m.hr-1) for the relevant métier, and the ‘time fished’, e, (hour) to get

an estimate of area covered per gear (Equation 1). The final output is calculated as

area swept per cell, per annum (m2.cell-1.yr-1). This is aggregated across métiers for

each gear class (Otter trawl, Beam trawl, Dredge and Demersal seine) per year and

across years. Two data layers are produced for each gear class; one for ‘surface’

abrasion and one for ‘subsurface’ abrasion.

Equation 1 – Swept Area calculation

SA = ∑ 𝒆vw

Where SA is the swept area, e is the number of minutes between pings, w is total width of fishing gear (m) causing abrasion, v is average speed vessel (m/min)

A swept-area ratio, SAr, was then calculated to account for the varying cell size of the GCS WGS84 grid. To produce the swept-area ratio (i.e. area swept in terms of the proportion of the cell fished within the time period, measured as a ratio of the area of the cell), SA was divided by the actual grid cell area, CA (Equation 2). Equation 2 – Swept-Area 3.4.1 ratio calculation

𝑺𝑨𝒓 =𝑺𝑨

𝑪𝑨

Where SA is the swept area, CA is cell area and SAr is swept area ratio (number of times the

cell was swept).

For both datasets, two standardised GIS layers were created per year; one for ‘surface’ abrasion and one for ‘subsurface’ abrasion. Steps for the data analysis The method for the analysis of the VMS data includes several steps. These have been defined using the OSPAR data call through ICES on VMS and logbook data, but they are equally applicable to any datasets contains VMS data. For an overview of the steps see Figure A4.1.


5

Steps for the data analysis are:

Step 1 - Full QA of all data submitted. Any data which seems erroneous should be excluded (an overview of quality checks can be found on the ICES WGSFD report (2015)).

Step 2 - Assign all data for mobile bottom contacting gears to BENTHIS métiers where possible. The estimates of width of fishing gear causing abrasion (surface and subsurface) and speed for each BENTHIS Métier and JNCC gear group can be found in table A4.1. With the current information it is possible to assign over 80 % of available data to Benthis métiers and above 10 % to more general JNCC gear groups, however there are some gear groups that are un-assigned.2 .

Step 3 – Calculate fishing effort (hours). Based on the outputs of step 1, estimates of fishing effort were aggregated per c-square for each métier for each year.

Step 4 - Fishing speeds were based on average speed values for each métier submitted as part of the data call. Where average speed values were not included as part of the submission, a generalised estimate of speed was derived based on the distribution of speed values available for each gear. Here are the detailed steps of the data analysis that was performed to produce such generalised estimates of speed:

a. The average speed per gear groups was first explored in the VMS data. The

data quality was checked. The gear codes were cleaned up. For example,

"PS." was converted to "PS" and "Unk", "NULL", "NIL", and NA code

converted to "NK" (not known).

b. The average fishing speed histograms were plotted (see figures in ICES WG

SFD 2015)in ). Most of the histograms were unimodal with average speed

comprises in expected range of values for fishing speed. For example, most of

the 95% quantile were below 5 nm/h, except for the gears "TBS", "TBB", "PS",

"OTM" and "NK". For these gears, histograms were bimodal and the high

average fishing speeds were reported by Netherland for "TBB" and NEAFC for

the other ones. Such average fishing speed were expect for Netherland "TBB"

fishery, while there were some doubts about how the VMS data were

processed for the NEAFC area. These modes of high values could correspond

to steaming behaviour.

c. The missing average fishing speed data were populated with the mean for

each given gear, while excluding data from Netherlands for "TBB", and NEAFC

for any kind of gear. When it was not possible (e.g. a gear without any

average fishing speed values available to compute a mean), the mean of a

similar gear was used. For instance, average fishing speed of "LHX" was

populated with the mean of "LH", "LHM" and "LHP".

2 There is no BENTHIS metier or JNCC gear group category for hydraulic dredge gear. JNCC gear

groups do not include estimates of bottom contact for demersal seines.


6

d. Speed information was converted from knots to metres per minute to aid

further calculations.

Step 5 – Allocate values of bottom contact (m). Due to the fact that there were no available data on vessel length or power submitted through the data call, average vessel length/power values recorded for each métier class in the BENTHIS survey were assumed. As a result, estimates of gear spread could be extrapolated based on the relationship between vessel size/power and gear spread as published by Eigaard et al. (2015). Where it was not possible to assign to BENTHIS métier codes, a more generalised set of gear bottom contact and speed values were derived based on the review done by JNCC into bottom contact of mobile demersal gears as described in the WGSFD 2015 report (see Table 5 for estimated values for surface and subsurface contact).

o Surface and subsurface abrasion were assumed to be the same for dredges; o Surface and subsurface abrasion were assumed to be the same for most

beam trawl métiers except shrimp fishery where ground gear is lighter; o Surface and subsurface abrasion were assumed to be different for otter

trawling and demersal seine netting with subsurface abrasion generally only associated with a proportion of the gear.

Step 6 - Swept area calculation (see formulas above) Steps 7and 8 – Generation of effort and swept area shapefiles at appropriate scale.

All analysis was conducted using R (R Core Team, 2012). A workflow and an R-script were developed to calculate fishing intensity. ICES has created a sql script that extracts the data for the calculations of swept area ratio.


7

Table A4-1 – Estimates of width of fishing gear causing abrasion and speed for each BENTHIS Métier and JNCC gear group (from ICES WGSFD 2016) 3

Gear class Benthis metier Model Average gear width (m)

Subsurface proportion (%)

Fishing speed (knots)

Otter trawl

OT_CRU 5.1039*(kW0.4690) 78.92 32.1 2.5

OT_DMF 9.6054*(kW0.4337) 105.47 7.8 3.1

OT_MIX 10.6608*(kW0.2921) 61.37 14.7 2.8

OT_MIX_CRU 37.5272*(kW0.1490) 105.12 29.2 3.0

OT_MIX_DMF_BEN 3.2141*LOA+77.9812 156.31 8.6 2.9

OT_MIX_DMF_PEL 6.6371*(LOA0.7706) 76.21 22 3.4

OT_MIX_CRU_DMF 3.9273*LOA+35.8254 113.96 22.9 2.6

OT_SPF 0.9652*LOA+68.3890 101.58 2.8 2.9

Beam trawl

TBB_CRU 1.4812*(kW0.4578) 17.5 52.2 3

TBB_DMF 0.6601*(kW0.5078) 20.28 100 5.2

TBB_MOL 0.9530*(LOA0.7094) 4.93 100 2.4

Dredge DRB_MOL 0.3142*(LOA1.2454) 16.97 100 2.5

Demersal seines SDN_DMF 1948.8347*(kW0.2363) 6536.64 5 NA

SSC_DMF 4461.2700*(LOA0.1176) 6454.21 14 NA

JNCC gear group Gear width Subsurface proportion (%) Fishing speed (knots) Beam Trawl 18 100 4.5 Nephrops Trawl 60 3.33 3

3 It should be noted that values for the gear groups OT and SPR/PTB were changed by the ICES WG SFD and might be re-ajusted at the next meeting


8

Otter Trawl 60 5 3 Otter Trawl (Twin) 100 5 3 Otter Trawl (Other) 60 3.33 3 Boat Dredge 12 100 4 Pair Trawl and Seine 250 0.8 3


9

Figure A4-1 Workflow for production of fishing effort and swept area maps from aggregated (c-square) VMS data


10

Final products: The final analysis produces the surface and sub-surface layers classified according to their swept area ratios. The trawling effort is classified with an intensity scale ranging from ‘none’ to ‘very high’ (cell area swept more than 300 % or 3 times per year). The current proposed scales for the classification of the fishing intensity are showing on Table A4.2, and an example of the draft outputs on Figure A1 -2. The criteria for developing the pressure scale were:

- the distribution of fishing pressure in the OSPAR area should be adequately shown

- areas with lower fishing pressure (SAR <1) should be distinguished as well as areas with higher fishing pressure and

- it should be based on a scientifically justifiable approach.

Several data-driven approaches were tested (e.g. quantiles, standard deviation, natural breaks), however, they show a good differentiation either of the lower SAR values (SAR<1) or of the higher values. A data-driven approach would also produce different pressure scales for surface and subsurface abrasion, which is not useful. As a preliminary solution a scale with 5 categories is proposed with SAR>3 as the upper limit. An area which is fished more than three times per year is supposed to be very highly disturbed, therefore any differentiation of higher SAR values would not have different biological consequences. One fishing event per year is considered to have a high impact on species abundance (Schroeder et al. 2008). The range of values below SAR=1 is split up in three equal parts. These boundaries are based on calculations provided by van Loon (2015) for the development of a fisheries pressure gradient in the Dutch area. The results showed a significant benthic response from SAR 0.15 to 1. However, these analyses were preliminary and the proposed fisheries scale will probably have to be updated with new results. Still the proposed approach shows a good distribution of fisheries pressure in the lower and the higher SAR values and for both surface and subsurface abrasion. However, due to the limitations of the fishing data available, it is important to note that there is a likely overestimation of the results in some areas as each cell has a unique Swept Area Ratio value assuming a homogenous distribution of fishing. This can be addressed if more detailed information from gear types at level 6 (métiers) is available. For example, we need a better understanding of the effects of SumWing trawls, which use a central shoe to reduce gear interaction with the seafloor, and electrofishing / pulse fishing which uses lighter gear that results in lower abrasion pressure


11

Table A4.2. Classification of the swept area ratios per gird cell for a year.

None (0) 0

Very Low (1)

>0.00 – ≤0.33

Low (2) >0.33 - ≤0.66

Medium (3)

>0.66- ≤1.00

High (4) >1.00- ≤3

Very High (5)

> 3

Figure A1 – 2. Example of subsurface swept areas ration for 2013.


12

Aggregating across years

The assessment of fishing pressures across years is necessary in order to understand what is the aggregated results or the maximum pressure value across all years, and if any trends can be made apparent. We have investigated a number of ways in which to show this variability including standard deviation and linear regression. The trend analysis cannot be calculated using linear regression to assess any trends due to the small number of years. Therefore, an analysis of variance is then undertaken with the surface and sub-surface yearly results in order to differentiate between cells with low and high SARs variability to allow the distinction between areas where fishing intensity seems to be constant or at similar levels across years from those where fishing intensity levels fluctuates. A grid cell was considered to be ‘variable’ when the variance analysis showed a change on categories of three or above.

In order to produce a layer showing the aggregated surface and subsurface pressures, taking into account the variations on fishing pressures across years, a generic rule was used: for cells with low variability (constant fishing: i.e. constantly under a similar fishing pressure) the mean of SAR across all years is calculated, the maximum or 95-percentile was not chosen to avoid overestimating the results. For cells with high variability (i.e. fishing pressure variable) the highest SAR value is selected to define the pressure category as it represent the maximum level of exposure within the cycle (see Figure A4.3).

Figure A4.3. Aggregated Surface abrasion pressure using 2010-2015 data series. The hatched

area around the UK showing the areas where inshore fisheries activity from vessels < 12m is

higher than those >12m .


13

References

Barnard, S. & Boyes, S.J, (2013), Review of Case Studies and Recommendations for the Inclusion of Expert Judgement in Marine Biodiversity Status Assessments, JNCC Report 490, JNCC, Peterborough.

Bergman, M.J.N., van Santbrink, J.W. (2000) Fishing mortality of populations of megafauna in sandy sediments. In: Kaiser MJ, de Groot SJ (eds) Effects of fishing on non-target species and habitats: biological, conservation and socio-economic issues. Blackwell Science, Oxford, p 49–68. Church N.J., Carter A.J., Tobin D., Edwards D., Eassom A., Cameron A., Johnson G.E., Robson, L.M. & Webb K.E. (in prep) JNCC Recommended Pressure Mapping Methodology 1. Abrasion: Methods paper for creating a geo-data layer for the pressure ‘Physical Damage (Reversible Change) - Penetration and/or disturbance of the substrate below the surface of the seabed, including abrasion’. JNCC report No. 515, JNCC, Peterborough.

Dernie, K. M. (2003). Predicting Recovery of Soft Sediment Communities Following Physical Disturbance. School of Ocean Sciences, University of Wales, Bangor.

Dinmore, T., Duplisea, D. E., Rackham, B. D., Maxwell, D. L. & Jennings, S. (2003). Impact of a large-scale area closure on patterns of fishing disturbance and the consequences for benthic communities. ICES Journal of Marine Science, 60, 371-380.

Eastwood, P. D., Mills, C. M., Aldridge, J. N., Houghton, C. A. & Rogers, S. I. (2007). Human activities in UK offshore waters: an assessment of direct, physical pressure on the seabed. ICES Journal of Marine Science, 64, 453-463.).

Eigaard Ole R., Bastardie F., Breen M., Dinesen G. E. 1, Hintzen N. T., Laffargue P., Mortensen L. O., Nielsen J. R., Nilsson H. C., O’Neill F. G., Polet H., Reid D. G., Sala A., Skold M., Smith C., Sørensen T. K., Tully O., Zengin M., and Rijnsdorp A. D. (2015). Estimating seabed pressure from demersal trawls, seines, and dredges based on gear design and dimensions. ICES Journal of Marine Science, xxxxx

Foden, J., Rogers, S. I. & Jones, A. P. (2010). Recovery of UK seabed habitats from benthic fishing and aggregate extraction - towards a cumulative impact assessment. Marine Ecology Progress Series, 411, 259–270.

Foden, J., Rogers, S. I. & Jones, A. P. (2011). Human pressures on UK seabed habitats: a cumulative impact assessment. Marine Ecology Progress Series, 428, 33–47.

Gerritsen, H. D., Minto, C., & Lordan, C. (2013). How much of the seabed is impacted by mobile fishing gear? Absolute estimates from Vessel Monitoring System (VMS) point data. ICES Journal of Marine Science .


14

Halpern, B. S., Walbridge, S., Selkoe, K. A., Kappel, C. V., Micheli, F., D'Agrosa, F. et al. (2008). A Global Map of Human Impact on Marine Ecosystems. Science, 319, 948-952.

Hinz, H., Murray, L.G., Lambert, G.I.,Hiddink, J.G. & Kaiser, M.J. (2012). Confidentiality over fishing effort data threatens science and management progress. Fish and Fisheries, 14, 110–117. doi: 10.1111/j.1467-2979.2012.00475.x

ICES (2010). Manual for the International Bottom Trawl Surveys (revision VIII). Retrieved May 17, 2012, from ICES: http://datras.ices.dk/Documents/Manuals/Addendum_1_Manual_for_the_IBTS_Revision_VIII.pdf

ICES (2015). Report of the Working Group on Spatial Fisheries Data (WGSFD). http://www.ices.dk/sites/pub/Publication%20Reports/Expert%20Group%20Report/SSGEPI/2015/01%20WGSFD%20-%20Report%20of%20the%20Working%20Group%20on%20Spatial%20Fisheries%20Data.pdf

Jennings, S., Alvsvag, J., Cotter, A. J., Ehrish, S., Greenstreet, S. P., Jarre-Teichmann, A., et al. (1999). Fishing effects on the northeat Atlantic shelf seas:patterns in fishing effor, diversity and sommunity structure. III. International trawling effort in the North Sea: an analysis of spatial and temporal trends. Fisheries Research, 40, 125-134.

Jennings, S., Lee, J., & Hiddink, J. G. (2012). Assessing fishery footprints and trade-offs between landings value, habtat sensitivity, and fishing impacts to inform marine spatial planning and an ecosystem approach. ICES Journal of Marine Science, 1-11.

JNCC (2011). Review of methods for mapping anthropogenic pressures in UK waters in support of the Marine Biodiversity Monitoring R&D Programme. Briefing paper to UKMMAS evdience groups. Presented 06/10/2011.

Korpinen, S., Meski, L., Andersen, J. H., & Laamanen, M. (2012). Human pressures and their potential impact on the baltic sea ecosystem. Ecological Indicators,15, 105-114.

Krost, P., M. Bernhard, F. Werner, & W. Hukriede. 1990. Otter trawl tracks in Kiel Bay (Western Baltic) mapped by side-scan sonar. Meeresforsch, 32,344-353.

Lee, J., South, A., & Jennings, S. (2010). Developing reliable, repeatable, and accessible methods to provide high-resolution estimates of fishing-effort distributions from vessel monitoring system (VMS) data. ICES Journal of Marine Science, 67, 1260–1271.

Løkkeborg, S. (2004) Impacts of trawling and scallop dredging on benthic habitats and communities. FAO fisheries technical paper 472.

Mills, C. M., Townsend, S. E., Jennings, S., Eastwood, P. D., & Houghton, C. A. (2007). Estimating high resolution trawl fishing effort from satellite-based vessel monitoring system data. ICES Journal of Marine Science, 64, 248–255.

http://datras.ices.dk/Documents/Manuals/Addendum_1_Manual_for_the_IBTS_Revision_VIII.pdf

http://datras.ices.dk/Documents/Manuals/Addendum_1_Manual_for_the_IBTS_Revision_VIII.pdf


15

Paschen, M., U. Richter, W. Kopnick, U. Lorenzen, M. Zimmermann, R. Fonteyne, B. Van Marlen & S. de Groote (2000). Trawl Penetration of the Sea-bed (TRAPESE); contract number 96-006; [final report ; research project financed by the European Community, Directorate General XIV - Fisheries]. Rostock, Rostock Inst. für Maritime Systeme und Strömungstechnik.

Piet, G. J., Quirijns, F. J., Robinson, L., & Greenstreet, S. P. (2006). Potential pressure indicators for fishing, and their data requirements. ICES Journal of Marine Science, 64, 110–121.

Rose, C., Carr, A., Ferro, D., Fonteyne, R., & Macmullen, P. (2000). Using Gear Technology to Understand and Reduce Unintended Effects of fishing on the Seabed and Associated Communities: background and potential directions. Retrieved May 17, 2012, from ICES: http://www.ices.dk/products/CMdocs/2000/B/B0300.pdf

Schroeder, A., L. Gutow & M. Gusky (2008): FishPact. Auswirkungen von Grund-schleppnetzfischereien sowie von Sand- und Kiesabbauvorhaben auf die Meeres-bodenstruktur und das Benthos in den Schutzgebieten der deutschen AWZ der Nordsee (MAR 36032/15). Report for the Federal Agency for Nature Conservation. van Loon, W. (2015): BH2 Multimetric index assessment of benthos in the southern North Sea. Unpublished Report.

Valdemarsen, J., Jorgensen, T., & Engas, A. (2007). Options to Mitigate Bottom Habitat Impact of Dragged Gear. Retrieved May 17, 2012, from FOA: ftp://ftp.fao.org/docrep/fao/010/a1466e/a1466e.pdf

http://www.ices.dk/products/CMdocs/2000/B/B0300.pdf

ftp://ftp.fao.org/docrep/fao/010/a1466e/a1466e.pdf

OSPAR CEMP guidelines – Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications Annex 6: The calculation of disturbance values and trend analyses

1

Annex 6: The Calculation of disturbance values and trend analyses

The degree of disturbance of a habitat is a product of its sensitivity and the exposure to a

specific pressure. Information and data on state-pressure impacts are limited to local

studies, but the outputs have been used to indicate the amount of changed expected if

habitats are affected by pressures or if part of the species forming a habitat have been lost

(e.g. Rondinini 2010). In order to assess the level of disturbance the linkage of sensitivity

information with pressure data is required.

A matrix combining the extent of pressure and habitat sensitivity supports the classification

in ten categories of physical disturbance (Table A6.1). Each category should provide an

approximation of the relative impact on the habitat with regard to e.g. habitat structure,

species richness, abundance or biomass. For the production of the matrix determining the

disturbance categories from pressure and sensitivity classes, the following algorithm was

applied:

a=pressure; b= sensitivity

Table A6-1. Disturbance matrix combining extent of pressure and habitat sensitivity

Disturbance

matrix

Habitat sensitivity

1 2 3 4 5

0 0 0 0 0 0

Exte

nt

of

pre

ssu

re

1 1 2 3 4 6

2 1 2 4 6 7

3 1 3 5 7 9

4 1 4 6 8 9

5 2 4 7 9 9

The pressure map (see Annex 4) and the sensitivity map (see Annex 1) are joined in ArcGIS

using the ‘INTERSECT’ tool and then the above matrix applied to produce a disturbance

category. Figure A6.1 shows an example of the procedure and the resulting map with the

disturbance categories for the German EEZ of the North Sea.


2

Figure A6-1. Example for the combination of the pressure and sensitivity layer to create the

disturbance layer

Disturbance matrix for the pressure ‘abrasion’

Pressure-impact relationships may be described by various types of functions, e.g. linear

relation or logarithm function, and depend on the habitat or the life strategy of species. As a

first approach to set up a disturbance matrix for the pressure ‘abrasion’, the modelling

results of Schroeder et al. (2008) were used as a basis. Schroeder et al. (2008) modelled

fishery-induced mortality rates of selected benthic species with different ecotypes (r- and K-

selected species of in- and epifauna) for the fishing gears beam and otter trawl (Figure

A6.2).


3

Figure A6-2. Percentage decrease in abundance of the benthic species Nephtys hombergii,

Nucula nitidosa, Crangon crangon and Echinus esculentus induced by beam and otter

trawling with different intensities per year (Schroeder et al. 2008).

For the development of a disturbance values used in the matrix, the decrease in abundance

was averaged over the different species and gears to obtain a logarithmic curve for the

physical impact of bottom trawling (Figure A6.3).

Figure A6-3. Estimated physical impact on benthic habitats by bottom trawling, based on

decrease in abundance modelled by Schroeder et al. (2008).

The values derived from the function were applied to the disturbance matrix combining

sensitivity and extent of pressure. Habitat sensitivity was set at intermediate with the

respective temporal fishing intensities (e.g. moderate extent of pressure (3) means 100 % of

cell fished or 1 fishing event per year) and then extrapolated to the very low and very high

categories. The final matrix showing the combination of exposure to the pressure abrasion

and sensitivity of habitats is shown in Table 2.

The temporal elements of the impact analysis and the development of more accurate

impacts curves is a current knowledge gap that is being address as part of some EU funded

abundance (

% o

f an

undis

turb

ed p

opula

tion

)

Number of fishing events per year Number of fishing events per year

Beam trawl Otter trawl


4

projects. We are expecting to continue working on improving these curves which will allow a

more accurate spatial assessments of impacts within and across regions taking into account

the data available from the monitoring programmes and the use of models.

Table A6-2. Disturbance matrix with the disturbance values taken from Figure A6-3 applied.

Disturbance

matrix

Habitat sensitivity

1 2 3 4 5

0 0 0 0 0 0

Exte

nt

of

pre

ssu

re

1 0.5 2 6 29 57

2 0.5 2 29 57 65

3 0.5 6 40 65 100

4 0.5 29 57 80 100

5 2 29 65 100 100

Due to the different nature of the pressures ‘selective extraction’, ‘abrasion’ and ‘changes in

siltation’, for each of these physical damage pressures a separate disturbance matrix will be

required in future work with this indictor in order to include a weighting factor in the

assessment.


5

Figure 6-2. Disturbance map resulting in combination of pressure and sensitivity information using the matrix

Table A6-1.

Aggregating Surface Abrasion Disturbance and Sub-Surface Abrasion Disturbance

Applying the method thus far produces a disturbance map for the pressure ‘surface abrasion’ and a

disturbance map for ‘sub-surface abrasion’ which need to be combined into a single disturbance

map. This is done by taking the highest disturbance category per area.

Trend analyses are undertaken between years in areas identified as variable, in order to assess the

variation of disturbance across years per habitat type. The trend analyses are simple plots over the

six-year period. Linear regression models were trialled but not used due to the small number of data

points (trends over 6 years).

The final outputs of this modelled assessment are levels of disturbance per habitat type across a

region. These levels of disturbance shall be amalgamated into a ‘Physical Damage Index ‘value for

each benthic habitat. During indicator development a formula for calculating this index was tested

and found not to deliver satisfactory results to capture temporal changes in fisheries pressures.

Therefore, as a preliminary result the disturbance categories for each habitat were aggregated into

two groups: disturbance categories 0 to 4, representing lower levels of disturbance; and disturbance

categories 5-9, representing higher levels of disturbance.


6

It is important to note that some areas have already lost some of the sensitive species/biotopes due

to past human activities, which will result in a lower disturbance score. To address those, the

outputs of this indicator will be validated using the assessment results from other benthic condition

indicators.

It is also expected that during the next MSFD reporting cycle, disturbance matrices and the final

algorithm will be calibrated and, if required, modified using the outputs from site-scale condition

indicators.

References:

Rondinini, C. 2010. Meeting the MPA network design principles of representation and adequacy:

developing species-area curves for habitats. JNCC Report No. 439

Schroeder, A., L. Gutow & M. Gusky (2008): FishPact. Auswirkungen von Grund-

schleppnetzfischereien sowie von Sand- und Kiesabbauvorhaben auf die Meeres-bodenstruktur und

das Benthos in den Schutzgebieten der deutschen AWZ der Nordsee (MAR 36032/15). Report for the

Federal Agency for Nature Conservation.

OSPAR CEMP guideline – Agreement 2017-09 Common Biodiversity Indicators: Extent of Physical Damage to predominant and special habitats (BH3) Technical Specifications Annex 7: Development of confidence layers

1

Annex 7: Development of confidence layers

The method used to create the sensitivity layer used in this indicator (as outlined in Annex 5), relies on the combination of data from a number of different sources. These sources carry an associated confidence with the sensitivity that they carry.

Habitat data: For the creation of the composite maps only survey mapping data with a MESH confidence score of 58 % was used above a broadscale map. This ensures that the composite map created aligns with the standards set under MESH project and EMODnet1. Within the maps there are different types of data, point data from scientific surveys will have the higher confidence than an area where only modelled data are available.

Confidence within data source As well as the confidence with the data source, each data point/ polygon has a confidence attribute:

Survey point data

Data points taken from marine recorder have a certain/ uncertain attribute. Each point will also have attributes such as date collected etc. that will infer a confidence.

Extrapolated point data

This is Step 2 of the creation of the sensitivity maps; extrapolating point sensitivity data to habitat polygons. This step produces a varying degree of confidence depending on the number of points used and the size of the polygon that the sensitivity information is extrapolated to.

Survey derived/ Modelled habitat maps

Confidence in the sensitivity information The allocation of confidence scores is based on the level of data used and the quality of information on the resistance and resilience of habitats and species to physical damage pressures. For example, scores derived from experimental or field survey studies have high confidence, whereas scores based on expert judgments have low confidence. The main reports used for this indicators are:

1 The combined survey and modelled habitat map, produced by JNCC, uses the MESH confidence assessment methodology

(http://www.emodnet-seabedhabitats.eu/default.aspx?page=1635). This means that each polygon has a confidence

equating to a percentage.

http://www.emodnet-seabedhabitats.eu/default.aspx?page=1635


2

MB0102: As part of the sensitivity assessment (Tillin et al. 2010), each habitat/ species was given a confidence of high, medium or low.

Ecological Groups: As part of the sensitivity assessment (Tillin and Tyler-Walters 2014), each species was given a confidence score consisting of three parts; quality of evidence (Q of E), applicability of evidence (A of E) and degree of concordance (D of C) of H, M or L. These three scores can be combined to form one confidence score where the lowest of the three scores is taken.

Figure 7-1 shows a visual representation of these sources of data along with their associated confidences. Method for the calculation of overall confidence In order to calculate an overall numeric confidence score, a method adapted from OSPAR (2015) was used. The method multiplies relative measures of confidence on a scale of 0 to 1, where there is a difference in confidence between categories or classes used in a data layer. Zero is actually excluded from the scoring as this would compromise the multiplicative methodology. The final outcome is to be presented preferentially as a three- or five-band percentile seat (0.33; 0.66; 1.00 or 0.20; 0.40; 0.50; 0.60; 0.80; 1.00).


3

The steps below describe how the OSPAR (2015) method for calculating confidence was adapted to develop confidence layers for BH3. A numerical score (0.33, 0.66 or 1) was manually assigned by the assessor to each of the attributes used to create the sensitivity layer. A high categorical score of confidence was given a numeric value of 1, a medium categorical score of confidence was given a numeric value of 0.66 and a low categorical score of confidence was given a numeric value of 0.33. The different methods used to create the sensitivity layer were taken in turn and a numeric confidence score was assigned to each of the attributes: confidence in method type; confidence within data source (habitat (MESH, EUSeaMap), species data); and sensitivity confidence. To get an overall confidence score, the scores from each of the confidence attributes were multiplied together.

Analytical steps Step 1 - Point data

Confidence in method type - This method uses point data to create the sensitivity

layer, it has a high confidence associated with it, therefore, the confidence in the

method type was assigned 1.

Confidence within data source – In this step, the data source was species data. All

species for which information was uncertain (see Annex 1 and Annex 2) were already

removed, therefore, all species were assigned a confidence score of 1, equivalent to

high confidence.

Sensitivity confidence - A low confidence was assigned 0.33, a medium confidence

was assigned 0.66 and high confidence was assigned 1. If there was a range in

confidence scores, the lowest score was taken.

The overall confidence for step 1 was calculated by multiplying the confidence scores

from the three different confidence attributes.

Step 2 – Extrapolated point data

Confidence in method type - This method uses extrapolated point data to create the

sensitivity data, it has a medium confidence associated with it, therefore, the

confidence in the method was assigned 0.66.

Confidence within data source – In this step, the data source was species data. All

species for which information was uncertain were already removed, therefore, all

species were assigned a confidence score of 0.66, equivalent to medium confidence.






Step 3 – Modelled habitat data


4

Confidence in method type - This method uses modelled habitat data to create the

sensitivity data, it has a low confidence associated with it, therefore, the confidence

in the method was assigned 0.33.

Confidence within data source – In this step, the data source was habitat data. A

confidence was assigned to the underlying habitat data, as shown in Table 7-1:

Table 7-1. Numeric confidence score assigned to habitat information.

Origin of habitat

information

Confidence score

Survey map – high

confidence (MESH score

>80)

1

Survey map – low

confidence

(MESH score <80)

0.66

EUSeaMap – high

confidence (MESH score

>80)

0.66

EUSeaMap – low confidence

(MESH score <80)

0.33






The final step involved merging the confidence steps from these three different methods of obtaining sensitivity information into one layer, creating a single overall confidence layer (Figure 7-2).


5

Figure 7-2. Draft example of confidence map created using numeric confidence method using a small section of UK as an example.

Categorical method of assigning confidence

The numerical method of assigning confidence, described above, was sense checked using a categorical method to assign confidence.

In order to combine the confidence in the data source and the associated data confidences, we combine them as in the table below (Table 7-2) to produce an overall confidence score of between 1 and 8, where 8 is the highest confidence.

Table 7-2. Criteria used to combine confidence attributes into an overall confidence score of between 1 and 8

Confidence score (1= lower confidence, 8 = higher confidence)

Habitat/ species data source

Data source confidence

Sensitivity

1 Predicted habitat map

low confidence score (MESH score <80)

Low confidence

2 Predicted habitat map

high confidence score (MESH score >80)

Medium/ High confidence

3 Survey habitat map low confidence score (MESH score <80)

Low confidence

4 Survey habitat map high confidence score (MESH score >80)

Medium/ High confidence


6

Figure 7-3. Draft example of confidence map using a small section of UK as an example incorporating data source and confidence attributes.

The two different methods for assigning confidence show a strong similarity to each other. Any discrepancies between the two methods appear to arise because the use of the numeric method can result in the same overall confidence score from different scores in the underlying confidence attributes (i.e. a confidence of 1 from the confidence in method type, 0.33 from the confidence within data source and 0.66 from the sensitivity confidence would result in an overall confidence of 0.2178, as would a confidence of 1 from the confidence in method type, 0.66 from the confidence within data source and 0.33 from the sensitivity confidence). The categorical method distinguishes between different scores for the confidence attributes, which are only intermediary steps in the numeric method.

5 Extrapolated point data

low confidence score Low confidence

6 Extrapolated point data

high confidence score Medium/ High confidence

7 Species data low confidence score Low confidence

8 Species data high confidence score Medium/ High confidence


7

References

OSPAR (2015). Confidence and Uncertainty in assessing cumulative effects. Presented at the OSPAR Meeting of the Intercessional Correspondence Group on Cumulative Effects, ljmuiden (NL): 26-27 Feb 2015 "Bringing it all together". Tillin, H.M., S.C. Hull & H. Tyler-Walters (2010): Development of a Sensitivity Matrix (pressures-MCZ/MPA features). Defra Contract No. MB0102 Task 3A, Report No. 22. http://jncc.defra.gov.uk/pdf/MB0102_Sensitivity_Assessment%5B1%5D.pdf Tillin, H. & Tyler-Walters, H., (2014): Assessing the sensitivity of subtidal sedimentary habitats to pressures associated with marine activities - Phase 1 Report, JNCC Report 512A. http://jncc.defra.gov.uk/page-6790



Date post:	21-Jan-2022
Category:	Documents
Upload:	others
View:	10 times
Download:	0 times