An Appraisal and Synthesis of Data Identifying Areas of Ross Worm, Sabellaria Spinulosa

An appraisal and synthesis of data identifying areas of ross worm, Sabellaria spinulosa, reef in The Wash

1. Introduction

1.1. Background

A workshop to determine the location and extent of Sabellaria spinulosa reef in The Wash was held at Natural England in Peterborough in November 2009 (Burton et al. 2010). The main actions from the workshop were to:

• review and analyse the data held by ESFJC and Centrica and assess it using the Gubbay 2007 protocol;
• asses the level of confidence in each data set using MESH confidence assessment;
• overlay and map as the data to assist identification of core reef area.

At a second workshop in October 2010, Natural England presented results from a preliminary synthesis of available data. During this workshop it was decided that areas identified as reef had to be combined with an assessment of the amount of effort used to identify reef over time. This report presents the results of a more detailed appraisal of the available data and a synthesis combining areas of reef and survey effort.

1.2. Data Sources

The data available for this review (Table 1) were supplied primarily by Eastern Sea Fisheries Joint Committee (ESFJC) and Centrica Renewable Energy Ltd (CREL) and their consultants. Some older reports were also used that were carried out on behalf of either ESFJC or, then, English Nature. As such, data was available in a variety of forms including: hard copy; GIS; video and photographs. Data had also undergone a variety of interpretations or manipulations.

Table 1. Data sources available to Natural England that include data on the occurrence of Sabellaria spinulosa reef in The Wash

2. Methods

2.1. Data appraisal and assessment

Data was available from a variety of sources covering a 13 year period. The amount of data associated with each survey varied greatly, with much less data available from earlier surveys. Each data was initially audited for the type of data it contained in relation to Sabellaria reef. The quality of data was then assessed in relation to the quality of the techniques used to detect reef. The data was then subject to a more formal appraisal of its quality using the MESH confidence tool ( If data was collected post 2007 then it was assessed using the Gubbay Reef criteria (Gubbay 2007). It was not possible to re-assess older data using Gubbay criteria due to the amount and quality of data available. Given the Sabellaria spinulosa reef likely existed before 2007, this data has been included but has been synthesised separately from post 2007 data (see below).

A description of each data set, including the data used in the synthesis below, is provided in Appendix I.

2.2. Data synthesis

It is very difficult to combine data sets when they might be considered to consist of very different methodologies. The data sets in this project often differ markedly when trying to achieve the same aim. For example, many data sets rely on AGDS to detect reef, while others do not. Some data sets have used Gubbay criteria to assess reefiness, while others do not. To avoid comparing data sets that may not be combatable, a variety of syntheses have been carried out. This helps prevent one particular type of data set dominating an analysis when confidence in that data set differs from the others, thus avoiding unnecessary bias in the results but also reflects how, and when, many of the surveys were done (Table 2). This section therefore presents the results of synthesising the following data to reflect amount of data to synthesise, the similarity of the data, Gubbay (2007) guidelines and all available data:

1. Data that relies on AGDS as the primary method of reef detection

2. Data collect pre 2007

3. Data collected post 2007

4. All available data

For each data source the areas of reef, and survey effort, were mapped in GIS. The MESH score of each survey was included in the attributes table of each layer. Then, for each of the syntheses above, each reef layer was combined to form a single table for all areas of reef in that synthesis.

The resulting layer and table contained the following attributes for each polygon: number of times reef found; years in which reef was found; average MESH score and total MESH score.

Each survey effort layer was then also combined to form single table showing total survey effort in that synthesis. The resulting layer and table contained the following attributes for each polygon: number of times surveyed; years in which area was surveyed; average MESH score and total MESH score. In each of these two layers the average MESH score reflects the average confidence of all data sets that contribute to a polygon. The total MESH score combines the MESH scores of each individual layer contributing to a polygon and gives an indication of the overall confidence in the data set as well as survey effort.

Table 2. Frequency of occurrence of survey types grouped by method and time of survey

To complete the synthesis the survey effort layer and reef areas layer were combined to form a final ‘Reef Index’ layer. In order to present areas of reef in relation to the amount of survey effort the following method was applied.

Reef Index =

If Number of times reef found = 0 then,

Reef Index = - number of times surveyed.

This method ensures that reef index reflects the survey effort required to detect reef. If reef is consistently found in an area then it will have a high reef index. If reef is found on only one or two occasions in an area that has been surveyed many times then the reef index will be lower. The method also ensures that if reef has never been found in an area then the reef index will be negative (Table 3).

Table 3. Possible values of reef index that could be assigned to a polygon depending on the number of times it has been surveyed compared to the number of times reef has been mapped.

3. Results

Maps of the following syntheses are presented in separate ‘pdf’ files supplied with this document. Layers in these maps can then be turned on and off to allow better interpretation of the data. For each synthesis there are 4 maps: one showing areas of reef index, one showing areas of positive reef index only (i.e. areas without reef are excluded) one showing average MESH score for each polygon and one showing total MESH score for each polygon. The output of this project has also created a GIS workspace where all this data can be displayed and queried at whatever scale. The raw data behind reef index can also be accessed from the GIS workspace. The GIS workspace will be supplied if requested as supplementary material to this report and requires access to Arc GIS software.

3.1. Synthesis of AGDS data

There were nine surveys that used AGDS as the primary method of reef detection. Values returned from AGDS (E1 and E2 scores) were used to predict where areas of reef are likely to occur and this data was then extrapolated to show areas of reef. This method evolved over time as knowledge of which E1/E2 scores likely to be reef increased (pers. comm. Ron Jessop).

These data sets also are not extensively ground truthed due to nature conservation advice at the time of collection. However, there are many data sets covering many years that use this technique and historical data of this type is of value to this exercise. These data form a distinct group as the primary method of detection and, in this instance, was different to that from later ESFJC surveys (one relied on ground truthing) and those by Centrica (3 relying on side scan sonar). The data synthesised were (from Table 1):

Foster-Smith & Sotheran 1999

Foster-Smith 2000

Foster-Smith & White 2001

Foster-Smith 2001

Jessop & Stoutt 2006

Jessop et al. 2006

Woo 2008

Jessop et al. 2008

The maps (see Figure M1, M2, M3 and M4) show areas of highest reef index around Lynn Knock and the Well, with higher average confidence in the areas around the Well. Those areas with highest reef index found reef in 1999, 2005, 2006 and 2007. The figure legends for the maps are:

• Figure M1. Map showing reef index for Sabelleria spinulosa reef in The Wash which has been identified using AGDS as the primary source of detection. Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Negative reef index indicates reef has never been found. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M2. Map showing positive reef index for Sabelleria spinulosa reef in The Wash which has been identified using AGDS as the primary source of detection. Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M3. Map showing average MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash using AGDS between 1999 and 2008.

• Figure M4. Map showing cumulative MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash using AGDS between 1999 and 2008.

3.2. Synthesis of pre Gubbay 2007 data

One of the aims of this synthesis was to re-asses data collected before 2007 using the criteria outlined in Gubbay (2007). It was not possible to do this fully, given the amount of data involved and the way in which it was collected and presented pre 2007. However, many techniques pre 2007 were largely the same as those used post 2007 and many of the ways in which reef was assessed was not dissimilar to the 2007 guidelines. This is not unexpected as the 2007 guidelines were produced during a workshop involving many of the people and organisations who were collecting data for Sabellaria spinulosa reef before 2007. The 2007 guidelines do however represent current best practice and therefore, not wanting to discount all the data collected before 2007 we synthesised the data that was not assessed using the guidelines regardless of the methods used. From table 1 these data sets were:

Foster-Smith & Sotheran 1999

Foster-Smith 2000

Foster-Smith & White 2001

Foster-Smith 2001

Jessop & Stoutt 2006

Jessop et al. 2006

Osiris Projects and Emu Ltd. 2008

The maps (Figure M5,M6, M7 and M8) show reef index was still highest around the Well with and area in Lynn Deep also achieving reasonably high index but yet again with relatively low confidence. Those areas with highest reef index found reef in 1999, 2005 and 2006. The figure legends for the maps are:

• Figure M5. Map showing reef index for Sabelleria spinulosa reef in The Wash which was identified prior to established guidelines for defining reef (Gubbay 2007). Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Negative reef index indicates reef has never been found. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M6. Map showing positive reef index for Sabelleria spinulosa reef in The Wash which was identified prior to established guidelines for defining reef (Gubbay 2007). Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M7. Map showing average MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash which was identified prior to established guidelines for defining reef (Gubbay 2007).

• Figure M8. Map showing cumulative MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash which was identified prior to established guidelines for defining reef (Gubbay 2007).

3.3. Synthesis of post Gubbay 2007 data

Five data sets were available that analysed data using the Gubbay (2007) guidelines. From Table 1 these were:

Woo 2008

Jessop et al. 2008

Jessop et al. 2009

Emu 2008

Emu 2009

The maps (Figure M9, M10, M11 and M12) Areas of highest reef index comprised of reef found on two separate years in two of the 2007, 2008 or 2009 surveys (Figure M9 and Figure M10). Most were concentrated around Lynn Knock with relatively low average MESH scores (confidence). There was one small area of highest reef index in The Well where MESH confidence is relatively high (Figure M11). The figure legends for the maps are:

• Figure M9. Map showing reef index for Sabelleria spinulosa reef in The Wash which was identified using guidelines for defining reef (Gubbay 2007). Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Negative reef index indicates reef has never been found. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M10. Map showing positive reef index for Sabelleria spinulosa reef in The Wash which was identified using established guidelines for defining reef (Gubbay 2007). Higher values of reef index are indicative of the amount of times reef has been described relative to survey effort. Real data from which reef index is derived can be obtained from the GIS supplied with this report.

• Figure M11. Map showing average MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash which was identified using established guidelines for defining reef (Gubbay 2007).

• Figure M12. Map showing cumulative MESH confidence scores for surveys carried out to detect Sabellaria spinulosa reef in The Wash which was identified using to established guidelines for defining reef (Gubbay 2007).

3.4. Synthesis of all available data

A final synthesis of all data was carried out to bring all the available evidence together. The maps (Figure M13, M14, M15 and M16) show highest reef index on the east of The Well. There are also some areas further south towards Seal Sands (Figure M13, M14). These areas all have relatively good MESH confidence between 60 and 68. Data for these areas reveals reef was found on three or four occasions in 1999, 2005, 2006, 2007 and 2009. A better understanding of these locations can be obtained from the GIS package supplied as supplementary material to this report. The figure legends for the individual maps are: