This indicator tests a novel method for assessing changes in the distribution pattern of marine birds. The associated distribution target has been met for all UK Marine Strategy Framework Directive marine areas, but with very low confidence, because estuaries and offshore regions, where marine bird roosting and feeding are common, have not been included.

Background

UK target on bird distribution

This indicator on distribution of breeding and non-breeding marine birds was developed to assess progress against the following target, which is set in the UK Marine Strategy Part 1 (HM Government, 2012): “At the scale of the Marine Strategy Framework Directive sub-regions, distribution of marine birds is not significantly affected by human activities. No major shifts or shrinkage in the population distribution of marine birds in 75% of species monitored.”

Key pressures and impacts

Changes in the distribution of waterfowl and waders along non-estuarine coasts could be caused by coastal developments and flood defences impacting breeding habitats and intertidal feeding areas. This indicator may also reflect prevailing climatic conditions. For example, outside the breeding season warmer winters in recent decades are thought to have caused a north-eastward shift in the UK distribution of some wader and waterfowl species.

Measures taken to address the impacts

Measures to protect waterfowl and waders from human activities leading to habitat loss are set out in the UK Marine Strategy Part 3 (HM Government, 2015). These include those taken under the EU Birds Directive (European Commission, 2009) and the planning, assessment, and licensing of coastal developments.

Further information

Species suitability

This indicator is only suitable for certain species of marine and coastal birds that tend to aggregate in inshore areas or at onshore breeding or roosting sites, for which distributional patterns can be accurately measured and are likely to indicate anthropogenic impacts.

The species included in the current assessment rely on inter-tidal areas for foraging, and their distribution can be affected by the loss of habitat to coastal developments or as a result of disturbance.

Two separate assessments of distributional change were carried out for the following:

  • coastal-breeding distribution of five waders and two waterfowl species (Anseriformes) using data on the presence or absence of breeding pairs nesting within 1 km of the highest astronomical tide.
  • non-breeding distribution of seven wader species (Charadriiformes), using data on their presence or absence on the shore during the winter when birds are either feeding on exposed intertidal areas including coastal lagoons and saltmarsh; or roosting just above the highest astronomical tide.

The restriction of this assessment to non-estuarine coasts, was assumed because the Marine Strategy Framework Directive does not apply to transitional waters such as estuaries This restriction greatly reduced the number of marine bird species that could be assessed. The inclusion of estuarine sites would enable many more species to be included in future assessments.

In the future, this indicator could also be derived from information on breeding colonies of terns, gulls, and cormorants. Like waders and waterfowl, the breeding distribution of these species would be affected by coastal habitat loss and by the introduction of non-native and native predatory mammals to island colonies. Currently, species distribution indicators for breeding terns, gulls, and cormorants cannot be constructed for this assessment because of insufficient data availability. There was an incomplete matching up of sites used by the two most recent seabird censuses (the 1985 to 1988 Seabird Colony Register and the 1998 to2002 Seabird 2000), so changes in occupancy could not be assessed at all sites. Assessments were also not possible for non-breeding gulls due to similar limitations in matching data from the 2003/04 and the 2005/06 Winter Gull Roost Survey and equivalent past surveys, along with the lack of data at the tetrad scale for the first winter atlas. These issues will hopefully be overcome, and these groups of species can be included in future assessments.

Anthropogenic impacts from renewable energy developments, dredging, and shipping, are known to affect the distribution of inshore aggregations of sea ducks, divers and possibly grebes. However, there are no systematic UK-wide monitoring programmes for these inshore waterbirds. Changes in the distribution of large cliff-nesting seabird colonies such as guillemot, razorbill, kittiwake, and fulmar do not provide useful indicators of anthropogenic impacts. The distribution of these colonies is largely determined by the distribution of suitable sea cliffs. Impacts on seabirds at large cliff-colonies, for example from fishing, on the birds’ food supply, are likely to be detected by abundance indicators long before any changes in distribution would become evident.

Assessment method

This is the first quantitative assessment of changes in the distribution of waders and waterfowl around the UK. The methods used in this assessment were developed and implemented by Humphreys and others (2015).

Measuring distribution

Data collation

The presence or absence of species was determined within each of the cells in a pre-defined grid of 2km by 2km survey squares aligned to the Ordnance Survey national grid, called tetrads’, around the coast of the UK. Tetrads were already used in the surveys that provided data for this assessment, data on coastal-breeding waders and waterfowl (Breeding Bird Survey) and by the Bird Atlas of Britain and Ireland in 1988 to 1991 (Gibbons and others, 1993) and 2007 to 2011 (Balmer and others, 2013) and during the winter by the Wetland Bird Survey, and the Non-Estuarine Waterbird Survey. Humphreys and others (2015) also trialled a 1km by 1km grid and a 10km by 10km grid, but neither proved as effective as the tetrads for this assessment.

Data were restricted to only coastal sites that are adjacent to ‘marine waters’ as defined under the Marine Strategy Framework Directive (European Commission, 2008: Article 3(1)).This consists of non-estuarine shores (below highest Astronomical tide), including coastal lagoons, saltmarshes, and inshore non-transitional waters. Sites that are coincident with transitional waters were excluded from the indicator.

Coastal-breeding waders and waterfowl

Data for coastal-breeding waders and waterfowl were obtained from the breeding atlas carried out over the period 1988 to1991 (Gibbons and others 1993) and the most recent atlas from 2007 to 2011 (Balmer and others 2013). For each atlas period, evidence of occupancy was limited to the presence or absence (1 or 0) of a species recorded in surveyed tetrads. Breeding evidence cannot be reliably assigned to tetrads in both survey periods as such data were not routinely collected in the first atlas period of 1988 to1991. Additionally, the timed tetrad visits, used as the main survey methodology, were only one hour in duration in the survey period of 1988 to1991 and therefore, to standardise effort, only the first hour of 2007 to 2011 timed terad visits were used

Non-breeding waders

Data on non-breeding waders were available from two sources: The Wetland Bird Survey and the Non-Estuarine Waterbird Survey. The Wetland Bird Survey conduct systematic monthly counts of the largest aggregations of wintering and migrating waterbirds in the UK, which are mostly in estuaries and not along the open coast. The Non-Estuarine Waterbird Survey was introduced in 1985 (referred to at the time as the ‘Winter Shorebird Count’) to collect data from sites on the open coast. Since then it has been conducted during the winters of 1997/98, 2006/07 and 2015/16. Data from the latter survey were not available for this assessment. Changes in the distribution of species along the open coast, as featured in this indicator, could be assessed only in those winters when both surveys were conducted. Therefore, data from both surveys were combined for each of the winters in 1997/98 and 2006/07.

The spatial distribution of sites from the two surveys was compared to ensure there were no overlaps. Where overlaps were identified, the Non-Estuarine Waterbird Survey count sites were excluded from the dataset in favour of the more frequently surveyed Wetland Bird Survey to maximise temporal coverage. The combined spatial distribution of the two surveys was then compared against the spatial distribution of coastal and transitional waters, and sites that are coincident with transitional waters were excluded from further analysis. However, 34 of the Wetland Bird Survey Core Count sites overlapped both coastal and transitional waters. To have excluded these entirely would have resulted in a substantial loss of data. These sites had sufficient spatial resolution to allow them to be allocated to either coastal or transitional waters.

Figure 1 illustrates data selection for three example areas. These allocations have been defined within the Wetland Bird Survey database to allow routine data extraction and processing in the future. This spatial resolution is at present only available for data collected from the winter of 1993/94 onwards, but in a few years, will be available for the entire time series beginning from the mid-1970s. The Wetland Bird Survey data collected during the first Non-Estuarine Waterbird Survey in the winter of 1985 was, therefore, not available digitally on a spatial scale small enough to transcribe into tetrads. Hence the 1985 winter was not included in this assessment.

Examples of the spatial overlap of transitional waters with the Wetland Bird Survey count sections which make up individual sites. a) the Wash, b) the Tay, Eden, and Firth of Forth, c) Anglesey and the Lleyn Peninsula.

Figure 1. Examples of the spatial overlap of transitional waters with the Wetland Bird Survey count sections which make up individual sites a) the Wash, b) the Tay, Eden, and Firth of Forth, c) Anglesey and the Lleŷn Peninsula.

Parameter/metric

‘Occupancy rate’ - the percentage of tetrads where the species was present.

Measuring change in distribution

Time-period

Data were available to measure changes in distribution between the following two time-periods, specific to each group of species:

  • coastal-breeding waders and waterfowl: breeding seasons of the periods 1988 to 1991 and 2007 to 2011
  • non-breeding waders: winters of 1997 to 1998 and 2006 to 2007

Spatial scope

Changes in ‘occupancy’ and ‘shift index’ (defined below) were calculated for each species within the UK Marine Strategy Framework Directive sub-region to assess progress towards the UK target for population distribution in marine birds.

Species selection

Sufficient data were available to obtain estimates of occupancy rates in each sub-region in more than one time period for the following species:

  • coastal-breeding waders: oystercatcher, curlew, redshank, ringed plover, and avocet
  • coastal-breeding waterfowl: shelduck and eider
  • non-breeding waders: oystercatcher, ringed plover, sanderling, purple sandpiper, curlew, redshank, and turnstone

Distribution could, therefore, be described during both breeding and non-breeding seasons for four species: oystercatcher, curlew, redshank, and ringed plover. It is worth noting that while there are breeding and non-breeding distribution data for these four species, the birds present along UK shores during the winter will not necessarily be the same individuals or from the same populations that bred in the UK and will include individuals that bred outside the UK. Sanderling, purple sandpiper and turnstone do not breed in the UK. Avocet, eider, and shelduck do over winter in the UK, but insufficient ‘winter’ data were available on their wintertime distribution throughout each sub-region.

For each species in each sub-region, the changes in distribution were quantified by measuring two parameters:

  1. ‘occupancy rate’: - equal to the percentage of available tetrads where the species was present
  2. ‘shift index’ - the extent to which the species’ distribution has shifted from one area to another. If the shift index is 1, there has been a complete shift in distribution, but if the shift index is 0, there has been no shift in distribution, and the same sites were occupied in both periods.

Parameters/metrics

Changes in distribution were expressed as two separate metrics:

  1. Change in occupancy rate – describes the extent of shrinkage or expansion in a species’ range.
  2. Shift Index - the extent to which the species’ distribution had shifted from one area to another.

Calculating the change in occupancy rate

The change in occupancy rate is calculated as the occupancy rate (%) in the first period (see time periods above) minus the occupancy rate in the subsequent period. This must not be confused with calculating the percentage difference in the absolute number of tetrads occupied. This latter approach has a lack of reference to the total number of tetrads in the assessment area, without which it is not possible to assess the statistical significance of any apparent differences. The approach used in this assessment, reflects the percentage occupancy of the species concerned. For example, if a species occupied only 10% of an area, a significant change in occupancy would be less than 10%. Conversely, if a species occupied 90% of an area, a significant change in occupancy may be much larger, for example 20%.

Calculations:

Step 1: calculate the occupancy rate in each period.

occupancy rate in period A, OA = 100 * (NsoA / Nst), and

occupancy rate in period B, OB = 100 * (NsoB / Nst), where

NsoA is the number of tetrads occupied by a given species in period A, NsoB is the number of tetrads occupied by a given species in period B and Nst is the total number of tetrads assessed for a given species.

Step 2: calculate the change in occupancy rate between periods

change in occupancy rate ΔOAB = OB - OA, where

OA is the occupancy rate in period A and OB is the occupancy rate in subsequent period B.

Example:

Number of tetrads assessed for species x - Nst = 560

Number of tetrads occupied in survey period A= 40, so OA = (40/560) = 7%

Number of tetrads occupied in survey period B = 28, so OB = (28/560) = 5%

Therefore, the change in percentage occupancy = 5%-7% = -2%

The statistical significance of changes in occupancy rate was assessed using a general additive model (Wood 2006), with a binomial error distribution comparing presence or absence (1,0) at the region level. Latitude and longitude were included as a smoothed interaction term as a proxy to account for spatial autocorrelation, with latitude calibrated to the same scale as longitude. Further details are given in Humphreys and others. (2015).

Calculating the shift index

A change in occupancy rate may not necessarily detect a shift in distribution when the overall percentage occupancy of the assessment areas remains unchanged. Hence the need for the shift index. The shift index is a measure of how much a species has moved from areas it once occupied to new areas that it had not previously occupied. This metric is based on the number of tetrads occupied in both survey periods in relation to the total number of tetrads occupied in either period. If the shift index = 1, there has been a complete shift in distribution; but if the shift index = 0, there has been no shift in distribution, meaning that the same sites are occupied in both periods. 

Calculations:

Shift Index  = 1 - (2*Nso_A&B /[Nso_A + Nso_B ]), where

Nso_A is the number of tetrads occupied by a given species during the period A, and Nso_B  is the number of tetrads occupied in the subsequent period B, and  Nso_A&B  is the number of tetrads occupied in both periods.

Example:

Number of tetrads occupied in survey period A, Nso_A = 40

Number of tetrads occupied in survey period B, Nso_B  = 28

Number of tetrads occupied in both periods Nso_A&B  =  5

Shift Index = 1 - (2*5 /[40 + 28 ]) = 0.85

Performance of the shift index

Humphreys and others (2015) evaluated the performance of the shift indicator by examining the sensitivity of the index and the permutation test significance levels (P-values) obtained in relation to sample size (number of tetrads), and overall occupancy (mean proportion of tetrads occupied). This test was conducted on curlew (Figures 2 and 3) showing that the shift index was reasonably independent of sample size (number of tetrads) but declined with higher rates of occupancy (Figure 2). The maximum possible value of the shift index is mathematically constrained by the occupancy rate during both time periods. For instance, there can never be a complete shift in distribution (for example shift index = 1) if more than 50 % of tetrads are occupied in both time periods. The maximum possible shift index will decrease as the mean occupancy rate across the two time periods increases (see Figure 2). The significance (P-value) declines with sample size as expected and shows a curvilinear relationship with occupancy (Figure 3). Therefore, significant shifts mainly occur for species with low-medium occupancy, as a direct consequence of the construction of both the index and the permutation test.

It is worth noting that the shift index will only detect shifts within the area being assessed. In this instance, it will not detect shifts from coastal feeding to estuarine feeding areas or from coastal breeding sites to inland breeding sites. It will also not detect shifts in distribution to areas outside the UK, such as to other countries bordering the North Sea.

The shift index plotted against the mean occupancy across the two time periods (of a given species in a given spatial unit).

Figure 2. The shift index plotted against the mean occupancy across the two time periods (of a given species in a given spatial unit). The dots show observed values for curlew at the scale of the composite administrative region based on breeding atlas data. The dashed line shows the maximum possible value of the index for a given mean level of occupancy. From Humphreys and others (2015).

The P value of the shift index plotted against mean occupancy across the two time periods, here for curlew at the scale of the composite administrative region based on breeding atlas data.

Figure 3. The P value of the shift index plotted against mean occupancy across the two time periods, here for curlew at the scale of the composite administrative region based on breeding atlas data. The line shows a non-linear curve fit, for illustration only. From Humphreys and others (2015).

Calculating the statistical significance of the shift index

The statistical significance of distributional shifts was assessed using a permutation test comparing the shift index as calculated from the observed data to 9,999 randomised permutations where the same number of tetrads were occupied. This one-tailed test evaluates the likelihood that the observed value of the shift index would occur by chance if birds occupy tetrads randomly. The statistical significance (P-value) of a given shift index value will vary depending on the number of tetrads in the area being assessed: statistical significance will be higher in larger areas with more tetrads (further details in Humphreys and others 2015).

Accounting for cases with no previous record of species’ occupancy

When considering the presence/absence of a species, it is important to consider whether the area is completely outside the potential range of that species or if the habitat within the area is completely unsuitable. For example, an area that is 100 % sandy beaches will ‘never’ be occupied by rocky shore specialists, such as purple sandpiper, and a coastline that is 100 % cliff and boulder beach will ‘never’ be occupied by soft-shore specialist, such as sanderling. If large stretches of unsuitable habitat are included in an assessment area, the significance of any changes in the occupancy rate of suitable habitat will be underestimated. Humphreys and others (2015) therefore applied a species-specific ‘mask’ to remove tetrads where a given species had never been recorded by any of the monthly Wetland Bird Survey counts since 1993 or by the Non-Estuarine Waterbird Survey (further details in Humphreys and others 2015).

Assessing change in distribution

The progress towards the UK target for population distribution of marine birds was assessed by comparing the two indices of change against respective thresholds, to detect ‘major shifts’ and ‘major shrinkages’ in the distribution of each species. ‘Major shrinkage’ occurs if ‘occupancy rate’ decreases, with statistical significance by 10% or more. ‘Major shift’ occurs when the shift index’ equals 0.7 or more. The target is not met for a species if there is either a statistically significant major shrinkage or a statistically significant major shift, or both.

The thresholds for occupancy rate and for the shift index were developed and tested by Humphreys and others (2012), as was the threshold of 75% of species in the UK target. They also trialed the assessment at smaller spatial scales such as administrative areas to detect local changes in distribution that may have resulted from disturbance or displacement of the birds by human activities.

Three different thresholds were tested for the shift index: significant shifts ≥ 0.7, ≥ 0.8 and ≥ 0.9. The thresholds for the shift index were set moderately high because at large scales like UK Marine Strategy Framework Directive sub-regions there is more scope for shifts to occur and the thresholds are meant to detect major shifts only. Figure 2 shows that at mean occupancy rates of greater than 60-70% the shift index will always be less than 0.7 because of mathematical constraints.

Three different thresholds were also tested for changes in occupancy rate: significant decreases of ≥ 10%, ≥ 20%, and ≥ 30%. This resulted in nine possible combinations of thresholds for how to assess if any major shifts or shrinkage in the population distribution of marine birds had occured. The most appropriate combination of thresholds was investigated separately for coastal-breeding waders and waterfowl and non-breeding waders. It was found that the lowest thresholds for both changes in occupancy rate (≥ 10 %) and shift index (≥ 0.7) were most sensitive to changes in distribution at a variety of spatial scales and they were therefore adopted for both coastal-breeding waders and waterfowl and non-breeding waders.​

Results

Findings from the 2012 UK Initial Assessment

This indicator was not considered as part of the Initial UK Assessment (HM Government, 2012).

Latest findings

Status assessment

The assessment in Table 1 shows that in both UK sub-regions, the UK target was met for breeding distribution and non-breeding distribution: all the species assessed showed no major shift or shrinkage in breeding distribution between the summers of 1988 to 1991 and 2007 to 2011 and in non-breeding distribution between the winters of 1997/98 and 2006/07. Note that there was a major shift in the breeding distribution of Avocet in both sub-regions, but this was due to a natural expansion in their range rather than displacement of the species from existing breeding sites due to human impacts.

Table 1. Assessment of major shifts or shrinkage in the breeding and non-breeding distribution of waders and waterfowl during the summers of 1998 to 1991 and 2007 to 2011 and the winters of 1997/98 and 2006/07.

 

UK Celtic Seas

UK Greater North Sea

Species

Breeding distribution change between 1988 to 1991 and 2007 to 2011

 

Non-breeding distribution change between winters 1997/98 and 2006/07

Breeding distribution change between 1988 to 1991 and 2007 to 2011

 

Non-breeding distribution change between winters 1997/98 and 2006/07

 

Shelduck

No change

Insufficient data

No change

Insufficient data

 

Eider

No change

Insufficient data

No change

Insufficient data

 

Oystercatcher

No change

No change

No change

No change

 

Avocet

No change*

Insufficient data

No change*

Insufficient data

 

Ringed plover

No change

No change

No change

No change

 

Sanderling

Does not breed in UK

No change

Does not breed in UK

No change

 

Purple sandpiper

Does not breed in UK

No change

Does not breed in UK

No change

 

Curlew

No change

No change

No change

No change

 

Redshank

No change

No change

No change

No change

 

Turnstone

Does not breed in UK

No change

Does not breed in UK

No change

 

% of species with no change in distribution

100%

100%

100%

100%

 

Trend assessment

The trend is stable.

The parameter values for change in occupancy rate and for the shift index are shown in Tables 2 and 3. These values were used to conduct the assessments against the UK target as shown in Table 1.

Table 2: UK Celtic Seas: assessment of major shifts or shrinkage in the breeding and non-breeding distribution of waders and waterfowl during the summers of 1998 to 1991 and 2007 to 2011 and the winters of 1997 to 1998 and 2006 to 2007. Shift index values greater than 0.7 with statistical significance are indicated with a *.

 

Breeding distribution change between 1988 to 1991 and 2007 to 2011

Non-breeding distribution change between winters 1997 to 1998 and 2006 to 2007

Species

change in occupancy rate %

Shift index

change in occupancy rate %

Shift index

Shelduck

0

0.517

Insufficient data

Insufficient data

Eider

-7

0.509

Insufficient data

Insufficient data

Oystercatcher

-1

0.261

-1

0.092

Avocet

5

1

Insufficient data

Insufficient data

Ringed plover

-3

0.472

1

0.133

Sanderling

Does not breed in UK

Does not breed in UK

0

0.157

Purple sandpiper

Does not breed in UK

Does not breed in UK

-1

0.196

Curlew

-6

0.566

-1

0.129

Redshank

-3

0.481

-1

0.111

Turnstone

Does not breed in UK

Does not breed in UK

2

0.172

Table 3: UK Greater North Sea: assessment of major shifts or shrinkage in the breeding and non-breeding distribution of waders and waterfowl during the summers of 1998 to 1991 and 2007 to 2011 and the winters of 1997 to 1998 and 2006 to 2007. Shift index values greater than 0.7 with statistical significance are indicated with a *.

 

Breeding distribution change between 1988 to 1991 and 2007 to 2011  

Non-breeding distribution change between winters 1997 to 1998 and 2006 to 2011

Species

change in occupancy rate %

Shift index

change in occupancy rate %

Shift index

Shelduck

0

0.407

Insufficient data

Insufficient data

Eider

-2

0.439

Insufficient data

Insufficient data

Oystercatcher

4

0.213

2

0.069

Avocet

7

0.778*

Insufficient data

Insufficient data

Ringed plover

-7

0.491

1

0.133

Sanderling

Does not breed in UK

Does not breed in UK

7

0.137

Purple sandpiper

Does not breed in UK

Does not breed in UK

-1

0.243

Curlew

3

0.289

5

0.146

Redshank

-6

0.376

-1

0.09

Turnstone

Does not breed in UK

Does not breed in UK

4

0.091

Further information

The parameter values for change in occupancy rate and for the shift index are shown in Tables 2 and 3. These values were used to conduct the assessments against the UK target as shown in Table 1.

Table 2: UK Celtic Seas: assessment of major shifts or shrinkage in the breeding and non-breeding distribution of waders and waterfowl during the summers of 1998 to 1991 and 2007 to 2011 and the winters of 1997/98 and 2006/07. Shift index values greater than 0.7 with statistical significance are indicated with a *.

 

Breeding distribution change between 1988 to 1991 and 2007 to 2011

Non-breeding distribution change between winters 1997/98 and 2006/07

Species

change in occupancy rate %

Shift index

change in occupancy rate %

Shift index

Shelduck

0

0.517

Insufficient data

Insufficient data

Eider

-7

0.509

Insufficient data

Insufficient data

Oystercatcher

-1

0.261

-1

0.092

Avocet

5

1

Insufficient data

Insufficient data

Ringed plover

-3

0.472

1

0.133

Sanderling

Does not breed in UK

Does not breed in UK

0

0.157

Purple sandpiper

Does not breed in UK

Does not breed in UK

-1

0.196

Curlew

-6

0.566

-1

0.129

Redshank

-3

0.481

-1

0.111

Turnstone

Does not breed in UK

Does not breed in UK

2

0.172

Table 3: UK Greater North Sea: assessment of major shifts or shrinkage in the breeding and non-breeding distribution of waders and waterfowl during the summers of 1998 to 1991 and 2007 to 2011 and the winters of 1997/98 and 2006/07. Shift index values greater than 0.7 with statistical significance are indicated with a *.

 

Breeding distribution change between 1988 to 1991 and 2007 to 2011  

Non-breeding distribution change between winters 1997/98 and 2006/07.

Species

change in occupancy rate %

Shift index

change in occupancy rate %

Shift index

Shelduck

0

0.407

Insufficient data

Insufficient data

Eider

-2

0.439

Insufficient data

Insufficient data

Oystercatcher

4

0.213

2

0.069

Avocet

7

0.778*

Insufficient data

Insufficient data

Ringed plover

-7

0.491

1

0.133

Sanderling

Does not breed in UK

Does not breed in UK

7

0.137

Purple sandpiper

Does not breed in UK

Does not breed in UK

-1

0.243

Curlew

3

0.289

5

0.146

Redshank

-6

0.376

-1

0.09

Turnstone

Does not breed in UK

Does not breed in UK

4

0.091

Conclusions

This assessment has successfully demonstrated that changes in the distribution of marine birds can be measured using a new method and used to assess progress against the UK target for bird distribution. The results show that this has been met, but with low confidence, so the findings were not used for the assessment of Good Environmental Status of marine birds. The inclusion of estuarine bird data will aid interpretation in assessments going forward and will improve confidence levels and provide further clarity in determining progress toward Good Environmental Status for marine birds.

Knowledge gaps

This indicator is currently lacking data on the internationally important numbers of sea ducks, divers, and grebes that over-winter in the shallow inshore waters of the UK. Filling this gap will help to detect impacts caused by the displacement of birds from inshore areas by offshore renewable developments, dredging, aggregate extraction, and shipping.  

Further information

Future inclusion of colonial-breeding terns, gulls, and cormorants

This indicator would also be relevant to assessing impacts on the distribution of colonial-breeding terns, gulls, and cormorants, which all feed in inshore areas and could be affected by human impacts differently to the intertidal-feeders currently assessed. But before such species could be included, problems over incompatibility between data from different surveys need to be addressed.

Humphreys and others (2015) attempted to include data for these groups of species into the current assessment. However, they were unable to match up all sites used by the two most recent seabird censuses (the 1995 to 1988 Seabird Colony Register and the 1998 to 2002 Seabird 2000). This is not an insurmountable problem but requires considerable effort to resolve. Nor were they possible for non-breeding gulls due to similar limitations in matching up data from the 2003/04 and 2005/06 Winter Gull Roost Survey equivalent past surveys, and the lack of data at the tetrad scale for the first winter atlas.

Smaller assessment scales

Humphreys and others (2012) also conducted parallel assessments at smaller spatial scales within each country – within smaller subdivisions of each sub-region using existing boundaries like administrative areas.  These are not reported here because of the issues discussed arising from the restriction to non-estuarine sites. However, future assessments would benefit from the addition of small-scale assessments, alongside sub-regional assessments, in order to provide a watching brief on the potential local anthropogenic impacts on marine bird distribution. These parallel assessments can be used to inform local measures and provide early warning of potential impacts to population-level.

References

Balmer D.E., Gillings, S, Caffrey, BJ, Swann RL, Downie IS, Fuller Rj (eds) (2013) 'Bird Atlas 2007–11: the breeding and wintering birds of Britain and Ireland' BTO Books, Thetford. (viewed on 11 October 2018)

European Commission (2008) 'Directive 2008/56/EC of the European Parliament and of the Council of 17 June 2008 establishing a framework for community action in the field of marine environmental policy (Marine Strategy Framework Directive)' Official Journal of the European Union L 164, 25.6.2008, pages 19-40 (viewed 21 September 2018)

European Commission (2009) 'Directive 2009/147/EC of the European Parliament and of the Council of 30 November 2009 on the conservation of wild birds' Official Journal of the European Union L 20, 26.1.2010, pages 7–25 (viewed on 10 October 2018)

Gibbons DW, Reid JB, Chapman RA (1993) 'The New Atlas of Breeding Birds in Britain and Ireland: 1988–1991' T. & A.D. Poyser, London.

HM Government (2012) 'Marine Strategy Part One: UK Initial Assessment and Good Environmental Status' December 2012. (viewed on 5 July 2018)

HM Government (2015) 'Marine Strategy Part Three: UK Programme of Measures' (viewed on 5 July 2018)

Humphreys EM Austin GE, Thaxter C, Johnston A, Risely, K Frederiksen, M Burton NHK (2015) 'Development of MSFD Indicators, Baselines, and Targets for Population Size and Distribution of Marine Birds in the UK' Report to JNCC.

Wood SN (2006) 'Generalized Additive Models: An Introduction with R' Chapman & Hall/CRC.

Acknowledgements

Assessment Metadata

Please contact marinestrategy@defra.gov.uk for metadata information

Recommended reference for this indicator assessment

Ian Mitchell1, Liz Humphreys6, Andrew Douse,2 Simon Foster,2 Melanie Kershaw,3 Neil McCulloch4, Matty Murphy5, & Jane Hawkridge1 2018. Marine bird distribution. UK Marine Online Assessment Tool, available at: https://moat.cefas.co.uk/biodiversity-food-webs-and-marine-protected-areas/birds/distribution/

1Joint Nature Conservation Committee

2Scottish Natural Heritage

3Natural England

4Department of Environment, Agriculture & Rural Affairs, Northern Ireland

5Natural Resources Wales

6British Trust for Ornithology