Abstract from BTO Research Report No 406:

Baylis, S.E., Austin, G.E., Musgrove, A.J. & Rehfisch, M.M. (Aug 2005)

Estimating Great Cormorant Phalacrocorax carbo Population Change as an Aid to Management.

Executive Summary

1. Currently, the only long-term data available on wintering Cormorant numbers come from the Wetland Bird Survey (WeBS). They have been used to produce indices of population change for England, Wales, Scotland, Northern Ireland and Great Britain. Data from WeBS have also been extracted and used in a population model to suggest that the proposed level of control should not threaten the sustainability of wintering Cormorant populations in England and Wales.

2. This study addresses the issues relating to the applicability and validity of the data used, especially during the early years of Cormorant counting by WeBS between the winters of 1986/1987 and 1991/1992. It aims to produce meaningful indices of wintering Cormorant population change.

3. Original WeBS core count recording forms from the above 5 years were scrutinised and checked against values found in the WeBS database. A range of scenarios encountered during this process was described and, in particular, zero values were judged to be either ‘null counts’ or ‘true zeros’. Where counts occurred at complex multi-sector sites, attention was given to the history of the sectors covered as well as the counts themselves, which provided an assessment of completeness.

4. Revised indices with consistency intervals were produced for England, Wales, Scotland, Northern Ireland and Great Britain using the validated data and compared with the originals. This followed the Underhill indexing method.

5. As an addition to this report, and after the data checking process, data were separately supplied to CSL.

6. Data from 1986/1987 were found to be unsuitable for inclusion in the production of the indices. Data from 1987/1988 onwards were used to produce revised indices for England and Wales (indices for Scotland, Northern Ireland and Great Britain are included in Appendices for reference). They show that the increase in wintering Cormorant in England and Wales has been less rapid than suggested by the original indices.

7. A list of around 1500 WeBS sites that currently hold the largest numbers of wintering Cormorant in England and Wales was extracted from the database. An assessment of the 370 sites that held the most Cormorants between 1999/2000 and 2003/2004 (the latest years for which WeBS count data have been analysed) showed which of these potentially provided long and near complete runs of data. Of the 370 sites, 123 single count unit sites and 10 multiple-count unit sites were defined as having been well counted from the early years.

8. The interpretation of WeBS data and indices are subject to certain caveats. In particular, coverage of estuaries and large freshwater bodies is far better than that of small freshwater lakes and ponds and riverine stretches. Currently, there is no random sampling of sites and no stratification of different types of wetland habitat. Wetland coverage by WeBS is not distributed evenly and, in general, a higher proportion of wetlands in southern and eastern England are counted than those in northern England and Wales.

9. The unrepresentative coverage of wetlands by WeBS makes it essential to treat any non-stratified sample with much caution. Without stratification, WeBS estimates of Cormorant population change will be inaccurate unless the population change is broadly constant through time in different wetland types or habitats, which is very unlikely (see Section 4.1 for discussion of issues). Thus, the model used to evaluate the effects of the control on the Cormorant population that is based on determining relationships between population size and change should not be treated as definitive.

10. It has been difficult to avoid an element of subjectivity in the data validation process, despite measures being taken to the contrary. This subjectivity may have affected the estimated population increase, probably to a relatively minor extent, but not in a predictable direction.

11. The introduction of the planned WeBS online system of data submission over the next 2 years may assist with the early reporting of wintering Cormorant population change. The applicability of this system as an early warning system or “tripwire” will be directly affected by the popularity of on-line reporting amongst observers in general and, at key sites in particular. The uptake rate and site coverage are difficult to predict.

Back to Research Reports 345 onwards