The Scottish Government

« Previous | Contents | Next »

Listen

ANNEX 6 AN EVALUATION OF THE VALUE FOR MONEY OF THE WINTER LOCAL GOOSE MANAGEMENT SCHEMES

EXECUTIVE SUMMARY

Goose populations have expanded greatly over the past 20-30 years. Scotland currently supports over 400,000 geese for at least part of the winter. The growth in populations has resulted from reductions in mortality rather than enhanced breeding productivity. The geese have benefited from reduced hunting pressure, improved protection with the creation of reserves and changes in agricultural practices. Geese are very loyal to their winter feeding areas, right down to individual fields. Consequently they concentrate onto particular farms and, as a result, they are having an increasingly adverse impact on agriculture due to the negative effect they have on farm incomes. A significant problem is that whilst farmers usually bear all the costs of accommodating geese on their land, they rarely enjoy any of the benefits. The range of measures available to farmers for dealing with grazing competition from geese is restricted. Reducing goose numbers by shooting would probably be the cheapest way of controlling geese numbers and limiting the costs they impose but the UK is restricted in the actions it can take by international agreements.

The challenge facing policy makers, therefore, was how to reconcile the needs of conserving geese with the needs of farmers who were being financially disadvantaged by them. The option of population control by means of shooting was not a practical or legal one on any scale so other management mechanisms have had to be used. The response from the government was to establish the National Goose Forum ( NGF) for Scotland in 1997, to provide Ministers with advice and guidance for developing a national framework for managing wild geese.

The NGF recommended the introduction of Local Goose Management Schemes ( LGMS) in five areas and set out a number of principles for managing the geese. The areas chosen were Islay, Solway, Strathbeg, Kintyre and South Walls in Orkney. They were selected on the basis of five criteria: (i) the protected status of the geese involved; (ii) the number of geese in the area; (iii) the density of geese in the area; (iv) the presence of reserves & other managed areas for geese in the area; (v) the level of impact of geese on agriculture.

Objectives of The Goose Management Schemes

The three aims of the policy set out by the National Goose Forum are:

• to meet UK's nature conservation obligations
• to minimise economic loss to farmers
• to maximise the value for money ( VFM) of public expenditure.

In order to achieve these objectives and to ensure that policy delivered Value for Money ( VFM), a number of overriding management principles were established and followed in each scheme. The management methods used to minimise loss have been based on creating three different types of zone in each Scheme area:

1 The feeding zones are the areas of heaviest goose usage. Geese are allowed to graze undisturbed on them with no scaring taking place.

2 Buffer zones receive less but still significant & predictable goose usage & generally lie close to the most intensively used feeding zones. Geese can be actively scared from all fields within the buffer zone but the scaring must not cause disturbance to any nearby feeding areas. It is accepted that scaring is unlikely to be completely effective and that some damage or loss of income will occur.

3 All land in the scheme but outwith the feeding and buffer zones is treated as a scaring zone. In certain cases shooting may be allowed under licence.

Areas are allocated into either Feeding or Buffer Zones on the basis of their proximity to areas of known high goose feeding density, sward height & field condition, size of fields and distance from known source of disturbance. Outwith these areas the geese are left to roost undisturbed.

The basic principle is to concentrate the geese onto sacrificial feeding areas and then to minimise the economic cost elsewhere by scaring them into these feeding areas. The Scheme minimises losses to farmers by providing direct payments to them from the taxpayer to reflect the costs, including profit foregone, of managing the land in a goose friendly manner. To enhance VFM such management payments are targeted at those areas where the highest densities of geese occur.

This paper considers the three policy objectives in turn and assesses whether or not they have been achieved.

Objective 1: To meet the UK's nature conservation obligations

In order to conclude that this objective has been met it is necessary to show that goose numbers have been maintained at ecologically and scientifically sound levels and that special measures have been taken to conserve the winter habitat of the geese. In practice, responsibility for conserving rare and protected species of geese is shared between the UK, where they spend the winter, and those countries where they breed in summer or use for feeding during migration.

Goose Numbers

Islay :Islay has 50% of the world's population of Greenland White-fronted geese (about 10,000 geese) & 70% (about 40,000) of the Greenland Barnacle geese. Since the Scheme started the numbers of white fronted geese have declined slightly but the barnacle geese have increased by a further 25% although numbers tend to fluctuate from year to year.

Solway: The Scheme is targeted at the management of the Svalbard Barnacle Goose population on the Solway. The population has increased from about 1,000 just after the Second World War to about 25,000 in 2003/04. They are considered to be rare on an international scale. Since the current scheme was introduced geese numbers have been relatively stable although there are annual fluctuations.

Strathbeg :The Loch of Strathbeg is internationally recognised for the numbers of wintering & migrating geese, mainly Pink-footed & Greylag Geese. Numbers average around 18,000 in the spring but these are not rare species and can be shot in the open season. The loch forms the centre of a large reserve but is not within the area of the Goose Scheme because it is managed for birds rather than agriculture.

South Walls :This Scheme supports about 4% of the world's population of the Greenland race of Barnacle Geese. The numbers have increased since they first appeared in the 1960s & now average 1,400 birds.

Kintyre: The area has about 10% of the world's population of Greenland White-fronted geese plus Greylags & some Barnacle Geese. Geese numbers have steadily increased during the last 10 years.

Both the numbers of geese and the presence of protected areas indicate that this objective to meet the UK's international conservation obligations has been achieved. The goose numbers in each Scheme area have either increased or remained relatively stable.

Objective 2: Minimise THE Economic Loss to Farmers

The LGMS minimises losses to farmers by making direct payments to them from the taxpayer to reflect the costs, including profit foregone, of managing the land in a goose friendly manner by means of feeding and buffer zones. In effect, the cost of conserving geese is transferred from individual farmers to society as whole. The amount paid to farmers depends on the payment rates per hectare and the number of hectares eligible for the scheme.

Islay accounts for 70% of the hectares in the 5 Scheme areas that are designated as Feeding and Buffer Zones. Islay is also different in that nearly all (98%) its designation is as Feeding rather than Buffer Zones. In the other 4 Scheme areas the Feeding Zones account for about half (43%-52%) of the hectares designated as Feeding and Buffer Zones. The other main difference between the different Schemes is that only Islay & South Walls have designated Scaring Zones in which either area or standard payments are made.

The payments to farmers are set to reflect the situation in each area and vary as a result of differences in agricultural production and in goose numbers and densities. The latter two factors principally explain why the rates differ between feeding and buffer zones. The rate for the arable or rotational grass in the feeding zone is considerably higher in Islay than elsewhere although the rate for non-rotational grass is between the rates payable in the feeding areas in the Solway, Kintyre and South Walls Schemes. The lowest rates in the feeding zones are found in Strathbeg. Islay also stands out because its payment rates in the buffer zone are between two and four times that in the other schemes. The special position in Islay reflects the fact that it attracts the largest concentrations of geese, densities are generally high across all of the cultivated land on the island and the geese have an impact on farming for virtually the whole winter and spring period.

Payment Levels by Zone and by Scheme

The level of payments in each zone will depend on the size of the zone and the payment rate per hectare. In 2003/04 Area payments under the Local Goose Management Schemes totalled £847,000. The distribution of payments between areas, unsurprisingly, reflects relative sizes of Feeding and to a lesser extent of Buffer Zones, with Islay receiving 72% of Area Payments. Strathbeg is notable in that its share of payments is only just over half that expected on the basis of its share of the hectares designated as either Feeding or Buffer Zones. The reason for this is that the payment rates in both zones are lower in Strathbeg than in any of the other Scheme areas. In contrast to Islay, the Strathbeg scheme is a spring only one so the pressure on agricultural land is considerably less.

In addition to area payments, there are other elements of Scheme expenditure relating to scaring activities and administration costs are also incurred. The total cost of area payments plus scaring activity (equipment and salaries of goose scarers) was about £887,000 in 2003/04. Not all LGMS areas provide cost figures on a comparable basis because administrative cost estimates, associated with the input from SNH staff, are sometimes excluded. If such additional administrative costs were to be included then the overall cost of the LGMS would be about £1,000,000 in 2003/04.

Scheme Payments and Actual Damage Costs

There are considerable problems in trying to measure the costs of managing the land in a goose friendly way. It is important to bear in mind all the qualifications associated with any cost estimates since they are prone to errors, uncertainties, the variability of the weather, the type of agricultural production and the prices prevailing during the period that any such calculations are made. This makes it difficult to know whether or not the payment rates that have been used in the scheme are realistic or not. It is unrealistic to be able to undertake detailed annual studies to re-estimate the costs and associated payment rates each year. Hence it is inevitable that there will need to be some broad brush estimates, based on what are considered to be the most realistic assumptions available at the time, but such assumptions could turn out to be incorrect on the basis of subsequent experience.

As part of a SEERAD funded study the consultants undertook detailed assessments of the likely costs attributable to managing geese in two case study areas: Islay, a livestock farming area with high goose densities and where the scheme area accommodates all of the geese on the island; and Strathbeg, a predominantly cereal cropping area where the geese spread out more widely and rely heavily on areas outwith the scheme to feed. A detailed farm survey involving in-depth interviews with a sample of farmers was used as the main research method. The survey approach that was adopted allowed total and marginal costs of goose damage to be estimated, and provided a contrast between the effects of quarry and non-quarry species, high and low goose density areas, and the effects of goose grazing on a predominately arable cropping area and one dominated by cattle and sheep. The sample farms were raised to the population of all farms in the areas under investigation.

The research study estimated an Islay-wide cost figure, attributable to managing geese, as being about £560,000 for 1999/00. This compares closely with the £596,000 actually paid for costs, including income foregone, on Islay under the Goose Scheme in the following year, 2000/01. In the case of the Strathbeg "catchment", the results reflect the predominance of arable-beef systems where there are large areas of stubble and good supplies of straw for winter feeding. In addition, goose densities tend to be lower than on Islay. Consequently, the agricultural costs of competition from geese are lower at Strathbeg and the study estimated that the costs per hectare were only 30% of the Islay level. The payment rate in Strathbeg for Feeding Zones (grassland) is about one-third of the rate on Islay for rotational grass in such zones so this seems to indicate that the rate paid in Strathbeg can be considered to be reasonable.

The sample farms accounted for 33% of the crops and grass area of the "catchment" of Strathbeg geese and 40% of the total goose population. Based on this, the economic costs from goose damage for the whole area were estimated at £219,000. It is misleading to compare this cost with compensation payments of £46,000 made under the Scheme because the consultants' estimate referred to the total area affected by geese roosting on the Loch of Strathbeg and not just the Scheme area. It is estimated that the Scheme area only supports about 40%-60% of the local winter goose population. However, the differences do suggest that the costs being incurred by farmers and conservation bodies in the whole area are considerably higher than the levels of compensation. Since the geese in this LGMS area are not rare and protected species they can be shot in the open season so one way for farmers to reduce their damage costs would be to undertake more shooting of the geese on their land.

OBJECTIVE 3: TO MAXIMISE THE VALUE FOR MONEY ( VFM) OF PUBLIC EXPENDITURE ON GOOSE MANAGEMENT SCHEMES

There are a number of approaches that can be used to assess the extent to which the LGMS provide VFM for the taxpayer. In general terms this question can be answered by comparing the costs and benefits of the Scheme. If the benefits exceed the costs then the expenditure provides sufficient VFM to justify the policy. This issue was also addressed by the SEERAD funded research project. Even if VFM is found to be acceptable there remains the question of whether or not VFM can be further improved. In essence, this means considering whether there are ways that benefits can be increased without an offsetting rise in costs or whether costs can be reduced without a commensurate loss of benefits. In order to consider this question, detailed information has been collected on each Scheme to allow comparisons to be made between them in order to see what lessons can be learned for further enhancing VFM.

Estimating Marginal Costs of Agricultural Damage Attributable to Goose Grazing

The research study estimated that the marginal economic costs per goose declined on Islay from £12.52 at low populations to less than a £1 at high populations, with a weighted average cost per goose of £12.74 costs. In practical terms, this relationship suggests that, for a limited area, with a fixed amount of herbage available, as on Islay, the more geese which graze, the less will be the consumption per head (possibly due to increased competition) and consequently the lower the damage per head. Another possible explanation is that the proportion of Barnacle Geese increases at higher densities. As these geese are smaller and eat less per bird than the larger Greenland White-fronted Goose, it might be expected that there would be less damage per goose at higher densities.

The Strathbeg survey data do not permit any reliable analysis of the effect of goose densities on agricultural costs, but each farm was classified as having "Low", "Medium" and "High" levels of geese on the basis of proximity to the roost. Although the analysis provides a clear indication that economic costs rise with increasing goose densities it is not possible to deduce whether the relationship is linear, or one of diminishing marginal costs.

Estimating Marginal Benefits Attributable to Wild Geese

The benefits from conserving wild geese need to be identified and set against such costs. People may positively value wild goose populations for recreational reasons such as shooting and viewing (use values), and from the pleasure of knowing they exist (non-use value). In the case of shooting, farmers and other landowners can sell the rights to shoot and hence directly benefit from the presence of wild geese. However, in the case of other recreational activities (bird-watching) and non-use values, markets are difficult to create. This means that even though people benefit from goose conservation, these benefits are not reflected in market prices or in financial returns. Valuing the benefits of conserving wild geese therefore means valuing non-market benefits.

The questionnaire used in the study for SEERAD was divided into five sections. People were first asked about their attitudes to nature conservation as part of rural policy, and their attitude to wild geese compared with other conservation issues. They were then asked how often they had visited nature reserves in the last year, whether they had seen wild geese in the UK, and whether they had visited Islay or Strathbeg in particular. The third section introduced information on the wild geese issue by means of coloured show cards. These contained simple explanations of the "goose problem", together with photographs and maps. Information was provided on current stocks of both endangered and non-endangered species, and of trends in Greenland Barnacle geese numbers since 1959. Respondents were told that the various goose management options covered would all imply higher costs to the government, and that therefore household taxes would have to rise to pay for them. Section four asked their willingness to pay for different conservation measures, whilst the last section collected socio-economic information.

Comparing Marginal Costs and Marginal Benefits

The willingness to pay ( WTP) estimates for each policy option allowed the value of the benefits from conserving wild geese to be estimated and compared with the costs of the different policy options. The results showed that a 10% increase in endangered goose numbers (an approximate increase of 5,000 from 50,000 to 55,000), will increase agricultural costs, including income foregone, on Islay by £1,020 per farm and in aggregate terms by £49,627 per annum. This compares with the estimate for aggregate annual benefits of £10 million. Benefits would therefore outweigh costs by a factor of 200. Hence the conservation of endangered geese species is shown to be worthwhile in cost-benefit terms.

The surveys provided additional insights into public preferences for wildlife conservation. It was evident that the public clearly care about policy outcomes in terms of number of geese. All groups (residents, visitors and the general public) favoured management policies that target endangered species and there was a stronger preference for policies that increased the numbers of endangered species than for policies that concentrated simply on preventing a decrease. There was little desire to pay for an increase in species that were not endangered.

Cost-Benefit Analysis shows that the payments made by the government to farmers for managing land in a goose friendly manner represent good value for money to the tax-payer. Although shooting is claimed to enhance the effectiveness of scaring there can be an apparent conflict between the different objectives of the Geese Management Scheme when the geese being shot are an endangered species. The SEERAD funded study showed that the public generally has an aversion to using shooting as a management measure and that they favoured schemes which supported endangered species. Consequently, the lowest cost management option, if it depends on shooting, does not necessarily represent the best value in economic terms. This is because the additional costs from a financial compensation scheme may be more than offset by the additional benefits from conserving endangered species of geese without resorting to shooting them. However, the SEERAD study was unable to explore all the policy scenarios involving the use of shooting, especially when used as a management measure under very restricted conditions.

A policy that results in some increase (up to 10%) in the species of geese that are considered endangered offers higher benefits than one that simply maintains existing numbers of such geese even if it costs more. Any increase that is much beyond 10% will, however, cease to offer additional benefits and hence would not provide VFM from further compensation payments. This indicates that the public has a diminishing marginal value for geese and that the value attached to conserving each extra goose falls as the number of geese increases.

CONCLUSION

The evidence suggests that overall the winter Local Geese Management Schemes are providing good VFM since the benefits from conserving the geese comfortably exceed the damage costs and the associated compensation payments to farmers.

This is not to say that VFM could not be improved further although by targeting the Scheme to increase benefits and to reduce costs policy has sought to enhance VFM. With this in mind the evaluation has led to a number of recommendation. These are:

Recommendation: Each LGMS should ensure that consideration of VFM is explicitly included as one of the key objectives for the local scheme. This will bring each scheme into line with the overall objectives identified by the National Goose Forum.

Recommendation: The winter Schemes should continue to concentrate on conserving rare and protected species of wild geese by paying farmers to manage their land to support such geese. Since Strathbeg does not support rare and protected species of geese the case for retaining it as part of the winter Schemes should be reviewed along with reconsideration of the summer schemes as necessary.

Recommendation: In approving arrangements for local goose management schemes, the NGMRG should take account of developments in European and domestic agricultural policy and process to ensure that these schemes are consistent with the general approach to management schemes designed to protect and/or enhance habitats and species.

Recommendation: There is a need for further work to improve the accuracy of the relationship between payment rates and density of geese. Goose density factors are used to define Feeding Zones and in some cases, such as Islay, influence the calculation of individual payments.

Recommendation: The NGMRG should consider/identify the population levels of geese that are consistent with meeting the UK's nature conservation obligations and identify the value of the LGMSs in assisting in meeting those obligations.

Recommendation: Since the size of the managed areas per goose varies considerably between the Schemes there should be a review of the zones used in each area. This will allow a comparison to be made of the criteria being used and should help to ensure consistency and will provide a check as to whether the correct areas have been included in each Scheme.

Recommendation: There should be work undertaken to review the basis for calculating the current payment rates in the different Scheme areas in light of current agricultural circumstances. This will determine whether the rates are still valid and check whether or not some farmers are being over or under-compensated for managing their land to support geese.

Recommendation: The quality of the count data on goose numbers in each Scheme area should be reviewed and consideration given to obtaining information on additional criteria such as the length of time geese are present in the area. The concept of goose days, for instance, measures both numbers and frequency of goose usage.

Recommendation: There should be a goose scaring strategy in each Scheme area with a co-ordinated scaring programme and clear objectives and which is publicly available.

Recommendation: There should be a presumption against the shooting of endangered species of geese.

Recommendation: Monitoring of the Schemes needs to be more consistent, comprehensive and appropriately targeted a minimum level of information and reporting, including the provision of maps, should be identified by SNH/ SEERAD.

Recommendation: When changes from the agreed budget occur, each of the causes should be clearly identified and quantified. The internal controls on expenditure should conform to the Scottish Executive requirements of regularity, propriety and good financial management in order to ensure that expenditure is undertaken in a transparent and publicly accountable manner. It would be helpful if the NGMRG could provide approval, in principle, of the budget for 3 years (to fit in with RDR criteria), for future schemes on the basis that: (a) parameters for altering future funding arrangements and operations are agreed by NGMRG; (b) scheme payments and the basis for any operations fall within those agreed parameters: and (c) LGMGs continue to provide annual reports to the NGMRG.

Recommendation: Each Scheme should identify and record all administrative costs incurred by SNH and SEERAD as a result of the Scheme, excluding SNH costs already required to meet other SNH commitments for protected areas monitoring, and should be included in scheme annual reports. Such costs are additional to the Area and Scaring payments and although they are not included in the budgeted costs they are an integral part of the Scheme even though they were not originally recognised as such. This will help to ensure that each Scheme has adequate support, that there is greater consistency between Schemes and that relative costs of administration can be compared.

ANNEX 6 : AN EVALUATION OF THE VALUE FOR MONEY OF THE WINTER LOCAL GOOSE MANAGEMENT SCHEMES

Introduction

1. An evaluation of a policy should focus on assessing its impact, its effectiveness in achieving its objectives, the costs and benefits involved, how worthwhile the expenditure has been and the implications for future management or policy decisions. In order to provide detailed evidence for assessing the value for money of the Goose Management Schemes a research study was commissioned by SEERAD²⁷. Its report has been published, along with an article in the academic literature ²⁸. Extensive use has been made of the findings from this study when evaluating the overall value for money of the expenditure on the Scheme. In order to determine both how VFM varies between the local schemes and to provide guidance on ways of enhancing VFM, additional detailed surveys have also been undertaken of each local scheme by the Review Group.

THE RATIONALE FOR LOCAL GOOSE MANAGEMENT SCHEMES

2. Goose populations have expanded greatly over the past 20-30 years. Scotland currently supports over 400,000 geese for at least part of the winter. These include almost the entire world population of Icelandic Pink-footed (230,000), Icelandic Greylag (73,000), native Scottish Greylag (10,000) and the Svalbard population of Barnacle Geese (27,000). Scotland also supports about 80% (47,000) of the Greenland Barnacle Goose and 66% (17,500) of the Greenland White-fronted Goose populations. Most migratory goose populations arrive in September/October and depart in March/April. Some of these populations such as the Barnacle and White-Fronted geese are heavily concentrated into a few areas. Most geese are highly gregarious and roost in large flocks at high densities.

3. The growth in populations has generally resulted from reductions in mortality rather than enhanced breeding productivity. The geese have benefited from reduced hunting pressure, improved protection with the creation of reserves and changes in agricultural practices such as a shift to winter grown cereals, greater use of fertiliser on grassland and more grassland being cut for storage as winter feed. These have increased the supply of suitable food for the geese when they appear in Scotland for the winter.

4. Geese tend to select sites that offer food availability, safety, a good all-round view and minimum disturbance. Some are very loyal to their winter feeding areas, right down to individual fields. Consequently they tend to concentrate in particular areas of Scotland, down to specific farms.

5. Farmers traditionally tolerated grazing geese when numbers were lower but the rising goose population has brought goose conservation into conflict with farmers in some areas. Grazing by geese, especially in areas where goose numbers are highly concentrated, can cause damage to spring-sown cereals and grass, thereby reducing yields and leading to increased expenditure on fertiliser. Geese may compete with stock for food, leading to farmers having to purchase winter feed. They can also cause problems with soil puddling and compaction. A significant problem has been that whilst farmers usually bear all the costs of accommodating geese on their land, they rarely enjoy any of the benefits.

6. Estimates of the agricultural costs, attributable to grazing competition from geese, vary widely and studies have differed in terms of the species involved, locations, farmland characteristics and management. Consequently it has proved difficult to interpret the results. Goose grazing has accounted for between 10-45% of the variation in yield loss, with other factors such as crop type, weather, soil and drainage conditions and stage of plant development all playing a role. There appears to be no model available that can accurately simulate the effects of goose grazing on vegetation. To build such a model would be time consuming and costly, requiring a considerable amount of new information from a long term study of yield losses

7. The range of measures available to farmers, for controlling the financial losses attributable to geese, is restricted. A variety of scaring measures would probably be the normal way of restricting the impact, from geese but the UK is restricted in the actions it can take by international agreements. The UK is party to the Ramsar Convention, EU Birds Directive, the Bonn and Berne Conventions and the African-Eurasian Migratory Water Bird Agreement ( AEWA), all of which have duties with regard to the protection of geese, their habitats or the key sites that they use. The Wildlife and Countryside Act 1981, as amended, provides the legal framework for the conservation and management of wild geese in Britain.

8. The EU Birds Directive and the 1981 Act make an important distinction between quarry and non-quarry species. The former can be hunted at any time outside the close season because their numbers are judged to be such that they can be exploited on a sustainable basis. Icelandic Greylag and Pink-footed geese are legal quarry species. Non-quarry species, which include Greenland White-fronted and Barnacle Geese are those whose populations are considered especially vulnerable and under threat because of the combination of their migratory nature, climatic conditions on their breeding rounds and their heavy concentration on their wintering grounds. They are protected at all times, although shooting licences can be issued under derogation from the EU Birds Directive to prevent serious damage to crops where non-lethal approaches to goose management have been tried unsuccessfully and no other satisfactory solution exists. Methods for mass destruction of geese, such as poisoning, capture and destruction and egg control, either on winter or breeding grounds, are likely to be politically and logistically unachievable. Also they are most unlikely to be cost-effective.

9. The problem facing policy makers, therefore, was how to reconcile the needs of conserving geese with the needs of farmers who were being financially disadvantaged by them. The option of shooting was not a practical or legal one on any scale so other management mechanisms have had to be used. The response from the government was to establish the National Goose Forum ( NGF) for Scotland in 1997, to provide Ministers with advice and guidance for developing a national framework for managing wild geese. The Forum was superseded in 2000 by the National Goose Management Review Group ( NGMRG) which brings together representatives from farming, hunting and nature conservation interests and as part of this partnership approach a number of Local Goose Management Groups have been set up to develop Local Goose Management Schemes ( LGMS). There is a general presumption against management schemes for goose populations not requiring protection and scaring is expected to be the main approach to preventing agricultural losses. Where this is not possible either because of the sheer numbers of geese or because any associated shooting would be detrimental to a population requiring protection, a management scheme should be introduced. Whilst employing similar overall integrated management strategies, the details vary between the schemes to reflect local conditions so that they are appropriate to the conservation status of the local goose population and the nature and scale of the problems encountered.

10. The NGF recommended the introduction of LGMS in five areas and set out a number of principles for managing the geese in order to achieve the conservation objective whilst minimising economic losses to farmers. The areas chosen were Islay, Solway, Strathbeg, Kintyre and South Walls in Orkney. They were selected on the basis of five criteria: (i) the protected status of the geese involved; (ii) the number of geese in the area; (iii) the density of geese in the area; (iv) the presence of reserves & other managed areas for geese in the area; (v) the level of impact of geese on agriculture.

Objectives of The GOOse Management Scheme

11. All the schemes are voluntary, share the same objectives, targeted on specific populations, operate in defined areas and are based on a common framework. The three aims of the policy set out by the National Goose Forum are:

• to meet the UK's nature conservation obligations
• to minimise economic loss to farmers
• to maximise the value for money ( VFM) of public expenditure.

12. In order to achieve these objectives and to ensure that policy delivered VFM, a number of overriding management principles were established and followed in each scheme. In each area a number of different zones were identified, within which different management measures were to be employed.

13. The management methods used to minimise loss have been based on creating three different types of zone in each Scheme area:

• The feeding zones are the areas of heaviest goose usage. Geese are allowed to graze undisturbed on them with no scaring taking place. The fields are maintained in a condition that is suitable for geese grazing throughout the goose season e.g. by the application of fertiliser and a regular programme of reseeding.
• Buffer zones receive less, but still significant & predictable, goose usage & generally lie close to the most intensively used feeding zones. Geese can be actively scared from all fields within the buffer zone but the scaring must not cause disturbance to any nearby feeding areas. The aim is to make these zones less attractive to geese over time. It is accepted that scaring is unlikely to be completely effective and that some damage or loss of income will occur.
• All land in the scheme but outwith the feeding and buffer zones is treated as a scaring zone where farmers are encouraged to actively scare geese from these areas into feeding areas. Such land experiences less regular and less predictable geese usage. In order to prevent economic loss from grazing by geese, any legal, non-lethal scaring methods can be used. Such methods can include the use of a human scarer, funded by SNH. For certain species, in certain cases, lethal scaring by shooting may be allowed under licence. Some schemes limit the area covered by the scarer/use of equipment and therefore have a more defined scaring area.

14. Areas are allocated into either Feeding, Buffer or Scaring Zones on the basis of their proximity to areas of known high goose feeding density, sward height & field condition, size of fields and distance from known source of disturbance.

15. The basic principle of the schemes is to concentrate the geese onto feeding areas and minimise the economic cost of damage elsewhere by scaring them into these feeding areas. In the case of Buffer Zones the scaring must be done to avoid disturbing geese in nearby feeding areas - hence visual not noisy methods are used. In contrast, in the Scaring Zones methods involving noise can be used because the aim is to make these zones unattractive to geese in order to protect areas with more vulnerable or higher value crops. In this way, the costs of the damage, and hence of the compensation payments, are reduced.

16. The Scheme minimises losses to farmers by providing direct payments to them from the taxpayer to reflect the costs, including profit foregone, of managing the land in a goose friendly manner within the feeding and buffer areas. To enhance VFM such management payments are targeted at those areas where the highest densities of geese occur.

17. The rest of this paper considers the three policy objectives in turn and assesses whether or not they have been achieved. In order to do this a number of surveys and a specially commissioned research study have been undertaken for this evaluation. In addition to determining whether or not the objectives have been achieved, consideration is also given to possible ways of further refining the Schemes to improve VFM.

Objective 1: To meet the UK's nature conservation obligations

18. In order to conclude that this objective has been met it is necessary to show that geese numbers have, so far as is possible, been maintained at favourable conservation status, which includes taking measures to conserve the winter habitat of the geese. Since the conservation of the geese also depends on actions taken in other countries where they breed or feed on migration to/from their winter feeding grounds in Scotland it is evident that numbers could decline regardless of policies in Scotland.

Geese Numbers

19. Islay: Islay has 40% of the world's population of Greenland White-fronted geese (about 10,000 geese) & 70% (about 40,000) of the Greenland Barnacle geese. Since the Scheme started the numbers of White fronted geese have declined slightly but the Barnacle geese have increased by a further 25% although numbers tend to fluctuate from year to year. There are seven SSSIs which are all Special Protection Areas ( SPAs) on Islay: Gruinard Flats (Barnacle goose roost); Bridgend Flats (Barnacle goose roost); Eilean na Duidhe (Greenland White-fronted goose roost); Rinns of Islay, Glac na Criche and Feur Lochain (Greenland White-fronted geese over winter here); and Laggan Peninsula and Bay (Barnacle goose over winter here).

20. Solway: The Scheme is targeted at the management of the Svalbard Barnacle Goose population on the Solway. These birds winter almost exclusively on the Inner Solway Firth, with about 70%-80% of them occurring on the Scottish side. They are present from late September to early May. The population has increased from about 1,000 just after the Second World War to about 25,000 in 2003/04. They are considered to be rare on an international scale. Since the current scheme was introduced geese numbers have been relatively stable although there are annual fluctuations. There are reserves in the area at Caerlaverock ( WWT) and Mersehead ( RSPB) and the geese roost mainly on areas of inter-tidal sand and mud flats which are included within the Upper Solway Flats & Marshes SSSI, which is also a Special Protection Area. Part of the mud flats is also within the Caerlaverock National Nature Reserve. The Solway Merse Management Scheme ( SMMS) is run by SNH and covers 1,050 ha, of which 400 ha are used by geese & 300 ha are owned by RSPB & WWT.

21. Strathbeg: The Loch of Strathbeg is internationally recognised for the numbers of wintering & migrating geese, mainly Pink-footed & Greylag Geese. The loch is designated as a SSSI, a SPA & a Ramsar site and forms the centre of a large reserve of about 930 hectares which is managed by the RSPB. Although the reserve is within the area of the Goose Scheme it is not eligible for payments because it is managed for birds rather than agriculture. Numbers of geese peak at up to 50,000 in autumn but when the Scheme operates in the early spring, before the geese head north to their breeding areas, the numbers are generally around 20,000 geese at the peak in March. The geese are highly mobile in their feeding behaviour & flocks appear to leave the area for other feeding grounds whilst other birds move into the area.

22. The Scheme supported 40% (28% in 2003) of the local roost population of Pink-footed geese in March & 68% (46% in 2003) in April 2004. The lower figure in March reflects the larger number of geese present in the area at that time. The increase in 2004 reflected the drop in goose numbers at the roost but a higher percentage of the population was retained on the feeding zones.

23. South Walls: This Scheme supports about 4% of the world's population of the Greenland race of Barnacle Geese. The numbers have increased since they first appeared in the 1960s & now average about 1,400 birds. The geese roost on the island of Switha, which is a SSSI/ SPA, and feed on the improved grassland of South Walls during the day.

24. Kintyre: The area has about 10% of the world's population of Greenland White-fronted geese plus Greylags & some Barnacle Geese. Geese numbers have steadily increased during the last 10 years. The Kintyre roost is a SPA.

25. Both the numbers of geese and the presence of protected areas indicate that this objective to meet the UK's international conservation obligations has been achieved. The goose numbers in each Scheme area have either increased or remained relatively stable, subject to annual fluctuations. Most of the locations covered by the goose schemes include, either within or adjacent to them, areas designated as Special Protection Areas, under the Habitats Directive, or Sites of Special Scientific Interest. Some are also Ramsar Sites.

Objective 2: Minimise THE Economic Loss to Farmers

26. The LGMS minimises financial losses to farmers by making direct payments to them from the taxpayer to reflect the costs associated with grazing competition from geese in the feeding or buffer areas. In effect the cost of conserving geese is transferred from individual farmers to society as whole. The amount paid to farmers depends on the payment rates per hectare and the number of hectares eligible for the scheme.

Size of Zones in each Scheme Area

27. The size of each zone and their relative importance in each area is shown for 2003/04 in Table 1;

Table 1: Size & Relative Importance of the Different Zones in Each Scheme Area

28. It is clear from the Table that Islay dominates the Schemes. It accounts for 70% of the hectares in the 5 Scheme areas that are designated as Feeding and Buffer Zones. Islay is also different in that nearly all (98%) its designation is as Feeding rather than Buffer Zones. In the other 4 Scheme areas the Feeding Zones account for about half (43%-52%) of the hectares designated as Feeding and Buffer Zones. The other main difference between the different Schemes is that only Islay & South Walls have designated Scaring Zones in which either area or standard payments are made.

Payments Rates in each Zone by Scheme

29. The economic losses to farmers, caused by geese grazing their land, are minimised by transferring the costs of goose damage from the farmers to the taxpayer on behalf of society as a whole. Farmers receive payments to compensate them for both the additional costs and for income foregone through the loss of agricultural production. This raises the question of whether the costs to the taxpayer are also being minimised. This means that farmers should be paid what is considered a fair, and not excessive, compensation that reflects the costs imposed on them by geese.

30. Compensation payments are set to reflect the levels of financial loss in each area and vary as a result of differences in agricultural production and in geese numbers and densities. The latter two factors principally explain why the rates differ between feeding and buffer zones. The rates are shown in Table 2 and reasons for the differences are discussed below.

Table 2: Payment Rates in each Scheme Area

* Payments of £100 per farm

31. The rate for the arable or rotational grass in the feeding zone is considerably higher in Islay than elsewhere although the rate for non-rotational grass is between the rates payable in the feeding areas in the Solway, Kintyre and South Walls Schemes. The lowest rates in the feeding zones are found in Strathbeg. This is because the scheme is a spring only one so the pressure on agricultural land occurs for a considerably shorter period resulting in a lower financial loss for farmers. Islay stands out because its payment rates in the buffer zone are between two and four times that in the other schemes whilst it is the only Scheme that provides payments in the scaring zone. The special position in Islay reflects the fact that it attracts the largest concentrations of geese, densities are generally high across all of the cultivated land on the island and the geese have an impact on agriculture for virtually all of the winter and spring period. Hence, the area susceptible to grazing pressure from geese, in both absolute and relative terms, is considerably greater. The size of the payments and reasons for differences in each Scheme are considered further below.

Islay

32. The geese use almost all the improved land on Islay. Studies on Islay have found that by early February the plots grazed by Barnacle Geese, on average, yielded 38% less than the ungrazed plots & by late March the sward height in grazed fields was on average 63% shorter than in ungrazed ones. Geese accounted for 85% of the consumed dry matter, with the rest eaten by the likes of deer & rabbits. Farm animals were excluded from the fields for the study.

33. Financial losses arise from a variety of additional costs: (a) more frequent reseeding, typically moving from reseeding every 8 years to reseeding every 5 years. The additional costs were estimated to be £39/ha pa (this was based on the Scottish Agricultural College ( SAC) Handbook and/or on grant claims e.g. under CCAGS & ABIS); (b) extra fertiliser, based on typical costs per ha per annum. Costs were estimated as being £37.50/ha pa, with an additional 15% costs added for the islands to reflect additional costs of transport, contracting etc to give an estimate of £43.12/ha pa.

34. Income foregone costs are reflected in: (a) grazing loss from more frequent reseeding, based on the equivalent of 4 months lost grazing in a year and reseeding costs at £90/ha per month. Taking account of that fact means that reseeding occurs 3 years earlier than it would do, otherwise the cost is equivalent to £27/ha pa; (b) loss of sheep grazing from November to end January, based on 3 hogs/ha, the cost is estimated at £26.25/ha pa; (c) loss of cattle grazing for one month, which requires additional hay to be bought in, as well as a 3 weeks delay in conception that results in lower sales weight. Using assumptions about the amount of extra hay required, the amount of lost weight, prices per kg of weight & number of cows per hectare leads to estimated losses of £130/ha on rotational grazing land & £86.67/ha for permanent pasture; (d) silage ground losses cause a reduced silage yield plus a loss of aftermath grazing due to late silage cut leads to estimated losses of £253/ha.

35. Combining the above estimates gives a rate of £269.94/ha payable on all rotational grass or arable ground and £169.00/ha payable on all non-rotational ground calculated to be eligible for feeding payments. However, the situation is more complex than this because the Islay scheme differs from all the others in that the payment per farmer is also calculated to reflect the density of geese. The maximum per hectare rates are only paid if a farmer has at least 15 geese per hectare on his farm. Farmers may have fields in both Feeding and Buffer Zones and payments depend on the average density of geese on the farm as a whole rather than in particular fields. The rates are reduced proportionately so that a farmer with an average of 10 geese per hectare will receive £177 not £269 per hectare ( i.e. 10/15*269).

36. Buffer Zone were limited to 20% of the area of improved and permanent grass and payments for income foregone are £200.45/ha for rotational ground (. i.e. improved pasture) and £140.25/ha for permanent pasture. These figures are based on income foregone in proportion to the assumed reduction in geese density of 22% due to scaring.

37. However, Table 2 shows that in most cases the actual payments in the Feeding Zones are lower than the maximum payments because goose densities are generally below 15/ha. The average Feeding Zone payments are £106/ha whilst the average Buffer Zone payments are actually higher at £195/ha. This contrasts with all the other Scheme areas. The explanation for this is that the areas that make up the relatively small Buffer Zone on Islay tend to be on those farms which have arable or rotational grass and consequently attract the highest densities of geese. Hence payment rates are close to the maximum. On the other hand, the actual payments in the Feeding Zones are well below the maximum possible because these zones have a higher proportion of land that is on non-rotational ground and consequently tends to attract lower densities of geese.

38. Payments in the scaring zone are linked to scaring activity. Scaring Zone payments are £19/ha on both improved & permanent pasture in the Feeding and Buffer Zones for those farmers carrying out their own scaring. Around 24 land managers with very low numbers of geese (1 goose per hectare or less) receive a minimum "goodwill" payment of £100 pa per farm in recognition of their involvement in the overall scaring strategy pursued by the scheme. Islay is the only Scheme that provides a per hectare payment for those farmers undertaking their own scaring. The reason for this is that the numbers and densities of geese are so much higher than in other areas that greater effort needs to be devoted to scaring to restrict damage. In addition, SNH pay for a goose scarer to undertake a co-ordinated scaring programme on reseeds or buffer areas where farmers do not wish to carry out scaring themselves. The true scaring zone on Islay is everything not paid for within the Feeding and Buffer Zones. In other Scheme areas, with the exception of South Walls, scaring payments are restricted to grants for scaring equipment and the employment of a goose scarer. On Islay licensed shooting of some Barnacle Geese is allowed to enhance the effectiveness of scaring and this can be done by either the farmers themselves or the SNH marksman.

Solway

39. Farmers used to be able to limit their financial losses by scaring the geese from vulnerable crops & grassland but as the number of geese increased so scaring became less effective and agricultural damage increased. This led to the introduction of more active integrated management measures, based on the three zone approach. The Feeding Zones are located around the reserves of Caerlaverock and Mersehead and were defined in 2000/01 on the basis of goose usage of 2,100 goose days per hectare. The fields used by the geese are little changed from the 1994 scheme. This means that changes in the Feeding Zones only occur when a farmer needs to reseed or alter his crop. Buffer Zones receive less heavy goose usage, averaging about 700 goose days per hectare. Scaring Zones experience less intensive and less regular use.

40. The payment rates for each zone are based on the national rates suggested by the NGF/ NGMRG rather than being based on local data. The rates paid are less than in Islay, above those in Kintyre and Strathbeg and the same as in the South Walls Scheme which also used the national average rate. The rates are £195/ha in the Feeding Zones, £50/ha in the Buffer Zones, with higher rates of £100/ha being paid on two farms close to the reserves and which are, therefore, heavily used by geese and require more frequent reseeding. The Feeding Zone rate of £195/ha was based on the assumption that reseeds would be required every 4 years rather than every 8 years (cost of £56.25/ha/pa), additional fertiliser to promote grass growth for geese (£37.50/ha). Income foregone through lost production comprised delayed turnout, conception delayed by three weeks with the associated loss of sale weight and silage losses (£100/ha). The Buffer Zone rate was based on the assumption that the income foregone would be half that of the Feeding Zone losses i.e. £50/ha. No payments are made in the Scaring Zones but scaring equipment is made available to farmers.

Strathbeg

41. In Strathbeg, the spring period causes the main conflict with agriculture because early bite is lost & there are delays in turning out stock & some damage to cereals. Each year geese are estimated to consume up to 60% of the spring grass yield & 20% of the silage yield. This deters farmers from growing more profitable winter cereals.

42. The payment rates for both the Feeding and Buffer Zones are the lowest of any of the Schemes. The reasons for this were discussed above in paragraph 31. In the Feeding Zone the rate for land managed as grass is £97/ha and is based on the estimated additional costs attributable to fertiliser (£25/ha) and to the delayed turnout of cattle (£72/ha). In the same zone, a lower rate of £25/ha, representing the cost of fertiliser, is paid on set aside land sown as grass. In the Buffer Zone where land is managed as grassland and may contain stock, the payment rate is £36/ha, based on 50% of the income foregone due to grazing by geese. There is no set aside land within the Buffer Zone. For scaring purposes one-off grants of up to £250 per applicant are paid for equipment. There are no dedicated scarers employed. This is because the geese covered by this Scheme are not rare and protected species and can be shot in the open season which lasts until the end of January. There is no shooting in the spring.

South Walls

43. The main costs arise from: (a) delayed/reduced spring bite; (b) reduced/delayed silage or hay crops; (c) reduced sward quality; (d) damaged/delayed reseeds. However, no specific estimate has been made of the likely financial costs to farmers in this Scheme and so the national payment rates, proposed by the NGF/ NGMRG have been used for the Feeding and Buffer Zones. These rates are 195/ha for the Feeding Zone and £50/ha for the Buffer Zone and are the same as those used in the Solway Scheme.

44. In the Scaring Zone of about 480 ha, flat rate payments of £100 per holding are made to farmers in order to encourage co-operation and to ensure everyone on the island is prepared to tolerate some goose use. As a result farmers are willing to allow the goose shepherd/scarer to have access to their land to scare & count geese, to carry out scaring themselves & to notify the goose scarer of significant numbers of geese on their land. No payments are made for scaring equipment.

Kintyre

45. Geese roost on local lochs but cause financial losses when they graze on agricultural fields. There is damage to barley & silage crops and also managed grassland as well as damage from puddling so there is a requirement for extra feed for stock and some loss of production income.

46. The payment rates for the Feeding Zone vary between stocked (£150/ha) and unstocked (£180/ha) fields. The lower rates for the former reflect the consumption of grass by livestock. These rates reflect the cost of 6 yearly reseeding of improved grassland, additional fertiliser application to promote grass growth, additional costs of cattle feed as a result of the loss of agricultural production. In the Buffer Zone the payment rate is £47/ha. As with the Feeding Zone payments these rates are slightly below the national rates recommended by the NGF/ NGMRG although it is not clear what special work has been undertaken to justify the rates used in this scheme. In the Scaring Zone there are some area payments made directly to farmers to allow access for the two scarers employed by SNH

47. The differences in payment rates between Feeding and Buffer Zones reflect the greater costs and loss of income in the former zones as a result of geese being concentrated into such areas as part of the management process. By compensating farmers for such costs and income loss the economic impact on them from the geese is minimised.

Payment Levels by Zone and by Scheme

48. The level of payments in each zone will depend on the size of the zone and the payment rate per hectare. The actual payments are shown in Table 3. In 2003/04, Area payments under the Local Goose Management Schemes totalled £847,000. The distribution of payments between areas, unsurprisingly, reflects the pattern shown in Table 1, with Islay receiving 72% of Area Payments. Strathbeg is notable in that its share of payments is only just over half that of its share of the hectares designated as either Feeding or Buffer Zones. The reason for this is revealed by Table 2, which shows that the payment rates in both zones are lower in Strathbeg than in any of the other Scheme areas. The reasons for this were discussed in paragraph 31.

Table 3: Area Payments by Zone & Scheme in 2003/04

* Includes scaring equipment & goose scarers but excludes goose counters & admin costs.

49. In addition to Area payments there are other elements of Scheme expenditure relating to scaring activities and administration. These, together with total Scheme costs are shown in Table 4. The total cost of area payments plus scaring activity (equipment and salaries of goose scarers) was about £887,000 in 2003/04. Not all LGMS areas provide cost figures on a comparable basis because there are no administrative cost estimates, associated with the input from SNH staff, for Solway and South Walls. If such additional administrative costs were to be included then the overall cost of the LGMS would be above the £972,000 per annum figure shown. An overall figure of about £1,000,000 in 2003/04 seems a reasonable approximation for the total cost in that year of the 5 winter Local Geese Management Schemes.

Table 4: Total Scheme Expenditure by Area

*Excluding administrative support of 14 days

** Excluding administrative support of 5 days

Scheme Payments and Actual Damage Costs

50. As paragraph 6 has pointed out, there are considerable problems in trying to measure the costs attributable to geese and there are, inevitably, uncertainties underlying the estimates. This makes it difficult to know whether or not the payment rates that have been used in the scheme are realistic or not. They have been calculated using, what were considered to be, the most realistic assumptions available at the time but such assumptions could turn out to be incorrect on the basis of subsequent experience.

51. It is important to bear in mind the qualifications associated with any estimates of the financial losses incurred by farmers as a result of grazing competition from geese. All such estimates are prone to errors, uncertainties, the variability of the weather, the type of agricultural production & the prices prevailing during the period that any such study is undertaken. It is unrealistic to be able to undertake detailed annual studies to re-estimate the costs and associated payment rates each year. Hence it is inevitable that there will need to be some broad brush estimates made of the damage costs and payment rates.

52. As part of the SEERAD funded research study, the consultants undertook detailed assessments of the damage costs attributable to geese in two case study areas: Islay, a livestock farming area with high goose densities and where the scheme area accommodates all of the geese on the island; and Strathbeg, a predominantly cereal cropping area where the geese spread out more widely and rely heavily on areas outwith the scheme to feed. A detailed farm survey involving in-depth interviews with a sample of farmers was used as the main research method. The survey approach that was adopted allowed total and marginal costs of goose damage to be estimated, and provided a contrast between the effects of quarry and non-quarry species, high and low goose density areas, and the effects of goose grazing on a predominately arable cropping area and one dominated by cattle and sheep.

53. The basis for estimating the costs of geese was farmers' recall of the situation during the previous winter of 1999/00. Farmers were questioned as to delays in turning cattle out to pasture in the previous spring due to the effects of geese, and the winter rations fed, whilst the researchers calculated the costs of delayed turnout using standard prices for feeds. For the effects on silage, the loss of yield and the number of weeks delay in cutting were obtained. For arable, farmers estimated reduced cereal yields due to grazing of winter crops or late planting of spring cereals, which were then costed using farm-gate prices. Other anticipated goose effects included the opportunity costs of a changed farm system to accommodate geese, and the effects of more frequent reseeding of grassland. The estimates for the sample farms were raised to the population of all farms in the areas under investigation.

Islay Sample

54. Goose numbers for Islay farms were based on the average of the goose fortnightly day count for the years 1997/98; 1998/99 and 1999/00 as recorded by Scottish Natural Heritage (and available for each farm). Densities per hectare were derived from this figure. The same data are used at Strathbeg but were recorded by the RSPB, and information on the distribution of geese and goose density within the area was derived from previous surveys.

55. Total costs for all 18 Islay farms in the sample were estimated to be £206,000, giving an average of £11,500 per farm. The main findings were:

• Loss of early grazing was the most costly goose effect, followed by losses of silage (including aftermath and hay)
• Including the relatively low cost of additional grass fertiliser (typically applied in the spring), some 67% of all costs measured were accounted for by losses of grass in the first part of the season.
• Additional reseeding costs and losses of winter grazing were also significant, but direct losses to cereal crops were relatively unimportant.

56. Based on the 37% share of Islay productive land and goose numbers included in the 18-farm sample, the research study estimated an Islay-wide cost figure of £560,000 for 1999/00. This compares closely with the £596,000 actually paid for damage costs and income foregone on Islay under the Goose Scheme in the following year, 2000/01. The minimal difference of 6% suggests that the Scheme's payments are not excessive and that it is being delivered almost at minimum cost, especially when allowance is made for all the estimating difficulties that have been referred to.

Strathbeg Sample

57. The Strathbeg "catchment" was represented by 15 farms covering 4,900 ha of which 4,500 ha were crops, grass and set-aside, and which accounted for 33% of the estimated area of significant goose grazing. Their average size was 272 ha of crops and grass (71% greater than the Islay sample). These sampled farms had experienced 1.5 geese per productive ha the previous winter (25% of the density on Islay). Total costs for the sample were estimated to be approximately £88,000, with an average of £5,800 per farm, and £22 per productive hectare (excluding rough grazing and set-aside). Main findings were that:

• costs were predominantly arable-related (40% cereal losses and additional inputs to cereals, and 37% relate to the effect of goose grazing in reducing winter cereal areas).
• Grass-related costs (22% of all costs) were dominated by the effect of lost early bite on cattle feeding.
• Silage losses are relatively unimportant and the costs of lost winter grazing are zero.

58. These results reflect the predominance of arable-beef systems where there are large areas of stubble and good supplies of straw for winter feeding. Compared with Islay, the costs per farm were 50% less, whilst costs per hectare are only 30% of the Islay level. This can mainly be attributed to the generally lower goose densities at Strathbeg. The payment rate in Strathbeg for Feeding Zones (grassland) is about one-third of the rates on Islay for rotational grass in such zones so this seems to indicate that the rates paid in Strathbeg could be reasonable.

59. The sample accounted for 33% of the crops and grass area of the "catchment" of Strathbeg geese and 40% of the total goose population. Based on this, the economic costs from goose damage for the whole area were estimated at £219,000. It is misleading to compare this cost with compensation payments of £46,000 made under the Scheme because the consultants' estimate referred to the total area affected by geese roosting on the Loch of Strathbeg and not just the Scheme area. However, the differences do suggest that the costs being incurred by farmers and conservation bodies in the whole area are between four and five times higher than the payments made under the LGMS. It is estimated that the Scheme area only supports about 40%-60% of the local winter goose population. This suggests that, unlike the other Scheme areas, farmers and conservation bodies rather than the taxpayer are bearing much of the costs. Since the geese in this LGMS area are not a protected species they can be shot in the open season, so one way for farmers to reduce their damage costs would be to undertake more shooting of the geese on their land.

Recommendation: There is a need for further work to improve the accuracy of the relationship between payment rates and density of geese. Goose density factors are used to define Feeding Zones and in some cases, such as Islay, influence the calculation of individual payments.

OBJECTIVE 3: TO MAXIMISE THE VALUE FOR MONEY ( VFM) OF PUBLIC EXPENDITURE ON GOOSE MANAGEMENT SCHEMES

60. There are a number of approaches that can be used to assess the extent to which the LGMS provide VFM for the taxpayer. In general terms this question can be answered by comparing the costs and benefits of the Scheme. If the benefits exceed the costs then the expenditure provides sufficient VFM to justify the policy. In order to examine this issue a research project was commissioned by SEERAD to examine this. Even if VFM is found to be acceptable there remains the question of whether or not VFM can be further improved. In essence, this means considering whether there are ways that benefits can be increased without an offsetting rise in costs or whether costs can be reduced without a commensurate loss of benefits. In order to answer such questions detailed information has been collected on each Scheme to allow comparisons to be made between them in order to see what lessons can be learned for further enhancing VFM.

61. A conservation policy that is optimal in economic terms is one that, in theory, balances marginal benefits and costs so that the value of conserving an additional goose just equals the extra costs of doing so.

Estimating Marginal Costs of Agricultural Damage Attributable to Geese Grazing

62. The availability of SNH's goose population data for individual farms on Islay allowed the consultants, who undertook the SEERAD funded study, to make an estimate of the relationship between goose numbers and densities, and economic costs, at the level of individual farms in the survey. The most interesting relationship is that linking the total of all economic costs measured per farm to the average number of geese per farm (1997/98 to 1999/00). The research study estimated that the marginal economic costs per goose declined from £12.52 at low populations to less than a £1 at high populations.

63. In practical terms, this relationship suggests that, for a limited area, with a fixed amount of herbage available, as on Islay, the more geese which graze, the less will be the consumption per head (possibly due to increased competition) and consequently the lower the damage per head. Another possible explanation is that the proportion of Barnacle Geese increases at higher densities. As these geese are smaller and eat less per bird than the larger Greenland White-fronted Goose, it might be expected that there would be less damage per goose at higher densities.

64. The Strathbeg survey data do not permit any reliable analysis of the effect of goose densities on damage, but each farm was classified as having "Low", "Medium" and "High" numbers of geese on the basis of proximity to the roost. Although the analysis provides a clear indication that economic costs rise with increasing goose densities it is not possible to deduce whether the relationship is linear, or one of diminishing marginal costs.

65. On both Islay and at Strathbeg, there is strong evidence that the total costs attributable to grazing by geese increase as the densities of geese rise. Hence it is reasonable for the Scheme on Islay, with the highest goose densities to make the highest payments per hectare. It is worth noting that on Islay, where there were good data on goose numbers for individual farms, the average costs per goose fell as density increased. At high densities, the effect of a marginal rise in geese is quite small, perhaps because competition for limited grass is increased and consumption per goose is reduced. Although these diminishing marginal effects are not marked at current average goose numbers and densities, this finding suggests that if numbers continue to rise in future, a situation may be reached where economic costs reach an absolute maximum (unless geese start to forage over greater areas, in which case costs will continue to rise). At Strathbeg, however, it was not possible to assess the existence of diminishing marginal damage, due to the lack of reliable count data.

Estimating Marginal Benefits Attributable to Wild Geese

66. The benefits from conserving wild geese need to be identified and set against the costs. People may positively value wild goose populations for recreational reasons such as shooting and viewing (use values), and from the pleasure of knowing they exist (non-use value). In the case of shooting, farmers and other landowners can sell the rights to shoot and hence directly benefit from the presence of wild geese. However, in the case of other recreational activities (bird-watching) and non-use values, markets are difficult to create. For example, in most cases, it would be either impractical or prohibitively expensive to charge people for watching geese and impossible to exclude them from caring about geese. This means that even though people benefit from goose conservation, these benefits do not show up in market prices or in financial returns. For instance, people who are concerned about the fate of wild geese would benefit (in terms of "increased satisfaction") if they felt a policy would increase the chances of conserving geese, even if these individuals never travel to Islay or Strathbeg to go bird watching. Valuing the benefits of conserving wild geese, therefore, means valuing non-market benefits.

67. Focus groups of members of the Scottish general public were held to discuss the issue of wild goose conservation in particular, and attitudes to conservation and farming in general. Based on the focus groups, questionnaire surveys were designed, and then administered to a quota-based random sample of the general public. This technique of sampling involves stratifying the target population - in this case, all households living in Scotland- by socio-economic characteristics, and ensuring that the sample is proportional to the population in terms of these characteristics ( e.g. age, sex). Scotland was chosen as the sampling frame since conservation policy is implemented by the Scottish Executive, and paid for, at least partly, through tax revenues from Scottish households.

68. The questionnaire was divided into five sections. People were first asked about their attitudes to nature conservation as part of rural policy, and their attitude to wild geese compared with other conservation issues. They were then asked how often they had visited nature reserves in the last year, whether they had seen wild geese in the UK, and whether they had visited Islay or Strathbeg in particular. The third section introduced information on the wild geese issue by means of coloured show cards. These contained simple explanations of the "goose problem", together with photographs and maps. Information was provided on current stocks of both endangered and non-endangered species, and of trends in Greenland Barnacle geese numbers since 1959. Respondents were told that the various goose management options covered would all imply higher costs to the government, and that, therefore, household taxes would have to rise to pay for them. Section four asked their willingness to pay ( WTP) for different conservation measures (described below), whilst the last section collected socio-economic information.

69 Only 3% of the sample had ever visited Islay or the Strathbeg areas. However, most people (53-66%, dependent on scenario) said they were willing to support policies for goose conservation, even if these were costly to them. Each respondent was then asked to state their WTP for one of four different goose conservation "projects" (respondents were randomly allocated to projects):

Option 1: Maintaining current population through habitat management, stopping all shooting of geese (many focus group respondents had exhibited a strong aversion to shooting).

Option 2: Measures to prevent a 10% fall in the population of endangered species only (White-fronted and Barnacle geese).

Option 3: Measures to achieve a 10% rise in the population of endangered species only.

Option 4: Measures to achieve a 10% rise in the population of all main migratory species (Pink-footed, Greylag, Greenland Barnacle and Greenland White-fronted).

Comparing Marginal Costs and Marginal Benefits

70. The WTP estimates for each policy option allowed the value of the benefits from conserving wild geese to be estimated and compared with the costs of the different policy options. The results showed that a 10% increase in endangered goose numbers (an approximate increase of 5,000 from 50,000 to 55,000), will increase damage costs on Islay by £1,020 per farm and in aggregate terms by £49,627 per annum. This compares with the estimate for aggregate annual benefits of £35 millions. Benefits would therefore outweigh costs by a factor of 700. Hence the conservation of endangered geese species is shown to be worthwhile in cost-benefit terms. However, unless compensation is paid through a management agreement, farmers would lose out in all scenarios.

71. There have been some concerns expressed that the values obtained through such surveys may overestimate WTP estimates. Consequently an alternative approach, known as a deliberative, group based approach was also used. This allows people to reflect on their responses, both personally, and through discussions with others. The logic is that in discussion and inner reflection, people become more aware of the opportunity costs of stating high WTP amounts; and of other worthy demands on their budget ( e.g. other conservation issues). This approach resulted in WTP estimates for goose conservation that were on average 3.5 times lower than the conventional survey based approach. However, even allowing for this "calibration", the results still point to benefits greatly in excess of costs for a policy to conserve wild geese. For example, for a 10% increase in endangered geese numbers, benefits would be around £10 million even with a downward calibration of 3.3:1, implying a benefit-cost ratio of 200/1.

72. The surveys provided additional insights into public preferences for wildlife conservation. It was evident that while the public clearly care about policy outcomes in terms of number of geese,

they also have preferences about the management methods that are employed. The main findings of note were:

• Different aspects of goose management policy were viewed differently by different groups. Visitors and the general public were strongly opposed to management based on shooting unlike local residents;
• All groups favoured management policies that target endangered species;
• There was a stronger preference for policies that increased the numbers of endangered species than for policies that concentrated simply on preventing a decrease. There was little desire to pay for an increase in species that were not endangered;
• There was some evidence to suggest that visitors, in particular, preferred geese to be protected in a range of different locations than just concentrated into a few reserves although most people did not show a strong preference one way or the other.

73. Cost-Benefit Analysis shows that the compensation payments made by the government to farmers for damage caused by wild geese represent goods value for money to the tax-payer. In addition, such a policy is preferable to the cheaper management option of controlling geese by shooting because the SEERAD research study found that the general public has an aversion to policies that rely on shooting and derived much higher benefits from policies that protected endangered species. Consequently the lowest cost management option does not necessarily represent the best value in economic terms. This is because the additional costs from a financial compensation scheme targeted at endangered species of geese are more than offset by the additional benefits from conserving such geese.

74. A policy that results in some increase (up to about 10%) in the species of geese that are considered endangered offers higher benefits than one that simply maintains existing numbers of such geese even if it costs more. However, the SEERAD study also found that the public's WTP rapidly diminished if any increase is much beyond a 10% figure. This reflects the fact that demand for geese, as with most products, is characterised by diminishing marginal utility so that the more one has of something the less one is willing to pay for further amounts. This means that the value attached to conserving each extra goose falls as the number of geese increases and beyond a certain point the extra benefits from additional geese are insufficient to offset the rising costs of compensating farmers for their financial losses due to grazing competition from geese.

Recommendation: Each LGMS should ensure that consideration of VFM is explicitly included as one of the key objectives for the local scheme. This will bring each scheme into line with the overall objectives identified by the National Goose Forum.

OTHER ISSUES RELATING TO THE VFM of Goose Management Schemes

75. Although the evidence shows that the Goose Schemes as a whole deliver very good VFM, thereby justifying continued expenditure on this policy, any evaluation also needs to consider whether or not the VFM can be improved further. It is unrealistic to undertake a detailed CBA of each local scheme because of the high costs involved so the best way to do this is to compare indicators that provide a proxy for the outputs (or benefits) and then compare them with the inputs (or costs) of each of the local schemes, for example the cost per goose of each Scheme. The reasons for any cost per unit differences can to be investigated to see if any lessons can be learned from them. Comparisons between the different Schemes can help to shed light on best practices that might be applied to other areas.

Should Particular Species be Targeted?

76. At the most straightforward level, benefits could be measured in terms of the number of geese protected under each scheme. However, benefits may not necessarily be perfectly correlated with geese numbers. One reason for this is that the public may prefer to protect some species more than others and the evaluation study does, indeed, show this to be the case. All groups of the population surveyed, as part of the SEERAD funded study, favoured management policies that target endangered species and there was a stronger preference for policies that increased the numbers of such species than for policies that concentrated simply on preventing a decrease. There was little desire to pay for an increase in species that were not endangered.

77. The inputs can be measured by the costs of each scheme and the costs can then be compared. Differences will arise from variations in the relative importance of each management option as well as from differences in payment rates within each management option. VFM of each local scheme would be reflected in costs per goose protected although a high cost per goose scheme could offer better VFM than a lower cost one if it protects particular species of geese to which the public attaches a relatively high value.

78. Four of the five winter Schemes target rare and protected species of geese so they are focussing on species that the public are WTP to conserve. There is a presumption against Management Schemes for geese populations not requiring special protection. An exception is where such species occur immediately surrounding large, well established roosts, where these form properly managed reserves & where significant damage to crops & grass is caused by heavy concentrations of geese. The Loch of Strathbeg is such an exception. The loch is designated as a SPA, SSSI and a Ramsar Site. This puts it into the top level of European nature conservation sites because of its importance in supporting large flocks of over-wintering geese. The fact that the populations of the types of geese found in this area are not under threat is reflected by the policy of allowing control by shooting in the open season. This is the only Scheme that has an open season for the geese concerned.

Recommendation: The winter Schemes should continue to concentrate on conserving rare and protected species of geese by paying farmers to manage their land to support such geese. Since Strathbeg does not support rare and protected species of geese the case for retaining it as part of the winter Schemes should be reviewed along with reconsideration of the summer schemes as necessary.

Recommendation: In approving arrangements for local goose management schemes, the NGMRG should take account of developments in European and domestic agricultural policy and process to ensure that these schemes are consistent with the general approach to management schemes designed to protect and/or enhance habitats and species.

How many geese should the taxpayer be paying to conserve?

79. Since the value of conserving extra geese declines as their numbers increase and the public is only willing to pay to support the conservation of rare and protected species, this implies that taxpayers' money should be used to conserve such geese only so long as their populations are considered to be vulnerable and under threat. Once this ceases to be the case there is minimal economic value to be derived from any further increase in their numbers so the costs involved will start to exceed the associated benefits from conserving more geese.

80. At the present state of knowledge, until new population viability analyses are completed for Greenland Barnacle Geese, Svalbard Barnacle Geese and Greenland White-Fronted Geese, it is not clear what population levels are consistent with meeting the UK's nature conservation obligations. Hence it is uncertain how many additional geese the schemes should be trying to conserve.

Recommendation: The NGMRG should consider/identify the population levels of geese that are consistent with meeting the UK's nature conservation obligations and identify the value of the LGMS in assisting in meeting those obligations.

How & why does cost per goose vary between Schemes?

81. One way of considering the VFM of different local Schemes is to compare the level of management payments with the average number of geese being conserved. This will provide an estimate of the cost per goose. Ideally, information on both the number of geese and the length of time they are in each of the Scheme areas, and hence able to compete for grazing, would be collected. It is possible that lower numbers of geese may actually impose greater costs than larger numbers if they are present in the area for a proportionately longer period.

82. Table 5 shows that there are marked variations in the cost per goose in the different areas. Costs per goose are lowest in Strathbeg (2.4p) and the Solway (4.4p) and highest in Kintyre (27.1p), with Islay and South Walls being in the middle at 12.9p. The average cost per goose of the 5 Schemes combined is 9p. If Strathbeg is excluded, on the basis that it does not focus on rare and protected species, the cost per goose increases to 11p. Kintyre accounts for 7% of the Area payments but only 3% of the geese whilst the Solway has only 13% of the payments but 26% of the geese. The lack of rare and protected geese at Strathbeg means that, despite the low cost per goose, the Scheme offers the least attractive VFM because the benefits are minimal. This is because most people are unwilling to pay to conserve such geese because their populations are not under threat.

83. Payments depend on both the rate per hectare and the number of hectares in the eligible area. Since Table 2 shows that the payments per hectare in Kintyre are not greatly different from those in other Scheme areas, this indicates that the Kintyre schemes has a relatively large managed area relative to the number of geese. Table 5 shows that Kintyre has 8% of the area designated as Feeding and Buffer Zone but only 3% of the geese numbers. This suggests that the area being zoned in Kintyre is more generous than elsewhere and this accounts for the higher cost per goose. In contrast, the Solway scheme accounts for 12% of the Feeding and Buffer Zones but 26% of the geese. The density of geese per hectare on these Zones appears to be about 6.4 times greater in the Solway Scheme than in the Kintyre one. After Kintyre, it is the Islay area that has the next highest ratio of managed area per goose whilst Strathbeg has the lowest ratio. This raises questions as to why there are such large variations in the ratio of managed areas to geese in the different schemes. One explanation is that whilst some Schemes, such as Islay, support virtually the entire population of geese in the local area, other Schemes, such as those for the Solway and Kintyre; support only a proportion of the geese in the area. This is because there are relatively more alternative feeding places, including bird reserves, outwith the Scheme boundary. Hence, in these cases, the estimated goose densities in the Scheme areas, which are shown in Table 5, are likely to be over stated. This also helps to explain the position in Strathbeg since a much larger proportion of the land grazed by geese is excluded from the Scheme than elsewhere. This reflects the fact that the type of geese in that area are not rare and protected species and scaring by shooting is an option available to farmers.

Table 5: Goose Densities and Costs per Goose

*Adopted Average Figures are used for Islay & Solway,

Recommendation: Since the size of the managed areas per goose varies considerably between the Schemes there should be a review of the zones used in each area. This will allow a comparison to be made of the criteria being used and should help to ensure consistency and will provide a check as to whether the correct areas have been included in each Scheme.

Recommendation: The quality of the count data on goose numbers in each Scheme area should be reviewed and consideration given to obtaining information on additional criteria such as the length of time geese are present in the area. The concept of goose days, for instance, measures both numbers and frequency of goose usage.

Is there any evidence that the Goose Schemes have reduced the costs of grazing competition from geese?

84. An evaluation needs to consider what might have happened in the absence of the policy being introduced i.e. did anything different happen because of the policy? Hence a relevant question is whether or not the Schemes reduced the costs attributable to geese.

85. One advantage of the Schemes is that they can encourage the use of management techniques to discourage goose use e.g. increasing activity in affected fields, leaving farm equipment in fields, leaving longer stubble, grazing with livestock, spreading slurry. These methods can be undertaken at no cost to the taxpayer although there has been no study of their effectiveness in decreasing damage costs and there may be some costs for the farmer. It also needs to be recognised that costs attributable to geese may vary from year to year depending on factors such as growing conditions, agricultural prices and geese numbers

86. Islay: Some attempt was made to manage geese during the 1980s with the establishment of protected areas in which geese had been allowed to graze undisturbed. Outwith the refuges, scaring was encouraged but due to the number of geese in the area this policy failed to reduce the costs imposed on farmers to any noticeable extent. In 1992, the Islay Voluntary Goose Management Scheme was set up and this resulted in farmers being paid compensation in return for allowing the geese to graze undisturbed. Payments were linked to the number of geese counted on individual farms & crofts. In 1998/99 payments to farmers were £407,000 and total Scheme costs were £454,000. The 5 year review paper claims the previous scheme was unsuccessful in reducing the impact of geese on agriculture but no detailed evidence appears to have been provided to support this claim. It also suggested that farmers felt that the level of compensation available was not a true reflection of their economic losses. In response to this the Management payments have increased to £636,000 under the current Scheme.

87. Solway: Since the NNR was established in 1957 farmers have been assisted by SNH, and its predecessors, to scare geese from vulnerable crops and grassland. A combination of changing agricultural practices and increasing geese numbers has resulted in the geese grazing farmland rather than the salt marshes. This has reduced the ability of scaring to prevent damage to crops and grass.

88. A pilot Barnacle Goose Management Scheme was introduced for the Caerlaverock area in 1994 and extended to include Southwick before being wound up prior to the introduction of the current scheme. In 1998/99, payments to farmers were £80,000 and total Scheme costs including monitoring and administration were £94,000. In 1995, SNH introduced a Merse Management Scheme, one aim of which was to improve the quality of traditionally used salt marsh grazings and ensure they were available for geese during the winter months. This was separate from the goose scheme. No estimates of the current costs attributable to geese in the area or the effectiveness of the pilot scheme appear to have been undertaken so it is not possible to determine the extent to which the current scheme has led to a reduction in such costs.

89. Strathbeg: The original Scheme was discontinued in 1997 but, at the maximum, in 1994/95 the payments to farmers totalled £18,00 and overall costs were £29,000. Since the current Scheme started there is evidence that there has been an increase in the use of feeding areas, indicating that geese learn the location of safe feeding areas and thereby increasing the impact of the scheme the longer it is in operation

90. South Walls: Prior to the current Scheme, payments to farmers totalled £10,000 in 1998/99 and total Scheme costs were nearly £13,000, excluding administration costs. There do not appear to be any detailed data on the costs imposed by geese either under the previous or current Schemes.

91. Kintyre: There only appear to be limited data on the cost attributable to geese prior to the current Scheme. Westmacott ²⁹ calculated that the costs averaged £27/ha on a sample of 14 farms in goose affected areas of Kintyre. This figure is well below the financial compensation payments made under the current LGMS although there are no estimates of the current level of damage.

Recommendation: There should be work undertaken to review the basis for calculating the current payment rates in the different Scheme areas in light of current agricultural circumstances. This will determine whether the rates are still valid and to check whether or not some farmers are being over or under-compensated for managing their land to support wild geese.

Scaring is the cheapest management method so why is it not used on its own?

92. Where geese are making use of agricultural land, the initial responsibility for tacking the potential damage to crops and grass rests with the farmers who are expected to take steps to minimise any adverse effects by scaring, and where appropriate, by shooting geese. The basic scaring strategy, which is employed in all the Schemes, was agreed by the NGF, and is to move geese onto feeding areas from more sensitive areas e.g. reseeds nearby in buffer & scaring zones. Mechanical & passive scaring equipment are either loaned out or purchased with grant assistance for use in scaring zones. In principle, scaring is the cheapest management option and if it could be used on its own then the VFM of the Schemes would be significantly enhanced. Unfortunately, this is not a practical option when the sheer numbers of geese mean that, in the absence of a concentrated shooting programme, scaring is unable to protect crops because it simply moves the geese to different areas.

93. One of the differences between schemes is their approach to scaring geese. In Islay, for instance, the scaring is undertaken both by SNH under a scaring programme and by some farmers, who receive area payments to cover the costs of their own scaring. In no other areas are farmers paid on a per hectare basis for their scaring efforts. Instead they receive grants for purchasing scaring equipment. Islay is treated as an exceptional case because the numbers of geese are so much larger than elsewhere.

94. It is difficult to gauge the effectiveness of scaring as there is no control over areas not in a Scheme. Consequently, it assumed that if farmers chose not to scare then the levels of damage are acceptable to them and benefits from scaring are less than the costs involved. There is evidence that the VFM is greater when scaring is undertaken by SNH rather than just by the farmers themselves. For instance, it seems that the attitudes of farmers has changed in those schemes where scaring effort is more organised since it appears that scaring programmes operated by SNH are more effective at dispersing geese than when done by individual farmers in an uncoordinated way. In the latter instance, whilst scaring by one farmer may benefit his activities by reducing damage on his own farm it is likely to disadvantage his neighbours by increasing costs on their farms. In economic terms, this means that the private benefits to one farmer will be more than offset by social costs to others because there are, what economists term, negative externalities associated with individual scaring programmes. Evidence that costs are reduced by co-ordinated scaring programmes comes from the fact that the area on Islay covered by farmers scaring geese has reduced from 511ha to 376ha, over the four years of the Scheme. This is almost in line with the area covered by the co-ordinated scaring programme.

95. Geese soon become accustomed to scaring equipment so most farmers/scarers avoid using scaring devices until they are absolutely necessary otherwise their effectiveness is reduced. Consequently, the equipment used requires to be varied & scaring needs to be focussed on the times of peak damage. Shooting is frequently used to try and increase the effectiveness of the scaring programme since the geese are perceived to be more wary after some are shot. The effectiveness of scaring is also perceived by the LGMGs to have improved in the light of experience although there appears to be a dearth of evidence as to which regimes have been the most successful.

96. Although shooting is claimed to enhance the effectiveness of scaring there can be an apparent conflict between the different objectives of the Geese Management Scheme when the geese being shot are an endangered species. The SEERAD funded study showed that the public generally has an aversion to using shooting as a management measure and that they favoured schemes which supported endangered species. However, the study was unable to explore all the policy scenarios involving the use of shooting, especially when used as a management measure under very restricted conditions. Other research has shown that the public's attitudes to the use of culling as a management measure is very sensitive to the scenario presented to them ³⁰. This suggests that any policy which sanctions the shooting of endangered species of geese should have a very strong justification, based on good evidence, to show that the alternative to shooting is likely to not only cost more but also to have a more detrimental impact on public preferences.

97. Islay: Farmers can opt to undertake their own scaring & agree on a programme with SNH. In return they are paid (£19/ha on Islay) or can receive grants for the cost of scaring equipment or borrow such equipment from SNH. If farmers do not wish to carry out the scaring then this can be undertaken by a goose scarer employed by SNH. In certain cases licences can be issued to shoot Barnacle Geese. The shooting can be done either by the farmer or the SNH marksman.

98. It is reported that the scaring regime of human scaring, shooting & deploying devices has had success in protecting crops & no changes have been made to the regime. Scaring has been concentrated on fields that are most vulnerable to goose damage & appears to have successfully moved geese to the feeding areas since farmers, with feeding zones, have reported increased usage due to geese being scared from neighbouring farms.

99. The SNH scaring regime seems to have been more successful than the farmers' own ones because several farmers have given up their own scaring and joined the SNH scheme since it saves them time and appears to have been more effective. However further work to examine and compare the effectiveness of different types of scaring