THE BICHEL COMMITTEE

6. Basic assumptions for assessing consequences

6.1 Choice of scenarios for pesticide phase-out
6.1.1 Preconditions and methods of farming
6.1.2 Choice of organic scenarios
6.2 Methods of controlling pests
6.2.1 Prevention and control when deciding crops and crop rotation regimes
6.2.2 Prevention and control of fungal diseases
6.2.3 Seed-born diseases
6.2.4 Prevention and control of pests
6.2.5 Prevention and control of weeds
6.2.6 Growth regulation
6.2.7 Biological control
6.2.8 Use of damage thresholds and decision support system
6.2.9 Genetically-modified crops
6.2.10 Possibility of reducing pesticide consumption through alternative spraying techniques
6.2.11 New pesticides
6.3 Changes in yield and contribution margin following a ban on pesticides
6.3.1 Methods of determining yield losses
6.3.2 Methods of estimating total loss magnitudes
6.3.3 Variations in yield level of current production
6.4 Methods for calculating economic consequences
6.5 Environmental and health-related assumptions
6.6 Review of analyses of the individual scenarios

The Committee's mandate stated that the Committee was to analyse scenarios for the total and partial phasing-out of pesticides. The more detailed choice of scenarios was, however, left to the Committee, and Chapter 6.1 describes these in greater detail.

One decisive prerequisite for calculating the consequences for crop yield when pesticides can no longer be used is a sound knowledge of the significance of pests for crop yield and of how yield losses can be prevented or controlled by non-chemical methods. In Chapter 6.2, the Committee has therefore assessed the feasibility of using alternative methods for the prevention and control of pests. In Chapter 6.3, the Committee has estimated yield losses for the different pests in the most important crops, when pesticides can no longer be used.

We have estimated the economic and environmental consequences of phasing out pesticides. Different models, which are described in Chapters 6.4 and 6.5, have been used in this context.

6.1 Choice of scenarios for pesticide phase-out

Different scenarios for total and partial phase-out

The Committee has assessed a number of scenarios for the total or partial phasing out of pesticides in farming over a 10-year period, and for restructuring for organic production within 30 years.

These scenarios have the following designations:

• Total pesticide phase-out (indicated by "0");
• Almost total pesticide phase-out (indicated by "0+");
• Limited use of pesticides (indicated by "+");
• Optimised use of pesticides (indicated by "++");
• Complete restructuring for organic farming.

Due to the lack of input data for market gardening, fruit growing and private forestry, no estimate has been made of the consequences of a partial phase-out.

6.1.1 Preconditions and methods of farming

Total phase-out of pesticides

For this scenario, consideration has been given to as many alternative methods of control as possible, including experience drawn from organic farming. Biological control in farming was, however, not considered as it was not expected to attain any practical significance within a 10-year period.

We have assumed that exemption would be granted for controlling diseases spread while sowing the early seed generations, as the consequences of the uncontrollable spread of such diseases would be enormous and would result in incalculable losses.

Point of departure in 12 different farm types

The point of departure of this analysis is the 12 farm types listed in Chapter 5.1. An agronomic review was done for 10 of these types, and we have proposed adjustments to crop-rotation regimes for when pesticides have been phased out.

The point of departure of the proposed crop-rotation regimes has been that the current production and structure of the farms be largely maintained, from the standpoints of livestock units and crop types, and that overall animal production also be maintained. In compensation for the drop in coarse-feed production, the land used for this purpose has been expanded slightly, at the expense of the acreage used for grain. Crop-rotation regimes using potatoes, sugar beet and seed grass have been retained, without investigating the practicability of such crop rotation in a total phase-out scenario. The present proportion of set-aside has been maintained.

The analysis also assumes identical mineral soils on the 10 farm types.

In extension of this scenario (which was proposed on the basis of agronomic considerations), another scenario was based on a combination of production optimised from the standpoints of agronomy and operating economy. In the latter case, a maximum of 30% set-aside was allowed, at the property level.

Partial phase-out of pesticides

The mandate did not specify partial phase-out scenarios, and the Committee has therefore set three scenarios for the partial phasing out of pesticide use. These three scenarios presume a combination of the prevention and control of pests.

Almost total pesticide phase-out (0+)

The scenario for the almost total phasing out of pesticides assumes that pesticides are only used where the crop would not be able to satisfy specific legislative requirements on purity, where there are requirements on the control of pests subject to quarantine regulations, as defined in statutory orders issued by the Plant Department, or for dressing first-generation seed corn. Thus, it would still be permissible to use pesticides for:

• protecting all seed corn up to and including the first-generation;

• controlling problematical weed species in seed grass;

• seed potatoes. Use of desiccants and products for controlling potato blight;

• controlling wild oat-grass in stands that make weeding impossible;

• control of Colorado beetles in seed potatoes;

• control of specific pests in pot plants and nurseries.

Limited use of pesticides (+)

Greater use of pesticides is permitted in this scenario than in the above scenarios. We have assumed that pesticide use is permitted in crops where there are large yield losses, or where it is considered that it would not be possible to sustain the profitable production of specific crops. The requirements were, a) significant average loss (>15-20%) due to pests, or b) that production would be so uncertain that it could be expected to fail or that it would interfere with crop rotation. This scenario makes no allowance for the fact that, at individual localities and in individual years, crop losses in excess of 15-20% can occur, since it is not possible to predict how often such a situation will arise for most crops.

Economic optimisation for crop rotation has been done in this scenario.
It would be possible to use pesticides in the following situations:

• minimum control of potato blight in potatoes for food and starch potatoes;
• dressing and band spraying in beets and sugar beet;
• controlling specific weeds in grain (e.g., camomile and charlock);
• controlling weeds in peas;
• controlling patches of perennial weeds, such as thistles, etc.;
• controlling weed grasses in particularly infested areas;
• controlling significant attacks of aphids in wheat and winter barley, when warnings are issued;
• controlling pollen beetles in spring rape, when the crop cannot compensate for the attack;
• band spraying of herbicides in maize;
• chemical control of couch grass every 10th year;
• controlling poisonous weeds, such as spring groundsel, in coarse feed;
• controlling aphids in grain and peas, when the damage threshold is exceeded;
• controlling snails and flea beetles in rape, when the damage threshold is exceeded;
• controlling clover weevil in clover seed production;
• controlling diseases and pests in market gardening and fruit growing on the basis of the intensity of the attack.

Optimised use of pesticides (++)

One assumption made in this scenario was that no significant economic losses would be caused by pests and that production would, therefore, remain at the same level as current production. The scenario is based on a proposal from the Danish Institute of Agricultural Sciences which, in 1996, assessed the realistic potential for reducing treatment frequency without affecting the present operating economy.

The scenario assumes the use of all of the present damage thresholds and mechanical weed control, where these methods can compete with the chemical methods. Crop rotation is presumed to be very similar to that in use today, where optimisation is practised for economy and also for minimising pesticide use as far as possible.

6.1.2 Choice of organic scenarios

Assumptions

We have attempted to clarify the consequences for the overall production of 100% organic farming in a number of scenarios, which differ in the quantities of imported fodder and in the yield levels of farming crops. The time horizon of the scenarios is 30 years. This long view was adopted because we were considering a situation that would mean significant structural changes in farming and which were considered possible with a 30-year period. Thus, we have assumed that livestock manure, and therefore the animals, can be distributed evenly over the entire country.

Three levels of fodder imports

According to the applicable rules on organic farming, such farms must purchase conventional fodder in quantities corresponding to 15-20% of the animals' daily fodder intake (measured as the energy in the fodder), and a certain percentage of conventional livestock manure. In a 100% organic Denmark, there would be no conventional farms from which to purchase livestock manure or fodder, although it would be possible to import fodder from abroad.

Three levels of fodder imports to Denmark are used in the scenarios:

• no imports, complete self-sufficiency in fodder;
• 15% imported for ruminants and 25%, for non-ruminants;
• unlimited imports of fodder and maintenance of the present level of animal production (1996).

Two different yield levels

The scenarios use two different yield levels in the predominant crops, i.e., grain and grass. The "present yield level", based on current organic practice, and an "improved yield level", in which it was assumed that corn production could be increased by 15% and clover grass production by 10%. The improved yield level was based on a more goal-oriented effort to increase corn production and better use of grass fields as a result of the lower yield of the individual dairy cow, as compared to present organic practice. The three levels of fodder imports and two yield levels are expressed in six different organic scenarios. An overview of these analyses is shown in Table 6.4.

Assumptions on production

In all scenarios, the production of milk and eggs corresponds to present production. The production of milk is limited by milk quotas and could be greater without causing agronomic difficulties. Beef is produced in proportion to the number of dairy cows, in the form of dry cows for slaughtering, heifers and bullocks. The remaining feed is used in producing pork, since meat from poultry is included in the model as pork. The production of pork varies, therefore, in proportion to the produced and imported quantities of fodder, and no vegetables were exported in the scenarios.

Greenhouse production, ornamental plants and fur-bearing animals not included

Production in greenhouses and the production of ornamental plants, etc., (a total of about 4,000 ha), as well as the production of fur-bearing animals, were not included in the scenarios. The assumptions and restrictions of the scenarios are described in more detail in the background report on organic scenarios, from the interdisciplinary group.

6.2 Methods of controlling pests

The Committee has assessed the feasibility of applying existing non-chemical methods of preventing and controlling pests, not only because this was requested in the Parliamentary resolution of 15 March 1997, but also because it was included in assessing the feasibility of a total or partial pesticide phase-out within a 10-year period.

In addition, the Committee has assessed the consequences of increasing the use of non-chemical methods.

6.2.1 Prevention and control when deciding crops and crop rotation regimes

Crop rotation is very important to the overall level of pests and their significance to crops. Thus, it is well-known that the need for pesticides is significantly lower in crop rotation in animal husbandry, with a high coarse-feed production, as compared to the need for specialised plant production, such as sugar beet or potatoes.

We do not consider that, under present economic conditions, there are any realistic prospects for changing crop choices to the production of fodder and vegetable foodstuffs. We do, however, consider that there are certain viable alternatives in the use of wholecrop for sows and biomass production for non-food purposes.

The cultivation of multiple crops, in the form of mixed seed, is not considered to offer any great crop-rotation potential in plant production, although its potential could be greater in organic production, where the inclusion of nitrogen-fixing species affects the yield.

6.2.2 Prevention and control of fungal diseases

Breeding and cultivation methods offer several options for preventing and controlling fungal diseases. At the time of writing, however, these methods are not sufficient to ensure that farming suffers no losses due to fungal diseases.

There is great potential for genetically reducing expected losses caused by leaf diseases. However, breeding probably cannot solve all problems simultaneously within a 10-year period. Since, apart from leaf diseases in grain, there is also a need to work with resistance to seed-born diseases and with better abilities for competing against weeds, it would be vital to set the right priorities in our breeding effort. We consider that there is a need for a major intensification of work on breeding and of research on breeding, if we are to see any noticeable change in the range of resistant varieties, as compared to the range available today.

Foreign plant breeding has great general significance to the Danish range of varieties and production; there is also a close collaboration between Danish and foreign breeders. The feasibility of changing Danish breeding priorities in favour of breeding for resistance would, thus, depend on the interests of foreign breeders.

There is not insignificant potential for the strategic use of resistance (e.g., increased use of mixed varieties, to reduce the losses that result from fungal diseases).

Several technical factors could be adjusted in present cultivation systems, such as the sowing time, fertilisation and quantities sown, which would improve the prospects of minimising the problems of pests. However, diseases can neither be prevented nor minimised solely by adjusting cultivation factors. Several of these technical changes would reduce the yield level.

Different methods, which can reduce attacks by diseases, are also available in market gardening and fruit growing. None of them can keep cultures free from all problems. Several of these methods are associated with an increased manpower effort.

The environmental advantage of developing and using resistant varieties is the (obviously) reduced consumption of pesticides, with the concomitantly reduced risk of polluting ground water and the surroundings. In the area of health, the gain would consist of the lowered exposure of spraying personnel or the farmer and smaller quantities of pesticide residues in the crops.

6.2.3 Seed-born diseases

85-90% of all grain seed is dressed today, as is a large part of other crops in Denmark. If dressing were to be generally omitted, we consider that there would either be a rapid proliferation of the significant seed-born diseases or that this part of our agricultural production would take place abroad, where dressing products are still permitted.

Continued dressing of the first generations of grain, followed by an assessment of the needs of subsequent consignments of seed, would be one way of reducing fungicide consumption - a way that should be examined more closely and tested. The assessment of needs would demand fast, certain analytical methods, the separation of seed consignments and probably the discarding of significant quantities of grain for multiplication. There could also be the question of considerable losses in beets, as a result of insect attacks and soil-born diseases, which could cause uncertainties when establishing a crop, if dressing were omitted.

Work is in progress today on several alternative methods of controlling seed-born diseases, including resistant varieties, the use of biological pesticides and technical control methods using hot water/air or brushes. None of these methods have been fully developed and major R&D work must be done before we can assess whether or not these methods could immediately replace the chemical methods.

The environmental advantage of developing alternative methods including, e.g., the use of resistant varieties, increased need assessment and biological pesticides, is the lower use of pesticides, even though the consumption of dressing is already low. Seed dressing entails a risk to the birds and small mammals that feed on the seed.

In the health area, the primary gain from omitting dressing would be the cessation of exposure during production, which is often carried out at large dressing plants, and during handling in connection with sowing, as well as smaller quantities of systemic pesticide residues in crops (i.e., the pesticides absorbed by plants).

6.2.4 Prevention and control of pests

There is only limited expertise on the resistance of Danish varieties to insects, which has so far been a largely unexplored area. Simple screening for receptivity to pests could prove this to be an unused resource. We consider it unlikely that the use of plants genetically-modified to promote insect resistance will become widespread within the next 10 years.

It is well-known that natural field fauna, such as ground beetles and spiders, influence pest populations. In some years, they make a significant contribution to controlling, e.g., aphids, whereas they are insufficient in other years, because of high proliferation rates. We lack specific knowledge of the effects occurring in this area.

The course of development of pest attacks is strongly affected by the climate and losses will be caused at regular intervals by major attacks, which cannot be prevented - typically in seasons of hot weather, when the proliferation rate is high.

Technical factors, such as the sowing time, fertilisation and soil preparation, can affect the populations of certain pests and such methods should be included to the extent practicable, to reduce the losses caused by pests.

Several alternative methods of controlling certain pests are available in market gardening and fruit growing. These methods include the placing of crops in satisfactory crop-rotation regimes, the adjustment of sowing times, the use of netting and of watering.

The use of insect-resistant crops would reduce the burden on the environment, due to the (obviously) reduced use of pesticides and the concomitantly reduced risk of polluting ground water and the surroundings.

In the area of health, the gain would consist of the reduced exposure of employees and smaller quantities of pesticide residues in the crops.

6.2.5 Prevention and control of weeds

Any total or partial pesticide phase-out would necessitate the combination of preventive, technical and mechanical methods, in order to attain a sufficient level of weed control. Experimental results have shown that there are potential options for mechanical weed control in almost all crops. In such crops as rape, mechanical methods can already compete with chemical methods. There is, however, some uncertainty as to how a conversion to mechanical weed control would affect the seed pool in the soil.

Mechanical weed control could be problematical in certain situations, e.g., special soil types, unstable weather conditions or poor crop establishment. Crop damage after harrowing and a generally lower level of weed control would increase losses and would be linked to increased costs, when crop choice and cultivation practice would need to be adjusted for the sake of weed control. The capacity of mechanical methods is generally less than that of chemical methods, which could be problematical in conjunction with unstable weather. We consider that there is great potential for improving present mechanical methods, including methods to replace manual weeding. A conversion to non-chemical methods would demand considerable retraining and supplementary training, and most farms would need to invest in new machines.

Under present cultivation conditions, the occurrence of poisonous plants in Danish farm produce presents no health problems for humans. There are occasional problems with mortalities in domestic animals caused by poisoning. In Denmark, spring groundsel and deadly nightshade are considered to be the two most significant poisonous species. It cannot be precluded that restructuring for organic/pesticide-free agriculture would allow these species to proliferate. There would hardly be any question of an increased poisoning risk for humans. But it cannot be precluded that there could be more cases of poisoning among domestic animals, which would cause a certain production loss in the form of reduced milk yields, reduced growth and suchlike.

Pursuant to the act on wild oat-grass, seed-bearing wild oat-grass must not occur during the growth season. When producing grain without pesticides, it would be necessary to replace the chemical control of wild oat-grass with manual weeding. This is a realistic method of controlling relatively small populations of wild oat-grass, although it is not realistic for large populations. In such cases, it would be necessary to change crop rotation in favour of coarse-feed production, in order to reduce the population.

The growing of grass and clover seed, as well as vegetable and flower seeds, covers a broad range of cultures. Denmark is the world's largest exporter of grass seed. Over 90% of our production is exported. Common factors of these cultures are that the crop is destined for sowing and that the primary price criteria are high purity, high germination capacity and that the seed contain no or only very few seeds of other cultures or weeds. These criteria set very high requirements on the cleanliness of the crops - requirements that, for the greater part of production, would be difficult to sustain without the use of herbicides, given our present level of expertise.

couch grass can be controlled without pesticides on most land. Comparisons of the necessity of controlling couch grass by mechanical harrowing after harvesting or by spraying glyphosate in plant-production crop-rotation regimes have been assessed in several studies. Mechanical harrowing after harvesting (as a substitute for treatment with glyphosate every four years) is necessary every year in such crop-rotation regimes. We have reasonably good experience of controlling couch grass in organic cattle farming, in which crop rotation is, however, very different to that practised in the individual types of plant production. Experience from organic farming shows that thistles can be a major problem. Variations in the quantities of root weeds from field to field will become greater without access to pesticides, as it can take several years to attain effective control of large stands of such weeds.

We must expect the costs of harvesting and drying to increase if chemical weed control were to be abandoned.

Provided that mechanical weed control is very efficient, the quantity of weeds would not differ significantly from that in fields treated with pesticides, so that the environmental gain for flora would be absent. Moreover, mechanical weed control has significant detrimental effects on the soil's meso- and macrofauna, especially on earthworms and springtails, and harrowing can damage crops.

On the other hand, increased mechanical weed control in farming would have generally beneficial effects on the environment, since it does not entail any risk of polluting ground water or of spreading pesticides to adjoining areas. Careful fertilisation is also considered to be an attractive option for improving the competitiveness of crops over weeds. All other things being equal, it would improve the utilisation of fertiliser and, thus, reduce fertiliser loss to the surroundings.

In the area of health, the most significant change resulting from the prevention of weed problems and the use of mechanical weed control, as compared to the use of pesticides, would be the reduced exposure of spraying personnel or the farmer and the reduced quantities of pesticide residues in the crops. However, broader use of manual weed control would increase the monotony of this repetitive work and would be a physical burden, despite the fact that it is only necessary for a brief period.

6.2.6 Growth regulation

Growth regulators are used in about 10% of our winter cereals, especially barley. Measured as the quantity of active ingredient, about a third is used in the market-gardening sector (ornamental plants) and about two-thirds is used in grain and seed grass. The use of growth regulators in winter wheat is diminishing. No identical trend has been discerned in rye, as this is more exposed to lodged corn. At the time of writing, there is no statistical information to demonstrate any drop in the use of growth regulators in seed grass.

Growth regulators are used to protect against lodged corn, quality reduction and increased difficulty of harvesting. In winter wheat, there are attractive options for the use of alternative methods of reducing the risk of lodged corn. Thus, this risk is small when cultivating varieties of good stalk stiffness and reduced plant counts. If varieties of lower stalk stiffness are cultivated, it could be necessary to reduce the applied quantity of nitrogen by 10-13 kg/ha. There is a considerable risk of lodged corn occurring in rye grown in the better soils, although this risk is lower in sandy soils. No rye varieties are free from the risk of lodged corn, although the risk is lower in some varieties. This risk can also be reduced by postponing sowing until the beginning of October and by reducing the quantities of seed sown and of applied nitrogen. This would, however, reduce the net yield by 6-7 hkg/ha.

The use of alternative methods of growth regulation in seed grass will only be reviewed to a limited extent. We can expect a reduction of cultivation stability in certain soils, until light is cast on the potential of alternative methods of growth regulation.

Growth regulators are primarily used in pot-plant cultivation to promote the especially richly-blossoming and compact plants that have the best sales values. No methods of replacing chemical growth regulators are immediately available for pot plants. Certain alternatives do exist, such as photo-induction or the reduction of phosphorus, but there is a great need for research on this, to determine whether or not these alternative methods can replace chemical methods in the wide variety of pot-plant cultures.

If the use of growth regulators were to cease, an increased risk of lodged corn would be expected in certain soils and on certain farms. This would cause major problems when drying the crop for harvesting. It would also cause increased pollution of grain by soil-born fungi, including by several species of Penicillium and Fusarium. The greatest problem in this context would probably be Fusarium culmorum, which is very common in soil and which can form a number of mycotoxins.

The mycotoxins constitute a general problem in conventional and organic farming, as they can proliferate under climate conditions that favour high humidity. They can also proliferate if grain drying takes too long.

6.2.7 Biological control

Biological methods (which include useful organisms and microbiological methods) of pest control have great potential in production in greenhouses, where they are already used to a significant extent in vegetable production, whereas there is still unutilised potential when producing ornamental plants in greenhouses. Effective methods for controlling diseases biologically in greenhouses are still limited. We consider that there is a certain potential in field cultures for the biological control of pests in special crops whereas, in the short term, biological disease control only appears to hold possible potential against seed-born diseases and fungi that damage germinating sprouts in spring-sown cereals.

There is a need for a major research effort in the development of field methods in this area, and in the improvement and dissemination of the use of biological control of diseases and pests in ornamental plants cultivated in greenhouses.

As is the case for the use of disease-resistant or insect-resistant crops, biological control would result in reduced environmental load, due to the reduced use of pesticides. In turn, this could be expected to reduce the risk of polluting ground water and surrounding areas.

Corresponding gains in the area of health would be the reduced exposure of employees and reduced quantities of pesticide residues in crops.

The utilisation of useful organisms and microbiological methods would, however, present a significant risk of proliferation of alien organisms, which could have a detrimental effect on the environment. Theoretically speaking, the proliferation of local species could also disturb natural ecological balances. The use of microbiological methods entails a risk of industrial injuries, in the form of allergies or bronchial diseases.

6.2.8 Use of damage thresholds and decision support system

In recent years, warning and decision support systems have been developed for several of our major crops, to support the farmer's assessment of the need for pesticides. Such decision support systems have made significant contributions to reducing and adjusting doses, not only by direct use of the programs, but also through advisory services and newsletters from the Danish Agricultural Advisory Service. Even though damage thresholds and decision support systems have attained a certain popularity, it has not been possible to reach all farmers. Damage-threshold systems are still lacking for a large number of crops, and there is great potential for the improvement of several of the present systems. We consider that it would be possible to attain a 20-50% reduction in many crops, by combining decision support systems with chemical and non-chemical methods.

Integrated production, which is described, e.g., by the Plant Protection Products Directive, is based on the use of the current body of knowledge to minimise dependency on pesticides.

Tests and research have shown that purposeful use of fertiliser, pesticides and other factors related to the phase-out can contribute to satisfying environmental requirements and simultaneously optimise production economically. The use of decision support systems offers clear potential for reducing the exposure of the environment and people.

6.2.9 Genetically-modified crops

Danish developments in genetically-modified crops, which it will be possible to market within a few years, have made the greatest progress in herbicide-resistant plants. When genetically-modified, herbicide-tolerant beet varieties are introduced, herbicide consumption is expected to be reduced significantly, by about 2 kg/ha active ingredient. We do not consider that there will be any significant reduction in the case of herbicide-tolerant rape and maize. Based on the current body of knowledge, we do not consider it possible to forecast the effects of genetically-modified plants on pesticide consumption in Danish agriculture during the next 10 years. All over the world, major research initiatives are under way in this area, which will without any doubt change our culture plants significantly. We must assume that it will be possible to establish a basis for reducing losses resulting from diseases, especially if techniques are developed for the fast improvement of genetically-modified, disease-resistant plants.

Genetically-modified plants offer an opportunity for reducing the use of pesticides and, therefore, the exposure of the environment and people. Some of these crops could, however, cause inadvertent spreading and concomitant damage to the environment. This is especially true of plants whose abilities for establishing themselves in competition with the natural fauna are improved.

Furthermore, plants that are resistant to insects could affect species other than the pests. In this context, we are particularly thinking of predatory insects and birds that feed on plant eaters associated with the genetically-modified crop. They could either be affected directly by the toxicity of their prey or indirectly, by changes in their food supply. Such effects can also be caused by the use of spray products. However, the potential effects of insect-resistant plants differ from those of spray products by virtue of the fact that they can occur throughout the growing season. It is to be expected, however, that a number of non-target organisms would be less affected by genetically-modified plants than by the conventional use of spray products.

In relation to current spraying techniques, the introduction of known alternative spraying techniques could only offer limited prospects of reducing the quantities of pesticides used. Exceptions to this are, however, techniques for positional treatment, which could offer the possibility of varied treatment patterns at the field level with the aid of GPS (Global Positioning System) technology.

There are good prospects for reducing spray drift by the use of new nozzles, which minimise the proportion of spray drops available for drifting. Some of these new nozzles increase capacity in comparison to earlier spray types, which also improves the prospects of carrying out spraying in calm weather. We also consider that, in fruit growing, new shielded sprays (which collect spray residues) could offer good prospects for reducing the impact on the surrounding environment.

We consider that there are good prospects for minimising the point-source pollution of ground water, own wells, borings and watercourses in connection with the filling, washing and cleaning of sprays, by reinforcing efforts in information and guidelines for the farmer, on the filling and cleaning of sprays. These alternatives would only require limited financial investments.

As far as health is concerned, the reduced need for spraying would immediately reduce the exposure of the spray operators.

6.2.11 New pesticides

New pesticides are constantly being developed to replace the old products and new products are also being developed that offer new control options, for instance, for take-all disease. These products are generally used in smaller quantities than has previously been the case, and there is an increasing tendency to use, e.g., certain insecticide products, as dressings. The search is being intensified for active ingredients derived from nature's own substances which, however, often need significant modification to become stable and suitable pesticides. The success rate for finding products has fallen, as a result of the more stringent environmental and health requirements now set on such products.

As resistance to many products is continuously increasing, the constant development of products that act through other mechanisms is vital, if we are to ensure continued effective pest control.

The data used when estimating losses resulting from disease and pest attacks were mostly obtained from pesticide studies conducted by the Landøkonomiske Foreninger (within the Danish Farmers' Union) and the Danish Institute of Agricultural Sciences. Several test-related factors are of significance to the magnitude and uncertainty, when estimating loss percentages resulting from pest attacks. Thus, although the magnitudes of losses obtained from such studies cannot be said to be representative of the losses that would occur in different farm types, there are no better sources on which to estimate losses caused by pests.

For some of the most significant pests, such as aphids in grain, it has been possibility to split losses into loss percentages in sandy and clay soils, whereas it has only been possible to use the national average for other pests.

Certain losses can only be calculated with difficulty, including effects on quality parameters. It is especially true of potatoes that losses resulting from poorer storage properties can only be calculated with difficulty. Consignments of potatoes with blighted tubers are particularly difficulty to store in loosely-layered clamps. Potato consignments with more than 2% infected tubers are considered high-risk consignments from the storage standpoint.

The production of malting barley is another area where crop quality can be significantly affected by whether or not pesticides are used. In some years, sorting (accounts are affected by the grain size) could be adversely affected by fungus or aphid attacks which could, in turn, make it impossible to sell the grain as malting barley.

Losses caused by harvesting difficulties and drying costs, which are particularly common where there are large weed stands, are also difficult to quantify and it is difficult to forecast the acreage that must be abandoned if they become overgrown with weeds.

There are large uncertainties in the loss magnitudes that would result from a switch to mechanical weed control. Only a few studies have been conducted, in which the effects on weeds and yield under a regime of mechanical weed control are compared with the effects of standard herbicide treatment. In some cases, it would not be possible to differentiate between the effects of remaining weeds on the yield and the crop damage that would be caused by mechanical weed control, itself. These studies were conducted under conditions of conventional farming, which means that the size and composition of the weed population would probably be lower than could be expected under herbicide-free crop rotation.

The weed problem in present crop-rotation regimes is due to the fact that the relevant species are well-adapted to the crops in question and that they are controlled less effectively with present pesticides. It is to be expected that, in a crop-rotation regime in which weed controlled is solely based on mechanical means, there would be an increase in highly competitive weed species, which are difficult to control mechanically.

The point of departure for loss estimates in grain was numerical data from the whole-year organic test farms where weeds were recorded, after mechanical weed control was carried out. The intensity of mechanical weed control at the organic farms is more limited than is generally recommended for mechanical weed control in grain. We therefore propose that the yield losses found as a result of the direct effects of weeds on the whole-year organic farms be halved, as it is assumed that increased mechanical weed control would reduce losses to below the present level.

This reduction would, however, be cancelled by the considerable losses resulting from the significant crop damage associated with mechanical weed control. Mechanical crop damage has, thus, been estimated on the basis of the comparison studies of herbicide treatment. As far as the other crops (which are not cultivated on the whole-year organic farms) are concerned, the loss estimates are based on the few studies available and on educated guesses.

The alternative methods that can be used to prevent and minimise the problem of pests are described in Chapter 6.2. The application of some of these alternative methods is, however, linked to yield drawbacks. To minimise weed problems, postponed sowing of wheat and winter barley is recommended. To ensure a healthy, competitive crops, however, sowing in the second half of September is recommended. Any additional postponement of sowing would reduce the prospects of establishing winter cereals in many years, and cannot, therefore be recommended. A minor yield loss (3-7%) can be expected as a result of the proposed postponement.

One of the other parameters that can reduce yield is the choice of the most resistant wheat varieties. Such varieties have a lower yield potential than the highest-yield varieties and, according to the tests of 1995-1997, giving resistance priority over yield would cost 4-5 hkg/ha. Since, in the case of losses resulting from disease, the average for all varieties during the period 1992-1997 (thus not making any special allowance for losses in the resistant varieties) was used, we have elected to halve the 4-5 hkg/ha loss due to the choice of resistant varieties, as the realistic additional yield of the more resistant varieties is considered to be 2 hkg/ha below that of the receptive varieties. 3% has been added for the choice of resistant wheat varieties, but there is nothing to indicate such a loss in the other species of grain.

6.3.2 Methods of estimating total loss magnitudes

Five different loss magnitudes have been used

The total loss has been pieced together from five different loss magnitudes; see Table 6.1. Loss 1 covers losses due to cultivation practice changed to minimise the risk of pest attacks, such as postponed sowing time and the choice of resistant varieties. Loss 2 covers losses resulting for disease attacks. Loss 3 covers losses resulting from pest attacks. Loss 4 covers losses resulting from the crop damage caused by mechanical weed control and Loss 5 covers losses resulting from the fact that more weeds remain after mechanical weed control than after the application of herbicides.

Loss magnitudes are multiplied together

The various loss magnitudes can either be added or multiplied. For our task, we have chosen multiplication. This ensures, for instance, that there is no risk of obtaining negative yields in extreme situations. In the studies, losses are usually expressed as hkg/ha, which we have converted to loss percentages. It has not generally been possible to differentiate loss magnitudes according to crop yields. This was only possible for diseases in wheat.

Indication of maximum loss

We have also calculated the maximum loss (loss, max.), which covers the situation in which one of the five loss functions gives the maximum loss and will, thus, establish a basis for the worst possible loss in the relevant crop. The maximum losses are often about twice as large as the average losses. Such losses can occur, e.g., if a potato blight attack develops very early in the growing season, or if wheat suffers a severe attack of stripe rust or Septoria. It is difficult to estimate the frequency at which such maximum losses will occur as they usually depend on the climate.

Table 6.1 : Please look here

Estimated loss percentages resulting from pests, etc., in different crops in the 0-scenario. Only direct yield losses have been included. Losses due to the increased costs of weeding are not included in this table.

The loss percentages shown are, however, considered relatively optimistic, e.g., as a result of the following:

• the expected loss resulting from weeds is halved, in comparison to what is observed on organic farms today. On the other hand, a larger loss has been added as a result of crop damage in connection with mechanical weed control;

• it is unknown whether, in a situation without control, disease epidemics would cause faster proliferation rates and reduced durability of the built-in resistance genes;

• no correction has been made for the situations in which the set conditions are not valid (e.g., low-level land);

• allowance has not been made to any great extent for the fact that production management would not be optimum in all situations;

• the fact that the latest fungicides (strobilurines) make it possible to harvest a greater yield than the traditionally-used products has not been taken into account.

6.3.3 Variations in yield level of current production

Scatter in yield level

In conventional cultivation the yield level generally exhibits very high scatter because of local cultivation conditions, climatic factor in the individual years and variations in pest levels. Taking matters at face value, greater variation is to be expected if pesticides are not available, as pests would have greater "freedom" to cause yield losses. This is reflected, e.g., in the many studies in which increased yields were harvested after spraying.

Figures from the National Department of Plant Production test database, which has collected data for the period 1992-1998, show very large annual yield variations for the tests in which pesticide treatment was generally carried out.

There is no data to document the variations in yield level in areas that have not been treated with pesticides at all. Data is only available for individual factors (disease or pests or weeds). If we consider the variations in the tests without fungicide treatment, and how they change over the years, the average differences between the treated and untreated areas are simply parallel level shifts in most years. In years of severe disease attack, such as 1998, there is a clear tendency for the curves to drop, which indicates reduced certainty of cultivation.

Variations at farm level

The yield variations in some crops are quite small. Large yield variations can occur in other crops, such as grass seed and potatoes. With several crops in crop rotation, the total yield of a farm will always vary less than with the individual crops. But types of farm with only a few crops will be strongly affected by the variations.

6.4 Methods for calculating economic consequences

The calculated yield losses and other agricultural adjustments (including the need for mechanical weed control) have been transferred for use in the analyses of operating economy in the different scenarios, The calculated results are, thus, expressions of the agronomic and partial operating-economic optimum situation.

With the aid of the Danish Institute of Agricultural and Fisheries Economics’ accounting statistics for working year 1995/1996, we have estimated harvest yields, product prices, subsidies and cost structures for each individual crop in the crop-rotation regime, for each of the 10 types of farm. A total of 2,000 accounts were included and, within each farm type, this number varies between 27 and 170 accounts. The accounts were specially selected and are considered to be representative of Danish farming.

As the prices and yields of rape were extraordinarily low in 1995/1996, we have elected to increase the production value of rape by DKK 900/ha in our calculations. It was also possible to assess the farms' costs for pesticides. We have distinguished between sandy and clay soils.

A linear programming model ("DØP", Danish acronym for "farm-economy pesticide model") was developed to calculate the economically optimum acreage and pesticide utilisation. For each scenario and farm type, the model calculated the acreage utilisation that gives the greatest total contribution margin II from field cultivation (contribution margin II is the amount available for covering the costs of building, land, etc., when all other costs, including wages, have been deducted). For each scenario and each farm type, the model calculated the acreage utilisation that gives the greatest possible return for land and buildings. A number of assumptions were built into the model, such as restrictions on pesticide use, as well as diverse crop-rotation restrictions, early crop effects, fodder balances, working capacity, etc. Only the crops used in Present Cropping were included in the calculations. Cattle farming was kept unchanged in the model and the aim was to sustain unchanged levels of feed production on the farms, throughout the entire phase-out.

Accounts were not kept of plant nutrients, although adjustments were made for the lower fertiliser costs in relation to the lower yields of the 0-scenario and the intermediate scenarios. One restriction was that set-aside land constituted a minimum of 10% and a maximum of 33% of the area with reform crops, including set-aside. 6% late crops were established, cf. the action plan on the aquatic environment II. The production of sugar beet, grass seed and clover grass was limited to the maximum quantity produced in 1995/1996 and a number of other assumptions were built into the model.

When correcting the contribution margin, allowance was made for yield losses and increases, changed costs for the purchasing and application of pesticides and changed costs for mechanical weed control. The value of the saved costs of pesticide application and the increased costs of mechanical weed control were determined on the basis of machine pool rates.

The costs of mechanical weed control were calculated for normal weed pressure. In the case of sugar beet and fodder beet, the costs were increased by the wages for 2x50 hours manual hoeing of weeds per ha. The costs of increased difficulty of harvesting and the increased need for drying were not included in the model and neither were costs of a more individual nature, such as difficulties with wild oat-grass or special weed problems on low-level land, etc. It was assumed that the increased costs of pest monitoring would amount to about DKK 150/ha/year.

The chemical control of couch grass was administered in the model as an independent use of pesticides common to the entire crop-rotation regime. The mechanical control of couch grass demands a regime with several spring crops, where late crops or winter cereals could otherwise be cultivated. For the control of couch grass without chemicals, the model demands space for thorough mechanical couch grass control every three years and winter cereals could only be cultivated on a maximum of 40% of the land.

The price of grain, in particular, has dropped considerably since 1995/1996 and higher surcharges have been imposed on pesticides. Price sensitivity analyses were undertaken for a single type of farm, i.e., plant cultivation on clay soil, to clarify ways in which these price changes could affect the operating-economic consequences. The full effect on the ad valorem tax on pesticides in the retail trade was not included in the calculations. In other words, the model has only calculated with a 25% increase in the prices of herbicides and fungicides and 50%, for insecticides, together with a 30% reduction in the grain price as compared with 1995/1996.

The basis of the analyses at the levels of the sector and society was Danish Institute of Agricultural and Fisheries Economics’ AAGE model (the agricultural applied general equilibrium model for the Danish economy). In principle, this model covers all Danish trades, industries and households, which are assumed to minimise production costs and maximise utility. Apart from the trades' and industries' demand for semi-finished products and primary production factors (work force, capital and land), this model also describes the available range of goods and services and, to a lesser extent, the public sector. The model also processes the trades' and industries' range of goods for export and their imports of goods and services for consumption and manufacturing. The model is characterised by the fact that all markets are in equilibrium by virtue of the assumption of totally flexible price and wage adjustment.

The model is based on a constant return in manufacturing which, in combination with the assumption of total competition on the markets, means that there is zero profit in the companies. Changes in consumer preferences and technology must basically be established outside the model.

This model makes possible the systematic description of the entire economy, as it captures the most important interactions and reactions in the economic system. The model shows adjustments in the economy in the long view, i.e., weight is attached to structural relationships in the economy. The model also makes it possible to illuminate the effects of changes in price conditions in manufacturing and factor utilisation, and the derived macroeconomic effects on consumption, employment, foreign trade, etc., which makes it suitable for quantifying the effects of changes in politico-economic initiatives.

It must be stated that the model cannot handle unbalanced aspects and the formation of expectations in the economy, for which reason it can tell us nothing of the extent and duration of adjustments from one state of equilibrium to another. From the standpoint of this analysis, this means that the model says nothing of the possible costs of adjustment that will confront the sector in the short term, if the use of pesticides were to be banned. It should also be noted that, in similarity to most economic models, this model is based on unchanged technology.

The model's input data is Danmarks Statistik's input-output table for 1992, in which the agricultural sector is split up into eight primary production sectors and five sectors of the agricultural manufacturing industry. Thus, primary agriculture is treated as an average farm, with eight production sectors, i.e., the model does not enable us to identify adjustment barriers within the sector, such as structural limitations and regional barriers to the adjustment of production. The output of the model must be interpreted as a result for the long term, where such barriers are negligible.

Subdivision of the operating-economic analyses into sandy and clay soils was retained in the sector and socioeconomic analyses, by weighting the sand and clay farms in the averages that were exported to the equilibrium model.

It was necessary to adjust the model on a number of points before it could be used for analysing the industrial and socioeconomic consequences of phasing out pesticides.

In the first place, the standard version of the model describes the agricultural consumption of pesticides as an aggregated item. We have therefore adjusted the specification of the model on a number of points, so that the consumption of different types of pesticide is specified for different crops. We have also built in options for substituting other types of factor related to the pesticide phase-out, which is necessary when modelling adjustments in pesticide consumption.

In the second place, it was necessary to extend the model with a description of pesticide-free production. In reality, this was a question of a new technology, for which the input data offers no description basis. As a novelty in general equilibrium analysis, the model was therefore extended so that, for each vegetable sector, a corresponding sector was defined to give the same production, but using a technology/combination of factors in which there are no pesticides (0-scenario), or in which pesticides are included to a limited extent (+-scenario). The combinations of factors in the alternative sectors were determined according to calculations made with the operating-economy model.

The data were converted for export by calculating for each plant-cultivation sector in the DØP model the percentage change in factor use on the changeover to pesticide-free production (this calculation was performed for each of the above types of farm and to give the weighted average for all agriculture). The percentages calculated in this way formed the basis on which to correct the use of factors in the AAGE model. As an example, Table 6.2 shows the correction of factor use in grain production, in the 0- and +-scenarios.

As can be seen from the table, a roughly 28%-larger area is needed to produce the same quantity of grain in the 0-scenario as in Present Cropping, which corresponds to a drop in yield of 23% per ha. In the +-scenario, an area increase of 16% is needed, which corresponds to a 14%-lower yield. It can also be seen that the contributions, e.g., of the machine pool, manpower and fertiliser in grain produce, would be about 18% greater than in Present Cropping, in contrast to 9-11% greater in the +-scenario.

In both cases, the scenarios were implemented by imposing a prohibition on the production of vegetable produce by present production technology. From the technical standpoint, production in the traditional sectors was "subsidised away", after which land, capital and the work force would be freed resulting, e.g., in a falling land interest rate. In such a situation, the land would be re-allocated to the alternative vegetable industries in agriculture (i.e., to the branches of the industries that do not use pesticides, or which only use them to a limited extent). In the new equilibrium, the land would be re-allocated between the active vegetable industries, so that the return to agricultural land would be identical in the individual branches of the industries). Capital and manpower would be re-allocated to the alternative vegetable sectors and to the other industries in the Danish economy.

The theoretical substitution options would not be used in the 0- and +-scenarios, as the consumption of pesticides would only occur within a fixed, rule-directed framework. The restricted use of pesticides is shown as permissible crop-dependent quantities and, for instance, in a fixed relationship to the amount of land (a fixed treatment quantity/ha). This would ensure that the use of pesticides in the +-scenario could not exceed the fixed framework of that scenario.

The analyses we have performed were based on a set of economic models (adjusted to the needs of analysis), which made it possible to clarify the economic consequences of phasing out pesticides at the levels of the farm, sector and society. The analytical concept was securely founded on economic theory and parts of the model have been used in calculating consequences in connection with the assessment of other political measures.

Basing the analyses on model calculations has meant that the results obviously reflect the fixed assumptions of the models. At the farm level, for instance, the calculations presumed complete knowledge and openness in the decision-making process, which probably reflected what the best managers could attain. Further, the analyses performed at the farm level focused on relatively short-term adjustments whereas, in the analyses carried out at the sector and socioeconomic levels, weight was attached to the long-term effects on agriculture and the Danish economy. These results must therefore be applied with caution, when planning policy in the short- and mid-term. It should also be noted that, as there are by definition no adjustment costs in the long term, the results of the equilibrium model underestimate the adjustment costs in the short term. Conversely, it is to be expected that the farm model's results overestimate the adjustment costs in the longer term, where there is greater potential for adjustment.

Finally, it should be noted that the models do not make it possible to describe technological changes. Thus, no consideration was given in the analyses to the fact that R&D could make it possible to develop crops and production methods that would be mode competitive in pesticide-free farming.

The above circumstances must, of course, be taken into consideration when assessing the results. Even though the exercise was a question of an idealised description of the situation in the three perspectives we consider that, despite their limitations, the analyses gave credible indications of the magnitudes and, in particular, the directions, of the effects of the scenarios analysed.

6.5 Environmental and health-related assumptions

Estimates were made of the environmental effects on different intermediate scenarios, in four main areas:

• effects of pesticides on the aquatic environment - algae and daphnia;
• effects of pesticides on the lower fauna - spring tails and earthworms;
• effects of pesticides on the avian populations of arable land - nine bird species;
• effects of pesticides on the soil's seed pool - five weed species.

The results of the model calculations should not be seen as an exact expression of the consequences for the relevant organisms. We should stress in this context that the model calculations cannot be considered to be an expression of the consequences for all other organisms in the terrestrial and aquatic environments, either.

Aquatic environment

The point of departure of the model calculations was a well-defined pond of between 30 and 450 m², with an average depth of 0.9 m, fed by the surface runoff from fields of 2-3 ha. In the model, pesticides are only transported by wind drift and surface runoff. The applicable requirements on distances from the various pesticides are built in and spraying is never carried out closer than 2 m to the aquatic environment. Wind drift constitutes a maximum of 1% of the acreage dose, whereas surface runoff is only considered to occur when the precipitation exceeds 10 mm, whereby the pond receives 0.2% of the pesticide content from the nearest 2-ha fields.

Models were set up for the main crops, i.e., grain, rape, potatoes, beets and peas. The doses and times of spraying of the individual pesticides conformed to the instructions for use. Only the most-used pesticides were included in the model calculations. The absorption of pesticides by crops, and the crops' eventual effects on pesticide degradation were not included in the model. The model does include the temperature dependency of pesticide degradation.

Birds on arable land

A number of bird species are characteristic of arable land in Denmark. Their population trends and distributions in the landscape were studied as part of research into the effects of pesticides on nature and the environment. Based on adjustments of this information, it was possible to undertake simple estimates of how avian populations would develop in the different scenarios of acreage and pesticide use.

The algorithm was based on data taken from three years' bird counts during breeding time on 54 large Danish farms, for which data was available on all crop and biotope conditions and all pesticide treatment. The distributions of the different species in relation to biotope conditions, crops and treatment frequencies were estimated and tested using covariance analyses. The treatment frequencies of pesticides that have exhibited statistically significant effects, as well as the mutual acreage relationships of the crops, were varied in the calculations.

It was assumed that the average field size would not change, that there would be no general changes in the number of hedges and other marginal vegetation and that other natural content of arable land would not change. It was also assumed that the each species' population density could be calculated independently of those of other species. It was assumed that, in the event that the effects of pesticides and herbicides occur simultaneously, the overall effect would be the product of the effects, i.e., as a mutually reinforcing effect. Finally, it was assumed that the estimated population densities could be extrapolated to the national level without any consideration for what the localities could actually support. The estimated success of populations can therefore be interpreted as the upper limits for the changes that can be expected.

Effects on fauna

To soil organisms, pesticide use as we know it in Denmark does not have much affect on the welfare of these species. The lower fauna are affected directly by treatment with insecticides, and indirectly, by the removal of plants and micro-organisms that form their food supply. The latter effect can be caused by the use of herbicides and fungicides. Other elements, such as soil treatment and the use of organic fertiliser, also have significant effects.

When assessing the effects on spring tails (Collembola) and earthworms, we have thus taken our point of departure solely in the effects that would be caused by the changed compositions of crops in the different scenarios. In assessing the effects on arthropods, the treatment frequency was used as an indicator of the actual effects.

Effects on flora

25-year plant trends were estimated by the application of two different types of mathematical model, i.e., the "seed-pool model" and the "crop-rotation model". These models have diverse limitations and have not been completely validated, but they can give preliminary estimates of development trends. The seed-pool model was developed for unrotated spring barley in sandy soil and, thus, does not include crop rotation. It used plant species that are frequently encountered as weeds. The model was validated over three years.

The crop-rotation model was developed for simulating crop-rotation regimes using genetically-spliced sugar beet and rape, respectively. The sugar beet model used a crop-rotation regime of sugar beet - barley - winter wheat - winter wheat, whereas the rape model used winter rape - winter wheat - winter wheat - winter barley. This model can test 4-6 wild plants as well as volunteer plants that occur as weeds. A rough estimate of the weed biomass trend was also calculated. Two seed-pool levels were tested in all three models. The first level was an average seed content of 6,900 seeds/m², which corresponds to the median value for Danish fields found in the latest study. The second level was 22,000 seeds/m², which corresponds to the upper limit for 80% of the fields.

The 0-scenario calculations were not the same as in the 0-scenario described by the sub-committee on agriculture, but covered a scenario in which no mechanical weed control or adjustments of crop rotation were practised. This permitted the wild plants to proliferate "freely". This scenario was compared with Present Cropping. An intermediate scenario was also estimated, which roughly corresponded to the +-scenario, as band spraying was carried out in beet crops, couch grass was controlled every 10 years, mechanical weed control was used and resistant varieties were cultivated.

6.6 Review of analyses of the individual scenarios

Table 6.3 shows the descriptions and calculations performed for the total or partial prohibition of pesticides in agriculture. This table shows whether or not economic optimisation of crop rotation was undertaken, estimates of farming and social economy and whether or not descriptions of the environmental and health-related effects were carried out.

Table 6.3: Please look here
Description of the analyses performed on the four scenarios for the total and partial prohibition of pesticides.

Due to the lack of data, no extensive estimates of the consequences of banning pesticides were performed for market gardening, fruit growing and private forestry. The areas in which a total ban would be problematical were described for the individual production areas. The economic consequences were also estimated in that connection. The environmental and health-related effects are only described qualitatively, as it was not possible to carry out calculations. No analyses were made for the intermediate scenarios.

Table 6.4 reviews the analyses that were done for the organic scenario.

Table 6.4: Please look here
Analyses performed on restructuring for organic production.