Pesticide use in agriculture (aei_pestuse)

National Reference Metadata in ESS Standard for Quality Reports Structure (ESQRS)

Compiling agency: Federal Public Service (FPS) Economy, SMEs, Self-Employed and Energy

Eurostat metadata
Reference metadata
1. Contact
2. Statistical presentation
3. Statistical processing
4. Quality management
5. Relevance
6. Accuracy and reliability
7. Timeliness and punctuality
8. Coherence and comparability
9. Accessibility and clarity
10. Cost and Burden
11. Confidentiality
12. Comment
Related Metadata
Annexes (including footnotes)

For any question on data and metadata, please contact: EUROPEAN STATISTICAL DATA SUPPORT


1. Contact Top
1.1. Contact organisation

Federal Public Service (FPS) Economy, SMEs, Self-Employed and Energy

1.2. Contact organisation unit


1.5. Contact mail address

North Gate II

Bd du Roi Albert II, 16 - 1000 Bruxelles

2. Statistical presentation Top

See sub-concepts below.

2.1. Data description

See sub-categories below.

2.1.1. Main characteristics of statistics

The data set comprises the quantity of actives substances contained in plant protection products used on selected crops for a choosen year in a five-year period and the area treated with each of these substances, according to the Regulation (EC) No 1185/2009 as amended by Commission Regulation (EU) 2017/269 of 16 February 2017 as regards the list of active substances.

2.1.2. Reference period of data collection

The data collection refers to 2017 in the 2015-2019 five year period.

2.1.3. National legislation
No Name of national legislation

not applicable Link to national legislation

not applicable Responsible organisation for national legislation

not applicable Year of entry into force of national legislation

not applicable Coverage of variables required under EU regulation

not applicable Divergence national definitions from EU regulation

not applicable Legal obligation for respondents
2.1.4. Additional comments data description

No additional comments.

2.2. Classification system

The classification used for pesticides corresponds to Annex III of Regulation (EC) No 1185/2009 ( of the European Parliament and of the Council.
The classification system for crops derives from the Annual crop statistics Handbook 2019 (

2.3. Coverage - sector

Agriculture is the main economic sector covered by the data set.

The selected crops are the principal crops in Belgium. These crops cover 95% of the Belgian Utilised Agricultural Area (UAA) in 2017.

All the active substances used by the Belgian agricultural holdings on the selected crops should be included.

2.3.1. Crops covered by the statistics

See the attached Excel file in the Annexes.

2.3.2. Commercial non-agricultural uses of pesticides

In the past, Belgium participated in the project "Pilot Studies on estimating nonagricultural use of pesticides”.

The common objectives for all pilot studies were to:

1) Identify methodologies and technical solutions for carrying out surveys on non-agricultural use of pesticides, with the aim of separating commercial and non-commercial use. This work should take into account also the work required to collect data on the use of pesticides in agriculture, in order to ensure coherence between the statistics.

2) Compare current experiences of Member States. This will be done by participating in coordination meetings and potential Task Forces.

3) Carry out pilot surveys, using the chosen methodology.

4) Use the collected data to analyze the use of pesticides for non-agricultural purposes also in relation to agricultural uses, with the aim of describing the potential threats to environment and other issues linked to pesticides.

5) Streamline the methodology on the basis of experience gained in the pilot studies.

Additionally, the project aimed at compiling a common, non-binding, European methodology or alternatively a handbook on pesticide statistics, taking into account the lessons learned in the pilot studies and actual data collection exercises.

Results of this project are available in the document "Towards a common methodology for the collection of data on the use of plant protection products in nonagricultural sectors.", 2015, Pierre Nadin and Vincent Van Bol.

2.4. Statistical concepts and definitions

The data reported are the quantity of each active substances listed in Annex III of Regulation 1185/2009 contained in plant protection products used on a selected crop, expressed in kg. The area treated with each substance are expressed in hectares.

2.5. Statistical unit

agricultural holdings

2.6. Statistical population

All the agricultural holding being covered by the Regulation (EC) No 1166/2008 on farm structure surveys.

2.7. Reference area

See sub-categories below.

2.7.1. Geographical area covered

The entire territory of the country.

2.7.2. Inclusion of special territories

not applicable

2.8. Coverage - Time


2.9. Base period

Not applicable for Pesticide Use Statistics, because it is not based on an index number of time series.

3. Statistical processing Top

Please see the sub-categories below.

3.1. Source data

See the attached Excel file in the Annexes.

3.2. Frequency of data collection

The FADN source data are collected every year on a national level, but only one year out of the five-year reference period (2017) is provided to Eurostat.

3.3. Data collection

See the attached Excel file in the Annexes.

3.4. Data validation

3 stages of validation :

- on the level of each agricultural holding, by the accountant (on base of the software application and usual reference databases : pesticide market name, correspondence between the type of pesticide and the crop, containing of active substances, ...)

- on the regional level (regional agricultural administrations and research centres) : global applied amounts per crop (consistency against standards, coherency, ...)

- on the national level by the FPS Health, Food Chain Safety and Environment and the FPS Economy, SMEs, Self-Employed and Energy comparing agricultural use of pesticides with the sales.

3.4.1. Data validation measures
3.4.2. Target of data validation measures
Data flagging
3.4.3. Specification target of data validation

not applicable

3.5. Data compilation

  • over the 3 years 2015, 2016 and 2017, more than 3 agricultural holdings used the substance on the crop in the sample, then the dose considered is the average dose over these 3 years.
  • over the 3 years 2015, 2016 and 2017, 3 agricultural holdings or less used the substance on the crop in the sample, then the dose considered is the dose of the year 2012.
  • some cases have been estimated on the basis of data from the Walloon region.

For 97 combinations active substance*crop, a dosage could not be estimated. On the national level, we ensure the consistency between the regional data (with possible adjustments), we determine the regional and national UAA per selected crops, we extrapolate the sample results to all of each region. For that, we apply the dosage (kg/ha) of the active substance used for a crop calculated on the basis of each sample to the entire UAA of the crop in each region.

The total quantity of active substance used for a crop is equal to the sum of the extrapolated quantities for each of the regions.

Some approximations must be made. So, the sample data for Walloon region does not distinguish between C1500 (grain maize and corn-cob-mix) and G3000 (green maize). The phytotechnical practices for these two crops being similar, we consider that the actives substances and the quantities of these substances per hectare are the same for the two crops.

For the Walloon region, separate data are available for wheat and spelt. We extrapolate the sample data for the two separate crops. The quantities for each of the active substances are then agrgegated to obtain the results for the Walloon C1100 aggregate.

In the sample of the Walloon region, there are no data for triticale (C1600), flowers and ornamental plants (N0000) and tomatoes under glass or high accessible cover (V3100S) . As the UAA for these crops are not zero, we consider that the substances applied and the dosages are the same as those applied in the Flemish region.

From the list of active substances and the Commission Reglementation (EU)2017/269 of 16 February 2017, different hierarchical levels are been determined. By linking these hierarchical
levels by the substance codes, the quantities used could be aggregated.

3.6. Adjustment

not applicable

4. Quality management Top

Please see the sub-categories below

4.1. Quality assurance

See sub-categories below.

4.1.1. Quality management system in organisation
4.1.2. Specification of implementation

At the individual agricultural holding level, some consistency checks are carried out on : category of pesticides on the right crops, observed doses not exceeding standards,

name and type of pesticide correctly referenced.

The same quality checks are applied to the regional level (NUTS 1). Comparison with previous years are also carried on and also with the litterature.

In case of inconsistencies, adjustments are performed by accountants in collaboration with farmers.

At national level, results are checked against annual sales figures.

4.1.3. Peer review
4.1.4. Main conclusions peer review

not applicable

4.1.5. Future quality improvements
Improve data validation
4.1.6. Specification of quality improvements

Studies should take place to try to improve the accuracy of the results. Sharing experiences with other countries could help.

4.1.7. Additional comments quality assurance

No other information.

4.2. Quality management - assessment

One of the weaknesses of this statistic is that the strata used for the FADN are perhaps not fine enough to be able to give reliable results by crops and by substance. The use of pesticides may vary significantly depending on some elements that are not taken into account in the FADN samples : different local growing conditions, cropping systems, recommendations of agronomists, weather conditions, ....But these accountancy data are the only data we have and this approach provide undoubtedly a better view of the agricultural use of pesticides compared to the only sales figures. Further studies will have to be carried out in the future to try to improve this statistics.

4.2.1. Overall quality
4.2.2. Relevance
4.2.3. Accuracy and reliability
4.2.4. Timeliness and punctuality
4.2.5. Comparability
4.2.6. Coherence
4.2.7. Additional comments quality assessment

It is difficult to draw real conclusions on changes in the agricultural use of pesticides by comparing the results of the first exercice year 2012 and the results of this exercice year 2017 for different reasons :-the fact that at a detailed level, the samples are probably not enough representative -the fact that we are forced to make certain assumptions for certains crops due to a lack of data : for the Flemish sample, assumptions had to be made for certain dosages for certain crops due to confidentiality concerns; for the Walloon region, assumptions had to be made for certains crops. - the fact that for the Flemish region, the software changed. The allocation between G1000 temporary grasses and G3000 permanent grasland is not the same than the allocation in 2012.

5. Relevance Top

Please see the sub-categories below.

5.1. Relevance - User Needs

At regional level, plants protection products application's estimations calculated from FADN acccountabilities are used by environmental administrations for their annual reports. Some research teams utilize also the regional calculation (see i.e. Fevery D, Peeters B, Lenders S, Spanoghe P (2015). Adjustments of the Pesticide Risk Index Used in Environmental Policy in Flanders. PLoS ONE 10(6): e0129669. doi: 10.1371/journal.pone.0129669).

5.1.1. Unmet user needs

Accurate usage estimates are essential to all pesticide risk indicators calculations.

5.1.2. Plans for satisfying unfilled user needs

Researchs must continue.

5.1.3. Additional comments user needs

no additional comments.

5.2. Relevance - User Satisfaction

There is no satisfaction survey but  exchanges between different actors : as the regional entities, the FPS Health, Food Chain Safety and Environment, the FPS Economy, SMEs, Self-Employed and Energy.

In the arcticle : Fevery D, Peeters B, Lenders S, Spanoghe P (2015). Adjustments of the Pesticide Risk Index Used in Environmental Policy in Flanders. PLoS ONE 10(6): e0129669. doi: 10.1371/journal.pone.0129669), it is considered that the estimated use of pesticides based on accountancy data is more accurate compared to sales figures and that this approach results in a better view on pesticide use and its respective environmental impact in Flanders.

5.2.1. User satisfaction survey
5.2.2. Year of user satisfaction survey

not applicable

5.2.3. Satisfaction level
Not completely satisfied
5.2.4. Additional comments user satisfaction

The detailed results of this statistic are approximate since they are based on a small sample of the population and assumptions and approximates are made.

5.3. Completeness

See sub-category below.

5.3.1. Data completeness - rate

Not applicable for Pesticide Use Statistics because in this data collection, there is no target on the number of data. Member States are asked to collect data on representative crops without stipulating the number of crops.

6. Accuracy and reliability Top

Please see the sub-categories below

6.1. Accuracy - overall

The first source of error is that the strata used for the FADN are perhaps not fine enough to be able to give reliable results for the pesticides used by crops and by substance. The use of pesticides may vary significantly, depending on local growing conditions, cropping systems and recommendations of agronomists, weather conditions, from area to area. Furthemore, the distribution of crops across land area varies, even in time.All of these elements are not necessarily taken into consideration in the samples.The second source of error comes from the assumptions made to estimate the missing substance dosages (for confidentiality reasons because of enough observations) in the data of the Flemish region. A data exchange protocol should be developed to obtain these missing data and a review should take place but the inclusion of lines with insufficient observations a statically error could be made. If one company has a very high usage of one substance then this will be extended to the entire populaton and thus result is an overestimation of the use of this substance. The third source of error comes from the assumptions made to estimate the missing data for the use of pesticides for Walloon region.The fourth source of error is that for certain crops (potatoes, for example) phytotechnics are considered to be the same between plants and seed potatoes. The last source of error is that only plant protection products who are applied on the main crops are recorded. Pesticides applied on bare soil or on minor crops are not taked into account. In the same way, 'all-purpose' pesticide (not related to specific crops) are probably underestimated. All these sources can introduce a bias difficult to determine between the estimated results of agricultural pesticides use and the reality.

6.1.1. Grading of accuracy
6.1.2. Factors lowering accuracy
Sampling error
Model assumption error
6.1.3. Specification of factors

sampling error : Results are based on the FADN samples. These samples don't take into account different elements from area to area of the country

(the different local growing conditions, the different weather conditions, the different recommendations of agronomist, ...)

model assumption error : some assumptions are made on the use of pesticides. These assumptions can introduce bias.

6.1.4. Additional comments overall accuracy

no additionnal comments

6.2. Sampling error

See the attached Excel file in the Annexes.

6.3. Non-sampling error

See sub-categories below.

6.3.1. Coverage error

See the attached Excel file in the Annexes.

6.3.2. Measurement error

See the attached Excel file in the Annexes.

6.3.3. Non response error

See the attached Excel file in the Annexes.

6.3.4. Processing error

See the attached Excel file in the Annexes.

6.3.5. Model assumption error

no model assumption error

6.4. Seasonal adjustment

Seasonal adjustment is not applicable to pesticide use statistics since all plant protection treatments associated directly or indirectly with the crop during the reference period are reported.

6.5. Data revision - policy

For this statistic, not systematic policy concerning data revision. In the future, consultative meetings between the data providers (FADN regional responsibles, the FPS Health and the FPS Economy) will take place to analyse the results and draw lessons.

For this first sending of data, assumptions made to estimate the missing substance dosages (for confidentiality reasons) in the data of the Flemish region.A data exchange protocol should be developed to obtain these missing data and a revision should eventually take place. But, there is a risk that data are not representative. These could introduce overestimation if one company has a very high usage of one substance and that this usage will be extended to the entire population.

6.6. Data revision - practice

There have been no revisions yet.

6.6.1. Data revision - average size

not applicable

6.6.2. Data revisions - conceptual changes
6.6.3. Reason for revisions
6.6.4. Impact of revisions
6.6.5. Additional comments data revisions

7. Timeliness and punctuality Top
7.1. Timeliness

See sub-categories below.

7.1.1. Time lag - first result

not applicable

7.1.2. Time lag - final result

not applicable

7.1.3. Reasons for possible long production times?

not applicable

7.2. Punctuality

See sub-categories below.

7.2.1. Punctuality - delivery and publication

Please see the sub-categories below

7.2.2. Data release according to schedule
7.2.3. Data release on target date
7.2.4. Reasons for delays

Data are not released nationally. There is thus no pre-announced schedule and no release calendar.

8. Coherence and comparability Top
8.1. Comparability - geographical

Data are collected on a regional level (NUTS 1). These data are comparable on a regional level taking into account all the previous remarks.

However, regional data are not disseminated. The geographical comparability between countries is evaluated by Eurostat.

8.1.1. Asymmetry for mirror flow statistics - coefficient

Not applicable, because there are no mirror flows in Pesticide Use Statistics.

8.2. Comparability - over time

Not applicable for Pesticide Use Statistics, because it is not based on time series.

8.2.1. Length of comparable time series

Not applicable for Pesticide Use Statistics, because it is not based on time series.

8.3. Coherence - cross domain

Agricultural pesticide use statistics are compared with sale statistics. The sales may differ from the actual use by import, export and storage.

8.4. Coherence - sub annual and annual statistics

Not applicable for Pesticide Use Statistics, because the data collection is based on a five-year period.

8.5. Coherence - National Accounts

Not applicable, because it has no relevance for national accounts.

8.6. Coherence - internal

The is no problem with the internal coherence.

9. Accessibility and clarity Top
9.1. Dissemination format - News release

There is no regular or ad-hoc press releases linked to the data.

9.1.1. Publication of news releases
9.1.2. Link to news releases

not applicable.

9.2. Dissemination format - Publications

not applicable

9.2.1. Production of paper publication
9.2.2. English paper publication
9.2.3. Production of electronic publication
9.2.4. English electronic publication
9.2.5. Link to publications

not applicable

9.3. Dissemination format - online database

The data are not available on-line.

9.3.1. Data tables - consultations

not applicable

9.3.2. Accessibility of on-line database
9.3.3. Link to on-line database

not applicable

9.4. Dissemination format - microdata access

Data are not accessible as micro-data.

9.4.1. Accessibility of micro-data
9.4.2. Link to micro-data

not applicable

9.5. Dissemination format - other

The data are only disseminated to Eurostat.

9.6. Documentation on methodology

The data are not nationally published. There are thus no national metadata files. Principal documents are research papers and are described at 9.6.4.

9.6.1. Availability of national reference metadata
9.6.2. Link to national reference metadata

not applicable

9.6.3. Availability of methodological papers
9.6.4. Link to methodological papers

1) Pilot study on estimating non-agricultural use of pesticides in Belgium for Service Public fédéral Santé Publique, Sécurité de la Chaîne alimentaire et environnement-Direction générale  nimaux, Végétaux et Alimentation. 2012-2014. E. Lievens, D. Fevery, L. Janssens, C. Bragard and P. Spanoghe

2) Towards a common methodology for the collection of data on the use of plant protection products in nonagricultural sectors. Eurostat-2015. Pierre NADIN – Vincent VAN BOL

3) Methodology for monitoring and reporting pesticides use. 12 November 2019. Eurostat. Workshop Pesticide statistics and indicators.

4) Estimation quantitative des utilisations de produits phytopharmaceutiques par les différents secteurs d'activité. 2020. CORDER asbl-Comité régional PHYTO, Applied microbiology-Phytopathology (ELIM-ELI-UCL), Earth and Life Institute, Université catholique de Louvain. Projet financé par le Service Public de Wallonie, Direction Générale Opérationnelle de l'Agriculture, des Ressources Naturelles et de l'Environnement.

9.6.5. Availability of handbook
9.6.6. Link to handbook

not applicable

9.7. Quality management - documentation

There is no quality report.

9.7.1. Metadata completeness - rate

There is no national metadata.

9.7.2. Metadata - consultations

not applicable

9.7.3. Availability of quality report
9.7.4. Link to quality report

not applicable

10. Cost and Burden Top

All the costs and burdens are for administrative purpose. For the respondants (farmers), there is no burden in addition to their commitment to the FADN.

10.1. Efficiency gains
10.2. Specification efficiency gains

not applicable

10.3. Measures to reduce burden
10.4. Specification burden reduction

not applicable

11. Confidentiality Top

Please sea the sub-categories below.

11.1. Confidentiality - policy

Statbel is subject to the statistical law of July 1962 ( and must ensure the confidentiality of data.

11.1.1. Transmission of confidential national data to Eurostat
11.1.2. Confidentiality according to Regulation
11.1.3. Data confidentiality policy

Confidential data is not published. Primary confidentiality and secondary confidentiality are treated.

11.2. Confidentiality - data treatment

Usually, we identify the primary confidential cells (dominance rule or small count) at the most detailed level. Then, we apply these primary rules at the different hierarchical levels.

If in one aggregate, there is one condifential cell, then we sacrifice another cell (secondary confidentiality) to protect it.

11.2.1. Procedures for confidentiality

In this statistic, it is very difficult to find individual data because the data are extrapolated from a sample. It would take a lot of information to be able to try to approach the microdata.

11.2.2. Additional comments confidentiality - data treatment

In the future, if the regions so wish, we can set up a system which would make it possible not to disclose the information if there are less than x declarants in the sample, but this system should be considered at national level otherwise the risk is important to have an impressive number of confidential cells

12. Comment Top

Related metadata Top

Annexes Top