0 Administrative Information about the Statistical Product


0.1 Name
Harvest of cereals etc.

0.2 Subject Area
Agriculture

0.3 Responsible Authority, Office, Person, etc.
Division of Agriculture and Transport
Merete Iisager, phone + 45 39 17 33 87, e-mail: mpr@dst.dk
Ina Nylander, phone + 45 39 17 33 88, e-mail iny@dst.dk

0.4 Purpose and History
Statistics Denmark compiles as well provisional as final results of the harvesting of cereals, rape seeds and pulses, based on information from farmers. The purpose is to estimate the harvest of cereals, rape and pulses for maturity, while harvest of cereals for silage and green fodder are included in the survey of roughage harvesting. Final statistics of crop yields are available as far back as 1875.

As a part of the cereal harvest surveys, the production and use of straw are compiled. Annual surveys on production has been conducted since 1970 and the use since 1972. Furthermore, extended surveys of straw for fuelling has been conducted for 1992/93, 1994/95 and 1996/97 after request from the authority: Danish Energy Agency.

Finally the statistics on cereals production includes prognosis on sowed areas with "winter cereals" and winter rape. Since 2002 these estimates has been based on the amount of certified seeds for cereals and rape. The results, which are important to forecast the coming harvest, has been released since 1967.

Harvest of grass seeds are based on information from The Danish Plant Directorate

0.5 Users and Application
The most important users are the EU and agricultural organisations. The data on production are also used in compiling the Economic Accounts for Agriculture (EAA). The use of straw for fuel is used by the Danish Energy Agency.

0.6 Sources
The survey on harvest of cereals and production and use of straw is based on data collected via questionnaires to app. 3500 farmers. In the calculations to national and regional results areas from the Agricultural and Horticultural Census are used.
For the prognosis on sowing of "winter cereals" and winter rape data from the Plant Directorate on certified seed and previous harvest surveys are used.

0.7 Legal Authority to Collect Data
The Act on Statistics Denmark.

0.8 Response burden
1.1 man-years, app. 347.000 DKK

0.9 EU Regulation
Council Regulation 837/1990 as regards survey on harvest and Directive 1989/130 as regards production of straw, while the values are part of Economic Accounts for Agriculture.


1 Contents


1.1 Description of Contents
The statistics on harvest of cereals, rape seeds, pulses and straw, are first and foremost a part of the agricultural production statistics.

The part covering cereals, rape and pulses provides data on crops (winter wheat, -barley, -rape; spring wheat, -barley, -rape, rye, oats, triticale, mixed cereals and field peas). For each type of crop, questions are asked about arable land, total yield (hkg), average yield (hkg/ha) and moisture percentage. The last information is used to calculate production to standard moisture content (15 per cent for cereals and pulses, 9 per cent for rape seeds)

The straw-part shows production and use for the above mentioned crops. The production is estimated based on a relation between yield of cereals/rape/pulses and straw yield for the different crops, while the use is based on data from the questionnaires filled out by the farmers.

Furthermore information on grass as aftermath in fields with cereals is collected and are the results are included in the survey on roughage crops.

The results are until 2006 calculated for entire Denmark and for counties. However, the counties in the Capital region are seen as one geographical area within statistics, meaning that results for 12 regions are calculated.

From 2006 and onwards results are compiled according to new administrative structure of Denmark, meaning broken down by regions (NUTS2) and provinces (NUTS3). Copenhagen and North Sealand (3 provinces) are merged into one group and there are no breakdown in provinces in Region Sealand. Region Northern Jutland is also considered as a province.

In the first dessimination, only results for the whole country are compiled. In the final results, also results for regions are given.

Within statistics on areas sown during autumn for next year harvest, areas for winter-wheat, winter-barley, rye, triticale and winter-rape are estimated

The on-off survey on use for straw for fuel is based on a sample among farms which, according to the Farm Structure Survey, has a fuelling system using straw. The farmers within the sample are questioned on the amount of straw used within their own holding, broken down on use for farming and dwelling, and the amount of straw (or energy) provided to others.

The result and grass seeds (area, yield, production) is compiled for about 20 types af grass seeds

1.2 Statistical Concepts
The survey harvest of cereals, rape seeds and pulses including production and use of straw is a based on a sample and the results are calculated using the population of farmers, as described below.

Population: All agricultural and horticultural farms, which had a total arable land of at least 5 hectares or an agricultural production corresponding to this in the last survey, are included in the population. In 2006 the population comprised almost 40,000 farms.

Sample size: 3,500 farms are selected from the sample used for last years agricultural and horticultural survey which comprises 20,000-25,000 farms. The reason for the selection methods is to select only newly update farms to minimize the risk to ask farmers without cereals.

Stratification: The sample selection is made by stratified random sampling, in order to ensure the best quality of results. The process of stratification is undertaken on geographical basis (divided into counties), on the size of the farms according to arable land with cereals, pulses and rape seeds and on Standard Gross Margin (economic size indicator) as subgroup. The farms are from 2007 divided into 80 strata. The total sample number is allocated to the strata to ensure national results with lowest possible statistical errors.

Sample selection: The sample is selected by undertaking stratification, random sampling from the sample used for the agricultural and horticultural census. However, at least 80 farms are drawn from each county. New sample is selected every year.

Raising method: In the compilation for provisional results the average crop yields measured in terms of hkg per hectare in each stratum in the sample survey on the basis of data on yield and area for each farm, are calculated for each strata and combined by the areas in the population to arrive at totals for the crop yields. On use of straw, the same principle is used; even the use of straw is recorded in per cent for the different purposes.

However, at the time when the provisional statistics of crop yields are compiled, only the total areas with crops at national level are known for the year under survey, as the final figures from the agricultural and horticultural census are not yet available. The total areas for each stratum are thus not known. The known division of areas into strata from the previous year is therefore used in allocating the total area into strata for the year under survey. The resulting areas for strata are multiplied by the average crop yields, implying that crop yields in each stratum are given. The total crop yield can now be calculated by a summation over the strata.

In compiling the final statistics the areas for each stratum are obtained from the agricultural and horticultural census for the same year. However, in 2005 thise areas are adjusted to totals rom the IACS-system. The method used in compiling the provisional statistics could in principle be applied. The only necessary change is that the known area figures for the strata are used in multiplying the average crop yields.

However, one disadvantage is associated with this procedure. The problem involved is that area figures are based on data submitted by the farmers to the Directorate for Food, Fisheries and Agri Business, which were reported in connection with the farmers' application for EU subsidies. These data are reported in April and not at the time of harvesting. This time-difference can imply that areas with cereals for maturity may instead have been harvested as cereals for silage/green fodder and conversely.

The need for corrections is investigated by comparing the areas from the Farm Structure Survey with the areas from the harvest survey for holdings present in both samples. If there for holdings with cattle is systematic deviations from areas in the spring to areas at harvest time, corrections is introduced.

Before 2001 areas from the harvest survey was used directly (for cattle farms) in the calculation of final harvest areas. However, because the sample for the harvest survey was not selected by random, the correction by this method was not fully reliable.

It applies to both the provisional and the final statistics that in strata where only two farms or less appear, the average crop yield in the stratum in question is equal to the county average, subsidiary national average for the crop in question.

Finally, it applies to both statistics that the data on moisture percentage are used in order to convert yields of cereals into a moisture percentage of 15 for cereals and 9 for rape. Before 2001 the standard moisture on cereals was 16 per cent.

Calculations in connection to harvest results
The harvest volumes of cereals are also shown in Feed Units (1 FU is almost equivalent to feed value of 1 kg of barley). The conversion factors are 1 FU = 0.97 kg wheat, = 0.00 kg rye, = 0.98 kg triticale, = 1.05 kg byg, = 1.26 kg of oats. For peas 1 FU = 0.92 kg and for rape seed 0.59 kg. The factors was revised in 2002 and has been used for calculation for 2001 data and onwards.

The calculation of produced straw is based on a estimated relationship between volume of grain and volume of straw. Om winter wheat, winter barley and spring barley 55 kg straw per 100 kg grain is expected. The factor for rye and triticale i 80, for oats 60, for spring wheat and peases 55, while 90 kg is expected for 100 kg of rape seeds.

Furthermore the volume of straw is compiled into Feed Units, 1 FU = 5.5 kg straw from wheat, 7.5 kg straw from rye, 5.0 kg straw from barley, 5.5 kg for straw from oats, triticale and mixed cereals, 4.0 kg straw from rape seeds and 3.5 kg of straw from peas.

Prognosis for winter cereals
The prognosis for sowing of winter-cereals and winter-rape is based on the actual certification of seeds combined by the relation between amount of certified seeds and harvested areas three years prior to the prognosis-year. However, on winter-rape, figures provided by The Farmers Advisory Service are used.

Regarding the on-off survey on specification on use of straw for fuel, please contact the responsible persons, mentioned in part 0.3.


2 Time


2.1 Reference Period
Regarding the harvest of cereals, rape seeds and pulses, the reference period is the end of the harvest the respective year, normally the beginning of September.
Regarding the prognosis, the data October 15 is used, as this data normally is the latest possible time for sowing in the autumn.

2.2 Date of Publication
The statistics are published annually.
Provisional figures for harvest of cereals, rape and pulses are published in November and final figures in April the following year.
The prognosis on sowing is published in the beginning of November.

2.3 Punctuality
The statistics are published as scheduled.

2.4 Frequency
Annually.


3 Accuracy


3.1 Overall accuracy
Overall accuracy on the harvest on cereals and production and use of straw is considered to be good. However, on crops where the total areas, in particular on spring-rape, the errors may be high.
The accuracy on the prognosis is lower than the accuracy on the harvest.

3.2 Sources of inaccuracy
Regarding the harvest, where the sample survey comprises about 3,500 farms, corresponding to approximately 7 pct. of all farms, the are statistical errors, even not calculated before 2004. For a description of the estimation procedure of the harvest results etc. see item 1.2.

Concerning the data within the questionnaries, there are expected errors in the data on the use of straw, because press and transportation of straw often is taken care of from neighbouring farms or agricultural service compagnies. Therefore, the responsible farmer to answer the questionnaires have not have full information.

The calculation of production of cereals is based on an expected relation between grain and straw, and is expected to be correct looking on average for miore years. However, in the single years, the relation will deviate from the average because of weather conditions etc. The relation is updated after 5-6 years.

On the sample the answering rate is normally more than 90 per cent and is not expected to cause errors in the estimations.

Regarding the prognosis, is is based on the assumption, that certified seeds are used in the same year as certified. However, impact of the weather conditions can mean that this assumption is not fully correct, why stocks of seeds will interfere the estimations. Furthermore, sown areas can be damaged by hard weather during winter and substituted by spring cereals. Also changeable use of own produced cereals as seeds can influence the prognosis.

3.3 Measures on accuracy
Sampling errors are from 2004 calculated every year. The standard error on total harvest is about 30.000 tonnes compared to total production on 9-10 mio. tonnes, i.e. the statistical error is 0,3 per cent Regarding crops with small total areas, the standard error can be up to 5 per cent. The errors are within the accepted limits in the EU-regulations.


4 Comparability


4.1 Comparability over Time
The final figures on crop yields are in principle comparable as far back as 1900. However, changes in methodology must be taken into account, but the present compilation method has been used since 1971.

The provisional statistics are comparable as from 1987.

The results on regional level are only compable from and to 2006. For this year results are compiled according to as well the old and new administrative structure of Denmark.

4.2 Comparability with other Statistics
The provisional statistics of crop yields are comparable with the annual forecasts of crop yields prepared by the Danish Agricultural Advisory Centre.

The prognosis on sown areas can be compared with the later results on areas in the agricultural census.

4.3 Coherence between provisional and final statistics
The difference between provisional and final figures for harvest range around 3 pct.

The prognosis is estimated once a year and there are in principle no provisional figures.


5 Accessibility


5.1 Forms of dissemination
The provisional statistics of crop harvest appear in Nyt fra Danmarks Statistik (News from Statistics Denmark). The final statistics on harvest are published in Agriculture (Agriculture) appearing in the series Statistiske Efterretninger (Statistical News).
Annual publications: Statistisk Årbog (Statistical Yearbook), Statistisk Tiårsoversigt (Statistical Ten-Year Review) and Landbrugsstatistik (Agricultural Statistics).
Figures are also available on www.statbank.dk/hst6, /hst7 and /hst5 (prognosis)

5.2 Basic material: Storage and usability
The survey data are kept for some years and can be subject for further analyses.
For further information, please contact the responsible person, see item 0.3.

5.3 Documentation
Methodological descriptions are published in Landbrug (Agriculture) appearing in the series Statistiske Efterretninger (Statistical News) and in Landbrugsstatistik (Agricultural Statistics). However, the present description is the most comprehensive.

5.4 Other Information
Please contact Mr. Ole Olsen, Statistics Denmark.


Supplementary documentation
No supplementary documentation is available