1.The Survey Population Consists Ofagricultural Companies and Enterprises Cultivating Utilised

1.The survey population consists ofagricultural companies and enterprises cultivating utilised agricultural land and farmers’ and family farms in Lithuania having more than one hectare of utilised agricultural land or annual agricultural income of more than EUR 1520.

2.The sampling frame for agricultural companies and enterprises is selected from the Statistical Business Register.The sampling frame for farmers‘ and family farms is a list of farms from the Statistical Farm Register, administered by Statistics Lithuania.

3.The census survey is carried out for agricultural companies and enterprises and the sample survey is carried out for farmers‘ and family farms.

4.A stratified simple random sample selection method is used to farmers’ and family farms. The sample size is 45008.

5.Notations:

– number of strata by the farm size (without “Take all” strata);

– total number of strata by the farm size;

– total number of municipalities used for stratification;

– sample size (without “Take all” strata);

, - population and sample sizes in the -th stratum by the farm size;

, - population and sample sizes in the -th stratum by the farm size -th municipality.

6.Some units from the sampling frame are selected for the survey with a selection probability equal to 1. Those include farmers’ and family farms with annual agricultural income (i.e. standard output) of more than EUR 10000, farms cultivating large areas of specific crops, such as nut trees, nurseries, perennial plants for twining, weaving, energy purposes, flax, oil crops, grain maize, tobacco, hops, aromatic, medicinal and culinary plants, plants for seeds and seedlings, fibre plants, or growing ostriches, also farms having an organic certification. Organic farms are appointed to a „Take all“ stratum (stratum number ) of 2470 units. The rest of the above mentioned farms are appointed to another “Take all” stratum (stratum number ) of 24725 units.

7.A sample of units from the rest of the population was drawn using two stratification variables: farm size in terms of standard output and municipality. The standard output for every holding is calculated fromthe data of the statistical register before the stratification.

8.First of all, remaining farms from the sampling frame were divided into strata by the standard output.

9.The rule was used to define the boundaries of the strata. We divide values of the standard output variable into intervals of equal length. Denote as the relative frequency of the elements in the population belonging to the k-th interval, . Then, we calculate the total and divide this total into equal parts. Thus, we find the approximate division points such that:

The elements belonging to the interval are appointed to the first stratum, the elements belonging to the interval are appointed to the second stratum, and so on.The elements belonging to the interval are appointed to the last stratum. There are 6 strata in total (including both “Take all” strata).

10.The Neyman optimal allocation method was used to allocate sample size to strata by the farm size. Let us denote the variance of (standard output of a farm) in the -th stratum by the farm size:

Then, we can define sample sizes as follows:

.

11.Population units in each stratum are divided into groups by municipalities. The sample size is allocated to strata by municipalities proportionally to their sizes:

,, .