Environmental conditions and past geo-political partitions have made Poland's agricultural space extremely diverse and complex. It is important to recognise this diversity and explain the mechanisms of change when managing agricultural development, which should take into account local determinants and potential (Adamowicz, Zwolińska-Ligaj, 2009; Dixon et al., 2001; Mądry et al., 2011; Rudnicki, 2016ab; Wysocki, 2010).
In geography, this is achieved by research on the spatial structure of agriculture, in terms of the set of interrelated agricultural attributes, phenomena and processes in a given space (Falkowski, Kostrowicki, 2001). The most comprehensive method in geographical–agricultural research is the typology of agriculture (Bański, 2007), in which Polish geography has made significant achievements. The typology of agriculture developed by J. Kostrowicki's team at the Institute of Geography and Spatial Organization of the Polish Academy of Sciences [Instytut Geografii i Przestrzennego Zagospodarowania Polskiej Akademii Nauk – IGiPZ PAN] has been used in research conducted in, among others, the USA, the former USSR, Canada, Australia, Brazil, India and Serbia (Kulikowski, 2005). Polish socio-economic geographers have drawn up a number of typologies for the entire country (e.g. Kostrowicki, Szczęsny, 1978; Szczęsny, 1988), selected regions (e.g. Biegajło, 1973; Głębocki, 1973; Falkowski, 1977; Stola, 1970; Matusik, 1973) and other European countries (e.g. Stola, 1973, 1977; Szczęsny, 1977, 1982; Tyszkiewicz, 1977). The continuing popularity of the developed typological method can be demonstrated by its continued presence in publications (Alvarez et al., 2014; Report ... 2016) and in relation to various regions of the world (Chendrayudu et al., 2015).
The present research issue is related to the rich tradition of typological studies on agriculture in Poland, which, according to the literature review, has become less relevant in favour of more comprehensive research into rural areas (e.g. Stola, 1993; Bański, Stola, 2002; Bański, 2014; Rosner, Stanny, 2014). The latest studies to deal with the typology of agriculture (the first on the gmina scale since 1988) are the works of B. Głębocki's team (2018, 2019). Due to the scale of the research and the data used, the results of these typologies are the most important point of reference for the presented proposal.
As previously mentioned, due to its utilitarian and practical nature, typological research on agriculture is conducted in various parts of the world, both in developed EU countries (e.g. Álvarez-López et al., 2008; Velbuena et al., 2008; Andersen, 2010) where they are intended to help prepare optimal agricultural policy instruments, and in developing and less developed countries (Parthasarathy et al., 2014; Chopin et al., 2015; Chendrayudu et al., 2015; Torero, Maruyama, 2016), where they often support measures to reduce malnutrition or poverty.
The purpose of this work is to elaborate the conception of the typological studies which allow the spatial structure of Polish agriculture to be delimited. The results let us answer the following questions: What were the main types of agriculture in Poland? How were they spatially diversified at the local and regional level? What are the external determinants of agriculture development?
2 Research methods and materials
The main source of the data used in the study were 2010 results from the National Agricultural Census (PSR). The data were aggregated by the location of each farm's headquarters (the headquarters of the farm should be understood as the gmina in which the farm has its seat or headquarters. In the absence of a formal headquarters or seat, the headquarters was determined to be the gmina that hosts the largest part of the farm's area (Central Statistical Office website  accessed 5.07.2018)). It should be noted that, despite the passage of time, the data used are currently the most complete source of information on agriculture at the survey's adopted territorial level. The employed database contained 931 attributes, which, in light of the spatial extent of the research (2,478 gminas), yielded over 2.3 million numerical values in various units (items, hectares, square metres, cubic metres). A multi-stage typology procedure was developed, the essence of which is presented in Fig. 1.
The first step involved selecting the following 12 diagnostic attributes determining the level of agriculture (these attributes were selected based on their typogenic impact in 12 thematic typologies of agriculture made in previous research. The attributes were cleared of outliers. Gminy excluded on this basis, a total of 14, are indicated in Fig. 2. For a precise description of the methodology, see Wiśniewski, 2019):
- (1)farms’ extensive forms of land use (total area of permanent pasture, forest land and other land) as a proportion of the total area of farms (destimulant),
- (2)average farm area (in ha of agricultural land),
- (3)number of persons working in Annual Work Units (AWU) per 100 ha of agricultural land (destimulant),
- (4)share of managers aged 44 or less in the total number of managers running farms,
- (5)farm managers by general educational level – secondary, further/post-secondary, or higher – as a proportion of the total number of farm managers running a farming business,
- (6)number of mechanised items per active farming business, the following point bonitation was used: combine harvester (5 pts); potato harvester (3 pts); beet harvester (5 pts); self-propelled forage harvester (5 pts); towed forage harvester (3 pts); fertiliser and lime spreader (1 pt); manure spreader (2 pts); towed mower (1 pt); front loader gripper (1 pt); potato digger (1 pt); potato planter (1 pt); collector trailer (2 pts); collecting press (3 pts); motorised plough (1 pt); multi-function plough (2 pts); towed field sprayer (2 pts); towed orchard sprayer (2 pts); bucket milking machine (1 pt); stanchion milking machine (2 pts); bucket milk cooler (1 pt); bulk tank milk cooler (1 pt). Author's bonitation based on assessment of technical complexity of machines and equipment,
- (7)consumption of mineral fertilisers calculated on the pure ingredient in kg of NPK (nitrogen, phosphorus, potassium) per 1 ha of agricultural land,
- (8)crop production intensity (in points, calculated using the B. Kopeć point method (1984)),
- (9)intensity of animal production (in points) - Ibid,
- (10)farms with non-agricultural income (combined categories of paid employment, non-agricultural income, pensions, non-employment and other income) constituting over 50% of income, as a proportion of the total number of active farms (destimulant),
- (11)farming businesses consuming 25% or less of final agricultural production value of the business, as a proportion of the total number of active farms,
- (12)total agricultural production per 1 ha of agricultural land (in PLN). Production was calculated by multiplying the area of individual crops and farm animal populations by SO “2010” coefficients. SO – Standard Output: the 5-year average of the annual production value of a given crop or stock farm per ha or per animal in regionally average conditions (Bocian et al., 2014).
Then, by principal component method (Hotelling, 1933), these attributes were reduced to four meta-variables that explained over 78% of the variance of the original set of attributes. The variables were subjected to cluster analysis using the k-means algorithm (MacQueen, 1967). The final number and composition of clusters were determined based on the mutual relations between intra- and inter-class variance of objects, having considered various different population classification variants (classification was tested using k-parameter values from 8 to 12). The goal was to maximise differences between the centroids of individual clusters while minimising differences between individual objects within clusters.
Thus, a 10-group division was arrived at, which was then verified and corrected using a random forest algorithm (Breiman, 2001). This approach, put very simply, consists in building multiple classification trees from a series of randomly selected samples (gminas and attributes alike were random), with the set being divided into two subsets: a training set and a verification set. In the final stage, the trees “vote” on whether a given object (gmina) belongs to a class.
The method in question attempts to eliminate the shortcomings of classic binary classification trees, and its key strengths are: resistance to over-training, estimation of incorrect-classification costs, resistance to disturbances associated with outliers or lack of data, and the ability to detect relationships between explanatory variables (Breiman, 2001; Hastie et al., 2013).
As a result of the random forest method, 390 gminas were re-designated. It should be remembered that the changes affected “extreme” gminas, i.e. those most difficult to group – with indicator values that in some sense “fit” two neighbouring types. The high classification quality is indicated by receiver operating characteristic (ROC) curve charts and classification convergence matrices (not included, due to publication length restrictions). A similar methodology – a two-stage division using k-mean cluster analysis and random forests – was used by R. Perdał in classifying gminas by degree of socio-economic development (2018).
To establish the identity of the types, a substantive formal identification was used that consisted in determining the name of the type (based on analysis of centroid locations in relation to the arithmetic mean of the diagnostic attributes adopted in the study). A significant difference between these values justified associating a type name with a given attribute. Substantive identification was carried out using cluster centres and the range of diagnostic attribute values (Wysocki, 2010).
Individual subtypes were grouped into three main classes characterised by a low, medium or high level of agricultural development. The criterion for this was the number of attributes above the national average (out of the original 12 attributes). It was arbitrarily agreed that in order to classify a given subtype as belonging to a highly developed type, at least half of the variables (six or more) should be high (above the national average). Fewer high-level attributes classified a given subtype as being of average (four or five attributes above average) or low (three or fewer attributes above average) level of agricultural development (see Table 1).
External determinants of agricultural development
|Type of determinants||Basis for delimitation and method||Classifications|
|Environmental||value of the agricultural production space quality index* / Jenks natural breaks method (1967)||1 – unfavourable|
(up to 91.3 points; 979 gminas)
|2 – medium-favourable|
(91.4–117.6 points; 1,052 gminas)
|3 – favourable|
(over 117.7 points; 447 gminas)
|Historical and political||political borders, 1815–1914 and 1918–39 (Kozłowski, Rudnicki, 2003)||1 – lands in the former Austrian partition and interwar Poland (351 gminas)|
|2 – lands in the former Russian partition and interwar Poland (1,118 gminas)|
|3 – lands in the former Prussian partition and interwar Germany (634 gminas)|
|4 – lands in the former Prussian partition and interwar Poland (375 gminas)|
|Level of socio-economic development||11 diagnostic attributes** / zeroed unitarisation (Kukuła, 1999) / Jenks natural breaks method (1967)||1 – low (848 gminas)|
|2 – medium (1,149 gminas)|
|3 – high (481 gminas)|
from the agricultural production space quality index used by the Institute of Soil Science and Plant Cultivation in Puławy in constructing a comprehensive environmental index of the ecological production utility of areas (Stuczyński et al., 2007).
from data of the Local Data Bank of the Central Statistical Office used to characterise a given territorial unit in demographic, infrastructure, financial and economic terms.
The resulting system was referenced against the spatial variation of external conditions for agricultural development. To this end, the gminas were divided into a number of groups representing their diversity of natural conditions, former historical and political divisions, and level of socio-economic development (see Table 1).
The procedure discussed above allowed three main types of agriculture to be distinguished: highly developed, medium developed and low developed, with 10 sub-types. The average values of the 12 diagnostic attributes for individual types and subtypes are presented in Table 2.
Description of types and sub-types of Polish agriculture
|Breakdown of types and subtypes||Diagnostic attributes|
|I. Highly developed||27.7||12.7||12.6||40.8||41.8||6.5||138.2||132.4||188.0||54.6||65.0||7,287.8|
|II. Medium developed||35.0||7.3||12.9||31.6||54.8||2.8||95.2||116.8||135.5||80.6||58.4||6,814.4|
|III. Low developed||58.4||4.6||22.8||33.4||41.9||2.5||51.3||98.7||102.3||84.3||35.5||4,735.4|
Each of the subtypes brings together gminas in which the agriculture has similar values of diagnostic attributes (see Fig. 2). These types can be characterised as follows.
3.1 Type I. Highly developed agriculture
This type is represented by 1,155 gminas (46.6%). By the criterion of similarity of diagnostic attributes, five subtypes were distinguished:
- highly developed crop and animal production (diversely developed; 281 gminas – 11.3%),
- highly developed agriculture focused on large-scale crop production (areally extensive; 303 – 12.2%),
- high level of agricultural development focused on specialist crop production, mainly fruit and vegetables (crop farms; 207 – 8.4%),
- highly developed agriculture focused on animal production (stock farms; 343 – 13.8%),
- highly developed agriculture characterised by industrial animal farming (industrial farms; 21 – 0.8%) (see Table 2).
Gminas of this type are considered to be strategic areas of agricultural production (Bański, Kulikowski, 2009) of fundamental significance to the Polish food economy. They are characterised by a highly developed farming culture and production specialisation, and are relatively large compared to the rest of the country (see Table 2). Support for this type of area should be directed at further improving competitiveness (modernising fixed assets), transferring modern technologies (for precision, energy and ecological farming), stimulating collaborations (e.g. producer groups), and further diversifying farm activities in order to strengthen both their surrounding areas (e.g. agri-food processing, services) and the multifunctional development of rural areas. Further changes should be made in a spirit of sustainable development, so as not to miss the opportunity to create a model of environmentally friendly agriculture. In this regard, particular attention should be paid to both the negative and positive effects of maintaining animal production and mineral fertilisation at optimal levels.
3.2 Type II. Medium developed agriculture
This type of agriculture (400 gminas – 16.1%) consists of two subtypes (see Table 2):
- medium level of agricultural development (transitional; 248 – 10.0%),
- medium level of agricultural development focused on raising stock (semi-industrial; 152 – 6.1%).
Medium-level agriculture is determined as areas where agriculture has to choose an appropriate development path. The first, preferred path is development and evolution towards highly developed agriculture. This could lead to success by exploiting, on the one hand, the development potential already available to the farms of such areas (as confirmed by the relatively large size, low workload and good technical equipment) and, on the other, interventionism (both national and regional) aimed at planning development according to the principles of sustainable agricultural and rural development (Roszkowska-Mądra, 2014). The second path of change is towards low-developed agriculture.
3.3 Type III. Low-developed agriculture
This type comprises 923 municipalities (37.2%) and consists of three subtypes:
- low level of agricultural development determined by specific environmental conditions (special; 250 – 10.1%),
- low level of agricultural development determined by the subsistence nature of production (traditional; 362 – 14.6%)
- low level of agricultural development determined by agrarian overpopulation and fragmentation (problematic; 311 – 12.6%).
Areas of the low-level agricultural development type are associated with problem areas in Polish agriculture (Bański, 1999; Czapiewski, 2010). This is confirmed by the low-developed agriculture type overlapping with the delimitation of agricultural areas that lag behind in development (as designated for both 1996 and 2002). It is impossible to assess changes in this system for methodological reasons, while the distribution of the main “hotspots” indicates that the socio-economics of these areas is relatively persistent in Poland.
3.4 Agricultural sub-types and external determinants of agricultural development
The spatial arrangement of selected sub-types (see Fig. 2) is related to external determinants of agricultural development. These relationships were determined by what percentage of gminas of a given agricultural sub-type was characterised by each political and historical determinant type, by each environmental determinant type, and by each socio-economic level determinant type (see Table 3).
Assessment of the spatial diversity of agricultural development types and subtypes, according to the share of gminas of a given subtype within three categories of external determinants of agricultural development
It has been shown that the impact of natural determinants is spatially diverse. Gminas with unfavourable environmental conditions predominate in areas with both low-developed agriculture (mainly subtype III.1) and highly developed agriculture (subtypes I.4 and I.5). It has been shown that, for subtypes I.1 and I.2, gminas with average environmental conditions constitute the numerically largest share. This attests to agricultural culture strongly influencing agricultural production results.
Meanwhile, the spatial arrangement of types of Polish agriculture was heavily determined by historical and political factors. This is mainly due to the fact that in the 19th century – the period of transition from a feudal to capitalist economy – the territory of Poland was occupied by three partitioning powers, each of which differed in its pace and level of socio-economic development. The analysis focusing on this aspect showed that among low-developed agriculture, the numerical majority of gminas of sub-type III.1 (problematic) was found in the area of the former Austrian partition. Areas within the borders of the former Russian partition were associated with a significant share of type I gminas, including sub-types I.3 and I.4. Within this partition, the majority of gminas were of the low-developed agriculture type, including subtype III.2. Attributes were of a relatively higher level in the gminas of the former Prussian partition within the interwar borders of both Germany (type I.2) and Poland (type I.1).
The diversified territorial types were also referenced against level of socio-economic development. It was found that a group of gminas of low-level socio-economic development had a majority of highly developed agriculture subtypes – I.4. Elsewhere, gminas of medium socio-economic development level predominated in three highly-developed agricultural sub-types (I.1, I.2, I.5), while only sub-type II.1 had a majority of gminas of high socio-economic development.
The issues raised relate to a rich tradition of agricultural typological research that, after almost three decades of diminished popularity, is attracting renewed interest among researchers (Rudnicki, 2016ab; Głębocki et al., 2018, 2019; Wiśniewski, 2019). The presented method for delimiting the spatial structure of agriculture uses results from PSR 2010 developed according to the Eurostat methodology. It may be of particular interest to revisit and continue similar studies in other EU countries using data from further agricultural censuses. It should be remembered that an agricultural typological method should allow “the agricultures of all countries and times to be objectively compared against one another” (Kostrowicki, 1983, p. 607).
The results of the typology showed that the most devolped agriculture was in the Kujawsko-Pomorskie, Wielkopolskie and Opolskie voivodeships. By contrast, the lowest developed type dominated in the Małopolskie and Podkarpackie voivodeships. The spatial structure of the selected types and sub-types is connected with external determinants of agriculture development.
The obtained results may constitute an important and useful source of data for further analyses, especially those aiming to determine the directions of agricultural and rural development policy. In this regard, the scale of the research appears to be an important advantage, as it allows local diversity to be determined, in contrast to regional approaches, which do not. The developed proposal for a comprehensive approach to the spatial structure of agriculture synthesises information that traditional portrayals presented using a range of individual indicators. The advantage – in terms of the practical usability of the results – is their readability and ease of interpretation, which results from the relatively small number of types and their names, which clearly indicate the nature of the classification.
According to the authors, the research results display high utility, and the random forest method (not previously used in Polish agricultural geography) proved an effective classification tool that improved the quality of the division obtained by traditional methods (k-means).
The research has shown that Polish agriculture constitutes a very complex system. It is a proving ground of sorts, on which diametrically opposed types of agriculture function alongside one another. The harmonious development of such a complex system requires specialised knowledge and in-depth analyses, including spatial ones. Therefore, the continuation of this type of research should be considered an important task for geographers.
The article presents selected research results from the doctoral dissertation entitled “Struktura przestrzenna polskiego rolnictwa w pierwszej dekadzie XXI wieku” [The spatial structure of Polish agriculture in the first decade of the 21st century] (Wiśniewski, 2019).
Adamowicz, M. and Zwolińska-Ligaj, M. (2009). Koncepcja wielofunkcyjności jako element zrównoważonego rozwoju obszarów wiejskich (The Concept of Multifuncionality as an Element of Sustainable Development of Rural Areas). Zeszyty Naukowe SGGWw Warszawie, Polityki Europejskie, Finanse i Marketing, 2 (51): 11–38.
Alvarez, S. Paas, W. Descheemaeker, K. and Groot, J. (2014). Constructing typologies, a way to deal with farm diversity: General guideline for humid tropics. Draft version for SRT1-workshop, 11–13 March 2014, Nairobi.
Álvarez-López, C.J. Riveiro-Valiño, J.A. and Marey-Pérez, M.F. (2008). Typology, classification and characterization of farms for agricultural production planning, Spanish Journal of Agricultural Research, 6(1): 125–136. DOI: https://doi.org/10.5424/sjar/2008061-299
- Export Citation
)| false Álvarez-López, C.J. Riveiro-Valiño, J.A. and Marey-Pérez, M.F.(2008). Typology, classification and characterization of farms for agricultural production planning, Spanish Journal of Agricultural Research, 6(1): 125–136. DOI: https://doi.org/10.5424/sjar/2008061-299 10.5424/sjar/2008061-299
Andersen, E. (2010). Regional typologies of farming systems contexts, SEAMLESS Report No.53, SEAMLESS integrated project, EU 6th Framework Programme, contract no. 010036-2, www.SEAMLESS-IP.org, 41 pp, ISBN no. 978-90-8585-596-5.
Bański, J. (1999). Obszary problemowe w rolnictwie Polski (Problem areas in Polish agriculture). Prace Geograficzne, 172, IGiPZ PAN, Warszawa.
Bański, J. (2007). Geografia rolnictwa Polski (Geography of Polish agriculture). PWE, Warszawa.
Bański, J. (2014). Współczesne typologie obszarów wiejskich w Polsce – przegląd podejść metodologicznych (Contemporary typologies of rural areas in Poland – an overview of methodological approaches). Przegląd Geograficzny, 86(4): 441–470. DOI: http://dx.doi.org/10.7163/PrzG.2014.4.1
Bański, J. and Kulikowski, R. (2009). Analiza obszarów wiejskich pod kątem konkurencyjności i towarowości (Analysis of rural areas in terms of competitiveness and marketability). Studia Obszarów Wiejskich, 16: 39–54.
Bański, J. and Stola, W. (2002). Przemiany struktury przestrzennej i funkcjonalnej obszarów wiejskich w Polsce (Transformation of the spatial and functional structure of rural areas in Poland). Studia Obszarów Wiejskich, 3, Warszawa.
Biegajło, W. (1973). Typologia rolnictwa na przykładzie województwa białostockiego (Agricultural typology a study made on the example of the Białystok voivodship). Prace Geograficzne, 100, IG PAN, Warszawa.
Bocian, M. Cholewa, I. and Tarasiuk, R. (2014). Współczynniki Standardowej Produkcji „2010” dla celów Wspólnotowej Typologii Gospodarstw Rolnych (“2010” Standard Production Factors for the purposes of the Community Typology for Agricultural Holdings). IERiGŻ-PIB, Warszawa.
Chendrayudu, N. Mamatha, S. Chandrasekhar, Reddy, D. and Veerannachari, V. (2015). Agricultural typology for modernisation: a case study of Rayalaseema Region, Andhra Pradesh. Asia Pacific Journal of Research, 1(XXV): 170–176.
Chopin, P. Blazy, J.M. and Doré, T. (2015). A new method to assess farming system evolution at the landscape scale. Agron. Sustain. Dev., 35: 325–337. DOI: https://doi.org/10.1007/s13593-014-0250-5
Czapiewski, K. (2010). Obszary problemowe w rolnictwie w kontekście delimitacji obszarów ONW (Problem Areas in Agriculture in the Light of LFA Areas Delimitation). Studia Obszarów Wiejskich, 24: 89–98.
Dixon, J. Gulliver, A. Gibbon, D. and Hall, M. (2001). Farming systems and poverty: Improving farmers – livelihoods in a changing world, FAO and World Bank, Rome.
Falkowski, J. (1977). Próba typologii i regionalizacji rolnictwa aglomeracji bydgosko-toruńskiej (An attempt at the typology and regionalization of agriculture in the Bydgoszcz-Toruń agglomeration). Przegląd Geograficzny, 49(4): 713–730.
Falkowski, J. and Kostrowicki, J. (2001). Geografia rolnictwa świata (Agriculture geography of the world). PWN, Warszawa.
Głębocki, B. (1973). Rolnictwo województwa zielonogórskiego – struktura przestrzenna (Agriculture of the Zielona Góra Voivodship - spatial structure). PWRiL, Warszawa.
Głębocki, B. Kacprzak, E. and Kossowski, T. (2018). Identyfikacja typów rolnictwa Polski z uwzględnieniem zależności przestrzennych (Identification of Polish agriculture types taking into account spatial relationships). Studia KPZK, 183: 315–330.
Głębocki, B. Kacprzak, E. and Kossowski, T. (2019). Multicriterion typology of agriculture: a spatial dependence approach. QUAESTIONES GEOGRAPHICAE 38(2): 29–49. DOI: https://doi.org/10.2478/quageo-2019-0021
Hastie, T. Tibshirani, R. and Friedman, J. (2013). The Elements of Statistical Learning Data Mining, Inference, and Prediction, Second Edition, Springer.
Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components. J. Educ. Psychol., 24: 417–441. DOI: https://doi.org/10.1037/h0071325
Informacja o wynikach kontroli: Nadzór nad funkcjonowaniem ferm zwierząt (Information on the results of the control: Supervision over the functioning of animal farms). (2014). Najwyższa Izba Kontroli, Warszawa.
Jenks, G.F. (1967). The Data Model Concept in Statistical Mapping. International Yearbook of Cartography, 7: 186–190.
Kopeć, B. (1984). Intensywność organizacji w rolnictwie polskim w latach 1960–1980 (The intensity of the organization in Polish agriculture in 1960–1980). Roczniki Nauk Rolniczych, G, 84(1): 7–27.
Kostrowicki, J. (1983). Polska geografia rolnictwa. Mechanizmy rozwoju (Polish agriculture geography. Development mechanisms). Przegląd Geograficzny, 55(3–4): 597–631.
Kostrowicki, J. Szczęsny, R. (1978). Typy rolnictwa (Types of agriculture) [in:] J. Kostrowicki (ed.), Przemiany struktury przestrzennej rolnictwa Polski 1950–1970, Prace Geograficzne, 127: 428–478.
Kozłowski, L. and Rudnicki, R. (2003). Historical-economic basis of present regions in Poland. International Conference, Between Schengen Space and Enlargement to the East: The Territorial Recombining of the European Union, Metz and Schengen, International Geographical Union, Commission on Political Geography, 134–141.
Kukuła, K. (1999). Metoda unitaryzacji zerowanej na tle wybranych metod normowania cech diagnostycznych (Zero unitarization method against the background of selected methods for normalizing diagnostic features). Acta Scientifica Academiae Ostroviensis, 4: 5–31.
Kulikowski, R. (2005). Dorobek Instytutu Geografii i Przestrzennego Zagospodarowania PAN w zakresie geografii rolnictwa (The achievements of the Institute of Geography and Spatial Organization of the Polish Academy of Sciences in the field of agricultural geography). Studia Obszarów Wiejskich, 7: 17–52.
MacQueen, J. (1967). Some methods for classification and analysis of multivariate observations, Proc. Fifth Berkeley Symp. on Math. Statist. and Prob., Univ. of Calif. Press, 1: 281–297.
Matusik, M. (1973). Próba typologii i regionalizacji rolnictwa na obszarze Dolnego Powiśla (An attempt at an agricultural typology and regionalization of Lower Powiśle, Nothern Poland). Prace Geograficzne, 102, Warszawa.
Mądry, W. Gozdowski, D. Roszkowska-Mądra, B. and Hryniewski, R. (2011). Typologia systemów produkcji rolniczej: koncepcja, metodologia i zastosowanie (Typology of farming systems: concept, methodology and use). Fragm. Agron., 28(3): 70–81.
Parthasarathy, R.P. Lagesh, M. Deb, U.K. and Jagadeesh, E. (2014). Typology of Agriculture for India for Technology Targeting and Development Planning. Paper presented at the pre-workshop symposia session on ― Rapid Transformation of Rural Economies in South Asia: Insights from Village Dynamics Studies‖ at the 8th ASAE International Conference, 14 October 2014, Dhaka.
Perdał, R. (2018). Zastosowanie analizy skupień i lasów losowych w klasyfikacji gmin w Polsce na skali poziomu rozwoju społeczno-gospodarczego (Using cluster analisys and technique of random forests in the classification of communes in Poland on the scale of socio-economic development). Metody ilościowe w badaniach ekonomicznych, XIX/3: 263–273. DOI: https://doi.org/10.22630/mibe.2018.19.3.24
Report and proposal for an International Framework for Farm Typologies, (2016). Technical Report Series GO-16-2016.
Rosner, A. and Stanny, M. (2014). Monitoring rozwoju obszarów wiejskich. Etap I. Przestrzenne zróżnicowanie poziomu rozwoju społeczno-gospodarczego w 2010 r. (Monitoring of rural development. Stage I. Spatial differentiation of the level of socio-economic development in 2010). Fundacja Europejski Fundusz Rozwoju Wsi Polskiej, Warszawa.
Roszkowska-Mądra, B. (2009). Koncepcje rozwoju europejskiego rolnictwa i obszarów wiejskich (The Concepts of agricultural and rural development in the European Union). Gospodarka Narodowa, 10: 83–102. DOI: https://doi.org/10.33119/gn/101247
Rudnicki, R. (2016a). Rolnictwo Polski. Studium statystyczno-przestrzenne lata 2002–2010 (Polish agriculture. Statistical and spatial study). Wyd. Naukowe UMK, Toruń.
Rudnicki, R. (2016b). Spatial structure of Polish Agriculture conditioned by Common Agricultural Policy instruments, Wyd. Naukowe UMK, Toruń.
Stola, W. (1970). Próba typologii rolnictwa Ponidzia (Attempt at an agricultural typology of Ponidzie in Central Poland). Prace Geograficzne, 81, IGiPZ PAN, Warszawa.
Stola, W. (1973). Rolnictwo departamentu Vaucluse (Francja). Próba typologii (Agriculture of the department of Vaucluse (France). Typology attempt). Dokumentacja Geograficzna, 3, Warszawa.
Stola, W. (1977). Próba zastosowania metod typologicznych do badań porównawczych rozwoju rolnictwa Belgii i Polski (An attempt to apply typological method in a comparative study of agricultures in Belgium and Poland). Przegląd Geograficzny, 49(4): 757–771.
Stola, W. (1993). Struktura przestrzenna i klasyfikacja funkcjonalna obszarów wiejskich Polski (Spatial structure and functional classification of rural areas in Poland). Dokumentacja Geograficzna, z. 3, Warszawa.
Stuczyński, T. Jończyk, K. Korzeniowska-Pacułek, R. Kuś, J. and Terelak, H. (2007). Warunki przyrodnicze ekologicznej produkcji rolniczej a jej stan obecny na obszarze Polski (Natural conditions for organic agricultural production and its current state in Poland). Studia i Raporty IUNG-PIB, Puławy, 5: 55–78.
Szczęsny, R. (1977). Przemiany typów rolnictwa Austrii w latach 1960–1970 (Changes of agricultural types in Austria 1960–1970). Przegląd Geograficzny, 49(4): 741–755.
Szczęsny, R. (1982). Typy rolnictwa Szwajcarii (Types of agriculture in Switzerland). Przegląd Geograficzny, 54(4): 511–531.
Szczęsny, R. (1988). Przemiany struktury przestrzennej rolnictwa Polski w latach 1970–1980. Przestrzenne zróżnicowanie typów rolnictwa (Changes of the spatial structure of Polish agriculture in the years 1970–1980. Spatial differentiation of types of agriculture). Prace habilitacyjne IGiPZ PAN, Ossolineum.
Torero, M. and Maruyama, E. (2016). Progress report on Agricultural Typologies Work for the Program of Accompanying Research for Agricultural Innovation (PARI), http://research4agrinnovation.org/publication/progress-report/
Tyszkiewicz, W. (1977). Typy rolnictwa Macedonii jako przykład rolnictwa światowego (Agricultural types in Macedonia as a sample of the typology of world agriculture). Przegląd Geograficzny, 49(4): 781–804.
Valbuena, D. Verburg, P.H. and Bregt, A.K. (2008). A method to define a typology for agent-based analysis in regional land-use research, Agriculture, Ecosystems and Environment, 128: 27–36. DOI: https://doi.org/10.1016/j.agee.2008.04.015
Wiśniewski Ł. (2019). Struktura przestrzenna polskiego rolnictwa w pierwszej dekadzie XXI wieku (The spatial structure of Polish agriculture in the first decade of the 21st century). Doctoral dissertation type-script, Toruń.
Wysocki, F. (2010). Metody taksonomiczne w rozpoznawaniu typów ekonomicznych rolnictwa i obszarów wiejskich (The methods of taxonomy for recognition of economic types in agriculture and rural areas). Wyd. Uniwersytetu Przyrodniczego w Poznaniu, Poznań.
Central Statistical Office  accessed 5.07.2018 https://stat.gov.pl/metainformacje/slownik-pojec/pojecia-stosowane-w-statystyce-publicznej/1116,pojecie.html